pax_global_header00006660000000000000000000000064141670317650014524gustar00rootroot0000000000000052 comment=16743e96a219cf5c07856b8d7dc36900eecac6c5 Penlight-1.12.0/000077500000000000000000000000001416703176500133575ustar00rootroot00000000000000Penlight-1.12.0/.busted000066400000000000000000000002021416703176500146400ustar00rootroot00000000000000return { default = { verbose = true, output = "gtest", lpath = "./lua/?.lua;./lua/?/init.lua", } } -- vim: ft=lua Penlight-1.12.0/.editorconfig000066400000000000000000000006711416703176500160400ustar00rootroot00000000000000root = true [*] end_of_line = lf insert_final_newline = true trim_trailing_whitespace = true charset = utf-8 [*.lua] indent_style = space indent_size = 2 [kong/templates/nginx*] indent_style = space indent_size = 4 [*.template] indent_style = space indent_size = 4 [Makefile] indent_style = tab Penlight-1.12.0/.github/000077500000000000000000000000001416703176500147175ustar00rootroot00000000000000Penlight-1.12.0/.github/workflows/000077500000000000000000000000001416703176500167545ustar00rootroot00000000000000Penlight-1.12.0/.github/workflows/luacheck.yml000066400000000000000000000007341416703176500212620ustar00rootroot00000000000000name: Luacheck on: [push, pull_request] jobs: luacheck: runs-on: ubuntu-latest steps: - name: Checkout uses: actions/checkout@v2 - name: Setup Lua uses: leafo/gh-actions-lua@v8 with: luaVersion: 5.4 - name: Setup Lua Rocks uses: leafo/gh-actions-luarocks@v4 - name: Setup dependencies run: luarocks install luacheck - name: Run Code Linter run: | luacheck . luarocks lint *.rockspec Penlight-1.12.0/.gitignore000066400000000000000000000000301416703176500153400ustar00rootroot00000000000000luacov.stats.out *.rock Penlight-1.12.0/.luacheckrc000066400000000000000000000011131416703176500154600ustar00rootroot00000000000000unused_args = false redefined = false max_line_length = false globals = { "ngx", } not_globals = { "string.len", "table.getn", } include_files = { "**/*.lua", "*.rockspec", ".busted", ".luacheckrc", } files["spec/**/*.lua"] = { std = "+busted", } exclude_files = { "tests/*.lua", "tests/**/*.lua", -- Travis Lua environment "here/*.lua", "here/**/*.lua", -- GH Actions Lua Environment ".lua", ".luarocks", ".install", -- TODO: fix these files "examples/symbols.lua", "examples/test-symbols.lua", } Penlight-1.12.0/.luacov000066400000000000000000000001011416703176500146410ustar00rootroot00000000000000modules = { ["pl"] = "lua/pl/init.lua", ["pl.*"] = "lua" } Penlight-1.12.0/.travis.yml000066400000000000000000000010241416703176500154650ustar00rootroot00000000000000language: python sudo: false env: - LUA="lua 5.1" - LUA="lua 5.2" - LUA="lua 5.3" - LUA="lua 5.4" - LUA="luajit 2.0" - LUA="luajit 2.0 --compat 5.2" - LUA="luajit 2.1" - LUA="luajit 2.1 --compat 5.2" before_install: - pip install hererocks - hererocks here -r^ --$LUA - source here/bin/activate - luarocks install luacov-coveralls - luarocks install busted install: - luarocks make script: - busted -c -v - lua run.lua tests --luacov - lua run.lua examples after_success: - luacov-coveralls Penlight-1.12.0/CHANGELOG.md000066400000000000000000000651461416703176500152040ustar00rootroot00000000000000# Changelog Versioning is strictly according to [Semantic Versioning](https://semver.org/), see the [README.md](README.md#versioning) for details on version scoping and deprecation policy. see [CONTRIBUTING.md](CONTRIBUTING.md#release-instructions-for-a-new-version) for release instructions ## 1.12.0 (2022-Jan-10) - deprecate: module `pl.text` the contents have moved to `pl.stringx` (removal later) [#407](https://github.com/lunarmodules/Penlight/pull/407) - deprecate: module `pl.xml`, please switch to a more specialized library (removal later) [#409](https://github.com/lunarmodules/Penlight/pull/409) - feat: `utils.npairs` added. An iterator with a range that honours the `n` field [#387](https://github.com/lunarmodules/Penlight/pull/387) - fix: `xml.maptags` would hang if it encountered text-nodes [#396](https://github.com/lunarmodules/Penlight/pull/396) - fix: `text.dedent` didn't handle declining indents nor empty lines [#402](https://github.com/lunarmodules/Penlight/pull/402) - fix: `dir.getfiles`, `dir.getdirectories`, and `dir.getallfiles` now have the directory optional, as was already documented [#405](https://github.com/lunarmodules/Penlight/pull/405) - feat: `array2d.default_range` now also takes a spreadsheet range, which means also other functions now take a range. [#404](https://github.com/lunarmodules/Penlight/pull/404) - fix: `lapp` enums allow [patterns magic characters](https://www.lua.org/pil/20.2.html) [#393](https://github.com/lunarmodules/Penlight/pull/393) - fix: `text.wrap` and `text.fill` numerous fixes for handling whitespace, accented characters, honouring width, etc. [#400](https://github.com/lunarmodules/Penlight/pull/400) - feat: `text.wrap` and `text.fill` have a new parameter to forcefully break words longer than the width given. [#400](https://github.com/lunarmodules/Penlight/pull/400) - fix: `stringx.expandtabs` could error out on Lua 5.3+ [#406](https://github.com/lunarmodules/Penlight/pull/406) - fix: `pl` the module would not properly forward the `newindex` metamethod on the global table. [#395](https://github.com/lunarmodules/Penlight/pull/395) - feat: `utils.enum` added to create enums and prevent magic strings [#408](https://github.com/lunarmodules/Penlight/pull/408) - change: `xml.new` added some sanity checks on input [#397](https://github.com/lunarmodules/Penlight/pull/397) - added: `xml.xml_escape` and `xml.xml_unescape` functions (previously private) [#397](https://github.com/lunarmodules/Penlight/pull/397) - feat: `xml.tostring` now also takes numeric indents (previously only strings) [#397](https://github.com/lunarmodules/Penlight/pull/397) - fix: `xml.walk` now detects recursion (errors out) [#397](https://github.com/lunarmodules/Penlight/pull/397) - fix: `xml.clone` now detects recursion (errors out) [#397](https://github.com/lunarmodules/Penlight/pull/397) - fix: `xml.compare` now detects recursion (errors out) [#397](https://github.com/lunarmodules/Penlight/pull/397) - fix: `xml.compare` text compares now work [#397](https://github.com/lunarmodules/Penlight/pull/397) - fix: `xml.compare` attribute order compares now only compare if both inputs provide an order [#397](https://github.com/lunarmodules/Penlight/pull/397) - fix: `xml.compare` child comparisons failing now report proper error [#397](https://github.com/lunarmodules/Penlight/pull/397) ## 1.11.0 (2021-Aug-18) - fix: `stringx.strip` behaved badly with string lengths > 200 [#382](https://github.com/lunarmodules/Penlight/pull/382) - fix: `path.currentdir` now takes no arguments and calls `lfs.currentdir` without argument [#383](https://github.com/lunarmodules/Penlight/pull/383) - feat: `utils.raise_deprecation` now has an option to NOT include a stack-trace [#385](https://github.com/lunarmodules/Penlight/pull/385) ## 1.10.0 (2021-Apr-27) - deprecate: `permute.iter`, renamed to `permute.order_iter` (removal later) [#360](https://github.com/lunarmodules/Penlight/pull/360) - deprecate: `permute.table`, renamed to `permute.order_table` (removal later) [#360](https://github.com/lunarmodules/Penlight/pull/360) - deprecate: `Date` module (removal later) [#367](https://github.com/lunarmodules/Penlight/pull/367) - feat: `permute.list_iter` to iterate over different sets of values [#360](https://github.com/lunarmodules/Penlight/pull/360) - feat: `permute.list_table` generate table with different sets of values [#360](https://github.com/lunarmodules/Penlight/pull/360) - feat: Lua 5.4 'warn' compatibility function [#366](https://github.com/lunarmodules/Penlight/pull/366) - feat: deprecation functionality `utils.raise_deprecation` [#361](https://github.com/lunarmodules/Penlight/pull/361) - feat: `utils.splitv` now takes same args as `split` [#373](https://github.com/lunarmodules/Penlight/pull/373) - fix: `dir.rmtree` failed to remove symlinks to directories [#365](https://github.com/lunarmodules/Penlight/pull/365) - fix: `pretty.write` could error out on failing metamethods (Lua 5.3+) [#368](https://github.com/lunarmodules/Penlight/pull/368) - fix: `app.parse` now correctly parses values containing '=' or ':' [#373](https://github.com/lunarmodules/Penlight/pull/373) - fix: `dir.makepath` failed to create top-level directories [#372](https://github.com/lunarmodules/Penlight/pull/372) - overhaul: `array2d` module was updated, got additional tests and several documentation updates [#377](https://github.com/lunarmodules/Penlight/pull/377) - feat: `aray2d` now accepts negative indices - feat: `array2d.row` added to align with `column` - fix: bad error message in `array2d.map` - fix: `array2d.flatten` now ensures to deliver a 'square' result if `nil` is encountered - feat: `array2d.transpose` added - feat: `array2d.swap_rows` and `array2d.swap_cols` now return the array - fix: `aray2d.range` correctly recognizes `R` column in spreadsheet format, was mistaken for `R1C1` format. - fix: `aray2d.range` correctly recognizes 2 char column in spreadsheet format - feat: `array2d.default_range` added (previously private) - feat: `array2d.set` if used with a function now passes `i,j` to the function in line with the `new` implementation. - fix: `array2d.iter` didn't properly iterate the indices [#376](https://github.com/lunarmodules/Penlight/issues/376) - feat: `array2d.columns` now returns a second value; the column index - feat: `array2d.rows` added to be in line with `columns` ## 1.9.2 (2020-Sep-27) - fix: dir.walk [#350](https://github.com/lunarmodules/Penlight/pull/350) ## 1.9.1 (2020-Sep-24) - released to superseed the 1.9.0 version which was retagged in git after some distro's already had picked it up. This version is identical to 1.8.1. ## 1.8.1 (2020-Sep-24) (replacing a briefly released but broken 1.9.0 version) ## Fixes - In `pl.class`, `_init` can now be inherited from grandparent (or older ancestor) classes. [#289](https://github.com/lunarmodules/Penlight/pull/289) - Fixes `dir`, `lexer`, and `permute` to no longer use coroutines. [#344](https://github.com/lunarmodules/Penlight/pull/344) ## 1.8.0 (2020-Aug-05) ### New features - `pretty.debug` quickly dumps a set of values to stdout for debug purposes ### Changes - `pretty.write`: now also sorts non-string keys [#319](https://github.com/lunarmodules/Penlight/pull/319) - `stringx.count` has an extra option to allow overlapping matches [#326](https://github.com/lunarmodules/Penlight/pull/326) - added an extra changelog entry for `types.is_empty` on the 1.6.0 changelog, due to additional fixed behaviour not called out appropriately [#313](https://github.com/lunarmodules/Penlight/pull/313) - `path.packagepath` now returns a proper error message with names tried if it fails ### Fixes - Fix: `stringx.rfind` now properly works with overlapping matches [#314](https://github.com/lunarmodules/Penlight/pull/314) - Fix: `package.searchpath` (in module `pl.compat`) [#328](https://github.com/lunarmodules/Penlight/pull/328) - Fix: `path.isabs` now reports drive + relative-path as `false`, eg. "c:some/path" (Windows only) - Fix: OpenResty coroutines, used by `dir.dirtree`, `pl.lexer`, `pl.permute`. If available the original coroutine functions are now used [#329](https://github.com/lunarmodules/Penlight/pull/329) - Fix: in `pl.strict` also predefine global `_PROMPT2` - Fix: in `pl.strict` apply `tostring` to the given name, in case it is not a string. - Fix: the lexer would not recognize numbers without leading zero; "-.123". See [#315](https://github.com/lunarmodules/Penlight/issues/315) ## 1.7.0 (2019-Oct-14) ### New features - `utils.quote_arg` will now optionally take an array of arguments and escape them all into a single string. - `app.parse_args` now accepts a 3rd parameter with a list of valid flags and aliasses - `app.script_name` returns the name of the current script (previously a private function) ### Changes - Documentation updates - `utils.quit`: exit message is no longer required, and closes the Lua state (on 5.2+). - `utils.assert_arg` and `utils.assert_string`: now return the validated value - `pl.compat`: now exports the `jit` and `jit52` flags - `pretty.write`: now sorts the output for easier diffs [#293](https://github.com/lunarmodules/Penlight/pull/293) ### Fixes - `utils.raise` changed the global `on_error`-level when passing in bad arguments - `utils.writefile` now checks and returns errors when writing - `compat.execute` now handles the Windows exitcode -1 properly - `types.is_empty` would return true on spaces always, independent of the parameter - `types.to_bool` will now compare case-insensitive for the extra passed strings - `app.require_here` will now properly handle an absolute base path - `stringx.split` will no longer append an empty match if the number of requested elements has already been reached [#295](https://github.com/lunarmodules/Penlight/pull/295) - `path.common_prefix` and `path.relpath` return the result in the original casing (only impacted Windows) [#297](https://github.com/lunarmodules/Penlight/pull/297) - `dir.copyfile`, `dir.movefile`, and `dir.makepath` create the new file/path with the requested casing, and no longer force lowercase (only impacted Windows) [#297](https://github.com/lunarmodules/Penlight/pull/297) - added a missing assertion on `path.getmtime` [#291](https://github.com/lunarmodules/Penlight/pull/291) - `stringx.rpartition` returned bad results on a not-found [#299](https://github.com/lunarmodules/Penlight/pull/299) ## 1.6.0 (2018-Nov-23) ### New features - `pl.compat` now provides `unpack` as `table.unpack` on Lua 5.1 ### Changes - `utils.unpack` is now documented and respects `.n` field of its argument. - `tablex.deepcopy` and `tablex.deepcompare` are now cycle aware (#262) - Installing through LuaRocks will now include the full rendered documentation ### Fixes - Fixed `seq.last` returning `nil` instead of an empty list when given an empty iterator (#253). - `pl.template` now applies `tostring` when substituting values in templates, avoiding errors when they are not strings or numbers (#256). - Fixed `pl.import_into` not importing some Penlight modules (#268). - Fixed version number stuck at 1.5.2 (#260). - Fixed `types.is_empty` returning `true` on tables containing `false` key (#267). - Fixed `types.is_empty` returning `false` if not a nil/table/string - Fixed `test.assertraise` throwing an error when passed an array with a function to call plus its arguments (#272). - Fixed `test.assertraise` not throwing an error when given function does not error but instead returns a string matching given error pattern. - Fixed placeholder expressions being evaluated with wrong precedence of binary and unary negation. - Fixed placeholder expressions being evaluated assuming wrong binary operator associativity (e.g. `_1-(_2+_3)` was evaluated as `(_1-_2)+_3`. - Fixed placeholder expressions being evaluated as if unary operators take precedence over power operator (e.g. `(-_1)^_2`) was evaluated as `-(_1^2)`). - Fixed vulnerable backtracking pattern in `pl.stringx.strip` (#275) ## 1.5.4 (2017-07-17) ### Fixes - Fixed `compat.execute` behaving differently on Lua 5.1 and 5.1+. - Fixed `lapp.process_options_string` setting global `success` variable. ## 1.5.3 (2017-07-16) ### Changes - Added `template.compile` function that allows caching compiled template and rendering it multiple times. - Added special `_debug` field to environment table argument in `template.substitute` for printing generated template code upon render error. ### Fixes - Fixed error (`attempt to concatenate a nil value (local 'vtype')`) in `lapp.process_options_string`. ## 1.5.2 (2017-04-08) ### Fixes - Removed leftover debug pring in `lapp.process_options_string`. ## 1.5.1 (2017-04-02) ### Fixes - Fixed `dir.getfiles` matching given pattern against full paths from base directory instead of file names. ## 1.5.0 (2017-04-01) ### Changes - `stringx.splitlines` considers `\r\n` a single line ending. - `stringx.splitlines` returns an empty list for an empty string. ### Fixes - `tablex.count_map` no longer raises an error. - `strict.module` correctly handles existing `__index` metamethod returning `false`. - `app.parse_args` accepts colon as a separator between option name and value, as advertised. - `pretty.load` handles case where a C hook is present. ' `os.execute` had issue with LuaJIT in 5.2 compat mode. ### Features - `template` supports customizing inline escape character and chunk name. - `seq` constructor supports iterators with a state object as the second argument. - `stringx.splitlines` has `keep_ends` argument. - `tablex.reduce` can take an optional initial value. ## 1.4.1 (2016-08-16) ### Changes - All functions that return instances of `pl.List`, `pl.Map` and `pl.Set` now require corresponding modules, so that their methods always work right away. ### Fixes - Fixed `dir.getallfiles` returning an empty array when called without `pattern` argument. ### Features ## 1.4.0 (2016-08-14) ### Changes ### Fixes - `pl.path` covers edge cases better (e.g `path.normpath` was broken) - `p.dir` shell patterns fixed - `os.tmpname` broken on modern Windows/MSVC14 - (likewise for `utils.executeex` which depends on it) - `pretty.write` more robust and does not lose floating-point precision; saves and restores debug hooks when loading. - `pl.lexer` fixes: `cpp` lexer now filters space by default - `tablex.sortv` no longer assumes that the values are all unique - `stringx.center` is now consistent with Python; `stringx.rfind` and `string.quote_string` fixed. - `data.write` had a problem with default delimiter, properly returns error now. - `pl.Set` `+` and `-` now have correct semantics ### Features - `pl.tablex` has `union` and `merge` convenience functions - `pl.lapp` understands '--' meaning end of parsed arguments - `utils.quote_arg` quotes command arguments for `os.execute`, correctly handling all special characters. - `utils.writefile` has optional `is_bin` argument - 'pl.lexer' supports line numbers with string argument - `stringx.endswith` may be passed an array of possible suffixes. - `data.read` - in CSV mode, assume empty fields are numerical zero ## 1.3.2 (2015-05-10) ### Changes - now works and passes tests with Lua 5.3 - utils.import will NOT override global symbols (import 'math' caused global type() to be clobbered) - Updated pl.dir.file_op to return true on success and false on failure... - workaround for issues with pl.lapp with amalg.lua - will look at global LAPP_SCRIPT if arg[0] is nil ### Fixes - func was broken: do NOT use ipairs to iterate if __index is overriden! - issue #133 pretty.read (naively) confused by unbalanced brackets - xml attribute underscore fix for simple parser - Fix path.normpath - lexer: fix parsing block comments/string. fix hang on empty string. - Fixed utils.execute returning different values for Lua 5.1 and Lua 5.2 - Issue #97; fixed attempt to put a month into a day - problem with tablex.count_map with custom comparison - tablex.pairmap overwrites result if key already exists; instead, upon detection that key already exists for a returned value, we modify the key's value to be a table and insert values into that table ### Features - Add Python style url module for quote and unquote. - stringx.quote_string, which scans for embedded long-string quote matches and escapes them by creating a long-string quote. - issue #117: tablex.range now works with decreasing numbers, consistent with numerical for loop - utils.import will NOT override global symbols (import 'math' caused global type() to be clobbered) - issue #125: DOCTYPE ignored in xml documents as well - Allow XML tostring() function to customize the default prefacing with `` - More Robust Quoted Strings - lapp: improved detection of unsupported short flags ## 1.3.1 (2013-09-24) ## 1.3.0 (2013-09-14) ### Changes - class: RIP base method - not possible to implement correctly - lapp: short flags can now always be followed directly by their value, for instance, `-I/usr/include/lua/5.1` - Date: new explicit `Date.Interval` class; `toUTC/toLocal` return new object; `Date.__tostring` always returns ISO 8601 times for exact serialization. `+/-` explicit operators. Date objects are explicitly flagged as being UTC or not. ### Fixes - class: super method fixed. - Date: DST is now accounted for properly. - Date: weekday calculation borked. ### Features - All tests pass with no-5.1-compatible Lua 5.2; now always uses `utils.load` and `utils.unpack` is always available. - types: new module containing `utils.is_xxx` methods plus new `to_bool`. - class: can be passed methods in a table (see `test=klass.lua`). This is particularly convenient for using from Moonscript. - general documentation improvements, e.g `class` ## 1.2.1 (2013-06-21) ### Changes - utils.set(get)fenv always defined (_not_ set as globals for 5.2 anymore!). These are defined in new module pl.compat, but still available through utils. - class.Frodo now puts 'Frodo' in _current environment_ ### Fixes - lapp.add_type was broken (Pete Kazmier) - class broke with classes that redefined __newindex - Set.isdisjoint was broken because of misspelling; default ctor Set() now works as expected - tablex.transform was broken; result now has same keys as original (CoolistheName007) - xml match not handling empty matches (royalbee) - pl.strict: assigning nil to global declares it, as God intended. (Pierre Chapuis) - tests all work with pl.strict - 5.2 compatible load now respects mode - tablex.difference thought that a value of `false` meant 'not present' (Andrew Starke) ### Features - tablex.sort(t) iterates over sorted keys, tablex.sortv(t) iterates over sorted values (Pete Kazmier) - tablex.readonly(t) creates a read-only proxy for a table (John Schember) - utils.is_empty(o) true if o==nil, o is an empty table, or o is an empty string (John Schember) - utils.executeex(cmd,bin) returns true if successful, return code, plus stdout and stderr output as strings. (tieske) - class method base for calling inherited methods (theypsilon) - class supports pre-constructor _create for making a custom self (used in pl.List) - xml HTML mode improvements - can parse non-trivial well-formed HTML documents. xml.parsehtml is a parse function, no longer a flag - if a LOM document has ordered attributes, use these when stringifying - xml.tostring has yet another extra parm to force prefacing with `` - lapp boolean flags may have `true` default - lapp slack mode where 'short' flags can be multi-char - test.asserteq etc take extra arg, which is extra level where error must be reported at - path.currentdir,chdir,rmdir,mkdir and dir as alias to lfs are exported; no dependencies on luafilesystem outside pl.path, making it easier to plug in different implementations. ## 1.2.0 (2013-05-28) ## 1.1.1 (2013-05-14) ## 1.1.0 (2013-03-18) ## 1.0.3 (2012-12-07) ## 1.0.2 (2012-05-12) ## 1.0.1 (2012-05-26) ## 1.0.0 (2012-04-26) ## 0.9.8 (2011-11-27) ## 0.9.7 (2011-11-27) ### Lua 5.2 compatibility (These are all now defined in pl.utils) - setfenv, getfenv defined for Lua 5.2 (by Sergey Rozhenko) ### Changes - array2d.flatten is new - OrderedMap:insert is new ### Fixes - seq.reduce re-implemented to give correct order (Carl Ã…dahl) - seq.unique was broken: new test - tablex.icopy broken for last argument; new test - utils.function_arg last parm 'msg' was missing - array2d.product was broken; more sensible implementation - array2d.range, .slice, .write were broken - text optional operator % overload broken for 'fmt % fun'; new tests - a few occurances of non-existent function utils.error removed ## 0.9.6 (2011-09-11) ### Lua 5.2 compatibility - Bad string escape in tests fixed ### Changes - LuaJIT FFI used on Windows for Copy/MoveFile functionality ### Fixes - Issue 13 seq.sort now calls seq.copy - issue 14 bad pattern to escape trailing separators in path.abspath - lexer: string tokens broken with some combinations - lexer: long comments broken for Lua and C - stringx.split behaves according to Python spec; extra parm meaning 'max splits' - stringx.title behaves according to Python spec - stringx.endswith broken for 2nd arg being table of postfixes - OrderedMap.set broken when value was nil and key did not exist in map; ctor throws error if unhappy ## 0.9.5 (2011-07-05) ### Lua 5.2 compatibility - defines Lua 5.2 beta compatible load() - defines table.pack() ### New functions - stringx.title(): translates "a dog's day" to "A Dog's Day" - path.normpath(): translates 'A//B','A/./B' and 'A/C/../B' to 'A/B' - utils.execute(): returns ok,return-code: compatible with 5.1 and 5.2 ### Fixes - pretty.write() _always_ returns a string, but will return also an error string if the argument is not a table. Non-integer indices between 1 and #t are no longer falsely considered part of the array - stringx.expandtabs() now works like the Python string method; it will expand each field up to the next tab stop - path.normcase() was broken, because of a misguided attempt to normalize the path. - UNC specific fix to path.abspath() - UNC paths recognized as absolute; dir.makedir() works here - utils.quit() varargs broken, e.g. utils.quit("answer was %d",42) - some stray globals caused trouble with 'strict' ## 0.9.4 (2011-04-08) ## 0.9.3 (2011-03-05) ## 0.9.2 (2011-02-16) ## 0.9.1 (2011-02-12) ## 0.9.0 (2010-12-20) ## 0.8.5 (2010-12-16) ### What's new with 0.8b ? #### Features: pl.app provides useful stuff like simple command-line argument parsing and require_here(), which makes subsequent require() calls look in the local directory by preference. p.file provides useful functions like copy(),move(), read() and write(). (These are aliases to dir.copyfile(),movefile(),utils.readfile(),writefile()) Custom error trace will only show the functions in user code. More robust argument checking. In function arguments, now supports 'string lambdas', e.g. `'|x| 2*x'` utils.readfile,writefile now insist on being given filenames. This will cause less confusion. tablex.search() is new: will look recursively in an arbitrary table; can specify tables not to follow. tablex.move() will work with source and destination tables the same, with overlapping ranges. #### Bug Fixes: dir.copyfile() now works fine without Alien on Windows dir.makepath() and rmtree() had problems. tablex.compare_no_order() is now O(NlogN), as expected. tablex.move() had a problem with source size ### What's New with 0.7.0b? #### Features: utils.is_type(v,tp) can say is_type(s,'string') and is_type(l,List). utils.is_callable(v) either a function, or has a `__call` metamethod. Sequence wrappers: can write things like this: seq(s):last():filter('<'):copy() seq:mapmethod(s,name) - map using a named method over a sequence. seq:enum(s) If s is a simple sequence, then for i,v in seq.enum(s) do print(i,v) end seq:take(s,n) Grab the next n values from a (possibly infinite) sequence. In a related change suggested by Flemming Madsden, the in-place List methods like reverse() and sort() return the list, allowing for method chaining. list.join() explicitly converts using tostring first. tablex.count_map() like seq.count_map(), but takes an equality function. tablex.difference() set difference tablex.set() explicit set generator given a list of values Template.indent_substitute() is a new Template method which adjusts for indentation and can also substitute templates themselves. pretty.read(). This reads a Lua table (as dumped by pretty.write) and attempts to be paranoid about its contents. sip.match_at_start(). Convenience function for anchored SIP matches. #### Bug Fixes: tablex.deepcompare() was confused by false boolean values, which it thought were synonymous with being nil. pretty.write() did not handle cycles, and could not display tables with 'holes' properly (Flemming Madsden) The SIP pattern '$(' was not escaped properly. sip.match() did not pass on options table. seq.map() was broken for double-valued sequences. seq.copy_tuples() did not use default_iter(), so did not e.g. like table arguments. dir.copyfile() returns the wrong result for \*nix operations. dir.makepath() was broken for non-Windows paths. ### What's New with 0.6.3? The map and reduce functions now take the function first, as Nature intended. The Python-like overloading of '\*' for strings has been dropped, since it is silly. Also, strings are no longer callable; use 's:at(1)' instead of 's(1)' - this tended to cause Obscure Error messages. Wherever a function argument is expected, you can use the operator strings like '+','==',etc as well as pl.operator.add, pl.operator.eq, etc. (see end of pl/operator.lua for the full list.) tablex now has compare() and compare_no_order(). An explicit set() function has been added which constructs a table with the specified keys, all set to a value of true. List has reduce() and partition() (This is a cool function which separates out elements of a list depending on a classifier function.) There is a new array module which generalizes tablex operations like map and reduce for two-dimensional arrays. The famous iterator over permutations from PiL 9.3 has been included. David Manura's list comprehension library has been included. Also, utils now contains his memoize function, plus a useful function args which captures the case where varargs contains nils. There was a bug with dir.copyfile() where the flag was the wrong way round. config.lines() had a problem with continued lines. Some operators were missing in pl.operator; have renamed them to be consistent with the Lua metamethod names. Penlight-1.12.0/CONTRIBUTING.md000066400000000000000000000047511416703176500156170ustar00rootroot00000000000000Contributing to Penlight ======================== So you want to contribute to Penlight? Fantastic! Here's a brief overview on how best to do so. ## What to change Here's some examples of things you might want to make a pull request for: * New features * Bugfixes * Inefficient blocks of code If you have a more deeply-rooted problem with how the library is built or some of the stylistic decisions made in the code, it's best to [create an issue](https://github.com/lunarmodules/Penlight/issues) before putting the effort into a pull request. The same goes for new features - it might be best to check the project's direction, existing pull requests, and currently open and closed issues first. ## Using Git appropriately Here's how to go about contributing to Penlight: 1. [Fork the repository](https://github.com/lunarmodules/Penlight/fork) to your Github account. 2. Create a *topical branch* - a branch whose name is succint but explains what you're doing, such as _"added-klingon-cloacking-device"_ - from `master` branch. 3. Make your changes, committing at logical breaks. 4. Push your branch to your personal account 5. [Create a pull request](https://help.github.com/articles/using-pull-requests) 6. Watch for comments or acceptance If you wanna be a rockstar; 1. Update the [CHANGELOG.md](https://github.com/lunarmodules/Penlight/blob/master/CHANGELOG.md) file 2. [Add tests](https://github.com/lunarmodules/Penlight/tree/master/tests) that show the defect your fix repairs, or that tests your new feature Please note - if you want to change multiple things that don't depend on each other, make sure you check out the `master` branch again and create a different topical branch before making more changes - that way we can take in each change separately. ## Release instructions for a new version - create a new release branch - update `./lua/pl/utils.lua` (the `_VERSION` constant) - update `./config.ld` with the new version number - create a new rockspec file for the version in `./rockspecs` - check the `./CHANGELOG.md` files for completeness - commit the release related changes with `release x.y.z` - render the documentation using `ldoc .` - commit the documentation as a separate commit with `release x.y.z docs` - push the release branch and create a PR - merge the PR - tag the release as `x.y.z` and push the tag to the github repo - upload the rockspec, and source rock files to LuaRocks - test installing through LuaRocks - announce the release on the Lua mailing list Penlight-1.12.0/LICENSE.md000066400000000000000000000020661416703176500147670ustar00rootroot00000000000000Copyright (C) 2009-2016 Steve Donovan, David Manura. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. Penlight-1.12.0/README.md000066400000000000000000000130121416703176500146330ustar00rootroot00000000000000# Penlight Lua Libraries [![Travis build status](https://img.shields.io/travis/lunarmodules/Penlight/master?logo=Travis)](https://travis-ci.org/lunarmodules/Penlight) [![AppVeyor build status](https://img.shields.io/appveyor/build/Tieske/penlight-ta1gi/master?logo=appveyor)](https://ci.appveyor.com/project/Tieske/penlight-ta1gi/branch/master) [![Coveralls code coverage](https://img.shields.io/coveralls/github/lunarmodules/Penlight?logo=coveralls)](https://coveralls.io/github/lunarmodules/Penlight) [![Luacheck](https://github.com/lunarmodules/Penlight/workflows/Luacheck/badge.svg)](https://github.com/lunarmodules/Penlight/actions) [![SemVer](https://img.shields.io/github/v/tag/lunarmodules/Penlight?color=brightgreen&label=SemVer&logo=semver&sort=semver)](CHANGELOG.md) ## Why a new set of libraries? Penlight brings together a set of generally useful pure Lua modules, focusing on input data handling (such as reading configuration files), functional programming (such as map, reduce, placeholder expressions, etc), and OS path management. Much of the functionality is inspired by the Python standard libraries. ## Module Overview ### Paths, Files and Directories * `path`: queries like `isdir`,`isfile`,`exists`, splitting paths like `dirname` and `basename` * `dir`: listing files in directories (`getfiles`,`getallfiles`) and creating/removing directory paths * `file`: `copy`,`move`; read/write contents with `read` and `write` ### Application Support * `app`: `require_here` to rebase `require` to work with main script path; simple argument parsing `parse_args` * `lapp`: sophisticated usage-text-driven argument parsing for applications * `config`: flexibly read Unix config files and Windows INI files * `strict`: check for undefined global variables - can use `strict.module` for modules * `utils`,`compat`: Penlight support for unified Lua 5.1/5.2 codebases * `types`: predicates like `is_callable` and `is_integer`; extended `type` function. ### Extra String Operations * `utils`: can split a string with a delimiter using `utils.split` * `stringx`: extended string functions covering the Python `string` type * `stringio`: open strings for reading, and creating strings using standard Lua IO methods * `lexer`: lexical scanner for splitting text into tokens; special cases for Lua and C * `text`: indenting and dedenting text, wrapping paragraphs; optionally make `%` work as in Python * `template`: small but powerful template expansion engine * `sip`: Simple Input Patterns - higher-level string patterns for parsing text ### Extra Table Operations * `tablex`: copying, comparing and mapping over * `pretty`: pretty-printing Lua tables, and various safe ways to load Lua as data * `List`: implementation of Python 'list' type - slices, concatenation and partitioning * `Map`, `Set`, `OrderedMap`: classes for specialized kinds of tables * `data`: reading tabular data into 2D arrays and efficient queries * `array2d`: operations on 2D arrays * `permute`: generate permutations ### Iterators, OOP and Functional * `seq`: working with iterator pipelines; collecting iterators as tables * `class`: a simple reusable class framework * `func`: symbolic manipulation of expressions and lambda expressions * `utils`: `utils.string_lambda` converts short strings like '|x| x^2' into functions * `comprehension`: list comprehensions: `C'x for x=1,4'()=={1,2,3,4}` ## Versioning Penlight is strictly versioned according to [Semantic Versioning](https://semver.org/). In scope of the version: * functionality provided by Penlight modules/classes * based on stock Lua PuC-Rio or LuaJIT Not in scope of the version: * Documentation * Error messages (textual changes) * Deprecation warnings (by default to `stderr`) ### Deprecating functionality Any version may deprecate functionality. So new deprecation notices may appear in major, minor, and patch releases. Final removal of functionality (assuming it is a breaking change) will only be done in a major version. It is strongly suggested to use the deprecation warning mechanism to test usage of deprecated functionalities when upgrading. This is done by enabling the warning system (in Lua 5.4, or the Penlight compatibility function for earlier versions): ```lua require "pl.compat" warn "@on" ``` See `pl.utils.raise_deprecation` for more info. ## License Penlight is distributed under the [MIT license](LICENSE.md). ## Installation Using [LuaRocks](https://luarocks.org): simply run `luarocks install penlight`. Manually: copy `lua/pl` directory into your Lua module path. It's typically `/usr/local/share/lua/5.x` on a Linux system and `C:\Program Files\Lua\5.x\lua` for Lua for Windows. ## Dependencies The file and directory functions depend on [LuaFileSystem](https://keplerproject.github.io/luafilesystem/), which is installed automatically if you are using LuaRocks. Additionally, if you want `dir.copyfile` to work elegantly on Windows, then you need [Alien](http://mascarenhas.github.io/alien/). Both libraries are present in Lua for Windows. ## Building the Documentation Requires [ldoc](https://github.com/stevedonovan/LDoc), which is available through LuaRocks. Then it's a simple matter of running `ldoc .` from the repo. ## Contributing Contributions are most welcome, please check the [contribution guidelines](CONTRIBUTING.md). ## Running tests Execute `lua run.lua tests` to run the tests. Execute `lua run.lua examples` to run examples. ## History For a complete history of the development of Penlight, please check the [changelog](CHANGELOG.md). Penlight-1.12.0/appveyor.yml000066400000000000000000000014761416703176500157570ustar00rootroot00000000000000shallow_clone: true environment: COVERALLS_REPO_TOKEN: secure: /0zDn4cz8xWKQuHfzvAUwyeFSQfLSP+toDR3lRApwWoIeg9O1OvJTnXurB8cssW0 matrix: - LUA: "lua 5.1" - LUA: "lua 5.2" - LUA: "lua 5.3" - LUA: "lua 5.4" - LUA: "luajit 2.0" - LUA: "luajit 2.0 --compat 5.2" - LUA: "luajit 2.1" - LUA: "luajit 2.1 --compat 5.2" before_build: - set PATH=C:\Python27\Scripts;%PATH% - pip install hererocks - hererocks here --%LUA% -rlatest - call here\bin\activate - luarocks install luacov-coveralls - luarocks install busted build_script: - luarocks make test_script: - busted -c -v - lua run.lua tests --luacov - lua run.lua examples on_success: # secure coveralls token not available on PR builds, only BRANCH builds - "if not \"%COVERALLS_REPO_TOKEN%\"==\"\" ( luacov-coveralls )" Penlight-1.12.0/config.ld000066400000000000000000000015521416703176500151500ustar00rootroot00000000000000project = 'Penlight' description = 'Penlight Lua Libraries 1.12.0' full_description = 'Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. Visit the GitHub project to review the code or file issues. Skip to the @{01-introduction.md|introduction}.' title = 'Penlight Documentation' dir = 'docs' style = '!fixed' template = true use_markdown_titles = true topics = 'docs_topics' examples = {'./examples','./tests/test-data.lua'} package = 'pl' format = 'discount' sort_modules=true file = './lua/pl' kind_names={topic='Manual',module='Libraries'} tparam_alias('array','array') tparam_alias('array2d','array') alias('ret',{'return',modifiers={type="$1"}}) Penlight-1.12.0/docs/000077500000000000000000000000001416703176500143075ustar00rootroot00000000000000Penlight-1.12.0/docs/classes/000077500000000000000000000000001416703176500157445ustar00rootroot00000000000000Penlight-1.12.0/docs/classes/pl.Date.html000066400000000000000000000563301416703176500201300ustar00rootroot00000000000000 Penlight Documentation

Class pl.Date

Date and Date Format classes.

See the Guide.

NOTE: the date module is deprecated! see https://github.com/lunarmodules/Penlight/issues/285

Dependencies: pl.class, pl.stringx, pl.utils

Functions

Date:set (t) set the current time of this Date object.
Date.tzone (ts) get the time zone offset from UTC.
Date:toUTC () convert this date to UTC.
Date:toLocal () convert this UTC date to local.
Date:year (y) set the year.
Date:month (m) set the month.
Date:day (d) set the day.
Date:hour (h) set the hour.
Date:min (min) set the minutes.
Date:sec (sec) set the seconds.
Date:yday (yday) set the day of year.
Date:year (y) get the year.
Date:month () get the month.
Date:day () get the day.
Date:hour () get the hour.
Date:min () get the minutes.
Date:sec () get the seconds.
Date:yday () get the day of year.
Date:weekday_name (full) name of day of week.
Date:month_name (full) name of month.
Date:is_weekend () is this day on a weekend?.
Date:add (t) add to a date object.
Date:last_day () last day of the month.
Date:diff (other) difference between two Date objects.
Date:__tostring () long numerical ISO data format version of this date.
Date:__eq (other) equality between Date objects.
Date:__lt (other) ordering between Date objects.
Date:__sub () difference between Date objects.
Date:__add (other) add a date and an interval.
Date.Interval (t) Date.Interval constructor
Date.Interval:__tostring () If it's an interval then the format is '2 hours 29 sec' etc.
Date.Format (fmt.) Date.Format constructor.
Date.Format:parse (str) parse a string into a Date object.
Date.Format:tostring (d) convert a Date object into a string.
Date.Format:US_order (yesno) force US order in dates like 9/11/2001

Methods

pl.Date:Date (t, ...) Date constructor.


Functions

Date:set (t)
set the current time of this Date object.

Parameters:

  • t int seconds since epoch
Date.tzone (ts)
get the time zone offset from UTC.

Parameters:

  • ts int seconds ahead of UTC
Date:toUTC ()
convert this date to UTC.
Date:toLocal ()
convert this UTC date to local.
Date:year (y)
set the year.

Parameters:

  • y int Four-digit year
Date:month (m)
set the month.

Parameters:

  • m int month
Date:day (d)
set the day.

Parameters:

  • d int day
Date:hour (h)
set the hour.

Parameters:

  • h int hour
Date:min (min)
set the minutes.

Parameters:

  • min int minutes
Date:sec (sec)
set the seconds.

Parameters:

  • sec int seconds
Date:yday (yday)
set the day of year.

Parameters:

  • yday int day of year
Date:year (y)
get the year.

Parameters:

  • y int Four-digit year
Date:month ()
get the month.
Date:day ()
get the day.
Date:hour ()
get the hour.
Date:min ()
get the minutes.
Date:sec ()
get the seconds.
Date:yday ()
get the day of year.
Date:weekday_name (full)
name of day of week.

Parameters:

  • full bool abbreviated if true, full otherwise.

Returns:

    string name
Date:month_name (full)
name of month.

Parameters:

  • full int abbreviated if true, full otherwise.

Returns:

    string name
Date:is_weekend ()
is this day on a weekend?.
Date:add (t)
add to a date object.

Parameters:

Returns:

    this date
Date:last_day ()
last day of the month.

Returns:

    int day
Date:diff (other)
difference between two Date objects.

Parameters:

  • other Date Date object

Returns:

    Date.Interval object
Date:__tostring ()
long numerical ISO data format version of this date.
Date:__eq (other)
equality between Date objects.

Parameters:

  • other
Date:__lt (other)
ordering between Date objects.

Parameters:

  • other
Date:__sub ()
difference between Date objects.
Date:__add (other)
add a date and an interval.

Parameters:

Date.Interval (t)
Date.Interval constructor

Parameters:

  • t int an interval in seconds
Date.Interval:__tostring ()
If it's an interval then the format is '2 hours 29 sec' etc.
Date.Format (fmt.)
Date.Format constructor.

Parameters:

  • fmt. string A string where the following fields are significant:

    • d day (either d or dd)
    • y year (either yy or yyy)
    • m month (either m or mm)
    • H hour (either H or HH)
    • M minute (either M or MM)
    • S second (either S or SS)

    Alternatively, if fmt is nil then this returns a flexible date parser that tries various date/time schemes in turn:

    • ISO 8601, like 2010-05-10 12:35:23Z or 2008-10-03T14:30+02
    • times like 15:30 or 8.05pm (assumed to be today's date)
    • dates like 28/10/02 (European order!) or 5 Feb 2012
    • month name like march or Mar (case-insensitive, first 3 letters); here the day will be 1 and the year this current year

    A date in format 3 can be optionally followed by a time in format 2. Please see test-date.lua in the tests folder for more examples.

Usage:

    df = Date.Format("yyyy-mm-dd HH:MM:SS")
Date.Format:parse (str)
parse a string into a Date object.

Parameters:

Returns:

    date object
Date.Format:tostring (d)
convert a Date object into a string.

Parameters:

  • d a date object, or a time value as returned by os.time

Returns:

    string
Date.Format:US_order (yesno)
force US order in dates like 9/11/2001

Parameters:

  • yesno

Methods

pl.Date:Date (t, ...)
Date constructor.

Parameters:

  • t

    this can be either

    • nil or empty - use current date and time
    • number - seconds since epoch (as returned by os.time). Resulting time is UTC
    • Date - make a copy of this date
    • table - table containing year, month, etc as for os.time. You may leave out year, month or day, in which case current values will be used.
    • year (will be followed by month, day etc)
  • ... true if Universal Coordinated Time, or two to five numbers: month,day,hour,min,sec
generated by LDoc 1.4.6
Penlight-1.12.0/docs/classes/pl.List.html000066400000000000000000001073271416703176500201710ustar00rootroot00000000000000 Penlight Documentation

Class pl.List

Python-style list class.

Please Note: methods that change the list will return the list. This is to allow for method chaining, but please note that ls = ls:sort() does not mean that a new copy of the list is made. In-place (mutable) methods are marked as returning 'the list' in this documentation.

See the Guide for further discussion

See http://www.python.org/doc/current/tut/tut.html, section 5.1

Note: The comments before some of the functions are from the Python docs and contain Python code.

Written for Lua version Nick Trout 4.0; Redone for Lua 5.1, Steve Donovan.

Dependencies: pl.utils, pl.tablex, pl.class

Functions

List.new ([t]) Create a new list.
List:clone () Make a copy of an existing list.
List:append (i) Add an item to the end of the list.
List:extend (L) Extend the list by appending all the items in the given list.
List:insert (i, x) Insert an item at a given position.
List:put (x) Insert an item at the begining of the list.
List:remove (i) Remove an element given its index.
List:remove_value (x) Remove the first item from the list whose value is given.
List:pop ([i]) Remove the item at the given position in the list, and return it.
List:index (x[, idx=1]) Return the index in the list of the first item whose value is given.
List:contains (x) Does this list contain the value?
List:count (x) Return the number of times value appears in the list.
List:sort ([cmp='<']) Sort the items of the list, in place.
List:sorted ([cmp='<']) Return a sorted copy of this list.
List:reverse () Reverse the elements of the list, in place.
List:minmax () Return the minimum and the maximum value of the list.
List:slice (first, last) Emulate list slicing.
List:clear () Empty the list.
List.range (start[, finish[, incr=1]]) Emulate Python's range(x) function.
List:len () list:len() is the same as #list.
List:chop (i1, i2) Remove a subrange of elements.
List:splice (idx, list) Insert a sublist into a list equivalent to 's[idx:idx] = list' in Python
List:slice_assign (i1, i2, seq) General slice assignment s[i1:i2] = seq.
List:join ([delim='']) Join the elements of a list using a delimiter.
List:concat ([delim='']) Join a list of strings.
List:foreach (fun, ...) Call the function on each element of the list.
List:foreachm (name, ...) Call the named method on each element of the list.
List:filter (fun[, arg]) Create a list of all elements which match a function.
List.split (s[, delim]) Split a string using a delimiter.
List:map (fun, ...) Apply a function to all elements.
List:transform (fun, ...) Apply a function to all elements, in-place.
List:map2 (fun, ls, ...) Apply a function to elements of two lists.
List:mapm (name, ...) apply a named method to all elements.
List:reduce (fun) 'reduce' a list using a binary function.
List:partition (fun, ...) Partition a list using a classifier function.
List:iter () return an iterator over all values.
List.iterate (seq) Create an iterator over a seqence.

metamethods

List:__concat (L) Concatenation operator.
List:__eq (L) Equality operator ==.
List:__tostring () How our list should be rendered as a string.


Functions

List.new ([t])
Create a new list. Can optionally pass a table; passing another instance of List will cause a copy to be created; this will return a plain table with an appropriate metatable. we pass anything which isn't a simple table to iterate() to work out an appropriate iterator

Parameters:

  • t An optional list-like table (optional)

Returns:

    a new List

See also:

Usage:

    ls = List();  ls = List {1,2,3,4}
List:clone ()
Make a copy of an existing list. The difference from a plain 'copy constructor' is that this returns the actual List subtype.
List:append (i)
Add an item to the end of the list.

Parameters:

  • i An item

Returns:

    the list
List:extend (L)
Extend the list by appending all the items in the given list. equivalent to 'a[len(a):] = L'.

Parameters:

  • L List Another List

Returns:

    the list
List:insert (i, x)
Insert an item at a given position. i is the index of the element before which to insert.

Parameters:

  • i int index of element before whichh to insert
  • x A data item

Returns:

    the list
List:put (x)
Insert an item at the begining of the list.

Parameters:

  • x a data item

Returns:

    the list
List:remove (i)
Remove an element given its index. (equivalent of Python's del s[i])

Parameters:

  • i int the index

Returns:

    the list
List:remove_value (x)
Remove the first item from the list whose value is given. (This is called 'remove' in Python; renamed to avoid confusion with table.remove) Return nil if there is no such item.

Parameters:

  • x A data value

Returns:

    the list
List:pop ([i])
Remove the item at the given position in the list, and return it. If no index is specified, a:pop() returns the last item in the list. The item is also removed from the list.

Parameters:

  • i int An index (optional)

Returns:

    the item
List:index (x[, idx=1])
Return the index in the list of the first item whose value is given. Return nil if there is no such item.

Parameters:

  • x A data value
  • idx int where to start search (default 1)

Returns:

    the index, or nil if not found.
List:contains (x)
Does this list contain the value?

Parameters:

  • x A data value

Returns:

    true or false
List:count (x)
Return the number of times value appears in the list.

Parameters:

  • x A data value

Returns:

    number of times x appears
List:sort ([cmp='<'])
Sort the items of the list, in place.

Parameters:

  • cmp func an optional comparison function (default '<')

Returns:

    the list
List:sorted ([cmp='<'])
Return a sorted copy of this list.

Parameters:

  • cmp func an optional comparison function (default '<')

Returns:

    a new list
List:reverse ()
Reverse the elements of the list, in place.

Returns:

    the list
List:minmax ()
Return the minimum and the maximum value of the list.

Returns:

  1. minimum value
  2. maximum value
List:slice (first, last)
Emulate list slicing. like 'list[first:last]' in Python. If first or last are negative then they are relative to the end of the list eg. slice(-2) gives last 2 entries in a list, and slice(-4,-2) gives from -4th to -2nd

Parameters:

  • first An index
  • last An index

Returns:

    a new List
List:clear ()
Empty the list.

Returns:

    the list
List.range (start[, finish[, incr=1]])
Emulate Python's range(x) function. Include it in List table for tidiness

Parameters:

  • start int A number
  • finish int A number greater than start; if absent, then start is 1 and finish is start (optional)
  • incr int an increment (may be less than 1) (default 1)

Returns:

    a List from start .. finish

Usage:

  • List.range(0,3) == List{0,1,2,3}
  • List.range(4) = List{1,2,3,4}
  • List.range(5,1,-1) == List{5,4,3,2,1}
List:len ()
list:len() is the same as #list.
List:chop (i1, i2)
Remove a subrange of elements. equivalent to 'del s[i1:i2]' in Python.

Parameters:

  • i1 int start of range
  • i2 int end of range

Returns:

    the list
List:splice (idx, list)
Insert a sublist into a list equivalent to 's[idx:idx] = list' in Python

Parameters:

  • idx int index
  • list List list to insert

Returns:

    the list

Usage:

    l = List{10,20}; l:splice(2,{21,22});  assert(l == List{10,21,22,20})
List:slice_assign (i1, i2, seq)
General slice assignment s[i1:i2] = seq.

Parameters:

  • i1 int start index
  • i2 int end index
  • seq List a list

Returns:

    the list
List:join ([delim=''])
Join the elements of a list using a delimiter. This method uses tostring on all elements.

Parameters:

  • delim string a delimiter string, can be empty. (default '')

Returns:

    a string
List:concat ([delim=''])
Join a list of strings.
Uses table.concat directly.

Parameters:

  • delim string a delimiter (default '')

Returns:

    a string
List:foreach (fun, ...)
Call the function on each element of the list.

Parameters:

  • fun func a function or callable object
  • ... optional values to pass to function
List:foreachm (name, ...)
Call the named method on each element of the list.

Parameters:

  • name string the method name
  • ... optional values to pass to function
List:filter (fun[, arg])
Create a list of all elements which match a function.

Parameters:

  • fun func a boolean function
  • arg optional argument to be passed as second argument of the predicate (optional)

Returns:

    a new filtered list.
List.split (s[, delim])
Split a string using a delimiter.

Parameters:

  • s string the string
  • delim string the delimiter (default spaces) (optional)

Returns:

    a List of strings

See also:

List:map (fun, ...)
Apply a function to all elements. Any extra arguments will be passed to the function.

Parameters:

  • fun func a function of at least one argument
  • ... arbitrary extra arguments.

Returns:

    a new list: {f(x) for x in self}

See also:

Usage:

    List{'one','two'}:map(string.upper) == {'ONE','TWO'}
List:transform (fun, ...)
Apply a function to all elements, in-place. Any extra arguments are passed to the function.

Parameters:

  • fun func A function that takes at least one argument
  • ... arbitrary extra arguments.

Returns:

    the list.
List:map2 (fun, ls, ...)
Apply a function to elements of two lists. Any extra arguments will be passed to the function

Parameters:

  • fun func a function of at least two arguments
  • ls List another list
  • ... arbitrary extra arguments.

Returns:

    a new list: {f(x,y) for x in self, for x in arg1}

See also:

List:mapm (name, ...)
apply a named method to all elements. Any extra arguments will be passed to the method.

Parameters:

  • name string name of method
  • ... extra arguments

Returns:

    a new list of the results

See also:

List:reduce (fun)
'reduce' a list using a binary function.

Parameters:

  • fun func a function of two arguments

Returns:

    result of the function

See also:

List:partition (fun, ...)
Partition a list using a classifier function. The function may return nil, but this will be converted to the string key ''.

Parameters:

  • fun func a function of at least one argument
  • ... will also be passed to the function

Returns:

    MultiMap a table where the keys are the returned values, and the values are Lists of values where the function returned that key.

See also:

List:iter ()
return an iterator over all values.
List.iterate (seq)
Create an iterator over a seqence. This captures the Python concept of 'sequence'. For tables, iterates over all values with integer indices.

Parameters:

  • seq a sequence; a string (over characters), a table, a file object (over lines) or an iterator function

Usage:

  • for x in iterate {1,10,22,55} do io.write(x,',') end ==> 1,10,22,55
  • for ch in iterate 'help' do do io.write(ch,' ') end ==> h e l p

metamethods

List:__concat (L)
Concatenation operator.

Parameters:

  • L List another List

Returns:

    a new list consisting of the list with the elements of the new list appended
List:__eq (L)
Equality operator ==. True iff all elements of two lists are equal.

Parameters:

  • L List another List

Returns:

    true or false
List:__tostring ()
How our list should be rendered as a string. Uses join().

See also:

generated by LDoc 1.4.6
Penlight-1.12.0/docs/classes/pl.Map.html000066400000000000000000000271741416703176500177740ustar00rootroot00000000000000 Penlight Documentation

Class pl.Map

A Map class.

> Map = require 'pl.Map'
> m = Map{one=1,two=2}
> m:update {three=3,four=4,two=20}
> = m == M{one=1,two=20,three=3,four=4}
true

Dependencies: pl.utils, pl.class, pl.tablex, pl.pretty

Fields

pl.Map.keys list of keys.
pl.Map.values list of values.

Methods

pl.Map:iter () return an iterator over all key-value pairs.
pl.Map:items () return a List of all key-value pairs, sorted by the keys.
pl.Map:setdefault (key, default) set a value in the map if it doesn't exist yet.
pl.Map:len () size of map.
pl.Map:set (key, val) put a value into the map.
pl.Map:get (key) get a value from the map.
pl.Map:getvalues (keys) get a list of values indexed by a list of keys.
pl.Map:update (table) update the map using key/value pairs from another table.

Metamethods

pl.Map:__eq (m) equality between maps.
pl.Map:__tostring () string representation of a map.


Fields

pl.Map.keys
list of keys.
pl.Map.values
list of values.

Methods

pl.Map:iter ()
return an iterator over all key-value pairs.
pl.Map:items ()
return a List of all key-value pairs, sorted by the keys.
pl.Map:setdefault (key, default)
set a value in the map if it doesn't exist yet.

Parameters:

  • key the key
  • default value to set

Returns:

    the value stored in the map (existing value, or the new value)
pl.Map:len ()
size of map. note: this is a relatively expensive operation!
pl.Map:set (key, val)
put a value into the map. This will remove the key if the value is nil

Parameters:

  • key the key
  • val the value
pl.Map:get (key)
get a value from the map.

Parameters:

  • key the key

Returns:

    the value, or nil if not found.
pl.Map:getvalues (keys)
get a list of values indexed by a list of keys.

Parameters:

  • keys a list-like table of keys

Returns:

    a new list
pl.Map:update (table)
update the map using key/value pairs from another table.

Parameters:

  • table tab

Metamethods

pl.Map:__eq (m)
equality between maps.

Parameters:

  • m Map another map.
pl.Map:__tostring ()
string representation of a map.
generated by LDoc 1.4.6
Penlight-1.12.0/docs/classes/pl.MultiMap.html000066400000000000000000000155621416703176500210050ustar00rootroot00000000000000 Penlight Documentation

Class pl.MultiMap

MultiMap, a Map which has multiple values per key.

Dependencies: pl.utils, pl.class, pl.List, pl.Map

Methods

pl.MultiMap:update (t) update a MultiMap using a table.
pl.MultiMap:set (key, val) add a new value to a key.


Methods

pl.MultiMap:update (t)
update a MultiMap using a table.

Parameters:

  • t either a Multimap or a map-like table.

Returns:

    the map
pl.MultiMap:set (key, val)
add a new value to a key. Setting a nil value removes the key.

Parameters:

  • key the key
  • val the value

Returns:

    the map
generated by LDoc 1.4.6
Penlight-1.12.0/docs/classes/pl.OrderedMap.html000066400000000000000000000271361416703176500212770ustar00rootroot00000000000000 Penlight Documentation

Class pl.OrderedMap

OrderedMap, a map which preserves ordering.

Derived from pl.Map.

Dependencies: pl.utils, pl.tablex, pl.class, pl.List, pl.Map

Methods

pl.OrderedMap:_init (t) construct an OrderedMap.
pl.OrderedMap:update (t) update an OrderedMap using a table.
pl.OrderedMap:set (key, val) set the key's value.
pl.OrderedMap:insert (pos, key, val) insert a key/value pair before a given position.
pl.OrderedMap:keys () return the keys in order.
pl.OrderedMap:values () return the values in order.
pl.OrderedMap:sort (cmp) sort the keys.
pl.OrderedMap:iter () iterate over key-value pairs in order.

Metamethods

pl.OrderedMap:__pairs () iterate over an ordered map (5.2).
pl.OrderedMap:__tostring () string representation of an ordered map.


Methods

pl.OrderedMap:_init (t)
construct an OrderedMap. Will throw an error if the argument is bad.

Parameters:

pl.OrderedMap:update (t)
update an OrderedMap using a table. If the table is itself an OrderedMap, then its entries will be appended. if it s a table of the form {{key1=val1},{key2=val2},...} these will be appended.

Otherwise, it is assumed to be a map-like table, and order of extra entries is arbitrary.

Parameters:

  • t tab a table.

Returns:

  1. the map, or nil in case of error
  2. the error message
pl.OrderedMap:set (key, val)
set the key's value. This key will be appended at the end of the map.

If the value is nil, then the key is removed.

Parameters:

  • key the key
  • val the value

Returns:

    the map
pl.OrderedMap:insert (pos, key, val)
insert a key/value pair before a given position. Note: if the map already contains the key, then this effectively moves the item to the new position by first removing at the old position. Has no effect if the key does not exist and val is nil

Parameters:

  • pos int a position starting at 1
  • key the key
  • val the value; if nil use the old value
pl.OrderedMap:keys ()
return the keys in order. (Not a copy!)

Returns:

    List
pl.OrderedMap:values ()
return the values in order. this is relatively expensive.

Returns:

    List
pl.OrderedMap:sort (cmp)
sort the keys.

Parameters:

  • cmp func a comparison function as for table.sort

Returns:

    the map
pl.OrderedMap:iter ()
iterate over key-value pairs in order.

Metamethods

pl.OrderedMap:__pairs ()
iterate over an ordered map (5.2).
pl.OrderedMap:__tostring ()
string representation of an ordered map.
generated by LDoc 1.4.6
Penlight-1.12.0/docs/classes/pl.Set.html000066400000000000000000000367751416703176500200210ustar00rootroot00000000000000 Penlight Documentation

Class pl.Set

A Set class.

> Set = require 'pl.Set'
> = Set{'one','two'} == Set{'two','one'}
true
> fruit = Set{'apple','banana','orange'}
> = fruit['banana']
true
> = fruit['hazelnut']
nil
> colours = Set{'red','orange','green','blue'}
> = fruit,colours
[apple,orange,banana]   [blue,green,orange,red]
> = fruit+colours
[blue,green,apple,red,orange,banana]
[orange]
> more_fruits = fruit + 'apricot'
> = fruit*colours
 =  more_fruits, fruit
banana,apricot,apple,orange]    [banana,apple,orange]

Dependencies: pl.utils, pl.tablex, pl.class, pl.Map, (pl.List if __tostring is used)

Methods

pl.Set:Set (t) create a set.
pl.Set:values (self) get a list of the values in a set.
pl.Set:map (self, fn, ...) map a function over the values of a set.
pl.Set:union (self, set) union of two sets (also +).
pl.Set:intersection (self, set) intersection of two sets (also *).
pl.Set:difference (self, set) new set with elements in the set that are not in the other (also -).
pl.Set:issubset (self, set) is the first set a subset of the second (also <)?.
pl.Set:isempty (self) is the set empty?.
pl.Set:isdisjoint (s1, s2) are the sets disjoint?
pl.Set:len (s) size of this set (also # for 5.2).

Metamethods

pl.Set:__tostring () string representation of a set.
pl.Set:__add () union of sets.
pl.Set:__mul () intersection of sets.
pl.Set:__sub () difference of sets.
pl.Set:__pow () symmetric difference of sets.
pl.Set:__lt () first set subset of second?
pl.Set:__len () cardinality of set (5.2).
pl.Set:__eq (s1, s2) equality between sets.


Methods

pl.Set:Set (t)
create a set.

Parameters:

  • t may be a Set, Map or list-like table.
pl.Set:values (self)
get a list of the values in a set.

Parameters:

  • self a Set

Returns:

    a list
pl.Set:map (self, fn, ...)
map a function over the values of a set.

Parameters:

  • self a Set
  • fn a function
  • ... extra arguments to pass to the function.

Returns:

    a new set
pl.Set:union (self, set)
union of two sets (also +).

Parameters:

  • self a Set
  • set another set

Returns:

    a new set
pl.Set:intersection (self, set)
intersection of two sets (also *).

Parameters:

  • self a Set
  • set another set

Returns:

    a new set

Usage:

    > s = Set{10,20,30}
    > t = Set{20,30,40}
    > = t
    [20,30,40]
    > = Set.intersection(s,t)
    [30,20]
    > = s*t
    [30,20]
pl.Set:difference (self, set)
new set with elements in the set that are not in the other (also -).

Parameters:

  • self a Set
  • set another set

Returns:

    a new set
pl.Set:issubset (self, set)
is the first set a subset of the second (also <)?.

Parameters:

  • self a Set
  • set another set

Returns:

    true or false
pl.Set:isempty (self)
is the set empty?.

Parameters:

  • self a Set

Returns:

    true or false
pl.Set:isdisjoint (s1, s2)
are the sets disjoint? (no elements in common). Uses naive definition, i.e. that intersection is empty

Parameters:

  • s1 a Set
  • s2 another set

Returns:

    true or false
pl.Set:len (s)
size of this set (also # for 5.2).

Parameters:

  • s a Set

Returns:

    size

Metamethods

pl.Set:__tostring ()
string representation of a set.
pl.Set:__add ()
union of sets.
pl.Set:__mul ()
intersection of sets.
pl.Set:__sub ()
difference of sets.
pl.Set:__pow ()
symmetric difference of sets.
pl.Set:__lt ()
first set subset of second?
pl.Set:__len ()
cardinality of set (5.2).
pl.Set:__eq (s1, s2)
equality between sets.

Parameters:

  • s1
  • s2
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/000077500000000000000000000000001416703176500161255ustar00rootroot00000000000000Penlight-1.12.0/docs/examples/seesubst.lua.html000066400000000000000000000203101416703176500214240ustar00rootroot00000000000000 Penlight Documentation

seesubst.lua

-- shows how replacing '@see module' in the Markdown documentation
-- can be done more elegantly using PL.
-- We either have something like 'pl.config' (a module reference)
-- or 'pl.seq.map' (a function reference); these cases must be distinguished
-- and a Markdown link generated pointing to the LuaDoc file.

local sip = require 'pl.sip'
local stringx = require 'pl.stringx'

local res = {}
local s = [[
(@see pl.bonzo.dog)
remember about @see pl.bonzo

]]

local _gsub_patterns = {}

local function gsub (s,pat,subst,start)
    local fpat = _gsub_patterns[pat]
    if not fpat then
        -- use SIP to generate a proper string pattern.
        -- the _whole thing_ is a capture, to get the whole match
        -- and the unnamed capture.
        fpat = '('..sip.create_pattern(pat)..')'
        _gsub_patterns[pat] = fpat
    end
    return s:gsub(fpat,subst,start)
end


local mod = sip.compile '$v.$v'
local fun = sip.compile '$v.$v.$v'

for line in stringx.lines(s) do
    line = gsub(line,'@see $p',function(see,path)
        if fun(path,res) or mod(path,res) then
            local ret = ('[see %s](%s.%s.html'):format(path,res[1],res[2])
            if res[3] then
                return ret..'#'..res[3]..')'
            else
                return ret..')'
            end
        end
    end)
    print(line)
end
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/sipscan.lua.html000066400000000000000000000156741416703176500212500ustar00rootroot00000000000000 Penlight Documentation

sipscan.lua

-- another SIP example, shows how an awkward log file format
-- can be parsed. It also prints out the actual Lua string
-- pattern generated:
-- SYNC%s*%[([+%-%d]%d*)%]%s*([+%-%d]%d*)%s*([+%-%d]%d*)

local sip = require 'pl.sip'
local stringx = require 'pl.stringx'

local s = [[
SYNC [1] 0 547 (14679 sec)
SYNC [2] 0 555 (14679 sec)
SYNC [3] 0 563 (14679 sec)
SYNC [4] 0 571 (14679 sec)
SYNC [5] -1 580 (14679 sec)
SYNC [6] 0 587 (14679 sec)
]]


local first = true
local expected
local res = {}
local pat = 'SYNC [$i{seq}] $i{diff} $i{val}'
print(sip.create_pattern(pat))
local match = sip.compile(pat)
for line in stringx.lines(s) do
  if match(line,res) then
    if first then
      expected = res.val
      first = false
    end
    print(res.val,expected - res.val)
    expected = expected + 8
  end
end
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/symbols.lua.html000066400000000000000000000546021416703176500212720ustar00rootroot00000000000000 Penlight Documentation

symbols.lua

require 'pl'
utils.import 'pl.func'
local ops = require 'pl.operator'
local List = require 'pl.List'
local append,concat = table.insert,table.concat
local compare,find_if,compare_no_order,imap,reduce,count_map = tablex.compare,tablex.find_if,tablex.compare_no_order,tablex.imap,tablex.reduce,tablex.count_map
local unpack = table.unpack

function bindval (self,val)
    rawset(self,'value',val)
end

local optable = ops.optable

function sexpr (e)
	if isPE(e) then
		if e.op ~= 'X' then
			local args = tablex.imap(sexpr,e)
			return '('..e.op..' '..table.concat(args,' ')..')'
		else
			return e.repr
		end
	else
		return tostring(e)
	end
end


psexpr = compose(print,sexpr)



function equals (e1,e2)
    local p1,p2 = isPE(e1),isPE(e2)
    if p1 ~= p2 then return false end  -- different kinds of animals!
    if p1 and p2 then -- both PEs
        -- operators must be the same
        if e1.op ~= e2.op then return false end
        -- PHs are equal if their representations are equal
        if e1.op == 'X' then return e1.repr == e2.repr
        -- commutative operators
        elseif e1.op == '+' or e1.op == '*' then
            return compare_no_order(e1,e2,equals)
        else
            -- arguments must be the same
            return compare(e1,e2,equals)
        end
    else -- fall back on simple equality for non PEs
        return e1 == e2
    end
end

-- run down an unbalanced operator chain (like a+b+c) and return the arguments {a,b,c}
function tcollect (op,e,ls)
    if isPE(e) and e.op == op then
        for i = 1,#e do
            tcollect(op,e[i],ls)
        end
    else
        ls:append(e)
        return
    end
end

function rcollect (e)
    local res = List()
    tcollect(e.op,e,res)
    return res
end


-- balance ensures that +/* chains are collected together, operates in-place.
-- thus (+(+ a b) c) or (+ a (+ b c)) becomes (+ a b c), order immaterial
function balance (e)
    if isPE(e) and e.op ~= 'X' then
        local op,args = e.op
        if op == '+' or op == '*' then
            args = rcollect(e)
        else
            args = imap(balance,e)
        end
        for i = 1,#args do
            e[i] = args[i]
        end
    end
    return e
end

-- fold constants in an expression
function fold (e)
    if isPE(e) then
        if e.op == 'X' then
            -- there could be _bound values_!
            local val = rawget(e,'value')
            return val and val or e
        else
            local op = e.op
            local addmul = op == '*' or op == '+'
            -- first fold all arguments
            local args = imap(fold,e)
            if not addmul and not find_if(args,isPE) then
                -- no placeholders in these args, we can fold the expression.
                local opfn = optable[op]
                if opfn then
                    return opfn(unpack(args))
                else
                    return '?'
                end
            elseif addmul then
                -- enforce a few rules for + and *
                -- split the args into two classes, PE args and non-PE args.
                local classes = List.partition(args,isPE)
                local pe,npe = classes[true],classes[false]
                if npe then -- there's at least one non PE argument
                    -- so fold them
                    if #npe == 1 then npe = npe[1]
                    else npe = npe:reduce(optable[op])
                    end
                    -- if the result is a constant, return it
                    if not pe then return npe end

                    -- either (* 1 x) => x or (* 1 x y ...) => (* x y ...)
                    if op == '*' then
                        if npe == 0 then return 0
                        elseif npe == 1 then -- identity
                            if #pe == 1 then return pe[1] else npe = nil end
                        end
                    else -- special cases for +
                        if npe == 0 then -- identity
                            if #pe == 1 then return pe[1] else npe = nil end
                        end
                    end
                end
                -- build up the final arguments
                local res = {}
                if npe then append(res,npe) end
                for val,count in pairs(count_map(pe,equals)) do
                    if count > 1 then
                        if op == '*' then val = val ^ count
                        else val = val * count
                        end
                    end
                    append(res,val)
                end
                if #res == 1 then return res[1] end
                return PE{op=op,unpack(res)}
            elseif op == '^' then
                if args[2] == 1 then return args[1] end -- identity
                if args[2] == 0 then return 1 end
            end
            return PE{op=op,unpack(args)}
        end
    else
        return e
    end
end

function expand (e)
    if isPE(e) and e.op == '*' and isPE(e[2]) and e[2].op == '+' then
        local a,b = e[1],e[2]
        return expand(b[1]*a) + expand(b[2]*a)
    else
        return e
    end
end

function isnumber (x)
    return type(x) == 'number'
end

-- does this PE contain a reference to x?
function references (e,x)
    if isPE(e) then
        if e.op == 'X' then return x.repr == e.repr
        else
            return find_if(e,references,x)
        end
    else
        return false
    end
end

local function muli (args)
    return PE{op='*',unpack(args)}
end

local function addi (args)
    return PE{op='+',unpack(args)}
end

function diff (e,x)
    if isPE(e) and references(e,x) then
        local op = e.op
        if op == 'X' then
            return 1
        else
            local a,b = e[1],e[2]
            if op == '+' then -- differentiation is linear
                local args = imap(diff,e,x)
                return balance(addi(args))
            elseif op == '*' then -- product rule
                local res,d,ee = {}
                for i = 1,#e do
                    d = fold(diff(e[i],x))
                    if d ~= 0 then
                        ee = {unpack(e)}
                        ee[i] = d
                        append(res,balance(muli(ee)))
                    end
                end
                if #res > 1 then return addi(res)
                else return res[1] end
            elseif op == '^' and isnumber(b) then -- power rule
                return b*x^(b-1)
            end
        end
    else
        return 0
    end
end
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/test-cmp.lua.html000066400000000000000000000135201416703176500213300ustar00rootroot00000000000000 Penlight Documentation

test-cmp.lua

local A = require 'pl.tablex'
print(A.compare_no_order({1,2,3},{2,1,3}))
print(A.compare_no_order({1,2,3},{2,1,3},'=='))
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/test-data.lua.html000066400000000000000000000467521416703176500214770ustar00rootroot00000000000000 Penlight Documentation

test-data.lua

local data = require 'pl.data'
local List = require 'pl.List'
local array = require 'pl.array2d'
local func = require 'pl.func'
local seq = require 'pl.seq'
local stringio = require 'pl.stringio'
local open = stringio. open
local asserteq = require 'pl.test' . asserteq
local T = require 'pl.test'. tuple

--[=[
dat,err = data.read(open [[
1.0 0.1
0.2 1.3
]])

if err then print(err) end

require 'pl.pretty'.dump(dat)
os.exit(0)
--]=]

-- tab-separated data, explicit column names
local t1f = open [[
EventID	Magnitude	LocationX	LocationY	LocationZ	LocationError	EventDate	DataFile
981124001	2.0	18988.4	10047.1	4149.7	33.8	24/11/1998 11:18:05	981124DF.AAB
981125001	0.8	19104.0	9970.4	5088.7	3.0	25/11/1998 05:44:54	981125DF.AAB
981127003	0.5	19012.5	9946.9	3831.2	46.0	27/11/1998 17:15:17	981127DF.AAD
981127005	0.6	18676.4	10606.2	3761.9	4.4	27/11/1998 17:46:36	981127DF.AAF
981127006	0.2	19109.9	9716.5	3612.0	11.8	27/11/1998 19:29:51	981127DF.AAG
]]

local t1 = data.read (t1f)
-- column_by_name returns a List
asserteq(t1:column_by_name 'Magnitude',List{2,0.8,0.5,0.6,0.2})
-- can use array.column as well
asserteq(array.column(t1,2),{2,0.8,0.5,0.6,0.2})

-- only numerical columns (deduced from first data row) are converted by default
-- can look up indices in the list fieldnames.
local EDI = t1.fieldnames:index 'EventDate'
assert(type(t1[1][EDI]) == 'string')

-- select method returns a sequence, in this case single-valued.
-- (Note that seq.copy returns a List)
asserteq(seq(t1:select 'LocationX where Magnitude > 0.5'):copy(),List{18988.4,19104,18676.4})

--[[
--a common select usage pattern:
for event,mag in t1:select 'EventID,Magnitude sort by Magnitude desc' do
    print(event,mag)
end
--]]

-- space-separated, but with last field containing spaces.
local t2f = open [[
USER PID %MEM %CPU COMMAND
sdonovan 2333  0.3 0.1 background --n=2
root 2332  0.4  0.2 fred --start=yes
root 2338  0.2  0.1 backyard-process
]]

local t2,err = data.read(t2f,{last_field_collect=true})
if not t2 then return print (err) end

-- the last_field_collect option is useful with space-delimited data where the last
-- field may contain spaces. Otherwise, a record count mismatch should be an error!
local lt2 = List(t2[2])
asserteq(lt2:join ',','root,2332,0.4,0.2,fred --start=yes')

-- fieldnames are converted into valid identifiers by substituting _
-- (we do this to make select queries parseable by Lua)
asserteq(t2.fieldnames,List{'USER','PID','_MEM','_CPU','COMMAND'})

-- select queries are NOT SQL so remember to use == ! (and no 'between' operator, sorry)
--s,err = t2:select('_MEM where USER="root"')
--assert(err == [[[string "tmp"]:9: unexpected symbol near '=']])

local s = t2:select('_MEM where USER=="root"')
assert(s() == 0.4)
assert(s() == 0.2)
assert(s() == nil)

-- CSV, Excel style. Double-quoted fields are allowed, and they may contain commas!
local t3f = open [[
"Department Name","Employee ID",Project,"Hours Booked"
sales,1231,overhead,4
sales,1255,overhead,3
engineering,1501,development,5
engineering,1501,maintenance,3
engineering,1433,maintenance,10
]]

local t3 = data.read(t3f,{csv=true})

-- although fieldnames are turned in valid Lua identifiers, there is always original_fieldnames
asserteq(t3.fieldnames,List{'Department_Name','Employee_ID','Project','Hours_Booked'})
asserteq(t3.original_fieldnames,List{'Department Name','Employee ID','Project','Hours Booked'})

-- a common operation is to select using a given list of columns, and each row
-- on some explicit condition. The select() method can take a table with these
-- parameters
local keepcols = {'Employee_ID','Hours_Booked'}

local q = t3:select { fields = keepcols,
    where = function(row) return row[1]=='engineering' end
    }

asserteq(seq.copy2(q),{{1501,5},{1501,3},{1433,10}})

-- another pattern is doing a select to restrict rows & columns, process some
-- fields and write out the modified rows.

local outf = stringio.create()

local names = {[1501]='don',[1433]='dilbert'}

t3:write_row (outf,{'Employee','Hours_Booked'})
q = t3:select_row {fields=keepcols,where=func.Eq(func._1[1],'engineering')}
for row in q do
    row[1] = names[row[1]]
    t3:write_row(outf,row)
end

asserteq(outf:value(),
[[
Employee,Hours_Booked
don,5
don,3
dilbert,10
]])

-- data may not always have column headers. When creating a data object
-- from a two-dimensional array, may specify the fieldnames, as a list or a string.
-- The delimiter is deduced from the fieldname string, so a string just containing
-- the delimiter will set it,  and the fieldnames will be empty.
local dat = List()
local row = List.range(1,10)
for i = 1,10 do
    dat:append(row:map('*',i))
end
dat = data.new(dat,',')
local out = stringio.create()
dat:write(out,',')
asserteq(out:value(), [[
1,2,3,4,5,6,7,8,9,10
2,4,6,8,10,12,14,16,18,20
3,6,9,12,15,18,21,24,27,30
4,8,12,16,20,24,28,32,36,40
5,10,15,20,25,30,35,40,45,50
6,12,18,24,30,36,42,48,54,60
7,14,21,28,35,42,49,56,63,70
8,16,24,32,40,48,56,64,72,80
9,18,27,36,45,54,63,72,81,90
10,20,30,40,50,60,70,80,90,100
]])

-- you can always use numerical field indices, AWK-style;
-- note how the copy_select method gives you a data object instead of an
-- iterator over the fields
local res = dat:copy_select '$1,$3 where $1 > 5'
local L = List
asserteq(L(res),L{
    L{6, 18},
    L{7,21},
    L{8,24},
    L{9,27},
    L{10,30},
})

-- the column_by_name method may take a fieldname or an index
asserteq(dat:column_by_name(2), L{2,4,6,8,10,12,14,16,18,20})

-- the field list may contain expressions or even constants
local q = dat:select '$3,2*$4 where $1 == 8'
asserteq(T(q()),T(24,64))

dat,err = data.read(open [[
1.0 0.1
0.2 1.3
]])

if err then print(err) end

-- if a method cannot be found, then we look up in array2d
-- array2d.flatten(t) makes a 1D list out of a 2D array,
-- and then List.minmax() gets the extrema.

asserteq(T(dat:flatten():minmax()),T(0.1,1.3))

local f = open [[
Time Message
1266840760 +# EE7C0600006F0D00C00F06010302054000000308010A00002B00407B00
1266840760 closure data 0.000000 1972 1972 0
1266840760 ++ 1266840760 EE 1
1266840760 +# EE7C0600006F0D00C00F06010302054000000408020A00002B00407B00
1266840764 closure data 0.000000 1972 1972 0
1266840764 ++ 1266840764 EE 1
1266840764 +# EE7C0600006F0D00C00F06010302054000000508030A00002B00407B00
1266840768 duplicate?
1266840768 +# EE7C0600006F0D00C00F06010302054000000508030A00002B00407B00
1266840768 closure data 0.000000 1972 1972 0
]]

-- the convert option provides custom converters for each specified column.
-- Here we convert the timestamps into Date objects and collect everything
-- else into one field
local Date = require 'pl.Date'

local function date_convert (ds)
    return Date(tonumber(ds))
end

local d = data.read(f,{convert={[1]=date_convert},last_field_collect=true})

asserteq(#d[1],2)
asserteq(d[2][1]:year(),2010)

d = {{1,2,3},{10,20,30}}
out = stringio.create()
data.write(d,out,{'A','B','C'},',')
asserteq(out:value(),
[[
A,B,C
1,2,3
10,20,30
]])

out = stringio.create()
d.fieldnames = {'A','B','C'}
data.write(d,out)

asserteq(out:value(),
[[
A	B	C
1	2	3
10	20	30
]])


d = data.read(stringio.open 'One,Two\n1,\n,20\n',{csv=true})
asserteq(d,{
    {1,0},{0,20},
    original_fieldnames={"One","Two"},fieldnames={"One","Two"},delim=","
})
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/test-listcallbacks.lua.html000066400000000000000000000137671416703176500234010ustar00rootroot00000000000000 Penlight Documentation

test-listcallbacks.lua

-- demonstrates how to use a list of callbacks
local List = require 'pl.List'
local utils = require 'pl.utils'
local actions = List()
local L = utils.string_lambda

actions:append(function() print 'hello' end)
actions:append(L '|| print "yay"')

-- '()' is a shortcut for operator.call or function(x) return x() end
actions:foreach '()'
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/test-pretty.lua.html000066400000000000000000000140241416703176500221000ustar00rootroot00000000000000 Penlight Documentation

test-pretty.lua

local pretty = require 'pl.pretty'

local tb = {
    'one','two','three',{1,2,3},
    alpha=1,beta=2,gamma=3,['&']=true,[0]=false,
    _fred = {true,true},
    s = [[
hello dolly
you're so fine
]]
}

print(pretty.write(tb))
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/test-symbols.lua.html000066400000000000000000000175621416703176500222530ustar00rootroot00000000000000 Penlight Documentation

test-symbols.lua

require 'pl'
-- force us to look in the script's directory when requiring...
app.require_here()
require 'symbols'

local MT = getmetatable(_1)

add = MT.__add
mul = MT.__mul
pow = MT.__pow


function testeq (e1,e2)
    if not equals(e1,e2) then
        print ('Not equal',repr(e1),repr(e2))
    end
end

sin = register(math.sin,'sin')

f = register(function(x,y,z) end)

--[[
testeq (_1,_1)
testeq (_1+_2,_1+_2)
testeq (_1 + 3*_2,_1 + 3*_2)
testeq (_2+_1,_1+_2)
testeq (sin(_1),sin(_1))
testeq (1+f(10,20,'ok'),f(10,20,'ok')+1)
--]]


function testexpand (e)
    print(repr(fold(expand(e)))) --fold
end

--[[
testexpand (a*(a+1))

testexpand ((x+2)*(b+1))
]]--

function testfold (e)
    print(repr(fold(e)))
end

a,b,c,x,y = Var 'a,b,c,x,y'

--~ testfold(_1 + _2)
--~ testfold(add(10,20))
--~ testfold(add(mul(2,_1),mul(3,_2)))
--[[
testfold(sin(a))
e = a^(b+2)
testfold(e)
bindval(b,1)
testfold(e)
bindval(a,2)
testfold(e)

bindval(a)
bindval(b)
]]



function testdiff (e)
    balance(e)
    e = diff(e,x)
    balance(e)
    print('+ ',e)
    e = fold(e)
    print('- ',e)
end


testdiff(x^2+1)
testdiff(3*x^2)
testdiff(x^2 + 2*x^3)
testdiff(x^2 + 2*a*x^3 + x^4)
testdiff(2*a*x^3)
testdiff(x*x*x)
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/testapp.lua.html000066400000000000000000000132651416703176500212620ustar00rootroot00000000000000 Penlight Documentation

testapp.lua

-- shows how a script can get a private file path
-- the output on my Windows machine is:
-- C:\Documents and Settings\steve\.testapp\test.txt
local app = require 'pl.app'
print(app.appfile 'test.txt')
generated by LDoc 1.4.6 Last updated 2018-11-23 21:07:42
Penlight-1.12.0/docs/examples/testclone.lua.html000066400000000000000000000166271416703176500216070ustar00rootroot00000000000000 Penlight Documentation

testclone.lua

--cloning a directory tree.
local lfs = require 'lfs'
local path = require 'pl.path'
local dir = require 'pl.dir'

local p1 = [[examples]]
local p2 = [[copy/of/examples]]

if not path.isfile 'examples/testclone.lua' then
	return print 'please run this in the penlight folder (below examples)'
end

-- make a copy of the examples folder
dir.clonetree(p1,p2,dir.copyfile)

assert(path.isdir 'copy')

print '---'
local t = os.time()
print(lfs.touch('examples/testclone.lua',t,t+10))

-- this should only update this file
dir.clonetree(p1,p2,
function(f1,f2)
  local t1 = path.getmtime(f1)
  local t2 = path.getmtime(f2)
  --print(f1,t1,f2,t2)
  if t1 > t2 then
	dir.copyfile(f1,f2)
	print(f1,f2,t1,t2)
  end
  return true
end)

-- and get rid of the whole copy directory, with subdirs
dir.rmtree 'copy'

assert(not path.exists 'copy')
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/testconfig.lua.html000066400000000000000000000165501416703176500217470ustar00rootroot00000000000000 Penlight Documentation

testconfig.lua

local stringio = require 'pl.stringio'
local config = require 'pl.config'

local function dump(t,indent)
    if type(t) == 'table' then
        io.write(indent,'{\n')
        local newindent = indent..'  '
        for k,v in pairs(t) do
            io.write(newindent,k,'=')
            dump(v,indent)
            io.write('\n')
        end
        io.write(newindent,'},\n')
    else
        io.write(indent,t,'(',type(t),')')
    end
end


local function testconfig(test)
    local f = stringio.open(test)
    local c = config.read(f)
    f:close()
    dump(c,'  ')
    print '-----'
end

testconfig [[
 ; comment 2 (an ini file)
[section!]
bonzo.dog=20,30
config_parm=here we go again
depth = 2
[another]
felix="cat"
]]

testconfig [[
# this is a more Unix-y config file
fred = 1
alice = 2
home = /bonzo/dog/etc
]]

testconfig [[
# this is just a set of comma-separated values
1000,444,222
44,555,224
]]
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/testglobal.lua.html000066400000000000000000000163511416703176500217410ustar00rootroot00000000000000 Penlight Documentation

testglobal.lua

-- very simple lexer program which looks at all identifiers in a Lua
-- file and checks whether they're in the global namespace.
-- At the end, we dump out the result of count_map, which will give us
-- unique identifiers with their usage count.
-- (an example of a program which itself needs to be careful about what
-- goes into the global namespace)

local utils = require 'pl.utils'
local file = require 'pl.file'
local lexer = require 'pl.lexer'
local List = require 'pl.List'
local pretty = require 'pl.pretty'
local seq = require 'pl.seq'
local path = require 'pl.path'

utils.on_error 'quit'

local txt = file.read(arg[1] or path.normpath('examples/testglobal.lua'))
local globals = List()
for t,v in lexer.lua(txt) do
	if t == 'iden' and rawget(_G,v) then
		globals:append(v)
	end
end

pretty.dump(seq.count_map(globals))
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/testinputfields.lua.html000066400000000000000000000140471416703176500230270ustar00rootroot00000000000000 Penlight Documentation

testinputfields.lua

local input = require 'pl.input'
local sum = 0.0
local count = 0
local text = [[
    981124001	2.0	18988.4	10047.1	4149.7
    981125001	0.8	19104.0	9970.4	5088.7
    981127003	0.5	19012.5	9946.9	3831.2
]]
for id,magn,x in input.fields(3,' ',text) do
  sum = sum + x
  count = count + 1
end
print('average x coord is ',sum/count)
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/testinputfields2.lua.html000066400000000000000000000135041416703176500231060ustar00rootroot00000000000000 Penlight Documentation

testinputfields2.lua

local input = require 'pl.input'
local seq = require 'pl.seq'
local text = [[
    981124001	2.0	18988.4	10047.1	4149.7
    981125001	0.8	19104.0	9970.4	5088.7
    981127003	0.5	19012.5	9946.9	3831.2
]]
local sum,count = seq.sum(input.fields ({3},' ',text))
print(sum/count)
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/testxml.lua.html000066400000000000000000000247411416703176500213030ustar00rootroot00000000000000 Penlight Documentation

testxml.lua

-- an example showing 'pl.lexer' doing some serious work.
-- The resulting Lua table is in the same LOM format used by luaexpat.
-- This is (clearly) not a professional XML parser, so don't use it
-- on your homework!

local lexer = require 'pl.lexer'
local pretty = require 'pl.pretty'

local append = table.insert
local skipws,expecting = lexer.skipws,lexer.expecting

local function parse_element (tok,tag)
	local tbl,t,v,attrib
	tbl = {}
	tbl.tag = tag  -- LOM 'tag' is the element tag
	t,v = skipws(tok)
	while v ~= '/' and v ~= '>' do
		if t ~= 'iden' then error('expecting attribute identifier') end
		attrib = v
		expecting(tok,'=')
		v = expecting(tok,'string')
		-- LOM: 'attr' subtable contains attrib/value pairs and an ordered list of attribs
		if not tbl.attr then tbl.attr = {} end
		tbl.attr[attrib] = v
		append(tbl.attr,attrib)
		t,v = skipws(tok)
	end
	if v == '/' then
		expecting(tok,'>')
		return tbl
	end
	-- pick up element data
	t,v = tok()
	while true do
		if t == '<' then
			t,v = skipws(tok)
			if t == '/' then -- element end tag
				t,v = tok()
				if t == '>' then return tbl end
				if t == 'iden' and v == tag then
					if tok() == '>' then return tbl end
				end
				error('expecting end tag '..tag)
			else
				append(tbl,parse_element(tok,v)) -- LOM: child elements added to table
				t,v = skipws(tok)
			end
		else
			append(tbl,v) -- LOM: text added to table
			t,v = skipws(tok)
		end
	end
end

local function parse_xml (tok)
	local t = skipws(tok)
	local v
	while t == '<' do
		t,v = tok()
		if t == '?' or t == '!' then
			-- skip meta stuff and commentary
			repeat t = tok() until t == '>'
			t = expecting(tok,'<')
		else
			return parse_element(tok,v)
		end
	end
end

local s = [[
<?xml version="1.0" encoding="UTF-8"?>
<sensor name="closure-meter-2" id="7D7D0600006F0D00" loc="100,100,0" device="closure-meter" init="true">
<detector name="closure-meter" phenomenon="closure" units="mm" id="1"
    vmin="0" vmax="5000" device="closure-meter" calib="0,0;5000,5000"
    sampling_interval="25000" measurement_interval="600000"
/>
</sensor>
]]

local tok = lexer.scan(s,nil,{space=false},{string=true})
local res = parse_xml(tok)
print(pretty.write(res))
generated by LDoc 1.4.6
Penlight-1.12.0/docs/examples/which.lua.html000066400000000000000000000166621416703176500207100ustar00rootroot00000000000000 Penlight Documentation

which.lua

-- a simple implementation of the which command. This looks for
-- the given file on the path. On windows, it will assume an extension
-- of .exe if no extension is given.
local List = require 'pl.List'
local path = require 'pl.path'
local app = require 'pl.app'

local pathl = List.split(os.getenv 'PATH',path.dirsep)

local function which (file)
    local res = pathl:map(path.join,file)
    res = res:filter(path.exists)
    if res then return res[1] end
end

local _,lua = app.lua()
local file = arg[1] or lua -- i.e. location of lua executable
local try

if not file then return print 'must provide a filename' end

if path.extension(file) == '' and path.is_windows then
    try = which(file..'.exe')
else
    try = which(file)
end

if try then print(try) else print 'cannot find on path' end
generated by LDoc 1.4.6
Penlight-1.12.0/docs/index.html000066400000000000000000000355531416703176500163170ustar00rootroot00000000000000 Penlight Documentation

Penlight Lua Libraries 1.12.0

Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. Visit the GitHub project to review the code or file issues. Skip to the introduction.

Libraries

pl Entry point for loading all PL libraries only on demand, into the global space.
pl.app Application support functions.
pl.array2d Operations on two-dimensional arrays.
pl.class Provides a reuseable and convenient framework for creating classes in Lua.
pl.compat Lua 5.1/5.2/5.3 compatibility.
pl.comprehension List comprehensions implemented in Lua.
pl.config Reads configuration files into a Lua table.
pl.data Reading and querying simple tabular data.
pl.dir Listing files in directories and creating/removing directory paths.
pl.file File manipulation functions: reading, writing, moving and copying.
pl.func Functional helpers like composition, binding and placeholder expressions.
pl.import_into PL loader, for loading all PL libraries, only on demand.
pl.input Iterators for extracting words or numbers from an input source.
pl.lapp Simple command-line parsing using human-readable specification.
pl.lexer Lexical scanner for creating a sequence of tokens from text.
pl.luabalanced Extract delimited Lua sequences from strings.
pl.operator Lua operators available as functions.
pl.path Path manipulation and file queries.
pl.permute Permutation operations.
pl.pretty Pretty-printing Lua tables.
pl.seq Manipulating iterators as sequences.
pl.sip Simple Input Patterns (SIP).
pl.strict Checks uses of undeclared global variables.
pl.stringio Reading and writing strings using file-like objects.
pl.stringx Python-style extended string library.
pl.tablex Extended operations on Lua tables.
pl.template A template preprocessor.
pl.test Useful test utilities.
pl.text Text processing utilities.
pl.types Dealing with Detailed Type Information
pl.url Python-style URL quoting library.
pl.utils Generally useful routines.
pl.xml XML LOM Utilities.

Classes

pl.Date Date and Date Format classes.
pl.List Python-style list class.
pl.Map A Map class.
pl.MultiMap MultiMap, a Map which has multiple values per key.
pl.OrderedMap OrderedMap, a map which preserves ordering.
pl.Set A Set class.

Manual

01-introduction.md
02-arrays.md
03-strings.md
04-paths.md
05-dates.md
06-data.md
07-functional.md
08-additional.md
09-discussion.md

Examples

seesubst.lua
sipscan.lua
symbols.lua
test-cmp.lua
test-data.lua
test-listcallbacks.lua
test-pretty.lua
test-symbols.lua
testclone.lua
testconfig.lua
testglobal.lua
testinputfields.lua
testinputfields2.lua
testxml.lua
which.lua
generated by LDoc 1.4.6
Penlight-1.12.0/docs/ldoc_fixed.css000066400000000000000000000147061416703176500171310ustar00rootroot00000000000000/* BEGIN RESET Copyright (c) 2010, Yahoo! Inc. All rights reserved. Code licensed under the BSD License: http://developer.yahoo.com/yui/license.html version: 2.8.2r1 */ html { color: #000; background: #FFF; } body,div,dl,dt,dd,ul,ol,li,h1,h2,h3,h4,h5,h6,pre,code,form,fieldset,legend,input,button,textarea,p,blockquote,th,td { margin: 0; padding: 0; } table { border-collapse: collapse; border-spacing: 0; } fieldset,img { border: 0; } address,caption,cite,code,dfn,em,strong,th,var,optgroup { font-style: inherit; font-weight: inherit; } del,ins { text-decoration: none; } li { margin-left: 20px; } caption,th { text-align: left; } h1,h2,h3,h4,h5,h6 { font-size: 100%; font-weight: bold; } q:before,q:after { content: ''; } abbr,acronym { border: 0; font-variant: normal; } sup { vertical-align: baseline; } sub { vertical-align: baseline; } legend { color: #000; } input,button,textarea,select,optgroup,option { font-family: inherit; font-size: inherit; font-style: inherit; font-weight: inherit; } input,button,textarea,select {*font-size:100%; } /* END RESET */ body { margin-left: 1em; margin-right: 1em; font-family: arial, helvetica, geneva, sans-serif; background-color: #ffffff; margin: 0px; } code, tt { font-family: monospace; font-size: 1.1em; } span.parameter { font-family:monospace; } span.parameter:after { content:":"; } span.types:before { content:"("; } span.types:after { content:")"; } .type { font-weight: bold; font-style:italic } body, p, td, th { font-size: .95em; line-height: 1.2em;} p, ul { margin: 10px 0 0 0px;} strong { font-weight: bold;} em { font-style: italic;} h1 { font-size: 1.5em; margin: 0 0 20px 0; } h2, h3, h4 { margin: 15px 0 10px 0; } h2 { font-size: 1.25em; } h3 { font-size: 1.15em; } h4 { font-size: 1.06em; } a:link { font-weight: bold; color: #004080; text-decoration: none; } a:visited { font-weight: bold; color: #006699; text-decoration: none; } a:link:hover { text-decoration: underline; } hr { color:#cccccc; background: #00007f; height: 1px; } blockquote { margin-left: 3em; } ul { list-style-type: disc; } p.name { font-family: "Andale Mono", monospace; padding-top: 1em; } pre { background-color: rgb(245, 245, 245); border: 1px solid #C0C0C0; /* silver */ padding: 10px; margin: 10px 0 10px 0; overflow: auto; font-family: "Andale Mono", monospace; } pre.example { font-size: .85em; } table.index { border: 1px #00007f; } table.index td { text-align: left; vertical-align: top; } #container { margin-left: 1em; margin-right: 1em; background-color: #ffffff; } #product { text-align: center; border-bottom: 1px solid #cccccc; background-color: #ffffff; } #product big { font-size: 2em; } #main { background-color:#FFFFFF; // #f0f0f0; border-left: 1px solid #cccccc; } #navigation { position: fixed; top: 0; left: 0; float: left; width: 14em; vertical-align: top; background-color:#FFFFFF; // #f0f0f0; border-right: 2px solid #cccccc; overflow: visible; overflow-y: scroll; height: 100%; padding-left: 1em; } #navigation h2 { background-color:#FFFFFF;//:#e7e7e7; font-size:1.1em; color:#000000; text-align: left; padding:0.2em; border-bottom:1px solid #dddddd; } #navigation ul { font-size:1em; list-style-type: none; margin: 1px 1px 10px 1px; } #navigation li { text-indent: -1em; display: block; margin: 3px 0px 0px 22px; } #navigation li li a { margin: 0px 3px 0px -1em; } #content { margin-left: 14em; padding: 1em; padding-left: 2em; width: 700px; border-left: 2px solid #cccccc; // border-right: 2px solid #cccccc; background-color: #ffffff; } #about { clear: both; padding-left: 1em; margin-left: 14em; // avoid the damn sidebar! border-top: 2px solid #cccccc; border-left: 2px solid #cccccc; background-color: #ffffff; } @media print { body { font: 12pt "Times New Roman", "TimeNR", Times, serif; } a { font-weight: bold; color: #004080; text-decoration: underline; } #main { background-color: #ffffff; border-left: 0px; } #container { margin-left: 2%; margin-right: 2%; background-color: #ffffff; } #content { padding: 1em; background-color: #ffffff; } #navigation { display: none; } pre.example { font-family: "Andale Mono", monospace; font-size: 10pt; page-break-inside: avoid; } } table.module_list { border-width: 1px; border-style: solid; border-color: #cccccc; border-collapse: collapse; } table.module_list td { border-width: 1px; padding: 3px; border-style: solid; border-color: #cccccc; } table.module_list td.name { background-color: #f0f0f0; ; min-width: 200px; } table.module_list td.summary { width: 100%; } table.function_list { border-width: 1px; border-style: solid; border-color: #cccccc; border-collapse: collapse; } table.function_list td { border-width: 1px; padding: 3px; border-style: solid; border-color: #cccccc; } table.function_list td.name { background-color: #f6f6ff; ; min-width: 200px; } table.function_list td.summary { width: 100%; } dl.table dt, dl.function dt {border-top: 1px solid #ccc; padding-top: 1em;} dl.table dd, dl.function dd {padding-bottom: 1em; margin: 10px 0 0 20px;} dl.table h3, dl.function h3 {font-size: .95em;} ul.nowrap { overflow:auto; whitespace:nowrap; } /* stop sublists from having initial vertical space */ ul ul { margin-top: 0px; } ol ul { margin-top: 0px; } ol ol { margin-top: 0px; } ul ol { margin-top: 0px; } /* make the target distinct; helps when we're navigating to a function */ a:target + * { background-color: #FF9; } /* styles for prettification of source */ pre .comment { color: #558817; } pre .constant { color: #a8660d; } pre .escape { color: #844631; } pre .keyword { color: #aa5050; font-weight: bold; } pre .library { color: #0e7c6b; } pre .marker { color: #512b1e; background: #fedc56; font-weight: bold; } pre .string { color: #8080ff; } pre .number { color: #f8660d; } pre .operator { color: #2239a8; font-weight: bold; } pre .preprocessor, pre .prepro { color: #a33243; } pre .global { color: #800080; } pre .user-keyword { color: #800080; } pre .prompt { color: #558817; } pre .url { color: #272fc2; text-decoration: underline; } Penlight-1.12.0/docs/libraries/000077500000000000000000000000001416703176500162635ustar00rootroot00000000000000Penlight-1.12.0/docs/libraries/pl.Set.html000066400000000000000000000360631416703176500203260ustar00rootroot00000000000000 Penlight Documentation

Module pl.Set

A Set class.

> Set = require 'pl.Set'
> = Set{'one','two'} == Set{'two','one'}
true
> fruit = Set{'apple','banana','orange'}
> = fruit['banana']
true
> = fruit['hazelnut']
nil
> colours = Set{'red','orange','green','blue'}
> = fruit,colours
[apple,orange,banana]   [blue,green,orange,red]
> = fruit+colours
[blue,green,apple,red,orange,banana]
[orange]
> more_fruits = fruit + 'apricot'
> = fruit*colours
 =  more_fruits, fruit
banana,apricot,apple,orange]    [banana,apple,orange]

Dependencies: pl.utils, pl.tablex, pl.class, pl.Map, (pl.List if __tostring is used)

Functions

Set (t) create a set.
values (self) get a list of the values in a set.
map (self, fn, ...) map a function over the values of a set.
union (self, set) union of two sets (also +).
intersection (self, set) intersection of two sets (also *).
difference (self, set) new set with elements in the set that are not in the other (also -).
issubset (self, set) is the first set a subset of the second (also <)?.
isempty (self) is the set empty?.
isdisjoint (s1, s2) are the sets disjoint?
len (s) size of this set (also # for 5.2).

metamethods

__tostring () string representation of a set.
__add () union of sets.
__mul () intersection of sets.
__sub () difference of sets.
__pow () symmetric difference of sets.
__lt () first set subset of second?
__len () cardinality of set (5.2).
__eq (s1, s2) equality between sets.


Functions

Set (t)
create a set.

Parameters:

  • t may be a Set, Map or list-like table.
values (self)
get a list of the values in a set.

Parameters:

  • self a Set

Returns:

    a list
map (self, fn, ...)
map a function over the values of a set.

Parameters:

  • self a Set
  • fn a function
  • ... extra arguments to pass to the function.

Returns:

    a new set
union (self, set)
union of two sets (also +).

Parameters:

  • self a Set
  • set another set

Returns:

    a new set
intersection (self, set)
intersection of two sets (also *).

Parameters:

  • self a Set
  • set another set

Returns:

    a new set

Usage:

    > s = Set{10,20,30}
    > t = Set{20,30,40}
    > = t
    [20,30,40]
    > = Set.intersection(s,t)
    [30,20]
    > = s*t
    [30,20]
difference (self, set)
new set with elements in the set that are not in the other (also -).

Parameters:

  • self a Set
  • set another set

Returns:

    a new set
issubset (self, set)
is the first set a subset of the second (also <)?.

Parameters:

  • self a Set
  • set another set

Returns:

    true or false
isempty (self)
is the set empty?.

Parameters:

  • self a Set

Returns:

    true or false
isdisjoint (s1, s2)
are the sets disjoint? (no elements in common). Uses naive definition, i.e. that intersection is empty

Parameters:

  • s1 a Set
  • s2 another set

Returns:

    true or false
len (s)
size of this set (also # for 5.2).

Parameters:

  • s a Set

Returns:

    size

metamethods

__tostring ()
string representation of a set.
__add ()
union of sets.
__mul ()
intersection of sets.
__sub ()
difference of sets.
__pow ()
symmetric difference of sets.
__lt ()
first set subset of second?
__len ()
cardinality of set (5.2).
__eq (s1, s2)
equality between sets.

Parameters:

  • s1
  • s2
generated by LDoc 1.4.6 Last updated 2018-11-23 21:07:42
Penlight-1.12.0/docs/libraries/pl.app.html000066400000000000000000000345401416703176500203510ustar00rootroot00000000000000 Penlight Documentation

Module pl.app

Application support functions.

See the Guide

Dependencies: pl.utils, pl.path

Functions

script_name () return the name of the current script running.
require_here (base) prefixes the current script's path to the Lua module path.
appfile (file) return a suitable path for files private to this application.
platform () return string indicating operating system.
lua () return the full command-line used to invoke this script.
parse_args (args, flags_with_values, flags_valid) parse command-line arguments into flags and parameters.


Functions

script_name ()
return the name of the current script running. The name will be the name as passed on the command line

Returns:

    string filename
require_here (base)
prefixes the current script's path to the Lua module path. Applies to both the source and the binary module paths. It makes it easy for the main file of a multi-file program to access its modules in the same directory. base allows these modules to be put in a specified subdirectory, to allow for cleaner deployment and resolve potential conflicts between a script name and its library directory.

Note: the path is prefixed, so it is searched first when requiring modules.

Parameters:

  • base string optional base directory (absolute, or relative path).

Returns:

    string the current script's path with a trailing slash
appfile (file)
return a suitable path for files private to this application. These will look like '~/.SNAME/file', with '~' as with expanduser and SNAME is the name of the script without .lua extension. If the directory does not exist, it will be created.

Parameters:

  • file string a filename (w/out path)

Returns:

  1. a full pathname, or nil
  2. cannot create directory error

Usage:

    -- when run from a script called 'testapp' (on Windows):
    local app = require 'pl.app'
    print(app.appfile 'test.txt')
    -- C:\Documents and Settings\steve\.testapp\test.txt
platform ()
return string indicating operating system.

Returns:

    'Windows','OSX' or whatever uname returns (e.g. 'Linux')
lua ()
return the full command-line used to invoke this script. It will not include the scriptname itself, see app.script_name.

Returns:

  1. command-line
  2. name of Lua program used

Usage:

    -- execute:  lua -lluacov -e 'print(_VERSION)' myscript.lua
    
    -- myscript.lua
    print(require("pl.app").lua())  --> "lua -lluacov -e 'print(_VERSION)'", "lua"
parse_args (args, flags_with_values, flags_valid)
parse command-line arguments into flags and parameters. Understands GNU-style command-line flags; short (-f) and long (--flag).

These may be given a value with either '=' or ':' (-k:2,--alpha=3.2,-n2), a number value can be given without a space. If the flag is marked as having a value, then a space-separated value is also accepted (-i hello), see the flags_with_values argument).

Multiple short args can be combined like so: ( -abcd).

When specifying the flags_valid parameter, its contents can also contain aliasses, to convert short/long flags to the same output name. See the example below.

Note: if a flag is repeated, the last value wins.

Parameters:

  • args {string} an array of strings (default is the global arg)
  • flags_with_values tab any flags that take values, either list or hash table e.g. { out=true } or { "out" }.
  • flags_valid tab (optional) flags that are valid, either list or hashtable. If not given, everything will be accepted(everything in flags_with_values will automatically be allowed)

Returns:

  1. a table of flags (flag=value pairs)
  2. an array of parameters

Raises:

if args is nil, then the global args must be available!

Usage:

    -- Simple form:
    local flags, params = app.parse_args(nil,
         { "hello", "world" },  -- list of flags taking values
         { "l", "a", "b"})      -- list of allowed flags (value ones will be added)
    
    -- More complex example using aliasses:
    local valid = {
        long = "l",           -- if 'l' is specified, it is reported as 'long'
        new = { "n", "old" }, -- here both 'n' and 'old' will go into 'new'
    }
    local values = {
        "value",   -- will automatically be added to the allowed set of flags
        "new",     -- will mark 'n' and 'old' as requiring a value as well
    }
    local flags, params = app.parse_args(nil, values, valid)
    
    -- command:  myapp.lua -l --old:hello --value world param1 param2
    -- will yield:
    flags = {
        long = true,     -- input from 'l'
        new = "hello",   -- input from 'old'
        value = "world", -- allowed because it was in 'values', note: space separated!
    }
    params = {
        [1] = "param1"
        [2] = "param2"
    }
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.array2d.html000066400000000000000000001131761416703176500211400ustar00rootroot00000000000000 Penlight Documentation

Module pl.array2d

Operations on two-dimensional arrays.

See The Guide

The size of the arrays is determined by using the length operator # hence the module is not nil safe, and the usual precautions apply.

Note: all functions taking i1,j1,i2,j2 as arguments will normalize the arguments using default_range.

Dependencies: pl.utils, pl.tablex, pl.types

Functions

size (a) return the row and column size.
column (a, j) extract a column from the 2D array.
row (a, i) extract a row from the 2D array.
map (f, a, arg) map a function over a 2D array
reduce_rows (f, a) reduce the rows using a function.
reduce_cols (f, a) reduce the columns using a function.
reduce2 (opc, opr, a) reduce a 2D array into a scalar, using two operations.
map2 (f, ad, bd, a, b, arg) map a function over two arrays.
product (f, t1, t2) cartesian product of two 1d arrays.
flatten (t) flatten a 2D array.
reshape (t, nrows, co) reshape a 2D array.
transpose (t) transpose a 2D array.
swap_rows (t, i1, i2) swap two rows of an array.
swap_cols (t, j1, j2) swap two columns of an array.
extract_rows (t, ridx) extract the specified rows.
extract_cols (t, cidx) extract the specified columns.
remove_row (t, i) remove a row from an array.
remove_col (t, j) remove a column from an array.
parse_range (s) parse a spreadsheet range or cell.
range (...) get a slice of a 2D array.
default_range (t[, i1=1[, j1=1[, i2=N[, j2=M]]]]) normalizes coordinates to valid positive entries and defaults.
slice (t[, i1=1[, j1=1[, i2=N[, j2=M]]]]) get a slice of a 2D array.
set (t, value[, i1=1[, j1=1[, i2=N[, j2=M]]]]) set a specified range of an array to a value.
write (t, f, fmt[, i1=1[, j1=1[, i2=N[, j2=M]]]]) write a 2D array to a file.
forall (t, row_op, end_row_op[, i1=1[, j1=1[, i2=N[, j2=M]]]]) perform an operation for all values in a 2D array.
move (dest, di, dj, src[, i1=1[, j1=1[, i2=N[, j2=M]]]]) move a block from the destination to the source.
iter (a, indices[, i1=1[, j1=1[, i2=N[, j2=M]]]]) iterate over all elements in a 2D array, with optional indices.
columns (a) iterate over all columns.
rows (a) iterate over all rows.
new (rows, cols, val) new array of specified dimensions


Functions

size (a)
return the row and column size. Size is calculated using the Lua length operator #, so usual precautions regarding nil values apply.

Parameters:

  • a array a 2d array

Returns:

  1. int number of rows (#a)
  2. int number of cols (#a[1])
column (a, j)
extract a column from the 2D array.

Parameters:

  • a array 2d array
  • j column index

Returns:

    1d array
row (a, i)
extract a row from the 2D array. Added in line with column, for read-only purposes directly accessing a[i] is more performant.

Parameters:

  • a array 2d array
  • i row index

Returns:

    1d array (copy of the row)
map (f, a, arg)
map a function over a 2D array

Parameters:

  • f func a function of at least one argument
  • a array 2d array
  • arg an optional extra argument to be passed to the function.

Returns:

    2d array
reduce_rows (f, a)
reduce the rows using a function.

Parameters:

  • f func a binary function
  • a array 2d array

Returns:

    1d array

See also:

reduce_cols (f, a)
reduce the columns using a function.

Parameters:

  • f func a binary function
  • a array 2d array

Returns:

    1d array

See also:

reduce2 (opc, opr, a)
reduce a 2D array into a scalar, using two operations.

Parameters:

  • opc func operation to reduce the final result
  • opr func operation to reduce the rows
  • a 2D array
map2 (f, ad, bd, a, b, arg)
map a function over two arrays. They can be both or either 2D arrays

Parameters:

  • f func function of at least two arguments
  • ad int order of first array (1 if a is a list/array, 2 if it is a 2d array)
  • bd int order of second array (1 if b is a list/array, 2 if it is a 2d array)
  • a tab 1d or 2d array
  • b tab 1d or 2d array
  • arg optional extra argument to pass to function

Returns:

    2D array, unless both arrays are 1D
product (f, t1, t2)
cartesian product of two 1d arrays.

Parameters:

  • f func a function of 2 arguments
  • t1 array a 1d table
  • t2 array a 1d table

Returns:

    2d table

Usage:

    product('..',{1,2},{'a','b'}) == {{'1a','2a'},{'1b','2b'}}
flatten (t)
flatten a 2D array. (this goes over columns first.)

Parameters:

  • t array 2d table

Returns:

    a 1d table

Usage:

    flatten {{1,2},{3,4},{5,6}} == {1,2,3,4,5,6}
reshape (t, nrows, co)
reshape a 2D array. Reshape the aray by specifying a new nr of rows.

Parameters:

  • t array 2d array
  • nrows int new number of rows
  • co bool use column-order (Fortran-style) (default false)

Returns:

    a new 2d array
transpose (t)
transpose a 2D array.

Parameters:

  • t array 2d array

Returns:

    a new 2d array
swap_rows (t, i1, i2)
swap two rows of an array.

Parameters:

  • t array a 2d array
  • i1 int a row index
  • i2 int a row index

Returns:

    t (same, modified 2d array)
swap_cols (t, j1, j2)
swap two columns of an array.

Parameters:

  • t array a 2d array
  • j1 int a column index
  • j2 int a column index

Returns:

    t (same, modified 2d array)
extract_rows (t, ridx)
extract the specified rows.

Parameters:

  • t array 2d array
  • ridx {int} a table of row indices

Returns:

    a new 2d array with the extracted rows
extract_cols (t, cidx)
extract the specified columns.

Parameters:

  • t array 2d array
  • cidx {int} a table of column indices

Returns:

    a new 2d array with the extracted colums
remove_row (t, i)
remove a row from an array.

Parameters:

  • t array a 2d array
  • i int a row index
remove_col (t, j)
remove a column from an array.

Parameters:

  • t array a 2d array
  • j int a column index
parse_range (s)
parse a spreadsheet range or cell. The range/cell can be specified either as 'A1:B2' or 'R1C1:R2C2' or for single cells as 'A1' or 'R1C1'.

Parameters:

  • s string a range (case insensitive).

Returns:

  1. int start row
  2. int start col
  3. int end row (or nil if the range was a single cell)
  4. int end col (or nil if the range was a single cell)
range (...)
get a slice of a 2D array. Same as slice.

Parameters:

  • ...

See also:

default_range (t[, i1=1[, j1=1[, i2=N[, j2=M]]]])
normalizes coordinates to valid positive entries and defaults. Negative indices will be counted from the end, too low, or too high will be limited by the array sizes.

Parameters:

  • t array a 2D array
  • i1 int or string start row or spreadsheet range passed to parse_range (default 1)
  • j1 int start col (default 1)
  • i2 int end row (default N)
  • j2 int end col (default M)

Returns:

    i1, j1, i2, j2

See also:

slice (t[, i1=1[, j1=1[, i2=N[, j2=M]]]])
get a slice of a 2D array. Note that if the specified range has a 1D result, the rank of the result will be 1.

Parameters:

  • t array a 2D array
  • i1 int or string start row or spreadsheet range passed to parse_range (default 1)
  • j1 int start col (default 1)
  • i2 int end row (default N)
  • j2 int end col (default M)

Returns:

    an array, 2D in general but 1D in special cases.

See also:

set (t, value[, i1=1[, j1=1[, i2=N[, j2=M]]]])
set a specified range of an array to a value.

Parameters:

  • t array a 2D array
  • value the value (may be a function, called as val(i,j))
  • i1 int or string start row or spreadsheet range passed to parse_range (default 1)
  • j1 int start col (default 1)
  • i2 int end row (default N)
  • j2 int end col (default M)

See also:

write (t, f, fmt[, i1=1[, j1=1[, i2=N[, j2=M]]]])
write a 2D array to a file.

Parameters:

  • t array a 2D array
  • f a file object (default stdout)
  • fmt string a format string (default is just to use tostring)
  • i1 int or string start row or spreadsheet range passed to parse_range (default 1)
  • j1 int start col (default 1)
  • i2 int end row (default N)
  • j2 int end col (default M)

See also:

forall (t, row_op, end_row_op[, i1=1[, j1=1[, i2=N[, j2=M]]]])
perform an operation for all values in a 2D array.

Parameters:

  • t array 2D array
  • row_op func function to call on each value; row_op(row,j)
  • end_row_op func function to call at end of each row; end_row_op(i)
  • i1 int or string start row or spreadsheet range passed to parse_range (default 1)
  • j1 int start col (default 1)
  • i2 int end row (default N)
  • j2 int end col (default M)

See also:

move (dest, di, dj, src[, i1=1[, j1=1[, i2=N[, j2=M]]]])
move a block from the destination to the source.

Parameters:

  • dest array a 2D array
  • di int start row in dest
  • dj int start col in dest
  • src array a 2D array
  • i1 int or string start row or spreadsheet range passed to parse_range (default 1)
  • j1 int start col (default 1)
  • i2 int end row (default N)
  • j2 int end col (default M)

See also:

iter (a, indices[, i1=1[, j1=1[, i2=N[, j2=M]]]])
iterate over all elements in a 2D array, with optional indices.

Parameters:

  • a array 2D array
  • indices bool with indices (default false)
  • i1 int or string start row or spreadsheet range passed to parse_range (default 1)
  • j1 int start col (default 1)
  • i2 int end row (default N)
  • j2 int end col (default M)

Returns:

    either value or i,j,value depending on the value of indices

See also:

columns (a)
iterate over all columns.

Parameters:

  • a array a 2D array

Returns:

    column, column-index
rows (a)
iterate over all rows. Returns a copy of the row, for read-only purposes directly iterating is more performant; ipairs(a)

Parameters:

  • a array a 2D array

Returns:

    row, row-index
new (rows, cols, val)
new array of specified dimensions

Parameters:

  • rows int number of rows
  • cols int number of cols
  • val initial value; if it's a function then use val(i,j)

Returns:

    new 2d array
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.class.html000066400000000000000000000261101416703176500206700ustar00rootroot00000000000000 Penlight Documentation

Module pl.class

Provides a reuseable and convenient framework for creating classes in Lua.

Two possible notations:

B = class(A)
class.B(A)

The latter form creates a named class within the current environment. Note that this implicitly brings in pl.utils as a dependency.

See the Guide for further discussion

Functions

_init (...) initializes an instance upon creation.
instance:is_a (some_class) checks whether an instance is derived from some class.
some_class:class_of (some_instance) checks whether an instance is derived from some class.
some_class:cast (some_instance) cast an object to another class.
class (base, c_arg, c) create a new class, derived from a given base class.


Functions

_init (...)
initializes an instance upon creation.

Parameters:

  • ... parameters passed to the constructor

Usage:

    local Cat = class()
    function Cat:_init(name)
      --self:super(name)   -- call the ancestor initializer if needed
      self.name = name
    end
    
    local pussycat = Cat("pussycat")
    print(pussycat.name)  --> pussycat
instance:is_a (some_class)
checks whether an instance is derived from some class. Works the other way around as class_of. It has two ways of using; 1) call with a class to check against, 2) call without params.

Parameters:

  • some_class class to check against, or nil to return the class

Returns:

    true if instance is derived from some_class, or if some_class == nil then it returns the class table of the instance

Usage:

    local pussycat = Lion()  -- assuming Lion derives from Cat
    if pussycat:is_a(Cat) then
      -- it's true, it is a Lion, but also a Cat
    end
    
    if pussycat:is_a() == Lion then
      -- It's true
    end
some_class:class_of (some_instance)
checks whether an instance is derived from some class. Works the other way around as is_a.

Parameters:

  • some_instance instance to check against

Returns:

    true if some_instance is derived from some_class

Usage:

    local pussycat = Lion()  -- assuming Lion derives from Cat
    if Cat:class_of(pussycat) then
      -- it's true
    end
some_class:cast (some_instance)
cast an object to another class. It is not clever (or safe!) so use carefully.

Parameters:

  • some_instance the object to be changed
class (base, c_arg, c)
create a new class, derived from a given base class. Supporting two class creation syntaxes: either Name = class(base) or class.Name(base). The first form returns the class directly and does not set its _name. The second form creates a variable Name in the current environment set to the class, and also sets _name.

Parameters:

  • base optional base class
  • c_arg optional parameter to class constructor
  • c optional table to be used as class
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.compat.html000066400000000000000000000401621416703176500210510ustar00rootroot00000000000000 Penlight Documentation

Module pl.compat

Lua 5.1/5.2/5.3 compatibility.

Injects table.pack, table.unpack, and package.searchpath in the global environment, to make sure they are available for Lua 5.1 and LuaJIT.

All other functions are exported as usual in the returned module table.

NOTE: everything in this module is also available in pl.utils.

Functions

execute (cmd) execute a shell command, in a compatible and platform independent way.
load (ld[, source[, mode[, env]]]) Load Lua code as a text or binary chunk (in a Lua 5.2 compatible way).
getfenv (f) Get environment of a function (in a Lua 5.1 compatible way).
setfenv (f, env) Set environment of a function (in a Lua 5.1 compatible way).

Fields

lua51 boolean flag this is Lua 5.1 (or LuaJIT).
jit boolean flag this is LuaJIT.
jit52 boolean flag this is LuaJIT with 5.2 compatibility compiled in.
dir_separator the directory separator character for the current platform.
is_windows boolean flag this is a Windows platform.

Global exported functions (for Lua 5.1 & LuaJIT)

table.pack (...) pack an argument list into a table.
table.unpack (t[, i[, j]]) unpack a table and return the elements.
package.searchpath (name, path[, sep[, rep]]) return the full path where a file name would be matched.

Global exported functions (for Lua < 5.4)

warn (...) raise a warning message.


Functions

execute (cmd)
execute a shell command, in a compatible and platform independent way. This is a compatibility function that returns the same for Lua 5.1 and Lua 5.2+.

NOTE: Windows systems can use signed 32bit integer exitcodes. Posix systems only use exitcodes 0-255, anything else is undefined.

NOTE2: In Lua 5.2 and 5.3 a Windows exitcode of -1 would not properly be returned, this function will return it properly for all versions.

Parameters:

  • cmd a shell command

Returns:

  1. true if successful
  2. actual return code
load (ld[, source[, mode[, env]]])
Load Lua code as a text or binary chunk (in a Lua 5.2 compatible way).

Parameters:

  • ld code string or loader
  • source name of chunk for errors (optional)
  • mode 'b', 't' or 'bt' (optional)
  • env environment to load the chunk in (optional)
getfenv (f)
Get environment of a function (in a Lua 5.1 compatible way). Not 100% compatible, so with Lua 5.2 it may return nil for a function with no global references! Based on code by Sergey Rozhenko

Parameters:

  • f a function or a call stack reference
setfenv (f, env)
Set environment of a function (in a Lua 5.1 compatible way).

Parameters:

  • f a function or a call stack reference
  • env a table that becomes the new environment of f

Fields

lua51
boolean flag this is Lua 5.1 (or LuaJIT).
  • lua51
jit
boolean flag this is LuaJIT.
  • jit
jit52
boolean flag this is LuaJIT with 5.2 compatibility compiled in.
  • jit52
dir_separator
the directory separator character for the current platform.
  • dir_separator
is_windows
boolean flag this is a Windows platform.
  • is_windows

Global exported functions (for Lua 5.1 & LuaJIT)

table.pack (...)
pack an argument list into a table.

Parameters:

  • ... any arguments

Returns:

    a table with field n set to the length
table.unpack (t[, i[, j]])
unpack a table and return the elements.

NOTE: this version does NOT honor the n field, and hence it is not nil-safe. See utils.unpack for a version that is nil-safe.

Parameters:

  • t table to unpack
  • i index from which to start unpacking, defaults to 1 (optional)
  • j index of the last element to unpack, defaults to #t (optional)

Returns:

    multiple return values from the table

See also:

package.searchpath (name, path[, sep[, rep]])
return the full path where a file name would be matched. This function was introduced in Lua 5.2, so this compatibility version will be injected in Lua 5.1 engines.

Parameters:

  • name string file name, possibly dotted
  • path string a path-template in the same form as package.path or package.cpath
  • sep string template separate character to be replaced by path separator. Default: "." (optional)
  • rep string the path separator to use, defaults to system separator. Default; "/" on Unixes, "\" on Windows. (optional)

Returns:

  1. on success: path of the file
  2. on failure: nil, error string listing paths tried

See also:

Global exported functions (for Lua < 5.4)

warn (...)
raise a warning message. This functions mimics the warn function added in Lua 5.4.

Parameters:

  • ... any arguments
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.comprehension.html000066400000000000000000000207441416703176500224430ustar00rootroot00000000000000 Penlight Documentation

Module pl.comprehension

List comprehensions implemented in Lua.

See the wiki page

local C= require 'pl.comprehension' . new()

C ('x for x=1,10') ()
==> {1,2,3,4,5,6,7,8,9,10}
C 'x^2 for x=1,4' ()
==> {1,4,9,16}
C '{x,x^2} for x=1,4' ()
==> {{1,1},{2,4},{3,9},{4,16}}
C '2*x for x' {1,2,3}
==> {2,4,6}
dbl = C '2*x for x'
dbl {10,20,30}
==> {20,40,60}
C 'x for x if x % 2 == 0' {1,2,3,4,5}
==> {2,4}
C '{x,y} for x = 1,2 for y = 1,2' ()
==> {{1,1},{1,2},{2,1},{2,2}}
C '{x,y} for x for y' ({1,2},{10,20})
==> {{1,10},{1,20},{2,10},{2,20}}
assert(C 'sum(x^2 for x)' {2,3,4} == 2^2+3^2+4^2)

(c) 2008 David Manura. Licensed under the same terms as Lua (MIT license).

Dependencies: pl.utils, pl.luabalanced

See the Guide



generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.config.html000066400000000000000000000216221416703176500210330ustar00rootroot00000000000000 Penlight Documentation

Module pl.config

Reads configuration files into a Lua table.

Understands INI files, classic Unix config files, and simple delimited columns of values. See the Guide

# test.config
# Read timeout in seconds
read.timeout=10
# Write timeout in seconds
write.timeout=5
#acceptable ports
ports = 1002,1003,1004

-- readconfig.lua
local config = require 'config'
local t = config.read 'test.config'
print(pretty.write(t))

### output #####
{
  ports = {
    1002,
    1003,
    1004
  },
  write_timeout = 5,
  read_timeout = 10
}

Functions

lines (file) like io.lines(), but allows for lines to be continued with ''.
read (file[, cnfg]) read a configuration file into a table


Functions

lines (file)
like io.lines(), but allows for lines to be continued with ''.

Parameters:

  • file a file-like object (anything where read() returns the next line) or a filename. Defaults to stardard input.

Returns:

  1. an iterator over the lines, or nil
  2. error 'not a file-like object' or 'file is nil'
read (file[, cnfg])
read a configuration file into a table

Parameters:

  • file either a file-like object or a string, which must be a filename
  • cnfg tab

    a configuration table that may contain these fields:

    • smart try to deduce what kind of config file we have (default false)
    • variabilize make names into valid Lua identifiers (default true)
    • convert_numbers try to convert values into numbers (default true)
    • trim_space ensure that there is no starting or trailing whitespace with values (default true)
    • trim_quotes remove quotes from strings (default false)
    • list_delim delimiter to use when separating columns (default ',')
    • keysep separator between key and value pairs (default '=')
    (optional)

Returns:

  1. a table containing items, or nil
  2. error message (same as config.lines
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.data.html000066400000000000000000000423101416703176500204740ustar00rootroot00000000000000 Penlight Documentation

Module pl.data

Reading and querying simple tabular data.

data.read 'test.txt'
==> {{10,20},{2,5},{40,50},fieldnames={'x','y'},delim=','}

Provides a way of creating basic SQL-like queries.

require 'pl'
local d = data.read('xyz.txt')
local q = d:select('x,y,z where x > 3 and z < 2 sort by y')
for x,y,z in q do
    print(x,y,z)
end

See the Guide

Dependencies: pl.utils, pl.array2d (fallback methods)

Functions

Data.column_by_name (name) return a particular column as a list of values (method).
Data.select (condn) return a query iterator on this data (method).
Data.select_row (condn) return a row iterator on this data (method).
Data.copy_select (condn) return a new data object based on this query (method).
Data.column_names () return the field names of this data object (method).
Data.write_row (f) write out a row (method).
Data.write (f) write data out to file (method).
read (file, cnfg) read a delimited file in a Lua table.
write (data, file[, fieldnames[, delim='\t']]) write 2D data to a file.
new (d[, fieldnames]) create a new dataset from a table of rows.
query (data, condn, context, return_row) create a query iterator from a select string.
filter (Q, infile, outfile, dont_fail) Filter input using a query.


Functions

Data.column_by_name (name)
return a particular column as a list of values (method).

Parameters:

  • name either name of column, or numerical index.
Data.select (condn)
return a query iterator on this data (method).

Parameters:

  • condn string the query expression

See also:

Data.select_row (condn)
return a row iterator on this data (method).

Parameters:

  • condn string the query expression
Data.copy_select (condn)
return a new data object based on this query (method).

Parameters:

  • condn string the query expression
Data.column_names ()
return the field names of this data object (method).
Data.write_row (f)
write out a row (method).

Parameters:

  • f file-like object
Data.write (f)
write data out to file (method).

Parameters:

  • f file-like object
read (file, cnfg)
read a delimited file in a Lua table. By default, attempts to treat first line as separated list of fieldnames.

Parameters:

  • file a filename or a file-like object
  • cnfg parsing options
    • delim string a string pattern to split fields
    • fieldnames array (i.e. don't read from first line)
    • no_convert bool (default is to try conversion on first data line)
    • convert tab table of custom conversion functions with column keys
    • numfields int indices of columns known to be numbers
    • last_field_collect bool only split as many fields as fieldnames.
    • thousands_dot int thousands separator in Excel CSV is '.'
    • csv bool fields may be double-quoted and contain commas; Also, empty fields are considered to be equivalent to zero.

Returns:

  1. data object, or nil
  2. error message. May be a file error, 'not a file-like object' or a conversion error
write (data, file[, fieldnames[, delim='\t']])
write 2D data to a file. Does not assume that the data has actually been generated with new or read.

Parameters:

  • data 2D array
  • file filename or file-like object
  • fieldnames {string} list of fields (optional) (optional)
  • delim string delimiter (default tab) (default '\t')

Returns:

    true or nil, error
new (d[, fieldnames])
create a new dataset from a table of rows. Can specify the fieldnames, else the table must have a field called 'fieldnames', which is either a string of delimiter-separated names, or a table of names.
If the table does not have a field called 'delim', then an attempt will be made to guess it from the fieldnames string, defaults otherwise to tab.

Parameters:

  • d the table.
  • fieldnames {string} optional fieldnames (optional)

Returns:

    the table.
query (data, condn, context, return_row)
create a query iterator from a select string. Select string has this format:
FIELDLIST [ where LUA-CONDN [ sort by FIELD] ]
FIELDLIST is a comma-separated list of valid fields, or '*'.

The condition can also be a table, with fields 'fields' (comma-sep string or table), 'sort_by' (string) and 'where' (Lua expression string or function)

Parameters:

  • data table produced by read
  • condn select string or table
  • context a list of tables to be searched when resolving functions
  • return_row if true, wrap the results in a row table

Returns:

  1. an iterator over the specified fields, or nil
  2. an error message
filter (Q, infile, outfile, dont_fail)
Filter input using a query.

Parameters:

  • Q string a query string
  • infile filename or file-like object
  • outfile filename or file-like object
  • dont_fail bool true if you want to return an error, not just fail
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.dir.html000066400000000000000000000461711416703176500203520ustar00rootroot00000000000000 Penlight Documentation

Module pl.dir

Listing files in directories and creating/removing directory paths.

Dependencies: pl.utils, pl.path

Soft Dependencies: alien, ffi (either are used on Windows for copying/moving files)

Functions

fnmatch (filename, pattern) Test whether a file name matches a shell pattern.
filter (filenames, pattern) Return a list of all file names within an array which match a pattern.
getfiles ([dirname='.'[, mask]]) return a list of all files in a directory which match a shell pattern.
getdirectories ([dirname='.']) return a list of all subdirectories of the directory.
copyfile (src, dest, flag) copy a file.
movefile (src, dest) move a file.
walk (root, bottom_up, follow_links) return an iterator which walks through a directory tree starting at root.
rmtree (fullpath) remove a whole directory tree.
makepath (p) create a directory path.
clonetree (path1, path2, file_fun, verbose) clone a directory tree.
dirtree (d) return an iterator over all entries in a directory tree
getallfiles ([start_path='.'[, shell_pattern='*']]) Recursively returns all the file starting at 'path'.


Functions

fnmatch (filename, pattern)
Test whether a file name matches a shell pattern. Both parameters are case-normalized if operating system is case-insensitive.

Parameters:

  • filename string A file name.
  • pattern string A shell pattern. The only special characters are '*' and '?': '*' matches any sequence of characters and '?' matches any single character.

Returns:

    bool

Raises:

dir and mask must be strings
filter (filenames, pattern)
Return a list of all file names within an array which match a pattern.

Parameters:

  • filenames tab An array containing file names.
  • pattern string A shell pattern (see fnmatch).

Returns:

    List(string) List of matching file names.

Raises:

dir and mask must be strings
getfiles ([dirname='.'[, mask]])
return a list of all files in a directory which match a shell pattern.

Parameters:

  • dirname string A directory. (default '.')
  • mask string A shell pattern (see fnmatch). If not given, all files are returned. (optional)

Returns:

    {string} list of files

Raises:

dirname and mask must be strings
getdirectories ([dirname='.'])
return a list of all subdirectories of the directory.

Parameters:

  • dirname string A directory. (default '.')

Returns:

    {string} a list of directories

Raises:

dir must be a valid directory
copyfile (src, dest, flag)
copy a file.

Parameters:

  • src string source file
  • dest string destination file or directory
  • flag bool true if you want to force the copy (default)

Returns:

    bool operation succeeded

Raises:

src and dest must be strings
movefile (src, dest)
move a file.

Parameters:

  • src string source file
  • dest string destination file or directory

Returns:

    bool operation succeeded

Raises:

src and dest must be strings
walk (root, bottom_up, follow_links)
return an iterator which walks through a directory tree starting at root. The iterator returns (root,dirs,files) Note that dirs and files are lists of names (i.e. you must say path.join(root,d) to get the actual full path) If bottom_up is false (or not present), then the entries at the current level are returned before we go deeper. This means that you can modify the returned list of directories before continuing. This is a clone of os.walk from the Python libraries.

Parameters:

  • root string A starting directory
  • bottom_up bool False if we start listing entries immediately.
  • follow_links bool follow symbolic links

Returns:

    an iterator returning root,dirs,files

Raises:

root must be a directory
rmtree (fullpath)
remove a whole directory tree. Symlinks in the tree will be deleted without following them.

Parameters:

  • fullpath string A directory path (must be an actual directory, not a symlink)

Returns:

  1. true or nil
  2. error if failed

Raises:

fullpath must be a string
makepath (p)
create a directory path. This will create subdirectories as necessary!

Parameters:

Returns:

    true on success, nil + errormsg on failure

Raises:

failure to create
clonetree (path1, path2, file_fun, verbose)
clone a directory tree. Will always try to create a new directory structure if necessary.

Parameters:

  • path1 string the base path of the source tree
  • path2 string the new base path for the destination
  • file_fun func an optional function to apply on all files
  • verbose bool an optional boolean to control the verbosity of the output. It can also be a logging function that behaves like print()

Returns:

  1. true, or nil
  2. error message, or list of failed directory creations
  3. list of failed file operations

Raises:

path1 and path2 must be strings

Usage:

    clonetree('.','../backup',copyfile)
dirtree (d)
return an iterator over all entries in a directory tree

Parameters:

Returns:

    an iterator giving pathname and mode (true for dir, false otherwise)

Raises:

d must be a non-empty string
getallfiles ([start_path='.'[, shell_pattern='*']])
Recursively returns all the file starting at 'path'. It can optionally take a shell pattern and only returns files that match 'shell_pattern'. If a pattern is given it will do a case insensitive search.

Parameters:

  • start_path string A directory. (default '.')
  • shell_pattern string A shell pattern (see fnmatch). (default '*')

Returns:

    List(string) containing all the files found recursively starting at 'path' and filtered by 'shell_pattern'.

Raises:

start_path must be a directory
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.file.html000066400000000000000000000213461416703176500205100ustar00rootroot00000000000000 Penlight Documentation

Module pl.file

File manipulation functions: reading, writing, moving and copying.

This module wraps a number of functions from other modules into a file related module for convenience.

Dependencies: pl.utils, pl.dir, pl.path

Functions

read () return the contents of a file as a string.
write () write a string to a file.
copy () copy a file.
move () move a file.
access_time () Return the time of last access as the number of seconds since the epoch.
creation_time () Return when the file was created.
modified_time () Return the time of last modification.
delete () Delete a file.


Functions

read ()
return the contents of a file as a string. This function is a copy of utils.readfile.
write ()
write a string to a file. This function is a copy of utils.writefile.
copy ()
copy a file. This function is a copy of dir.copyfile.
move ()
move a file. This function is a copy of dir.movefile.
access_time ()
Return the time of last access as the number of seconds since the epoch. This function is a copy of path.getatime.
creation_time ()
Return when the file was created. This function is a copy of path.getctime.
modified_time ()
Return the time of last modification. This function is a copy of path.getmtime.
delete ()
Delete a file. This function is a copy of os.remove.
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.func.html000066400000000000000000000325041416703176500205220ustar00rootroot00000000000000 Penlight Documentation

Module pl.func

Functional helpers like composition, binding and placeholder expressions.

Placeholder expressions are useful for short anonymous functions, and were inspired by the Boost Lambda library.

> utils.import 'pl.func'
> ls = List{10,20,30}
> = ls:map(_1+1)
{11,21,31}

They can also be used to bind particular arguments of a function.

> p = bind(print,'start>',_0)
> p(10,20,30)
> start>   10   20  30

See the Guide

Dependencies: pl.utils, pl.tablex

Functions

import (tname, context) wrap a table of functions.
register (fun[, name]) register a function for use in placeholder expressions.
tail (ls) all elements of a table except the first.
repr (e, lastpred) create a string representation of a placeholder expression.
instantiate (e) instantiate a PE into an actual function.
I (e) instantiate a PE unless it has already been done.
bind1 (fn, p) bind the first parameter of the function to a value.
compose (f, g) create a function which chains two functions.
bind (fn, ...) bind the arguments of a function to given values.


Functions

import (tname, context)
wrap a table of functions. This makes them available for use in placeholder expressions.

Parameters:

  • tname string a table name
  • context tab context to put results, defaults to environment of caller
register (fun[, name])
register a function for use in placeholder expressions.

Parameters:

  • fun func a function
  • name string an optional name (optional)

Returns:

    a placeholder functiond
tail (ls)
all elements of a table except the first.

Parameters:

  • ls tab a list-like table.
repr (e, lastpred)
create a string representation of a placeholder expression.

Parameters:

  • e a placeholder expression
  • lastpred not used
instantiate (e)
instantiate a PE into an actual function. First we find the largest placeholder used, e.g. 2; from this a list of the formal parameters can be build. Then we collect and replace any non-PE values from the PE, and build up a constant binding list. Finally, the expression can be compiled, and e.PEfunction is set.

Parameters:

  • e a placeholder expression

Returns:

    a function
I (e)
instantiate a PE unless it has already been done.

Parameters:

  • e a placeholder expression

Returns:

    the function
bind1 (fn, p)
bind the first parameter of the function to a value.

Parameters:

  • fn func a function of one or more arguments
  • p a value

Returns:

    a function of one less argument

Usage:

    (bind1(math.max,10))(20) == math.max(10,20)
compose (f, g)
create a function which chains two functions.

Parameters:

  • f func a function of at least one argument
  • g func a function of at least one argument

Returns:

    a function

Usage:

    printf = compose(io.write,string.format)
bind (fn, ...)
bind the arguments of a function to given values. bind(fn,v,_2) is equivalent to bind1(fn,v).

Parameters:

  • fn func a function of at least one argument
  • ... values or placeholder variables

Returns:

    a function

Usage:

  • (bind(f,_1,a))(b) == f(a,b)
  • (bind(f,_2,_1))(a,b) == f(b,a)
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.html000066400000000000000000000132061416703176500175660ustar00rootroot00000000000000 Penlight Documentation

Module pl

Entry point for loading all PL libraries only on demand, into the global space.

Requiring 'pl' means that whenever a module is implicitly accesssed (e.g. utils.split) then that module is dynamically loaded. The submodules are all brought into the global space. Updated to use pl.import_into



generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.import_into.html000066400000000000000000000136271416703176500221370ustar00rootroot00000000000000 Penlight Documentation

Module pl.import_into

PL loader, for loading all PL libraries, only on demand.

Whenever a module is implicitly accesssed, the table will have the module automatically injected. (e.g. _ENV.tablex) then that module is dynamically loaded. The submodules are all brought into the table that is provided as the argument, or returned in a new table. If a table is provided, that table's metatable is clobbered, but the values are not. This module returns a single function, which is passed the environment. If this is true, then return a 'shadow table' as the module See the Guide



generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.input.html000066400000000000000000000245541416703176500207340ustar00rootroot00000000000000 Penlight Documentation

Module pl.input

Iterators for extracting words or numbers from an input source.

require 'pl'
local total,n = seq.sum(input.numbers())
print('average',total/n)

source is defined as a string or a file-like object (i.e. has a read() method which returns the next line)

See here

Dependencies: pl.utils

Functions

alltokens (getter, pattern[, fn]) create an iterator over all tokens.
create_getter (f) create a function which grabs the next value from a source.
numbers (f) generate a sequence of numbers from a source.
words (f) generate a sequence of words from a source.
fields (ids, delim, f, opts) parse an input source into fields.


Functions

alltokens (getter, pattern[, fn])
create an iterator over all tokens. based on allwords from PiL, 7.1

Parameters:

  • getter func any function that returns a line of text
  • pattern string
  • fn string Optionally can pass a function to process each token as it's found. (optional)

Returns:

    an iterator
create_getter (f)
create a function which grabs the next value from a source. If the source is a string, then the getter will return the string and thereafter return nil. If not specified then the source is assumed to be stdin.

Parameters:

  • f a string or a file-like object (i.e. has a read() method which returns the next line)

Returns:

    a getter function
numbers (f)
generate a sequence of numbers from a source.

Parameters:

  • f A source

Returns:

    An iterator
words (f)
generate a sequence of words from a source.

Parameters:

  • f A source

Returns:

    An iterator
fields (ids, delim, f, opts)
parse an input source into fields. By default, will fail if it cannot convert a field to a number.

Parameters:

  • ids a list of field indices, or a maximum field index
  • delim string delimiter to parse fields (default space)
  • f a source @see create_getter
  • opts tab option table, {no_fail=true}

Returns:

    an iterator with the field values

Usage:

    for x,y in fields {2,3} do print(x,y) end -- 2nd and 3rd fields from stdin
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.lapp.html000066400000000000000000000276131416703176500205300ustar00rootroot00000000000000 Penlight Documentation

Module pl.lapp

Simple command-line parsing using human-readable specification.

Supports GNU-style parameters.

lapp = require 'pl.lapp'
local args = lapp [[
Does some calculations
  -o,--offset (default 0.0)  Offset to add to scaled number
  -s,--scale  (number)  Scaling factor
  <number> (number) Number to be scaled
]]

print(args.offset + args.scale * args.number)

Lines beginning with '-' are flags; there may be a short and a long name; lines beginning with '<var>' are arguments. Anything in parens after the flag/argument is either a default, a type name or a range constraint.

See the Guide

Dependencies: pl.sip

Functions

quit (msg, no_usage) quit this script immediately.
error (msg, no_usage) print an error to stderr and quit.
open (file[, opt]) open a file.
assert (condn, msg) quit if the condition is false.
add_type (name, converter[, constraint]) add a new type to Lapp.
process_options_string (str, args) process a Lapp options string.

Fields

show_usage_error controls whether to dump usage on error.


Functions

quit (msg, no_usage)
quit this script immediately.

Parameters:

  • msg string optional message
  • no_usage bool suppress 'usage' display
error (msg, no_usage)
print an error to stderr and quit.

Parameters:

  • msg string a message
  • no_usage bool suppress 'usage' display
open (file[, opt])
open a file. This will quit on error, and keep a list of file objects for later cleanup.

Parameters:

assert (condn, msg)
quit if the condition is false.

Parameters:

  • condn bool a condition
  • msg string message text
add_type (name, converter[, constraint])
add a new type to Lapp. These appear in parens after the value like a range constraint, e.g. ' (integer) Process PID'

Parameters:

  • name string name of type
  • converter either a function to convert values, or a Lua type name.
  • constraint func optional function to verify values, should use lapp.error if failed. (optional)
process_options_string (str, args)
process a Lapp options string. Usually called as lapp().

Parameters:

  • str string the options text
  • args {string} a table of arguments (default is _G.arg)

Returns:

    a table with parameter-value pairs

Fields

show_usage_error
controls whether to dump usage on error. Defaults to true
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.lexer.html000066400000000000000000000371341416703176500207120ustar00rootroot00000000000000 Penlight Documentation

Module pl.lexer

Lexical scanner for creating a sequence of tokens from text.

lexer.scan(s) returns an iterator over all tokens found in the string s. This iterator returns two values, a token type string (such as 'string' for quoted string, 'iden' for identifier) and the value of the token.

Versions specialized for Lua and C are available; these also handle block comments and classify keywords as 'keyword' tokens. For example:

> s = 'for i=1,n do'
> for t,v in lexer.lua(s)  do print(t,v) end
keyword for
iden    i
=       =
number  1
,       ,
iden    n
keyword do

See the Guide for further discussion

Functions

scan (s, matches[, filter[, options]]) create a plain token iterator from a string or file-like object.
insert (tok, a1, a2) insert tokens into a stream.
getline (tok) get everything in a stream upto a newline.
lineno (tok) get current line number.
getrest (tok) get the rest of the stream.
get_keywords () get the Lua keywords as a set-like table.
lua (s[, filter[, options]]) create a Lua token iterator from a string or file-like object.
cpp (s[, filter[, options]]) create a C/C++ token iterator from a string or file-like object.
get_separated_list (tok[, endtoken=')'[, delim=']]) get a list of parameters separated by a delimiter from a stream.
skipws (tok) get the next non-space token from the stream.
expecting (tok, expected_type, no_skip_ws) get the next token, which must be of the expected type.


Functions

scan (s, matches[, filter[, options]])
create a plain token iterator from a string or file-like object.

Parameters:

  • s string or file a string or a file-like object with :read() method returning lines.
  • matches tab an optional match table - array of token descriptions. A token is described by a {pattern, action} pair, where pattern should match token body and action is a function called when a token of described type is found.
  • filter tab a table of token types to exclude, by default {space=true} (optional)
  • options tab a table of options; by default, {number=true,string=true}, which means convert numbers and strip string quotes. (optional)
insert (tok, a1, a2)
insert tokens into a stream.

Parameters:

  • tok a token stream
  • a1 a string is the type, a table is a token list and a function is assumed to be a token-like iterator (returns type & value)
  • a2 string a string is the value
getline (tok)
get everything in a stream upto a newline.

Parameters:

  • tok a token stream

Returns:

    a string
lineno (tok)
get current line number.

Parameters:

  • tok a token stream

Returns:

    the line number. if the input source is a file-like object, also return the column.
getrest (tok)
get the rest of the stream.

Parameters:

  • tok a token stream

Returns:

    a string
get_keywords ()
get the Lua keywords as a set-like table. So res["and"] etc would be true.

Returns:

    a table
lua (s[, filter[, options]])
create a Lua token iterator from a string or file-like object. Will return the token type and value.

Parameters:

  • s string the string
  • filter tab a table of token types to exclude, by default {space=true,comments=true} (optional)
  • options tab a table of options; by default, {number=true,string=true}, which means convert numbers and strip string quotes. (optional)
cpp (s[, filter[, options]])
create a C/C++ token iterator from a string or file-like object. Will return the token type type and value.

Parameters:

  • s string the string
  • filter tab a table of token types to exclude, by default {space=true,comments=true} (optional)
  • options tab a table of options; by default, {number=true,string=true}, which means convert numbers and strip string quotes. (optional)
get_separated_list (tok[, endtoken=')'[, delim=']])
get a list of parameters separated by a delimiter from a stream.

Parameters:

  • tok the token stream
  • endtoken string end of list. Can be '\n' (default ')')
  • delim string separator (default ')

Returns:

    a list of token lists.
skipws (tok)
get the next non-space token from the stream.

Parameters:

  • tok the token stream.
expecting (tok, expected_type, no_skip_ws)
get the next token, which must be of the expected type. Throws an error if this type does not match!

Parameters:

  • tok the token stream
  • expected_type string the token type
  • no_skip_ws bool whether we should skip whitespace
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.luabalanced.html000066400000000000000000000136761416703176500220330ustar00rootroot00000000000000 Penlight Documentation

Module pl.luabalanced

Extract delimited Lua sequences from strings.

Inspired by Damian Conway's Text::Balanced in Perl.

  • [1] Lua Wiki Page
  • [2] http://search.cpan.org/dist/Text-Balanced/lib/Text/Balanced.pm

 local lb = require "pl.luabalanced"
 --Extract Lua expression starting at position 4.
  print(lb.match_expression("if x^2 + x > 5 then print(x) end", 4))
  --> x^2 + x > 5     16
 --Extract Lua string starting at (default) position 1.
 print(lb.match_string([["test\"123" .. "more"]]))
 --> "test\"123"     12
 
(c) 2008, David Manura, Licensed under the same terms as Lua (MIT license).



generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.operator.html000066400000000000000000000440521416703176500214230ustar00rootroot00000000000000 Penlight Documentation

Module pl.operator

Lua operators available as functions.

(similar to the Python module of the same name)

There is a module field optable which maps the operator strings onto these functions, e.g. operator.optable['()']==operator.call

Operator strings like '>' and '{}' can be passed to most Penlight functions expecting a function argument.

Functions

call (fn, ...) apply function to some arguments ()
index (t, k) get the indexed value from a table []
eq (a, b) returns true if arguments are equal ==
neq (a, b) returns true if arguments are not equal ~=
lt (a, b) returns true if a is less than b <
le (a, b) returns true if a is less or equal to b <=
gt (a, b) returns true if a is greater than b >
ge (a, b) returns true if a is greater or equal to b >=
len (a) returns length of string or table #
add (a, b) add two values +
sub (a, b) subtract b from a -
mul (a, b) multiply two values *
div (a, b) divide first value by second /
pow (a, b) raise first to the power of second ^
mod (a, b) modulo; remainder of a divided by b %
concat (a, b) concatenate two values (either strings or __concat defined) ..
unm (a) return the negative of a value -
lnot (a) false if value evaluates as true not
land (a, b) true if both values evaluate as true and
lor (a, b) true if either value evaluate as true or
table (...) make a table from the arguments {}
match (a, b) match two strings ~.
nop (...) the null operation.

Tables

optable Map from operator symbol to function.


Functions

call (fn, ...)
apply function to some arguments ()

Parameters:

  • fn a function or callable object
  • ... arguments
index (t, k)
get the indexed value from a table []

Parameters:

  • t a table or any indexable object
  • k the key
eq (a, b)
returns true if arguments are equal ==

Parameters:

  • a value
  • b value
neq (a, b)
returns true if arguments are not equal ~=

Parameters:

  • a value
  • b value
lt (a, b)
returns true if a is less than b <

Parameters:

  • a value
  • b value
le (a, b)
returns true if a is less or equal to b <=

Parameters:

  • a value
  • b value
gt (a, b)
returns true if a is greater than b >

Parameters:

  • a value
  • b value
ge (a, b)
returns true if a is greater or equal to b >=

Parameters:

  • a value
  • b value
len (a)
returns length of string or table #

Parameters:

  • a a string or a table
add (a, b)
add two values +

Parameters:

  • a value
  • b value
sub (a, b)
subtract b from a -

Parameters:

  • a value
  • b value
mul (a, b)
multiply two values *

Parameters:

  • a value
  • b value
div (a, b)
divide first value by second /

Parameters:

  • a value
  • b value
pow (a, b)
raise first to the power of second ^

Parameters:

  • a value
  • b value
mod (a, b)
modulo; remainder of a divided by b %

Parameters:

  • a value
  • b value
concat (a, b)
concatenate two values (either strings or __concat defined) ..

Parameters:

  • a value
  • b value
unm (a)
return the negative of a value -

Parameters:

  • a value
lnot (a)
false if value evaluates as true not

Parameters:

  • a value
land (a, b)
true if both values evaluate as true and

Parameters:

  • a value
  • b value
lor (a, b)
true if either value evaluate as true or

Parameters:

  • a value
  • b value
table (...)
make a table from the arguments {}

Parameters:

  • ... non-nil arguments

Returns:

    a table
match (a, b)
match two strings ~. uses string.find

Parameters:

  • a
  • b
nop (...)
the null operation.

Parameters:

  • ... arguments

Returns:

    the arguments

Tables

optable

Map from operator symbol to function. Most of these map directly from operators; But note these extras

Fields:

  • operator
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.path.html000066400000000000000000000727251416703176500205340ustar00rootroot00000000000000 Penlight Documentation

Module pl.path

Path manipulation and file queries.

This is modelled after Python's os.path library (10.1); see the Guide.

NOTE: the functions assume the paths being dealt with to originate from the OS the application is running on. Windows drive letters are not to be used when running on a Unix system for example. The one exception is Windows paths to allow both forward and backward slashes (since Lua also accepts those)

Dependencies: pl.utils, lfs

Functions

dir () Lua iterator over the entries of a given directory.
mkdir () Creates a directory.
rmdir () Removes a directory.
attrib () Gets attributes.
currentdir () Get the working directory.
link_attrib () Gets symlink attributes.
chdir () Changes the working directory.
isdir (P) is this a directory?
isfile (P) is this a file?
getsize (P) return size of a file.
exists (P) does a path exist?
getatime (P) Return the time of last access as the number of seconds since the epoch.
getmtime (P) Return the time of last modification as the number of seconds since the epoch.
getctime (P) Return the system's ctime as the number of seconds since the epoch.
splitpath (P) given a path, return the directory part and a file part.
abspath (P[, pwd]) return an absolute path.
splitext (P) given a path, return the root part and the extension part.
dirname (P) return the directory part of a path
basename (P) return the file part of a path
extension (P) get the extension part of a path.
isabs (P) is this an absolute path?
join (p1, p2, ...) return the path resulting from combining the individual paths.
normcase (P) normalize the case of a pathname.
normpath (P) normalize a path name.
relpath (P[, start]) relative path from current directory or optional start point
expanduser (P) Replace a starting '~' with the user's home directory.
tmpname () Return a suitable full path to a new temporary file name.
common_prefix (path1, path2) return the largest common prefix path of two paths.
package_path (mod) return the full path where a particular Lua module would be found.

Fields

is_windows are we running Windows?
sep path separator for this platform.
dirsep separator for PATH for this platform


Functions

dir ()
Lua iterator over the entries of a given directory. Implicit link to luafilesystem.dir
mkdir ()
Creates a directory. Implicit link to luafilesystem.mkdir
rmdir ()
Removes a directory. Implicit link to luafilesystem.rmdir
attrib ()
Gets attributes. Implicit link to luafilesystem.attributes
currentdir ()
Get the working directory. Implicit link to luafilesystem.currentdir
link_attrib ()
Gets symlink attributes. Implicit link to luafilesystem.symlinkattributes
chdir ()
Changes the working directory. On Windows, if a drive is specified, it also changes the current drive. If only specifying the drive, it will only switch drive, but not modify the path. Implicit link to luafilesystem.chdir
isdir (P)
is this a directory?

Parameters:

isfile (P)
is this a file?

Parameters:

getsize (P)
return size of a file.

Parameters:

exists (P)
does a path exist?

Parameters:

Returns:

    the file path if it exists (either as file, directory, socket, etc), nil otherwise
getatime (P)
Return the time of last access as the number of seconds since the epoch.

Parameters:

getmtime (P)
Return the time of last modification as the number of seconds since the epoch.

Parameters:

getctime (P)
Return the system's ctime as the number of seconds since the epoch.

Parameters:

splitpath (P)
given a path, return the directory part and a file part. if there's no directory part, the first value will be empty

Parameters:

Returns:

  1. directory part
  2. file part

Usage:

    local dir, file = path.splitpath("some/dir/myfile.txt")
    assert(dir == "some/dir")
    assert(file == "myfile.txt")
    
    local dir, file = path.splitpath("some/dir/")
    assert(dir == "some/dir")
    assert(file == "")
    
    local dir, file = path.splitpath("some_dir")
    assert(dir == "")
    assert(file == "some_dir")
abspath (P[, pwd])
return an absolute path.

Parameters:

  • P string A file path
  • pwd string optional start path to use (default is current dir) (optional)
splitext (P)
given a path, return the root part and the extension part. if there's no extension part, the second value will be empty

Parameters:

Returns:

  1. string root part (everything upto the "."", maybe empty)
  2. string extension part (including the ".", maybe empty)

Usage:

    local file_path, ext = path.splitext("/bonzo/dog_stuff/cat.txt")
    assert(file_path == "/bonzo/dog_stuff/cat")
    assert(ext == ".txt")
    
    local file_path, ext = path.splitext("")
    assert(file_path == "")
    assert(ext == "")
dirname (P)
return the directory part of a path

Parameters:

Returns:

    string everything before the last dir-separator

See also:

Usage:

    path.dirname("/some/path/file.txt")   -- "/some/path"
    path.dirname("file.txt")              -- "" (empty string)
basename (P)
return the file part of a path

Parameters:

Returns:

    string

See also:

Usage:

    path.basename("/some/path/file.txt")  -- "file.txt"
    path.basename("/some/path/file/")     -- "" (empty string)
extension (P)
get the extension part of a path.

Parameters:

Returns:

    string

See also:

Usage:

    path.extension("/some/path/file.txt") -- ".txt"
    path.extension("/some/path/file_txt") -- "" (empty string)
isabs (P)
is this an absolute path?

Parameters:

Usage:

    path.isabs("hello/path")    -- false
    path.isabs("/hello/path")   -- true
    -- Windows;
    path.isabs("hello\path")    -- false
    path.isabs("\hello\path")   -- true
    path.isabs("C:\hello\path") -- true
    path.isabs("C:hello\path")  -- false
join (p1, p2, ...)
return the path resulting from combining the individual paths. if the second (or later) path is absolute, we return the last absolute path (joined with any non-absolute paths following). empty elements (except the last) will be ignored.

Parameters:

Returns:

    string the combined path

Usage:

    path.join("/first","second","third")   -- "/first/second/third"
    path.join("first","second/third")      -- "first/second/third"
    path.join("/first","/second","third")  -- "/second/third"
normcase (P)

normalize the case of a pathname. On Unix, this returns the path unchanged, for Windows it converts;

  • the path to lowercase
  • forward slashes to backward slashes

Parameters:

Usage:

    path.normcase("/Some/Path/File.txt")
    -- Windows: "\some\path\file.txt"
    -- Others : "/Some/Path/File.txt"
normpath (P)
normalize a path name. A//B, A/./B, and A/foo/../B all become A/B.

An empty path results in '.'.

Parameters:

relpath (P[, start])
relative path from current directory or optional start point

Parameters:

  • P string a path
  • start string optional start point (default current directory) (optional)
expanduser (P)
Replace a starting '~' with the user's home directory. In windows, if HOME isn't set, then USERPROFILE is used in preference to HOMEDRIVE HOMEPATH. This is guaranteed to be writeable on all versions of Windows.

Parameters:

tmpname ()
Return a suitable full path to a new temporary file name. unlike os.tmpname(), it always gives you a writeable path (uses TEMP environment variable on Windows)
common_prefix (path1, path2)
return the largest common prefix path of two paths.

Parameters:

Returns:

    the common prefix (Windows: separators will be normalized, casing will be original)
package_path (mod)
return the full path where a particular Lua module would be found. Both package.path and package.cpath is searched, so the result may either be a Lua file or a shared library.

Parameters:

  • mod string name of the module

Returns:

  1. on success: path of module, lua or binary
  2. on error: nil, error string listing paths tried

Fields

is_windows
are we running Windows?
sep
path separator for this platform.
dirsep
separator for PATH for this platform
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.permute.html000066400000000000000000000266321416703176500212550ustar00rootroot00000000000000 Penlight Documentation

Module pl.permute

Permutation operations.

Dependencies: pl.utils, pl.tablex

Functions

order_iter (a) an iterator over all order-permutations of the elements of a list.
order_table (a) construct a table containing all the order-permutations of a list.
list_iter (...) an iterator over all permutations of the elements of the given lists.
list_table (...) construct a table containing all the permutations of a set of lists.
iter (...) deprecated.
table (...) deprecated.


Functions

order_iter (a)
an iterator over all order-permutations of the elements of a list. Please note that the same list is returned each time, so do not keep references!

Parameters:

  • a list-like table

Returns:

    an iterator which provides the next permutation as a list
order_table (a)
construct a table containing all the order-permutations of a list.

Parameters:

  • a list-like table

Returns:

    a table of tables

Usage:

    permute.order_table {1,2,3} --> {{2,3,1},{3,2,1},{3,1,2},{1,3,2},{2,1,3},{1,2,3}}
list_iter (...)
an iterator over all permutations of the elements of the given lists.

Parameters:

  • ... list-like tables, they are nil-safe if a length-field n is provided (see utils.pack)

Returns:

    an iterator which provides the next permutation as return values in the same order as the provided lists, preceeded by an index

Usage:

    local strs = utils.pack("one", nil, "three")  -- adds an 'n' field for nil-safety
    local bools = utils.pack(true, false)
    local iter = permute.list_iter(strs, bools)
    
    print(iter())    --> 1, one, true
    print(iter())    --> 2, nil, true
    print(iter())    --> 3, three, true
    print(iter())    --> 4, one, false
    print(iter())    --> 5, nil, false
    print(iter())    --> 6, three, false
list_table (...)
construct a table containing all the permutations of a set of lists.

Parameters:

  • ... list-like tables, they are nil-safe if a length-field n is provided

Returns:

    a list of lists, the sub-lists have an 'n' field for nil-safety

Usage:

    local strs = utils.pack("one", nil, "three")  -- adds an 'n' field for nil-safety
    local bools = utils.pack(true, false)
    local results = permute.list_table(strs, bools)
    -- results = {
    --   { "one, true, n = 2 }
    --   { nil, true, n = 2 },
    --   { "three, true, n = 2 },
    --   { "one, false, n = 2 },
    --   { nil, false, n = 2 },
    --   { "three", false, n = 2 },
    -- }
iter (...)
deprecated.

Parameters:

  • ...

See also:

table (...)
deprecated.

Parameters:

  • ...

See also:

generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.pretty.html000066400000000000000000000331551416703176500211210ustar00rootroot00000000000000 Penlight Documentation

Module pl.pretty

Pretty-printing Lua tables.

Also provides a sandboxed Lua table reader and a function to present large numbers in human-friendly format.

Dependencies: pl.utils, pl.lexer, pl.stringx, debug

Functions

read (s) Read a string representation of a Lua table.
load (s[, env[, paranoid]]) Read a Lua chunk.
write (tbl[, space[, not_clever]]) Create a string representation of a Lua table.
dump (t[, filename]) Dump a Lua table out to a file or stdout.
debug (...) Dump a series of arguments to stdout for debug purposes.
number (num[, kind[, prec]]) Format large numbers nicely for human consumption.


Functions

read (s)
Read a string representation of a Lua table. This function loads and runs the string as Lua code, but bails out if it contains a function definition. Loaded string is executed in an empty environment.

Parameters:

  • s string string to read in {...} format, possibly with some whitespace before or after the curly braces. A single line comment may be present at the beginning.

Returns:

    a table in case of success. If loading the string failed, return nil and error message. If executing loaded string failed, return nil and the error it raised.
load (s[, env[, paranoid]])
Read a Lua chunk.

Parameters:

  • s string Lua code.
  • env tab environment used to run the code, empty by default. (optional)
  • paranoid bool abort loading if any looping constructs a found in the code and disable string methods. (optional)

Returns:

    the environment in case of success or nil and syntax or runtime error if something went wrong.
write (tbl[, space[, not_clever]])
Create a string representation of a Lua table. This function never fails, but may complain by returning an extra value. Normally puts out one item per line, using the provided indent; set the second parameter to an empty string if you want output on one line.

NOTE: this is NOT a serialization function, not a full blown debug function. Checkout out respectively the serpent or inspect Lua modules for that if you need them.

Parameters:

  • tbl tab Table to serialize to a string.
  • space string The indent to use. Defaults to two spaces; pass an empty string for no indentation. (optional)
  • not_clever bool Pass true for plain output, e.g {['key']=1}. Defaults to false. (optional)

Returns:

  1. a string
  2. an optional error message
dump (t[, filename])
Dump a Lua table out to a file or stdout.

Parameters:

  • t tab The table to write to a file or stdout.
  • filename string File name to write too. Defaults to writing to stdout. (optional)
debug (...)
Dump a series of arguments to stdout for debug purposes. This function is attached to the module table __call method, to make it extra easy to access. So the full:

 print(require("pl.pretty").write({...}))

Can be shortened to:

 require"pl.pretty" (...)

Any nil entries will be printed as "<nil>" to make them explicit.

Parameters:

  • ... the parameters to dump to stdout.

Usage:

    -- example debug output
    require"pl.pretty" ("hello", nil, "world", { bye = "world", true} )
    
    -- output:
    {
      ["arg 1"] = "hello",
      ["arg 2"] = "<nil>",
      ["arg 3"] = "world",
      ["arg 4"] = {
        true,
        bye = "world"
      }
    }
number (num[, kind[, prec]])
Format large numbers nicely for human consumption.

Parameters:

  • num number a number.
  • kind string one of 'M' (memory in KiB, MiB, etc.), 'N' (postfixes are 'K', 'M' and 'B'), or 'T' (use commas as thousands separator), 'N' by default. (optional)
  • prec int number of digits to use for 'M' and 'N', 1 by default. (optional)
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.seq.html000066400000000000000000000547771416703176500203770ustar00rootroot00000000000000 Penlight Documentation

Module pl.seq

Manipulating iterators as sequences.

See The Guide

Dependencies: pl.utils, pl.types, debug

Functions

matching (s) given a string, return a function(y) which matches y against the string.
list (t) sequence adaptor for a table.
keys (t) return the keys of the table.
range (start, finish) create an iterator over a numerical range.
minmax (iter) return the minimum and the maximum value of the sequence.
sum (iter, fn) return the sum and element count of the sequence.
copy (iter) create a table from the sequence.
copy2 (iter, i1, i2) create a table of pairs from the double-valued sequence.
copy_tuples (iter) create a table of 'tuples' from a multi-valued sequence.
random (n, l, u) return an iterator of random numbers.
sort (iter, comp) return an iterator to the sorted elements of a sequence.
zip (iter1, iter2) return an iterator which returns elements of two sequences.
count_map (iter) Makes a table where the key/values are the values and value counts of the sequence.
printall (iter, sep, nfields, fmt) print out a sequence iter with a separator.
map (fn, iter, arg) return a sequence where every element of a sequence has been transformed by a function.
filter (iter, pred, arg) filter a sequence using a predicate function.
reduce (fn, iter, initval) 'reduce' a sequence using a binary function.
take (iter, n) take the first n values from the sequence.
skip (iter, n) skip the first n values of a sequence
enum (iter) a sequence with a sequence count and the original value.
mapmethod (iter, name, arg1, arg2) map using a named method over a sequence.
last (iter) a sequence of (last,current) values from another sequence.
foreach (iter, fn) call the function on each element of the sequence.
lines (f, ...) create a wrapped iterator over all lines in the file.


Functions

matching (s)
given a string, return a function(y) which matches y against the string.

Parameters:

  • s a string
list (t)
sequence adaptor for a table. Note that if any generic function is passed a table, it will automatically use seq.list()

Parameters:

  • t a list-like table

Usage:

  • sum(list(t)) is the sum of all elements of t
  • for x in list(t) do...end
keys (t)
return the keys of the table.

Parameters:

  • t an arbitrary table

Returns:

    iterator over keys
range (start, finish)
create an iterator over a numerical range. Like the standard Python function xrange.

Parameters:

  • start a number
  • finish a number greater than start
minmax (iter)
return the minimum and the maximum value of the sequence.

Parameters:

  • iter a sequence

Returns:

  1. minimum value
  2. maximum value
sum (iter, fn)
return the sum and element count of the sequence.

Parameters:

  • iter a sequence
  • fn an optional function to apply to the values
copy (iter)
create a table from the sequence. (This will make the result a List.)

Parameters:

  • iter a sequence

Returns:

    a List

Usage:

  • copy(list(ls)) is equal to ls
  • copy(list {1,2,3}) == List{1,2,3}
copy2 (iter, i1, i2)
create a table of pairs from the double-valued sequence.

Parameters:

  • iter a double-valued sequence
  • i1 used to capture extra iterator values
  • i2 as with pairs & ipairs

Returns:

    a list-like table

Usage:

    copy2(ipairs{10,20,30}) == {{1,10},{2,20},{3,30}}
copy_tuples (iter)
create a table of 'tuples' from a multi-valued sequence. A generalization of copy2 above

Parameters:

  • iter a multiple-valued sequence

Returns:

    a list-like table
random (n, l, u)
return an iterator of random numbers.

Parameters:

  • n the length of the sequence
  • l same as the first optional argument to math.random
  • u same as the second optional argument to math.random

Returns:

    a sequence
sort (iter, comp)
return an iterator to the sorted elements of a sequence.

Parameters:

  • iter a sequence
  • comp an optional comparison function (comp(x,y) is true if x < y)
zip (iter1, iter2)
return an iterator which returns elements of two sequences.

Parameters:

  • iter1 a sequence
  • iter2 a sequence

Usage:

    for x,y in seq.zip(ls1,ls2) do....end
count_map (iter)
Makes a table where the key/values are the values and value counts of the sequence. This version works with 'hashable' values like strings and numbers. pl.tablex.count_map is more general.

Parameters:

  • iter a sequence

Returns:

  1. a map-like table
  2. a table

See also:

printall (iter, sep, nfields, fmt)
print out a sequence iter with a separator.

Parameters:

  • iter a sequence
  • sep the separator (default space)
  • nfields maximum number of values per line (default 7)
  • fmt optional format function for each value
map (fn, iter, arg)
return a sequence where every element of a sequence has been transformed by a function. If you don't supply an argument, then the function will receive both values of a double-valued sequence, otherwise behaves rather like tablex.map.

Parameters:

  • fn a function to apply to elements; may take two arguments
  • iter a sequence of one or two values
  • arg optional argument to pass to function.
filter (iter, pred, arg)
filter a sequence using a predicate function.

Parameters:

  • iter a sequence of one or two values
  • pred a boolean function; may take two arguments
  • arg optional argument to pass to function.
reduce (fn, iter, initval)
'reduce' a sequence using a binary function.

Parameters:

  • fn func a function of two arguments
  • iter a sequence
  • initval optional initial value

Usage:

  • seq.reduce(operator.add,seq.list{1,2,3,4}) == 10
  • seq.reduce('-',{1,2,3,4,5}) == -13
take (iter, n)
take the first n values from the sequence.

Parameters:

  • iter a sequence of one or two values
  • n number of items to take

Returns:

    a sequence of at most n items
skip (iter, n)
skip the first n values of a sequence

Parameters:

  • iter a sequence of one or more values
  • n number of items to skip
enum (iter)
a sequence with a sequence count and the original value. enum(copy(ls)) is a roundabout way of saying ipairs(ls).

Parameters:

  • iter a single or double valued sequence

Returns:

    sequence of (i,v), i = 1..n and v is from iter.
mapmethod (iter, name, arg1, arg2)
map using a named method over a sequence.

Parameters:

  • iter a sequence
  • name the method name
  • arg1 optional first extra argument
  • arg2 optional second extra argument
last (iter)
a sequence of (last,current) values from another sequence. This will return S(i-1),S(i) if given S(i)

Parameters:

  • iter a sequence
foreach (iter, fn)
call the function on each element of the sequence.

Parameters:

  • iter a sequence with up to 3 values
  • fn a function
lines (f, ...)
create a wrapped iterator over all lines in the file.

Parameters:

  • f either a filename, file-like object, or 'STDIN' (for standard input)
  • ... for Lua 5.2 only, optional format specifiers, as in io.read.

Returns:

    a sequence wrapper
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.sip.html000066400000000000000000000271641416703176500203700ustar00rootroot00000000000000 Penlight Documentation

Module pl.sip

Simple Input Patterns (SIP).

SIP patterns start with '$', then a one-letter type, and then an optional variable in curly braces.

sip.match('$v=$q','name="dolly"',res)
==> res=={'name','dolly'}
sip.match('($q{first},$q{second})','("john","smith")',res)
==> res=={second='smith',first='john'}

Type names:

v     identifier
i     integer
f     floating-point
q     quoted string
([{<  match up to closing bracket

See the Guide

Functions

create_pattern (spec, options) convert a SIP pattern into the equivalent Lua string pattern.
compile (spec, options) convert a SIP pattern into a matching function.
match (spec, line, res, options) match a SIP pattern against a string.
match_at_start (spec, line, res) match a SIP pattern against the start of a string.
fields (spec, f) given a pattern and a file object, return an iterator over the results
pattern (spec, fun) register a match which will be used in the read function.
read (f, matches) enter a loop which applies all registered matches to the input file.


Functions

create_pattern (spec, options)
convert a SIP pattern into the equivalent Lua string pattern.

Parameters:

  • spec a SIP pattern
  • options a table; only the at_start field is currently meaningful and ensures that the pattern is anchored at the start of the string.

Returns:

    a Lua string pattern.
compile (spec, options)
convert a SIP pattern into a matching function. The returned function takes two arguments, the line and an empty table. If the line matched the pattern, then this function returns true and the table is filled with field-value pairs.

Parameters:

  • spec a SIP pattern
  • options optional table; {at_start=true} ensures that the pattern is anchored at the start of the string.

Returns:

    a function if successful, or nil,error
match (spec, line, res, options)
match a SIP pattern against a string.

Parameters:

  • spec a SIP pattern
  • line a string
  • res a table to receive values
  • options (optional) option table

Returns:

    true or false
match_at_start (spec, line, res)
match a SIP pattern against the start of a string.

Parameters:

  • spec a SIP pattern
  • line a string
  • res a table to receive values

Returns:

    true or false
fields (spec, f)
given a pattern and a file object, return an iterator over the results

Parameters:

  • spec a SIP pattern
  • f a file-like object.
pattern (spec, fun)
register a match which will be used in the read function.

Parameters:

  • spec string a SIP pattern
  • fun func a function to be called with the results of the match

See also:

read (f, matches)
enter a loop which applies all registered matches to the input file.

Parameters:

  • f a file-like object
  • matches array optional list of {spec,fun} pairs, as for pattern above.
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.strict.html000066400000000000000000000233041416703176500210750ustar00rootroot00000000000000 Penlight Documentation

Module pl.strict

Checks uses of undeclared global variables.

All global variables must be 'declared' through a regular assignment (even assigning nil will do) in a main chunk before being used anywhere or assigned to inside a function. Existing metatables __newindex and __index metamethods are respected.

You can set any table to have strict behaviour using strict.module. Creating a new module with strict.closed_module makes the module immune to monkey-patching, if you don't wish to encourage monkey business.

If the global PENLIGHT_NO_GLOBAL_STRICT is defined, then this module won't make the global environment strict - if you just want to explicitly set table strictness.

Functions

module ([name[, mod[, predeclared]]]) make an existing table strict.
make_all_strict (T) make all tables in a table strict.
closed_module (mod, name) make a new module table which is closed to further changes.


Functions

module ([name[, mod[, predeclared]]])
make an existing table strict.

Parameters:

  • name string name of table (optional)
  • mod tab the table to protect - if nil then we'll return a new table (optional)
  • predeclared tab
    • table of variables that are to be considered predeclared.
    (optional)

Returns:

    the given table, or a new table

Usage:

    local M = { hello = "world" }
    strict.module ("Awesome_Module", M, {
      Lua = true,  -- defines allowed keys
    })
    
    assert(M.hello == "world")
    assert(M.Lua == nil)       -- access allowed, but has no value yet
    M.Lua = "Rocks"
    assert(M.Lua == "Rocks")
    M.not_allowed = "bad boy"  -- throws an error
make_all_strict (T)
make all tables in a table strict. So strict.make_all_strict(_G) prevents monkey-patching of any global table

Parameters:

  • T tab the table containing the tables to protect. Table T itself will NOT be protected.
closed_module (mod, name)
make a new module table which is closed to further changes.

Parameters:

  • mod tab module table
  • name string module name
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.stringio.html000066400000000000000000000201571416703176500214260ustar00rootroot00000000000000 Penlight Documentation

Module pl.stringio

Reading and writing strings using file-like objects.


f = stringio.open(text)
l1 = f:read()  -- read first line
n,m = f:read ('*n','*n') -- read two numbers
for line in f:lines() do print(line) end -- iterate over all lines
f = stringio.create()
f:write('hello')
f:write('dolly')
assert(f:value(),'hellodolly')

See the Guide.

Functions

create () create a file-like object which can be used to construct a string.
open (s) create a file-like object for reading from a given string.


Functions

create ()
create a file-like object which can be used to construct a string. The resulting object has an extra value() method for retrieving the string value. Implements file:write, file:seek, file:lines, plus an extra writef method which works like utils.printf.

Usage:

    f = create(); f:write('hello, dolly\n'); print(f:value())
open (s)
create a file-like object for reading from a given string. Implements file:read.

Parameters:

Usage:

    fs = open '20 10'; x,y = f:read ('*n','*n'); assert(x == 20 and y == 10)
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.stringx.html000066400000000000000000001371501416703176500212700ustar00rootroot00000000000000 Penlight Documentation

Module pl.stringx

Python-style extended string library.

see 3.6.1 of the Python reference. If you want to make these available as string methods, then say stringx.import() to bring them into the standard string table.

See the Guide

Dependencies: pl.utils, pl.types

String Predicates

isalpha (s) does s only contain alphabetic characters?
isdigit (s) does s only contain digits?
isalnum (s) does s only contain alphanumeric characters?
isspace (s) does s only contain whitespace?
islower (s) does s only contain lower case characters?
isupper (s) does s only contain upper case characters?
startswith (s, prefix) does s start with prefix or one of prefixes?
endswith (s, suffix) does s end with suffix or one of suffixes?

Strings and Lists

join (s, seq) concatenate the strings using this string as a delimiter.
splitlines (s[, keep_ends]) Split a string into a list of lines.
split (s[, re[, n]]) split a string into a list of strings using a delimiter.
expandtabs (s, tabsize) replace all tabs in s with tabsize spaces.

Finding and Replacing

lfind (s, sub[, first[, last]]) find index of first instance of sub in s from the left.
rfind (s, sub[, first[, last]]) find index of first instance of sub in s from the right.
replace (s, old, new[, n]) replace up to n instances of old by new in the string s.
count (s, sub[, allow_overlap]) count all instances of substring in string.

Stripping and Justifying

ljust (s, w[, ch=' ']) left-justify s with width w.
rjust (s, w[, ch=' ']) right-justify s with width w.
center (s, w[, ch=' ']) center-justify s with width w.
lstrip (s[, chrs='%s']) trim any characters on the left of s.
rstrip (s[, chrs='%s']) trim any characters on the right of s.
strip (s[, chrs='%s']) trim any characters on both left and right of s.

Partitioning Strings

splitv (s[, re='%s']) split a string using a pattern.
partition (s, ch) partition the string using first occurance of a delimiter
rpartition (s, ch) partition the string p using last occurance of a delimiter
at (s, idx) return the 'character' at the index.

Text handling

indent (s, n[, ch=' ']) indent a multiline string.
dedent (s) dedent a multiline string by removing any initial indent.
wrap (s[, width=70[, breaklong=false]]) format a paragraph into lines so that they fit into a line width.
fill (s[, width=70[, breaklong=false]]) format a paragraph so that it fits into a line width.

Template

Template (tmpl) Creates a new Template class.
Template:substitute (tbl) substitute values into a template, throwing an error.
Template:safe_substitute (tbl) substitute values into a template.
Template:indent_substitute (tbl) substitute values into a template, preserving indentation.

Miscelaneous

lines (s) return an iterator over all lines in a string
title (s) inital word letters uppercase ('title case').
shorten (s, w, tail) Return a shortened version of a string.
quote_string (s) Quote the given string and preserve any control or escape characters, such that reloading the string in Lua returns the same result.
format_operator () Python-style formatting operator.
import () import the stringx functions into the global string (meta)table


String Predicates

isalpha (s)
does s only contain alphabetic characters?

Parameters:

isdigit (s)
does s only contain digits?

Parameters:

isalnum (s)
does s only contain alphanumeric characters?

Parameters:

isspace (s)
does s only contain whitespace? Matches on pattern '%s' so matches space, newline, tabs, etc.

Parameters:

islower (s)
does s only contain lower case characters?

Parameters:

isupper (s)
does s only contain upper case characters?

Parameters:

startswith (s, prefix)
does s start with prefix or one of prefixes?

Parameters:

  • s string a string
  • prefix a string or an array of strings
endswith (s, suffix)
does s end with suffix or one of suffixes?

Parameters:

  • s string a string
  • suffix a string or an array of strings

Strings and Lists

join (s, seq)
concatenate the strings using this string as a delimiter. Note that the arguments are reversed from string.concat.

Parameters:

  • s string the string
  • seq a table of strings or numbers

Usage:

    stringx.join(' ', {1,2,3}) == '1 2 3'
splitlines (s[, keep_ends])
Split a string into a list of lines. "\r", "\n", and "\r\n" are considered line ends. They are not included in the lines unless keepends is passed. Terminal line end does not produce an extra line. Splitting an empty string results in an empty list.

Parameters:

  • s string the string.
  • keep_ends bool include line ends. (optional)

Returns:

    List of lines
split (s[, re[, n]])
split a string into a list of strings using a delimiter.

Parameters:

  • s string the string
  • re string a delimiter (defaults to whitespace) (optional)
  • n int maximum number of results (optional)

Returns:

    List

Usage:

  • #(stringx.split('one two')) == 2
  • stringx.split('one,two,three', ',') == List{'one','two','three'}
  • stringx.split('one,two,three', ',', 2) == List{'one','two,three'}
expandtabs (s, tabsize)
replace all tabs in s with tabsize spaces. If not specified, tabsize defaults to 8. Tab stops will be honored.

Parameters:

  • s string the string
  • tabsize int [opt=8] number of spaces to expand each tab

Returns:

    expanded string

Usage:

  • stringx.expandtabs('\tone,two,three', 4)   == '    one,two,three'
  • stringx.expandtabs('  \tone,two,three', 4) == '    one,two,three'

Finding and Replacing

lfind (s, sub[, first[, last]])
find index of first instance of sub in s from the left.

Parameters:

  • s string the string
  • sub string substring
  • first int first index (optional)
  • last int last index (optional)

Returns:

    start index, or nil if not found
rfind (s, sub[, first[, last]])
find index of first instance of sub in s from the right.

Parameters:

  • s string the string
  • sub string substring
  • first int first index (optional)
  • last int last index (optional)

Returns:

    start index, or nil if not found
replace (s, old, new[, n])
replace up to n instances of old by new in the string s. If n is not present, replace all instances.

Parameters:

  • s string the string
  • old string the target substring
  • new string the substitution
  • n int optional maximum number of substitutions (optional)

Returns:

    result string
count (s, sub[, allow_overlap])
count all instances of substring in string.

Parameters:

  • s string the string
  • sub string substring
  • allow_overlap bool allow matches to overlap (optional)

Usage:

    assert(stringx.count('banana', 'ana') == 1)
    assert(stringx.count('banana', 'ana', true) == 2)

Stripping and Justifying

ljust (s, w[, ch=' '])
left-justify s with width w.

Parameters:

  • s string the string
  • w int width of justification
  • ch string padding character (default ' ')

Usage:

    stringx.ljust('hello', 10, '*') == '*****hello'
rjust (s, w[, ch=' '])
right-justify s with width w.

Parameters:

  • s string the string
  • w int width of justification
  • ch string padding character (default ' ')

Usage:

    stringx.rjust('hello', 10, '*') == 'hello*****'
center (s, w[, ch=' '])
center-justify s with width w.

Parameters:

  • s string the string
  • w int width of justification
  • ch string padding character (default ' ')

Usage:

    stringx.center('hello', 10, '*') == '**hello***'
lstrip (s[, chrs='%s'])
trim any characters on the left of s.

Parameters:

  • s string the string
  • chrs string default any whitespace character, but can be a string of characters to be trimmed (default '%s')
rstrip (s[, chrs='%s'])
trim any characters on the right of s.

Parameters:

  • s string the string
  • chrs string default any whitespace character, but can be a string of characters to be trimmed (default '%s')
strip (s[, chrs='%s'])
trim any characters on both left and right of s.

Parameters:

  • s string the string
  • chrs string default any whitespace character, but can be a string of characters to be trimmed (default '%s')

Usage:

    stringx.strip('  --== Hello ==--  ', "- =")  --> 'Hello'

Partitioning Strings

splitv (s[, re='%s'])
split a string using a pattern. Note that at least one value will be returned!

Parameters:

  • s string the string
  • re string a Lua string pattern (defaults to whitespace) (default '%s')

Returns:

    the parts of the string

See also:

Usage:

    a,b = line:splitv('=')
partition (s, ch)
partition the string using first occurance of a delimiter

Parameters:

  • s string the string
  • ch string delimiter (match as plain string, no patterns)

Returns:

  1. part before ch
  2. ch
  3. part after ch

Usage:

  • {stringx.partition('a,b,c', ','))} == {'a', ',', 'b,c'}
  • {stringx.partition('abc', 'x'))} == {'abc', '', ''}
rpartition (s, ch)
partition the string p using last occurance of a delimiter

Parameters:

  • s string the string
  • ch string delimiter (match as plain string, no patterns)

Returns:

  1. part before ch
  2. ch
  3. part after ch

Usage:

  • {stringx.rpartition('a,b,c', ','))} == {'a,b', ',', 'c'}
  • {stringx.rpartition('abc', 'x'))} == {'', '', 'abc'}
at (s, idx)
return the 'character' at the index.

Parameters:

  • s string the string
  • idx int an index (can be negative)

Returns:

    a substring of length 1 if successful, empty string otherwise.

Text handling

indent (s, n[, ch=' '])
indent a multiline string.

Parameters:

  • s string the (multiline) string
  • n integer the size of the indent
  • ch string the character to use when indenting (default ' ')

Returns:

    indented string
dedent (s)
dedent a multiline string by removing any initial indent. useful when working with [[..]] strings. Empty lines are ignored.

Parameters:

  • s string the (multiline) string

Returns:

    a string with initial indent zero.

Usage:

    local s = dedent [[
             One
    
           Two
    
         Three
    ]]
    assert(s == [[
        One
    
      Two
    
    Three
    ]])
wrap (s[, width=70[, breaklong=false]])
format a paragraph into lines so that they fit into a line width. It will not break long words by default, so lines can be over the length to that extent.

Parameters:

  • s string the string to format
  • width integer the margin width (default 70)
  • breaklong boolean if truthy, words longer than the width given will be forced split. (default false)

Returns:

    a list of lines (List object), use fill to return a string instead of a List.

See also:

fill (s[, width=70[, breaklong=false]])
format a paragraph so that it fits into a line width.

Parameters:

  • s string the string to format
  • width integer the margin width (default 70)
  • breaklong boolean if truthy, words longer than the width given will be forced split. (default false)

Returns:

    a string, use wrap to return a list of lines instead of a string.

See also:

Template

Template (tmpl)
Creates a new Template class. This is a shortcut to Template.new(tmpl).

Parameters:

  • tmpl string the template string

Returns:

    Template
Template:substitute (tbl)
substitute values into a template, throwing an error. This will throw an error if no name is found.

Parameters:

  • tbl table a table of name-value pairs.

Returns:

    string with place holders substituted
Template:safe_substitute (tbl)
substitute values into a template. This version just passes unknown names through.

Parameters:

  • tbl table a table of name-value pairs.

Returns:

    string with place holders substituted
Template:indent_substitute (tbl)
substitute values into a template, preserving indentation.
If the value is a multiline string or a template, it will insert the lines at the correct indentation.
Furthermore, if a template, then that template will be substituted using the same table.

Parameters:

  • tbl table a table of name-value pairs.

Returns:

    string with place holders substituted

Miscelaneous

lines (s)
return an iterator over all lines in a string

Parameters:

Returns:

    an iterator

Usage:

    local line_no = 1
    for line in stringx.lines(some_text) do
      print(line_no, line)
      line_no = line_no + 1
    end
title (s)
inital word letters uppercase ('title case'). Here 'words' mean chunks of non-space characters.

Parameters:

Returns:

    a string with each word's first letter uppercase

Usage:

    stringx.title("hello world") == "Hello World")
shorten (s, w, tail)
Return a shortened version of a string. Fits string within w characters. Removed characters are marked with ellipsis.

Parameters:

  • s string the string
  • w int the maxinum size allowed
  • tail bool true if we want to show the end of the string (head otherwise)

Usage:

  • ('1234567890'):shorten(8) == '12345...'
  • ('1234567890'):shorten(8, true) == '...67890'
  • ('1234567890'):shorten(20) == '1234567890'
quote_string (s)
Quote the given string and preserve any control or escape characters, such that reloading the string in Lua returns the same result.

Parameters:

  • s The string to be quoted.

Returns:

    The quoted string.
format_operator ()
Python-style formatting operator. Calling text.format_operator() overloads the % operator for strings to give Python/Ruby style formated output. This is extended to also do template-like substitution for map-like data.

Note this goes further than the original, and will allow these cases:

  1. a single value
  2. a list of values
  3. a map of var=value pairs
  4. a function, as in gsub

For the second two cases, it uses $-variable substituion.

When called, this function will monkey-patch the global string metatable by adding a __mod method.

See the lua-users wiki

Usage:

    require 'pl.text'.format_operator()
    local out1 = '%s = %5.3f' % {'PI',math.pi}                   --> 'PI = 3.142'
    local out2 = '$name = $value' % {name='dog',value='Pluto'}   --> 'dog = Pluto'
import ()
import the stringx functions into the global string (meta)table
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.tablex.html000066400000000000000000001670041416703176500210520ustar00rootroot00000000000000 Penlight Documentation

Module pl.tablex

Extended operations on Lua tables.

See the Guide

Dependencies: pl.utils, pl.types

Functions

size (t) total number of elements in this table.
index_by (tbl, idx) return a list of all values in a table indexed by another list.
transform (fun, t, ...) apply a function to all values of a table, in-place.
range (start, finish[, step=1]) generate a table of all numbers in a range.
reduce (fun, t, memo) 'reduce' a list using a binary function.
index_map (t) create an index map from a list-like table.
makeset (t) create a set from a list-like table.
union (t1, t2) the union of two map-like tables.
intersection (t1, t2) the intersection of two map-like tables.
count_map (t, cmp) A table where the key/values are the values and value counts of the table.
set (t, val[, i1=1[, i2=#t]]) set an array range to a value.
new (n, val) create a new array of specified size with initial value.
clear (t, istart) clear out the contents of a table.
removevalues (t, i1, i2) remove a range of values from a table.
readonly (t) modifies a table to be read only.

Copying

update (t1, t2) copy a table into another, in-place.
copy (t) make a shallow copy of a table
deepcopy (t) make a deep copy of a table, recursively copying all the keys and fields.
icopy (dest, src[, idest=1[, isrc=1[, nsrc=#src]]]) copy an array into another one, clearing dest after idest+nsrc, if necessary.
move (dest, src[, idest=1[, isrc=1[, nsrc=#src]]]) copy an array into another one.
insertvalues (t[, position], values) insert values into a table.

Comparing

deepcompare (t1, t2[, ignore_mt[, eps]]) compare two values.
compare (t1, t2, cmp) compare two arrays using a predicate.
compare_no_order (t1, t2, cmp) compare two list-like tables using an optional predicate, without regard for element order.

Finding

find (t, val, idx) return the index of a value in a list.
rfind (t, val, idx) return the index of a value in a list, searching from the end.
find_if (t, cmp, arg) return the index (or key) of a value in a table using a comparison function.
search (t, value[, exclude]) find a value in a table by recursive search.

MappingAndFiltering

map (fun, t, ...) apply a function to all values of a table.
imap (fun, t, ...) apply a function to all values of a list.
map_named_method (name, t, ...) apply a named method to values from a table.
map2 (fun, t1, t2, ...) apply a function to values from two tables.
imap2 (fun, t1, t2, ...) apply a function to values from two arrays.
mapn (fun, ..., fun) Apply a function to a number of tables.
pairmap (fun, t, ...) call the function with the key and value pairs from a table.
filter (t, pred, arg) filter an array's values using a predicate function

Iterating

foreach (t, fun, ...) apply a function to all elements of a table.
foreachi (t, fun, ...) apply a function to all elements of a list-like table in order.
sort (t, f) return an iterator to a table sorted by its keys
sortv (t, f) return an iterator to a table sorted by its values

Extraction

keys (t) return all the keys of a table in arbitrary order.
values (t) return all the values of the table in arbitrary order
sub (t, first, last) Extract a range from a table, like 'string.sub'.

Merging

merge (t1, t2, dup) combine two tables, either as union or intersection.
difference (s1, s2, symm) a new table which is the difference of two tables.
zip (...) return a table where each element is a table of the ith values of an arbitrary number of tables.


Functions

size (t)
total number of elements in this table. Note that this is distinct from #t, which is the number of values in the array part; this value will always be greater or equal. The difference gives the size of the hash part, for practical purposes. Works for any object with a __pairs metamethod.

Parameters:

  • t tab a table

Returns:

    the size
index_by (tbl, idx)
return a list of all values in a table indexed by another list.

Parameters:

  • tbl tab a table
  • idx array an index table (a list of keys)

Returns:

    a list-like table

Usage:

  • index_by({10,20,30,40},{2,4}) == {20,40}
  • index_by({one=1,two=2,three=3},{'one','three'}) == {1,3}
transform (fun, t, ...)
apply a function to all values of a table, in-place. Any extra arguments are passed to the function.

Parameters:

  • fun func A function that takes at least one argument
  • t tab a table
  • ... extra arguments passed to fun

See also:

range (start, finish[, step=1])
generate a table of all numbers in a range. This is consistent with a numerical for loop.

Parameters:

  • start int number
  • finish int number
  • step int make this negative for start < finish (default 1)
reduce (fun, t, memo)
'reduce' a list using a binary function.

Parameters:

  • fun func a function of two arguments
  • t array a list-like table
  • memo array optional initial memo value. Defaults to first value in table.

Returns:

    the result of the function

Usage:

    reduce('+',{1,2,3,4}) == 10
index_map (t)
create an index map from a list-like table. The original values become keys, and the associated values are the indices into the original list.

Parameters:

  • t array a list-like table

Returns:

    a map-like table
makeset (t)
create a set from a list-like table. A set is a table where the original values become keys, and the associated values are all true.

Parameters:

  • t array a list-like table

Returns:

    a set (a map-like table)
union (t1, t2)
the union of two map-like tables. If there are duplicate keys, the second table wins.

Parameters:

  • t1 tab a table
  • t2 tab a table

Returns:

    tab

See also:

intersection (t1, t2)
the intersection of two map-like tables.

Parameters:

  • t1 tab a table
  • t2 tab a table

Returns:

    tab

See also:

count_map (t, cmp)
A table where the key/values are the values and value counts of the table.

Parameters:

  • t array a list-like table
  • cmp func a function that defines equality (otherwise uses ==)

Returns:

    a map-like table

See also:

set (t, val[, i1=1[, i2=#t]])
set an array range to a value. If it's a function we use the result of applying it to the indices.

Parameters:

  • t array a list-like table
  • val a value
  • i1 int start range (default 1)
  • i2 int end range (default #t)
new (n, val)
create a new array of specified size with initial value.

Parameters:

  • n int size
  • val initial value (can be nil, but don't expect # to work!)

Returns:

    the table
clear (t, istart)
clear out the contents of a table.

Parameters:

  • t array a list
  • istart optional start position
removevalues (t, i1, i2)
remove a range of values from a table. End of range may be negative.

Parameters:

  • t array a list-like table
  • i1 int start index
  • i2 int end index

Returns:

    the table
readonly (t)
modifies a table to be read only. This only offers weak protection. Tables can still be modified with table.insert and rawset.

NOTE: for Lua 5.1 length, pairs and ipairs will not work, since the equivalent metamethods are only available in Lua 5.2 and newer.

Parameters:

  • t tab the table

Returns:

    the table read only (a proxy).

Copying

update (t1, t2)
copy a table into another, in-place.

Parameters:

  • t1 tab destination table
  • t2 tab source (actually any iterable object)

Returns:

    first table
copy (t)
make a shallow copy of a table

Parameters:

  • t tab an iterable source

Returns:

    new table
deepcopy (t)
make a deep copy of a table, recursively copying all the keys and fields. This supports cycles in tables; cycles will be reproduced in the copy. This will also set the copied table's metatable to that of the original.

Parameters:

  • t tab A table

Returns:

    new table
icopy (dest, src[, idest=1[, isrc=1[, nsrc=#src]]])
copy an array into another one, clearing dest after idest+nsrc, if necessary.

Parameters:

  • dest array a list-like table
  • src array a list-like table
  • idest int where to start copying values into destination (default 1)
  • isrc int where to start copying values from source (default 1)
  • nsrc int number of elements to copy from source (default #src)
move (dest, src[, idest=1[, isrc=1[, nsrc=#src]]])
copy an array into another one.

Parameters:

  • dest array a list-like table
  • src array a list-like table
  • idest int where to start copying values into destination (default 1)
  • isrc int where to start copying values from source (default 1)
  • nsrc int number of elements to copy from source (default #src)
insertvalues (t[, position], values)
insert values into a table. similar to table.insert but inserts values from given table values, not the object itself, into table t at position pos.

Parameters:

  • t array the list
  • position int (default is at end) (optional)
  • values array

Comparing

deepcompare (t1, t2[, ignore_mt[, eps]])
compare two values. if they are tables, then compare their keys and fields recursively.

Parameters:

  • t1 A value
  • t2 A value
  • ignore_mt bool if true, ignore __eq metamethod (default false) (optional)
  • eps number if defined, then used for any number comparisons (optional)

Returns:

    true or false
compare (t1, t2, cmp)
compare two arrays using a predicate.

Parameters:

  • t1 array an array
  • t2 array an array
  • cmp func A comparison function; bool = cmp(t1_value, t2_value)

Returns:

    true or false

Usage:

    assert(tablex.compare({ 1, 2, 3 }, { 1, 2, 3 }, "=="))
    
    assert(tablex.compare(
       {1,2,3, hello = "world"},  -- fields are not compared!
       {1,2,3}, function(v1, v2) return v1 == v2 end)
compare_no_order (t1, t2, cmp)
compare two list-like tables using an optional predicate, without regard for element order.

Parameters:

  • t1 array a list-like table
  • t2 array a list-like table
  • cmp A comparison function (may be nil)

Finding

find (t, val, idx)
return the index of a value in a list. Like string.find, there is an optional index to start searching, which can be negative.

Parameters:

  • t array A list-like table
  • val A value
  • idx int index to start; -1 means last element,etc (default 1)

Returns:

    index of value or nil if not found

Usage:

  • find({10,20,30},20) == 2
  • find({'a','b','a','c'},'a',2) == 3
rfind (t, val, idx)
return the index of a value in a list, searching from the end. Like string.find, there is an optional index to start searching, which can be negative.

Parameters:

  • t array A list-like table
  • val A value
  • idx index to start; -1 means last element,etc (default #t)

Returns:

    index of value or nil if not found

Usage:

    rfind({10,10,10},10) == 3
find_if (t, cmp, arg)
return the index (or key) of a value in a table using a comparison function.

NOTE: the 2nd return value of this function, the value returned by the comparison function, has a limitation that it cannot be false. Because if it is, then it indicates the comparison failed, and the function will continue the search. See examples.

Parameters:

  • t tab A table
  • cmp func A comparison function
  • arg an optional second argument to the function

Returns:

  1. index of value, or nil if not found
  2. value returned by comparison function (cannot be false!)

Usage:

    -- using an operator
    local lst = { "Rudolph", true, false, 15 }
    local idx, cmp_result = tablex.rfind(lst, "==", "Rudolph")
    assert(idx == 1)
    assert(cmp_result == true)
    
    local idx, cmp_result = tablex.rfind(lst, "==", false)
    assert(idx == 3)
    assert(cmp_result == true)       -- looking up 'false' works!
    
    -- using a function returning the value looked up
    local cmp = function(v1, v2) return v1 == v2 and v2 end
    local idx, cmp_result = tablex.rfind(lst, cmp, "Rudolph")
    assert(idx == 1)
    assert(cmp_result == "Rudolph")  -- the value is returned
    
    -- NOTE: this fails, since 'false' cannot be returned!
    local idx, cmp_result = tablex.rfind(lst, cmp, false)
    assert(idx == nil)               -- looking up 'false' failed!
    assert(cmp_result == nil)
search (t, value[, exclude])
find a value in a table by recursive search.

Parameters:

  • t tab the table
  • value the value
  • exclude array any tables to avoid searching (optional)

Returns:

    a fieldspec, e.g. 'a.b' or 'math.sin'

Usage:

    search(_G,math.sin,{package.path}) == 'math.sin'

MappingAndFiltering

map (fun, t, ...)
apply a function to all values of a table. This returns a table of the results. Any extra arguments are passed to the function.

Parameters:

  • fun func A function that takes at least one argument
  • t tab A table
  • ... optional arguments

Usage:

    map(function(v) return v*v end, {10,20,30,fred=2}) is {100,400,900,fred=4}
imap (fun, t, ...)
apply a function to all values of a list. This returns a table of the results. Any extra arguments are passed to the function.

Parameters:

  • fun func A function that takes at least one argument
  • t array a table (applies to array part)
  • ... optional arguments

Returns:

    a list-like table

Usage:

    imap(function(v) return v*v end, {10,20,30,fred=2}) is {100,400,900}
map_named_method (name, t, ...)
apply a named method to values from a table.

Parameters:

  • name string the method name
  • t array a list-like table
  • ... any extra arguments to the method

Returns:

    a List with the results of the method (1st result only)

Usage:

    local Car = {}
    Car.__index = Car
    function Car.new(car)
      return setmetatable(car or {}, Car)
    end
    Car.speed = 0
    function Car:faster(increase)
      self.speed = self.speed + increase
      return self.speed
    end
    
    local ferrari = Car.new{ name = "Ferrari" }
    local lamborghini = Car.new{ name = "Lamborghini", speed = 50 }
    local cars = { ferrari, lamborghini }
    
    assert(ferrari.speed == 0)
    assert(lamborghini.speed == 50)
    tablex.map_named_method("faster", cars, 10)
    assert(ferrari.speed == 10)
    assert(lamborghini.speed == 60)
map2 (fun, t1, t2, ...)
apply a function to values from two tables.

Parameters:

  • fun func a function of at least two arguments
  • t1 tab a table
  • t2 tab a table
  • ... extra arguments

Returns:

    a table

Usage:

    map2('+',{1,2,3,m=4},{10,20,30,m=40}) is {11,22,23,m=44}
imap2 (fun, t1, t2, ...)
apply a function to values from two arrays. The result will be the length of the shortest array.

Parameters:

  • fun func a function of at least two arguments
  • t1 array a list-like table
  • t2 array a list-like table
  • ... extra arguments

Usage:

    imap2('+',{1,2,3,m=4},{10,20,30,m=40}) is {11,22,23}
mapn (fun, ..., fun)
Apply a function to a number of tables. A more general version of map The result is a table containing the result of applying that function to the ith value of each table. Length of output list is the minimum length of all the lists

Parameters:

  • fun A function that takes as many arguments as there are tables
  • ... tab n tables
  • fun A function that takes as many arguments as there are tables

Usage:

  • mapn(function(x,y,z) return x+y+z end, {1,2,3},{10,20,30},{100,200,300}) is {111,222,333}
  • mapn(math.max, {1,20,300},{10,2,3},{100,200,100}) is    {100,200,300}
pairmap (fun, t, ...)
call the function with the key and value pairs from a table. The function can return a value and a key (note the order!). If both are not nil, then this pair is inserted into the result: if the key already exists, we convert the value for that key into a table and append into it. If only value is not nil, then it is appended to the result.

Parameters:

  • fun func A function which will be passed each key and value as arguments, plus any extra arguments to pairmap.
  • t tab A table
  • ... optional arguments

Usage:

  • pairmap(function(k,v) return v end,{fred=10,bonzo=20}) is {10,20} _or_ {20,10}
  • pairmap(function(k,v) return {k,v},k end,{one=1,two=2}) is {one={'one',1},two={'two',2}}
filter (t, pred, arg)
filter an array's values using a predicate function

Parameters:

  • t array a list-like table
  • pred func a boolean function
  • arg optional argument to be passed as second argument of the predicate

Iterating

foreach (t, fun, ...)
apply a function to all elements of a table. The arguments to the function will be the value, the key and finally any extra arguments passed to this function. Note that the Lua 5.0 function table.foreach passed the key first.

Parameters:

  • t tab a table
  • fun func a function on the elements; function(value, key, ...)
  • ... extra arguments passed to fun

See also:

foreachi (t, fun, ...)
apply a function to all elements of a list-like table in order. The arguments to the function will be the value, the index and finally any extra arguments passed to this function

Parameters:

  • t array a table
  • fun func a function with at least one argument
  • ... optional arguments
sort (t, f)
return an iterator to a table sorted by its keys

Parameters:

  • t tab the table
  • f func an optional comparison function (f(x,y) is true if x < y)

Returns:

    an iterator to traverse elements sorted by the keys

Usage:

    for k,v in tablex.sort(t) do print(k,v) end
sortv (t, f)
return an iterator to a table sorted by its values

Parameters:

  • t tab the table
  • f func an optional comparison function (f(x,y) is true if x < y)

Returns:

    an iterator to traverse elements sorted by the values

Usage:

    for k,v in tablex.sortv(t) do print(k,v) end

Extraction

keys (t)
return all the keys of a table in arbitrary order.

Parameters:

  • t tab A table
values (t)
return all the values of the table in arbitrary order

Parameters:

  • t tab A table
sub (t, first, last)
Extract a range from a table, like 'string.sub'. If first or last are negative then they are relative to the end of the list eg. sub(t,-2) gives last 2 entries in a list, and sub(t,-4,-2) gives from -4th to -2nd

Parameters:

  • t array a list-like table
  • first int An index
  • last int An index

Returns:

    a new List

Merging

merge (t1, t2, dup)
combine two tables, either as union or intersection. Corresponds to set operations for sets () but more general. Not particularly useful for list-like tables.

Parameters:

  • t1 tab a table
  • t2 tab a table
  • dup bool true for a union, false for an intersection.

See also:

Usage:

  • merge({alice=23,fred=34},{bob=25,fred=34}) is {fred=34}
  • merge({alice=23,fred=34},{bob=25,fred=34},true) is {bob=25,fred=34,alice=23}
difference (s1, s2, symm)
a new table which is the difference of two tables. With sets (where the values are all true) this is set difference and symmetric difference depending on the third parameter.

Parameters:

  • s1 tab a map-like table or set
  • s2 tab a map-like table or set
  • symm bool symmetric difference (default false)

Returns:

    a map-like table or set
zip (...)
return a table where each element is a table of the ith values of an arbitrary number of tables. It is equivalent to a matrix transpose.

Parameters:

  • ... array arrays to be zipped

Usage:

    zip({10,20,30},{100,200,300}) is {{10,100},{20,200},{30,300}}
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.template.html000066400000000000000000000303361416703176500214030ustar00rootroot00000000000000 Penlight Documentation

Module pl.template

A template preprocessor.

Originally by Ricki Lake

There are two rules:

  • lines starting with # are Lua
  • otherwise, $(expr) is the result of evaluating expr

Example:

#  for i = 1,3 do
   $(i) Hello, Word!
#  end
===>
1 Hello, Word!
2 Hello, Word!
3 Hello, Word!

Other escape characters can be used, when the defaults conflict with the output language.

> for _,n in pairs{'one','two','three'} do
static int l_${n} (luaState *state);
> end

See the Guide.

Dependencies: pl.utils

Functions

substitute (str[, env]) expand the template using the specified environment.
ct:render ([env[, parent[, db]]]) executes the previously compiled template and renders it.
compile (str[, opts]) compiles the template.


Functions

substitute (str[, env])

expand the template using the specified environment. This function will compile and render the template. For more performant recurring usage use the two step approach by using compile and ct:render. There are six special fields in the environment table env

  • _parent: continue looking up in this table (e.g. _parent=_G).
  • _brackets: bracket pair that wraps inline Lua expressions, default is '()'.
  • _escape: character marking Lua lines, default is '#'
  • _inline_escape: character marking inline Lua expression, default is '$'.
  • _chunk_name: chunk name for loaded templates, used if there is an error in Lua code. Default is 'TMP'.
  • _debug: if truthy, the generated code will be printed upon a render error

Parameters:

  • str string the template string
  • env tab the environment (optional)

Returns:

    rendered template + nil + source_code, or nil + error + source_code. The last return value (source_code) is only returned if the debug option is used.
ct:render ([env[, parent[, db]]])
executes the previously compiled template and renders it.

Parameters:

  • env tab the environment. (optional)
  • parent tab continue looking up in this table (e.g. parent=_G). (optional)
  • db bool if thruthy, it will print the code upon a render error (provided the template was compiled with the debug option). (optional)

Returns:

    rendered template + nil + source_code, or nil + error + source_code. The last return value (source_code) is only returned if the template was compiled with the debug option.

Usage:

    local ct, err = template.compile(my_template)
    local rendered , err = ct:render(my_env, parent)
compile (str[, opts])

compiles the template. Returns an object that can repeatedly be rendered without parsing/compiling the template again. The options passed in the opts table support the following options:

  • chunk_name: chunk name for loaded templates, used if there is an error in Lua code. Default is 'TMP'.
  • escape: character marking Lua lines, default is '#'
  • inline_escape: character marking inline Lua expression, default is '$'.
  • inline_brackets: bracket pair that wraps inline Lua expressions, default is '()'.
  • newline: string to replace newline characters, default is nil (not replacing newlines).
  • debug: if truthy, the generated source code will be retained within the compiled template object, default is nil.

Parameters:

  • str string the template string
  • opts tab the compilation options to use (optional)

Returns:

    template object, or nil + error + source_code

Usage:

    local ct, err = template.compile(my_template)
    local rendered , err = ct:render(my_env, parent)
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.test.html000066400000000000000000000324721416703176500205520ustar00rootroot00000000000000 Penlight Documentation

Module pl.test

Useful test utilities.

test.asserteq({1,2},{1,2}) -- can compare tables
test.asserteq(1.2,1.19,0.02) -- compare FP numbers within precision
T = test.tuple -- used for comparing multiple results
test.asserteq(T(string.find(" me","me")),T(2,3))

Dependencies: pl.utils, pl.tablex, pl.pretty, pl.path, debug

Functions

error_handler (file, line, got_text, needed_text, msg) error handling for test results.
complain (x, y, msg, where) general test complain message.
asserteq (x, y, eps, where) like assert, except takes two arguments that must be equal and can be tables.
assertmatch (s1, s2, where) assert that the first string matches the second.
assertraise (fn, e, where) assert that the function raises a particular error.
asserteq2 (x1, x2, y1, y2, where) a version of asserteq that takes two pairs of values.
tuple (...) encode an arbitrary argument list as a tuple.
timer (msg, n, fun, ...) Time a function.


Functions

error_handler (file, line, got_text, needed_text, msg)
error handling for test results. By default, this writes to stderr and exits the program. Re-define this function to raise an error and/or redirect output

Parameters:

  • file
  • line
  • got_text
  • needed_text
  • msg
complain (x, y, msg, where)
general test complain message. Useful for composing new test functions (see tests/tablex.lua for an example)

Parameters:

  • x a value
  • y value to compare first value against
  • msg message
  • where extra level offset for errors
asserteq (x, y, eps, where)
like assert, except takes two arguments that must be equal and can be tables. If they are plain tables, it will use tablex.deepcompare.

Parameters:

  • x any value
  • y a value equal to x
  • eps an optional tolerance for numerical comparisons
  • where extra level offset
assertmatch (s1, s2, where)
assert that the first string matches the second.

Parameters:

  • s1 a string
  • s2 a string
  • where extra level offset
assertraise (fn, e, where)
assert that the function raises a particular error.

Parameters:

  • fn a function or a table of the form {function,arg1,...}
  • e a string to match the error against
  • where extra level offset
asserteq2 (x1, x2, y1, y2, where)
a version of asserteq that takes two pairs of values. x1==y1 and x2==y2 must be true. Useful for functions that naturally return two values.

Parameters:

  • x1 any value
  • x2 any value
  • y1 any value
  • y2 any value
  • where extra level offset
tuple (...)
encode an arbitrary argument list as a tuple. This can be used to compare to other argument lists, which is very useful for testing functions which return a number of values. Unlike regular array-like tables ('sequences') they may contain nils. Tuples understand equality and know how to print themselves out. The # operator is defined to be the size, irrespecive of any nils, and there is an unpack method.

Parameters:

  • ...

Usage:

    asserteq(tuple( ('ab'):find 'a'), tuple(1,1))
timer (msg, n, fun, ...)
Time a function. Call the function a given number of times, and report the number of seconds taken, together with a message. Any extra arguments will be passed to the function.

Parameters:

  • msg string a descriptive message
  • n int number of times to call the function
  • fun func the function
  • ... optional arguments to fun
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.text.html000066400000000000000000000135711416703176500205560ustar00rootroot00000000000000 Penlight Documentation

Module pl.text

Text processing utilities.

This provides a Template class (modeled after the same from the Python libraries, see string.Template). It also provides similar functions to those found in the textwrap module.

IMPORTANT: this module has been deprecated and will be removed in a future version (2.0). The contents of this module have moved to the pl.stringx module.

See the Guide.

Dependencies: pl.stringx, pl.utils



generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.types.html000066400000000000000000000313241416703176500207320ustar00rootroot00000000000000 Penlight Documentation

Module pl.types

Dealing with Detailed Type Information

Functions

is_callable (obj) is the object either a function or a callable object?.
is_type (obj, tp) is the object of the specified type?.
type (obj) a string representation of a type.
is_integer (x) is this number an integer?
is_empty (o, ignore_spaces) Check if the object is "empty".
is_indexable (val) is an object 'array-like'?
is_iterable (val) can an object be iterated over with pairs?
is_writeable (val) can an object accept new key/pair values?
to_bool (o[, true_strs[, check_objs]]) Convert to a boolean value.


Functions

is_callable (obj)
is the object either a function or a callable object?.

Parameters:

  • obj Object to check.
is_type (obj, tp)
is the object of the specified type?. If the type is a string, then use type, otherwise compare with metatable.

NOTE: this function is imported from utils.is_type.

Parameters:

  • obj An object to check
  • tp The expected type

See also:

type (obj)
a string representation of a type. For tables and userdata with metatables, we assume that the metatable has a _name field. If the field is not present it will return 'unknown table' or 'unknown userdata'. Lua file objects return the type 'file'.

Parameters:

  • obj an object

Returns:

    a string like 'number', 'table', 'file' or 'List'
is_integer (x)
is this number an integer?

Parameters:

  • x a number

Returns:

    boolean

Raises:

error if x is not a number
is_empty (o, ignore_spaces)

Check if the object is "empty". An object is considered empty if it is:

  • nil
  • a table without any items (key-value pairs or indexes)
  • a string with no content ("")
  • not a nil/table/string

Parameters:

  • o The object to check if it is empty.
  • ignore_spaces If the object is a string and this is true the string is considered empty if it only contains spaces.

Returns:

    true if the object is empty, otherwise a falsy value.
is_indexable (val)
is an object 'array-like'? An object is array like if:

  • it is a table, or
  • it has a metatable with __len and __index methods

NOTE: since __len is 5.2+, on 5.1 is usually returns false for userdata

Parameters:

  • val any value.

Returns:

    true if the object is array-like, otherwise a falsy value.
is_iterable (val)
can an object be iterated over with pairs? An object is iterable if:

  • it is a table, or
  • it has a metatable with a __pairs meta method

NOTE: since __pairs is 5.2+, on 5.1 is usually returns false for userdata

Parameters:

  • val any value.

Returns:

    true if the object is iterable, otherwise a falsy value.
is_writeable (val)

can an object accept new key/pair values? An object is iterable if:

  • it is a table, or
  • it has a metatable with a __newindex meta method

Parameters:

  • val any value.

Returns:

    true if the object is writeable, otherwise a falsy value.
to_bool (o[, true_strs[, check_objs]])

Convert to a boolean value. True values are:

  • boolean: true.
  • string: 'yes', 'y', 'true', 't', '1' or additional strings specified by true_strs.
  • number: Any non-zero value.
  • table: Is not empty and check_objs is true.
  • everything else: Is not nil and check_objs is true.

Parameters:

  • o The object to evaluate.
  • true_strs optional Additional strings that when matched should evaluate to true. Comparison is case insensitive. This should be a List of strings. E.g. "ja" to support German. (optional)
  • check_objs True if objects should be evaluated. (optional)

Returns:

    true if the input evaluates to true, otherwise false.
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.url.html000066400000000000000000000162521416703176500203730ustar00rootroot00000000000000 Penlight Documentation

Module pl.url

Python-style URL quoting library.

Functions

quote (s, quote_plus) Quote the url, replacing special characters using the '%xx' escape.
unquote (s) Unquote the url, replacing '%xx' escapes and plus signs.


Functions

quote (s, quote_plus)
Quote the url, replacing special characters using the '%xx' escape.

Parameters:

  • s string the string
  • quote_plus bool Also escape slashes and replace spaces by plus signs.

Returns:

    The quoted string, or if s wasn't a string, just plain unaltered s.
unquote (s)
Unquote the url, replacing '%xx' escapes and plus signs.

Parameters:

Returns:

    The unquoted string, or if s wasn't a string, just plain unaltered s.
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.utils.html000066400000000000000000001307121416703176500207270ustar00rootroot00000000000000 Penlight Documentation

Module pl.utils

Generally useful routines.

See the Guide.

Dependencies: pl.compat, all exported fields and functions from pl.compat are also available in this module.

Functions

pack (...) pack an argument list into a table.
unpack (t[, i[, j]]) unpack a table and return its contents.
printf (fmt, ...) print an arbitrary number of arguments using a format.
fprintf (f, fmt, ...) write an arbitrary number of arguments to a file using a format.
import (t, T) take a table and 'inject' it into the local namespace.
choose (cond, value1, value2) return either of two values, depending on a condition.
array_tostring (t[, temp[, tostr]]) convert an array of values to strings.
is_type (obj, tp) is the object of the specified type?
npairs (t[, i_start=1[, i_end=t.n or #t[, step=1]]]) an iterator with indices, similar to ipairs, but with a range.

Tables

patterns Some standard patterns
stdmt Standard meta-tables as used by other Penlight modules

Error handling

assert_arg (n, val, tp, verify, msg, lev) assert that the given argument is in fact of the correct type.
enum (...) creates an Enum table.
function_arg (idx, f, msg) process a function argument.
assert_string (n, val) assert the common case that the argument is a string.
on_error (mode) control the error strategy used by Penlight.
raise (err) used by Penlight functions to return errors.

File handling

readfile (filename, is_bin) return the contents of a file as a string
writefile (filename, str, is_bin) write a string to a file
readlines (filename) return the contents of a file as a list of lines

OS functions

executeex (cmd, bin) execute a shell command and return the output.
quote_arg (argument) Quote and escape an argument of a command.
quit ([code], msg, ...) error out of this program gracefully.

String functions

escape (s) escape any Lua 'magic' characters in a string
split (s, re, plain, n) split a string into a list of strings separated by a delimiter.
splitv (s, re, plain, n) split a string into a number of return values.

Functional

memoize (func) 'memoize' a function (cache returned value for next call).
add_function_factory (mt, fun) associate a function factory with a type.
string_lambda (lf) an anonymous function as a string.
bind1 (fn, p) bind the first argument of the function to a value.
bind2 (fn, p) bind the second argument of the function to a value.

Deprecation

set_deprecation_func (func) Sets a deprecation warning function.
raise_deprecation (opts) raises a deprecation warning.


Functions

pack (...)
pack an argument list into a table.

Parameters:

  • ... any arguments

Returns:

    a table with field n set to the length

See also:

unpack (t[, i[, j]])
unpack a table and return its contents.

NOTE: this implementation differs from the Lua implementation in the way that this one DOES honor the n field in the table t, such that it is 'nil-safe'.

Parameters:

  • t table to unpack
  • i index from which to start unpacking, defaults to 1 (optional)
  • j index of the last element to unpack, defaults to t.n or else #t (optional)

Returns:

    multiple return values from the table

See also:

Usage:

    local t = table.pack(nil, nil, nil, 4)
    local a, b, c, d = table.unpack(t)   -- this unpack is NOT nil-safe, so d == nil
    
    local a, b, c, d = utils.unpack(t)   -- this is nil-safe, so d == 4
printf (fmt, ...)
print an arbitrary number of arguments using a format. Output will be sent to stdout.

Parameters:

  • fmt The format (see string.format)
  • ... Extra arguments for format
fprintf (f, fmt, ...)
write an arbitrary number of arguments to a file using a format.

Parameters:

  • f File handle to write to.
  • fmt The format (see string.format).
  • ... Extra arguments for format
import (t, T)
take a table and 'inject' it into the local namespace.

Parameters:

  • t The table (table), or module name (string), defaults to this utils module table
  • T An optional destination table (defaults to callers environment)
choose (cond, value1, value2)
return either of two values, depending on a condition.

Parameters:

  • cond A condition
  • value1 Value returned if cond is truthy
  • value2 Value returned if cond is falsy
array_tostring (t[, temp[, tostr]])
convert an array of values to strings.

Parameters:

  • t a list-like table
  • temp (table) buffer to use, otherwise allocate (optional)
  • tostr custom tostring function, called with (value,index). Defaults to tostring. (optional)

Returns:

    the converted buffer
is_type (obj, tp)
is the object of the specified type? If the type is a string, then use type, otherwise compare with metatable

Parameters:

  • obj An object to check
  • tp String of what type it should be

Returns:

    boolean

Usage:

    utils.is_type("hello world", "string")   --> true
    -- or check metatable
    local my_mt = {}
    local my_obj = setmetatable(my_obj, my_mt)
    utils.is_type(my_obj, my_mt)  --> true
npairs (t[, i_start=1[, i_end=t.n or #t[, step=1]]])
an iterator with indices, similar to ipairs, but with a range. This is a nil-safe index based iterator that will return nil when there is a hole in a list. To be safe ensure that table t.n contains the length.

Parameters:

  • t table the table to iterate over
  • i_start integer start index (default 1)
  • i_end integer end index (default t.n or #t)
  • step integer step size (default 1)

Returns:

  1. integer index
  2. any value at index (which can be nil!)

See also:

Usage:

    local t = utils.pack(nil, 123, nil)  -- adds an n field when packing
    
    for i, v in utils.npairs(t, 2) do  -- start at index 2
      t[i] = tostring(t[i])
    end
    
    -- t = { n = 3, [2] = "123", [3] = "nil" }

Tables

patterns
Some standard patterns

Fields:

  • FLOAT floating point number
  • INTEGER integer number
  • IDEN identifier
  • FILE file
stdmt
Standard meta-tables as used by other Penlight modules

Fields:

  • List the List metatable
  • Map the Map metatable
  • Set the Set metatable
  • MultiMap the MultiMap metatable

Error handling

assert_arg (n, val, tp, verify, msg, lev)
assert that the given argument is in fact of the correct type.

Parameters:

  • n argument index
  • val the value
  • tp the type
  • verify an optional verification function
  • msg an optional custom message
  • lev optional stack position for trace, default 2

Returns:

    the validated value

Raises:

if val is not the correct type

Usage:

    local param1 = assert_arg(1,"hello",'table')  --> error: argument 1 expected a 'table', got a 'string'
    local param4 = assert_arg(4,'!@#$%^&*','string',path.isdir,'not a directory')
         --> error: argument 4: '!@#$%^&*' not a directory
enum (...)
creates an Enum table. This helps prevent magic strings in code by throwing errors for accessing non-existing values.

Calling on the object does the same, but returns a soft error; nil + err.

The values are equal to the keys. The enum object is read-only.

Parameters:

  • ... strings that make up the enumeration.

Returns:

    Enum object

Usage:

  • -- accessing at runtime
    local obj = {}
    obj.MOVEMENT = utils.enum("FORWARD", "REVERSE", "LEFT", "RIGHT")
    
    if current_movement == obj.MOVEMENT.FORWARD then
      -- do something
    
    elseif current_movement == obj.MOVEMENT.REVERES then
      -- throws error due to typo 'REVERES', so a silent mistake becomes a hard error
      -- "'REVERES' is not a valid value (expected one of: 'FORWARD', 'REVERSE', 'LEFT', 'RIGHT')"
    
    end
  • -- validating user-input
    local parameter = "...some user provided option..."
    local ok, err = obj.MOVEMENT(parameter) -- calling on the object
    if not ok then
      print("bad 'parameter', " .. err)
      os.exit(1)
    end
function_arg (idx, f, msg)
process a function argument. This is used throughout Penlight and defines what is meant by a function: Something that is callable, or an operator string as defined by pl.operator, such as '>' or '#'. If a function factory has been registered for the type, it will be called to get the function.

Parameters:

  • idx argument index
  • f a function, operator string, or callable object
  • msg optional error message

Returns:

    a callable

Raises:

if idx is not a number or if f is not callable
assert_string (n, val)
assert the common case that the argument is a string.

Parameters:

  • n argument index
  • val a value that must be a string

Returns:

    the validated value

Raises:

val must be a string

Usage:

    local val = 42
    local param2 = utils.assert_string(2, val) --> error: argument 2 expected a 'string', got a 'number'
on_error (mode)

control the error strategy used by Penlight. This is a global setting that controls how utils.raise behaves:

  • 'default': return nil + error (this is the default)
  • 'error': throw a Lua error
  • 'quit': exit the program

Parameters:

  • mode either 'default', 'quit' or 'error'

See also:

raise (err)
used by Penlight functions to return errors. Its global behaviour is controlled by utils.on_error. To use this function you MUST use it in conjunction with return, since it might return nil + error.

Parameters:

  • err the error string.

See also:

Usage:

    if some_condition then
      return utils.raise("some condition was not met")  -- MUST use 'return'!
    end

File handling

readfile (filename, is_bin)
return the contents of a file as a string

Parameters:

  • filename The file path
  • is_bin open in binary mode

Returns:

    file contents
writefile (filename, str, is_bin)
write a string to a file

Parameters:

  • filename The file path
  • str The string
  • is_bin open in binary mode

Returns:

  1. true or nil
  2. error message

Raises:

error if filename or str aren't strings
readlines (filename)
return the contents of a file as a list of lines

Parameters:

  • filename The file path

Returns:

    file contents as a table

Raises:

error if filename is not a string

OS functions

executeex (cmd, bin)
execute a shell command and return the output. This function redirects the output to tempfiles and returns the content of those files.

Parameters:

  • cmd a shell command
  • bin boolean, if true, read output as binary file

Returns:

  1. true if successful
  2. actual return code
  3. stdout output (string)
  4. errout output (string)
quote_arg (argument)
Quote and escape an argument of a command. Quotes a single (or list of) argument(s) of a command to be passed to os.execute, pl.utils.execute or pl.utils.executeex.

Parameters:

  • argument (string or table/list) the argument to quote. If a list then all arguments in the list will be returned as a single string quoted.

Returns:

    quoted and escaped argument.

Usage:

    local options = utils.quote_arg {
        "-lluacov",
        "-e",
        "utils = print(require('pl.utils')._VERSION",
    }
    -- returns: -lluacov -e 'utils = print(require('\''pl.utils'\'')._VERSION'
quit ([code], msg, ...)
error out of this program gracefully.

Parameters:

  • code The exit code, defaults to -1 if omitted (optional)
  • msg The exit message will be sent to stderr (will be formatted with the extra parameters)
  • ... extra arguments for message's format'

See also:

Usage:

    utils.quit(-1, "Error '%s' happened", "42")
    -- is equivalent to
    utils.quit("Error '%s' happened", "42")  --> Error '42' happened

String functions

escape (s)
escape any Lua 'magic' characters in a string

Parameters:

  • s The input string
split (s, re, plain, n)
split a string into a list of strings separated by a delimiter.

Parameters:

  • s The input string
  • re optional A Lua string pattern; defaults to '%s+'
  • plain optional If truthy don't use Lua patterns
  • n optional maximum number of elements (if there are more, the last will remian un-split)

Returns:

    a list-like table

Raises:

error if s is not a string

See also:

splitv (s, re, plain, n)
split a string into a number of return values. Identical to split but returns multiple sub-strings instead of a single list of sub-strings.

Parameters:

  • s the string
  • re A Lua string pattern; defaults to '%s+'
  • plain don't use Lua patterns
  • n optional maximum number of splits

Returns:

    n values

See also:

Usage:

    first,next = splitv('user=jane=doe','=', false, 2)
    assert(first == "user")
    assert(next == "jane=doe")

Functional

memoize (func)
'memoize' a function (cache returned value for next call). This is useful if you have a function which is relatively expensive, but you don't know in advance what values will be required, so building a table upfront is wasteful/impossible.

Parameters:

  • func a function of at least one argument

Returns:

    a function with at least one argument, which is used as the key.
add_function_factory (mt, fun)
associate a function factory with a type. A function factory takes an object of the given type and returns a function for evaluating it

Parameters:

  • mt tab metatable
  • fun func a callable that returns a function
string_lambda (lf)
an anonymous function as a string. This string is either of the form '|args| expression' or is a function of one argument, '_'

Parameters:

  • lf function as a string

Returns:

    a function

Usage:

    string_lambda '|x|x+1' (2) == 3
    string_lambda '_+1' (2) == 3
bind1 (fn, p)
bind the first argument of the function to a value.

Parameters:

  • fn a function of at least two values (may be an operator string)
  • p a value

Returns:

    a function such that f(x) is fn(p,x)

Raises:

same as function_arg

See also:

Usage:

    local function f(msg, name)
      print(msg .. " " .. name)
    end
    
    local hello = utils.bind1(f, "Hello")
    
    print(hello("world"))     --> "Hello world"
    print(hello("sunshine"))  --> "Hello sunshine"
bind2 (fn, p)
bind the second argument of the function to a value.

Parameters:

  • fn a function of at least two values (may be an operator string)
  • p a value

Returns:

    a function such that f(x) is fn(x,p)

Raises:

same as function_arg

Usage:

    local function f(a, b, c)
      print(a .. " " .. b .. " " .. c)
    end
    
    local hello = utils.bind1(f, "world")
    
    print(hello("Hello", "!"))  --> "Hello world !"
    print(hello("Bye", "?"))    --> "Bye world ?"

Deprecation

set_deprecation_func (func)
Sets a deprecation warning function. An application can override this function to support proper output of deprecation warnings. The warnings can be generated from libraries or functions by calling utils.raise_deprecation. The default function will write to the 'warn' system (introduced in Lua 5.4, or the compatibility function from the compat module for earlier versions).

Note: only applications should set/change this function, libraries should not.

Parameters:

  • func a callback with signature: function(msg, trace) both arguments are strings, the latter being optional.

See also:

Usage:

    -- write to the Nginx logs with OpenResty
    utils.set_deprecation_func(function(msg, trace)
      ngx.log(ngx.WARN, msg, (trace and (" " .. trace) or nil))
    end)
    
    -- disable deprecation warnings
    utils.set_deprecation_func()
raise_deprecation (opts)
raises a deprecation warning. For options see the usage example below.

Note: the opts.deprecated_after field is the last version in which a feature or option was NOT YET deprecated! Because when writing the code it is quite often not known in what version the code will land. But the last released version is usually known.

Parameters:

  • opts options table

See also:

Usage:

    warn("@on")   -- enable Lua warnings, they are usually off by default
    
    function stringx.islower(str)
      raise_deprecation {
        source = "Penlight " .. utils._VERSION,                   -- optional
        message = "function 'islower' was renamed to 'is_lower'", -- required
        version_removed = "2.0.0",                                -- optional
        deprecated_after = "1.2.3",                               -- optional
        no_trace = true,                                          -- optional
      }
      return stringx.is_lower(str)
    end
    -- output: "[Penlight 1.9.2] function 'islower' was renamed to 'is_lower' (deprecated after 1.2.3, scheduled for removal in 2.0.0)"
generated by LDoc 1.4.6
Penlight-1.12.0/docs/libraries/pl.xml.html000066400000000000000000001220421416703176500203640ustar00rootroot00000000000000 Penlight Documentation

Module pl.xml

XML LOM Utilities.

This implements some useful things on LOM documents, such as returned by lxp.lom.parse. In particular, it can convert LOM back into XML text, with optional pretty-printing control. It is based on stanza.lua from Prosody

> d = xml.parse "<nodes><node id='1'>alice</node></nodes>"
> = d
<nodes><node id='1'>alice</node></nodes>
> = xml.tostring(d,'','  ')
<nodes>
   <node id='1'>alice</node>
</nodes>

Can be used as a lightweight one-stop-shop for simple XML processing; a simple XML parser is included but the default is to use lxp.lom if it can be found.

 Prosody IM
 Copyright (C) 2008-2010 Matthew Wild
 Copyright (C) 2008-2010 Waqas Hussain--
 classic Lua XML parser by Roberto Ierusalimschy.
 modified to output LOM format.
 http://lua-users.org/wiki/LuaXml
 
See the Guide

Dependencies: pl.utils

Soft Dependencies: lxp.lom (fallback is to use basic Lua parser)

Functions

new (tag[, attr={}]) create a new document node.
parse (text_or_filename, is_file, use_basic) parse an XML document.
elem (tag, items) Create a Node with a set of children (text or Nodes) and attributes.
tags (list) given a list of names, return a number of element constructors.
Doc:addtag (tag[, attrs={}]) Adds a document Node, at current position.
Doc:text (text) Adds a text node, at current position.
Doc:up () Moves current position up one level.
Doc:reset () Resets current position to top level.
Doc:add_direct_child (child) Append a child to the currrent Node (ignoring current position).
Doc:add_child (child) Append a child at the current position (without changing position).
Doc:set_attribs (t) Set attributes of a document node.
Doc:set_attrib (a, v) Set a single attribute of a document node.
Doc:get_attribs () Gets the attributes of a document node.
Doc.subst (template, data) create a substituted copy of a document,
Doc:child_with_name (tag) Return the first child with a given tag name (non-recursive).
Doc:get_elements_with_name (tag[, dont_recurse=false]) Returns all elements in a document that have a given tag.
Doc:children () Iterator over all children of a document node, including text nodes.
Doc:first_childtag () Return the first child element of a node, if it exists.
Doc:matching_tags ([tag=nil[, xmlns=nil]]) Iterator that matches tag names, and a namespace (non-recursive).
Doc:childtags () Iterator over all child tags of a document node.
Doc:maptags (callback) Visit child Nodes of a node and call a function, possibly modifying the document.
xml_escape (str) Escapes a string for safe use in xml.
xml_unescape (str) Unescapes a string from xml.
tostring (doc[, b_ind[, t_ind[, a_ind[, xml_preface]]]]) Function to pretty-print an XML document.
Doc:tostring ([b_ind[, t_ind[, a_ind[, xml_preface="<?xml version='1.0'?>"]]]]) Method to pretty-print an XML document.
Doc:get_text () get the full text value of an element.
clone (doc[, strsubst]) Returns a copy of a document.
Doc:filter ([strsubst]) Returns a copy of a document.
compare (t1, t2) Compare two documents or elements.
is_tag (d) is this value a document element?
walk (doc, depth_first, operation) Calls a function recursively over Nodes in the document.
parsehtml (s) Parse a well-formed HTML file as a string.
basic_parse (s, all_text, html) Parse a simple XML document using a pure Lua parser based on Robero Ierusalimschy's original version.
Doc:match (pat) does something...


Functions

new (tag[, attr={}])
create a new document node.

Parameters:

  • tag string the tag name
  • attr table attributes (table of name-value pairs) (default {})

Returns:

    the Node object

See also:

Usage:

    local doc = xml.new("main", { hello = "world", answer = "42" })
    print(doc)  -->  <main hello='world' answer='42'/>
parse (text_or_filename, is_file, use_basic)
parse an XML document. By default, this uses lxp.lom.parse, but falls back to basic_parse, or if use_basic is truthy

Parameters:

  • text_or_filename file or string representation
  • is_file whether textorfile is a file name or not
  • use_basic do a basic parse

Returns:

  1. a parsed LOM document with the document metatatables set
  2. nil, error the error can either be a file error or a parse error
elem (tag, items)
Create a Node with a set of children (text or Nodes) and attributes.

Parameters:

  • tag string a tag name
  • items table or string either a single child (text or Node), or a table where the hash part is the attributes and the list part is the children (text or Nodes).

Returns:

    the new Node

See also:

Usage:

    local doc = xml.elem("top", "hello world")                -- <top>hello world</top>
    local doc = xml.elem("main", xml.new("child"))            -- <main><child/></main>
    local doc = xml.elem("main", { "this ", "is ", "nice" })  -- <main>this is nice</main>
    local doc = xml.elem("main", { xml.new "this",
                                   xml.new "is",
                                   xml.new "nice" })          -- <main><this/><is/><nice/></main>
    local doc = xml.elem("main", { hello = "world" })         -- <main hello='world'/>
    local doc = xml.elem("main", {
      "prefix",
      xml.elem("child", { "this ", "is ", "nice"}),
      "postfix",
      attrib = "value"
    })   -- <main attrib='value'>prefix<child>this is nice</child>postfix</main>"
tags (list)
given a list of names, return a number of element constructors. If passing a comma-separated string, then whitespace surrounding the values will be stripped.

The returned constructor functions are a shortcut to xml.elem where you no longer provide the tag-name, but only the items table.

Parameters:

  • list string or table a list of names, or a comma-separated string.

Returns:

    (multiple) constructor functions; function(items). For the items parameter see xml.elem.

See also:

Usage:

    local new_parent, new_child = xml.tags 'mom, kid'
    doc = new_parent {new_child 'Bob', new_child 'Annie'}
    -- <mom><kid>Bob</kid><kid>Annie</kid></mom>
Doc:addtag (tag[, attrs={}])
Adds a document Node, at current position. This updates the last inserted position to the new Node.

Parameters:

  • tag string the tag name
  • attrs table attributes (table of name-value pairs) (default {})

Returns:

    the current node (self)

Usage:

    local doc = xml.new("main")
    doc:addtag("penlight", { hello = "world"})
    doc:addtag("expat")  -- added to 'penlight' since position moved
    print(doc)  -->  <main><penlight hello='world'><expat/></penlight></main>
Doc:text (text)
Adds a text node, at current position.

Parameters:

Returns:

    the current node (self)

Usage:

    local doc = xml.new("main")
    doc:text("penlight")
    doc:text("expat")
    print(doc)  -->  <main><penlightexpat</main>
Doc:up ()
Moves current position up one level.

Returns:

    the current node (self)
Doc:reset ()
Resets current position to top level. Resets to the self node.

Returns:

    the current node (self)
Doc:add_direct_child (child)
Append a child to the currrent Node (ignoring current position).

Parameters:

  • child a child node (either text or a document)

Returns:

    the current node (self)

Usage:

    local doc = xml.new("main")
    doc:add_direct_child("dog")
    doc:add_direct_child(xml.new("child"))
    doc:add_direct_child("cat")
    print(doc)  -->  <main>dog<child/>cat</main>
Doc:add_child (child)
Append a child at the current position (without changing position).

Parameters:

  • child a child node (either text or a document)

Returns:

    the current node (self)

Usage:

    local doc = xml.new("main")
    doc:addtag("one")
    doc:add_child(xml.new("item1"))
    doc:add_child(xml.new("item2"))
    doc:add_child(xml.new("item3"))
    print(doc)  -->  <main><one><item1/><item2/><item3/></one></main>
Doc:set_attribs (t)
Set attributes of a document node. Will add/overwite values, but will not remove existing ones. Operates on the Node itself, will not take position into account.

Parameters:

  • t table a table containing attribute/value pairs

Returns:

    the current node (self)
Doc:set_attrib (a, v)
Set a single attribute of a document node. Operates on the Node itself, will not take position into account.

Parameters:

  • a attribute
  • v its value, pass in nil to delete the attribute

Returns:

    the current node (self)
Doc:get_attribs ()
Gets the attributes of a document node. Operates on the Node itself, will not take position into account.

Returns:

    table with attributes (attribute/value pairs)
Doc.subst (template, data)
create a substituted copy of a document,

Parameters:

  • template may be a document or a string representation which will be parsed and cached
  • data a table of name-value pairs or a list of such tables

Returns:

    an XML document
Doc:child_with_name (tag)
Return the first child with a given tag name (non-recursive).

Parameters:

  • tag the tag name

Returns:

    the child Node found or nil if not found
Doc:get_elements_with_name (tag[, dont_recurse=false])
Returns all elements in a document that have a given tag.

Parameters:

  • tag string a tag name
  • dont_recurse boolean optionally only return the immediate children with this tag name (default false)

Returns:

    a list of elements found, list will be empty if none was found.
Doc:children ()
Iterator over all children of a document node, including text nodes. This function is not recursive, so returns only direct child nodes.

Returns:

    iterator that returns a single Node per iteration.
Doc:first_childtag ()
Return the first child element of a node, if it exists. This will skip text nodes.

Returns:

    first child Node or nil if there is none.
Doc:matching_tags ([tag=nil[, xmlns=nil]])
Iterator that matches tag names, and a namespace (non-recursive).

Parameters:

  • tag string tag names to return. Returns all tags if not provided. (default nil)
  • xmlns string the namespace value ('xmlns' attribute) to return. If not provided will match all namespaces. (default nil)

Returns:

    iterator that returns a single Node per iteration.
Doc:childtags ()
Iterator over all child tags of a document node. This will skip over text nodes.

Returns:

    iterator that returns a single Node per iteration.
Doc:maptags (callback)
Visit child Nodes of a node and call a function, possibly modifying the document. Text elements will be skipped. This is not recursive, so only direct children will be passed.

Parameters:

  • callback function a function with signature function(node), passed the node. The element will be updated with the returned value, or deleted if it returns nil.
xml_escape (str)
Escapes a string for safe use in xml. Handles quotes(single+double), less-than, greater-than, and ampersand.

Parameters:

  • str string string value to escape

Returns:

    escaped string

Usage:

    local esc = xml.xml_escape([["'<>&]])  --> "&quot;&apos;&lt;&gt;&amp;"
xml_unescape (str)
Unescapes a string from xml. Handles quotes(single+double), less-than, greater-than, and ampersand.

Parameters:

  • str string string value to unescape

Returns:

    unescaped string

Usage:

    local unesc = xml.xml_escape("&quot;&apos;&lt;&gt;&amp;")  --> [["'<>&]]
tostring (doc[, b_ind[, t_ind[, a_ind[, xml_preface]]]])
Function to pretty-print an XML document.

Parameters:

  • doc an XML document
  • b_ind string or int an initial block-indent (required when t_ind is set) (optional)
  • t_ind string or int an tag-indent for each level (required when a_ind is set) (optional)
  • a_ind string or int if given, indent each attribute pair and put on a separate line (optional)
  • xml_preface string or bool force prefacing with default or custom , if truthy then &lt;?xml version='1.0'?&gt; will be used as default. (optional)

Returns:

    a string representation

See also:

Doc:tostring ([b_ind[, t_ind[, a_ind[, xml_preface="<?xml version='1.0'?>"]]]])
Method to pretty-print an XML document. Invokes xml.tostring.

Parameters:

  • b_ind string or int an initial indent (required when t_ind is set) (optional)
  • t_ind string or int an indent for each level (required when a_ind is set) (optional)
  • a_ind string or int if given, indent each attribute pair and put on a separate line (optional)
  • xml_preface string force prefacing with default or custom (default "<?xml version='1.0'?>")

Returns:

    a string representation

See also:

Doc:get_text ()
get the full text value of an element.

Returns:

    a single string with all text elements concatenated

Usage:

    local doc = xml.new("main")
    doc:text("one")
    doc:add_child(xml.elem "two")
    doc:text("three")
    
    local t = doc:get_text()    -->  "onethree"
clone (doc[, strsubst])

Returns a copy of a document. The strsubst parameter is a callback with signature function(object, kind, parent).

Param kind has the following values, and parameters:

  • "*TAG": object is the tag-name, parent is the Node object. Returns the new tag name.

  • "*TEXT": object is the text-element, parent is the Node object. Returns the new text value.

  • other strings not prefixed with *: kind is the attribute name, object is the attribute value, parent is the Node object. Returns the new attribute value.

Parameters:

  • doc Node or string a Node object or string (text node)
  • strsubst function an optional function for handling string copying which could do substitution, etc. (optional)

Returns:

    copy of the document

See also:

Doc:filter ([strsubst])
Returns a copy of a document. This is the method version of xml.clone.

Parameters:

  • strsubst function an optional function for handling string copying (optional)

See also:

compare (t1, t2)
Compare two documents or elements. Equality is based on tag, child nodes (text and tags), attributes and order of those (order only fails if both are given, and not equal).

Parameters:

  • t1 Node or string a Node object or string (text node)
  • t2 Node or string a Node object or string (text node)

Returns:

    boolean true when the Nodes are equal.
is_tag (d)
is this value a document element?

Parameters:

  • d any value

Returns:

    boolean true if it is a table with property tag being a string value.
walk (doc, depth_first, operation)
Calls a function recursively over Nodes in the document. Will only call on tags, it will skip text nodes. The function signature for operation is function(tag_name, Node).

Parameters:

  • doc Node or string a Node object or string (text node)
  • depth_first boolean visit child nodes first, then the current node
  • operation function a function which will receive the current tag name and current node.
parsehtml (s)
Parse a well-formed HTML file as a string. Tags are case-insenstive, DOCTYPE is ignored, and empty elements can be .. empty.

Parameters:

  • s the HTML
basic_parse (s, all_text, html)
Parse a simple XML document using a pure Lua parser based on Robero Ierusalimschy's original version.

Parameters:

  • s the XML document to be parsed.
  • all_text if true, preserves all whitespace. Otherwise only text containing non-whitespace is included.
  • html if true, uses relaxed HTML rules for parsing
Doc:match (pat)
does something...

Parameters:

  • pat
generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/000077500000000000000000000000001416703176500155645ustar00rootroot00000000000000Penlight-1.12.0/docs/manual/01-introduction.md.html000066400000000000000000001231761416703176500220220ustar00rootroot00000000000000 Penlight Documentation

Introduction

Purpose

It is often said of Lua that it does not include batteries. That is because the goal of Lua is to produce a lean expressive language that will be used on all sorts of machines, (some of which don't even have hierarchical filesystems). The Lua language is the equivalent of an operating system kernel; the creators of Lua do not see it as their responsibility to create a full software ecosystem around the language. That is the role of the community.

A principle of software design is to recognize common patterns and reuse them. If you find yourself writing things like `io.write(string.format('the answer is %d ',42))` more than a number of times then it becomes useful just to define a function printf. This is good, not just because repeated code is harder to maintain, but because such code is easier to read, once people understand your libraries.

Penlight captures many such code patterns, so that the intent of your code becomes clearer. For instance, a Lua idiom to copy a table is {unpack(t)}, but this will only work for 'small' tables (for a given value of 'small') so it is not very robust. Also, the intent is not clear. So tablex.deepcopy is provided, which will also copy nested tables and and associated metatables, so it can be used to clone complex objects.

The default error handling policy follows that of the Lua standard libraries: if a argument is the wrong type, then an error will be thrown, but otherwise we return nil,message if there is a problem. There are some exceptions; functions like input.fields default to shutting down the program immediately with a useful message. This is more appropriate behaviour for a script than providing a stack trace. (However, this default can be changed.) The lexer functions always throw errors, to simplify coding, and so should be wrapped in pcall.

If you are used to Python conventions, please note that all indices consistently start at 1.

The Lua function table.foreach has been deprecated in favour of the for in statement, but such an operation becomes particularly useful with the higher-order function support in Penlight. Note that tablex.foreach reverses the order, so that the function is passed the value and then the key. Although perverse, this matches the intended use better.

The only important external dependence of Penlight is LuaFileSystem (lfs), and if you want dir.copyfile to work cleanly on Windows, you will need either alien or be using LuaJIT as well. (The fallback is to call the equivalent shell commands.)

To Inject or not to Inject?

It was realized a long time ago that large programs needed a way to keep names distinct by putting them into tables (Lua), namespaces (C++) or modules (Python). It is obviously impossible to run a company where everyone is called 'Bruce', except in Monty Python skits. These 'namespace clashes' are more of a problem in a simple language like Lua than in C++, because C++ does more complicated lookup over 'injected namespaces'. However, in a small group of friends, 'Bruce' is usually unique, so in particular situations it's useful to drop the formality and not use last names. It depends entirely on what kind of program you are writing, whether it is a ten line script or a ten thousand line program.

So the Penlight library provides the formal way and the informal way, without imposing any preference. You can do it formally like:

local utils = require 'pl.utils'
utils.printf("%s\n","hello, world!")

or informally like:

require 'pl'
utils.printf("%s\n","That feels better")

require 'pl' makes all the separate Penlight modules available, without needing to require them each individually.

Generally, the formal way is better when writing modules, since then there are no global side-effects and the dependencies of your module are made explicit.

Andrew Starks has contributed another way, which balances nicely between the formal need to keep the global table uncluttered and the informal need for convenience. require'pl.import_into' returns a function, which accepts a table for injecting Penlight into, or if no table is given, it passes back a new one.

local pl = require'pl.import_into'()

The table pl is a 'lazy table' which loads modules as needed, so we can then use pl.utils.printf and so forth, without an explicit `require' or harming any globals.

If you are using _ENV with Lua 5.2 to define modules, then here is a way to make Penlight available within a module:

local _ENV,M = require 'pl.import_into' ()

function answer ()
    -- all the Penlight modules are available!
    return pretty.write(utils.split '10 20  30', '')
end

return M

The default is to put Penlight into \_ENV, which has the unintended effect of making it available from the module (much as module(...,package.seeall) does). To satisfy both convenience and safety, you may pass true to this function, and then the module M is not the same as \_ENV, but only contains the exported functions.

Otherwise, Penlight will not bring in functions into the global table, or clobber standard tables like 'io'. require('pl') will bring tables like 'utils','tablex',etc into the global table if they are used. This 'load-on-demand' strategy ensures that the whole kitchen sink is not loaded up front, so this method is as efficient as explicitly loading required modules.

You have an option to bring the pl.stringx methods into the standard string table. All strings have a metatable that allows for automatic lookup in string, so we can say s:upper(). Importing stringx allows for its functions to also be called as methods: s:strip(),etc:

require 'pl'
stringx.import()

or, more explicitly:

require('pl.stringx').import()

A more delicate operation is importing tables into the local environment. This is convenient when the context makes the meaning of a name very clear:

> require 'pl'
> utils.import(math)
> = sin(1.2)
0.93203908596723

utils.import can also be passed a module name as a string, which is first required and then imported. If used in a module, import will bring the symbols into the module context.

Keeping the global scope simple is very necessary with dynamic languages. Using global variables in a big program is always asking for trouble, especially since you do not have the spell-checking provided by a compiler. The pl.strict module enforces a simple rule: globals must be 'declared'. This means that they must be assigned before use; assigning to nil is sufficient.

> require 'pl.strict'
> print(x)
stdin:1: variable 'x' is not declared
> x = nil
> print(x)
nil

The strict module provided by Penlight is compatible with the 'load-on-demand' scheme used by require 'pl.

strict also disallows assignment to global variables, except in the main program. Generally, modules have no business messing with global scope; if you must do it, then use a call to rawset. Similarly, if you have to check for the existence of a global, use rawget.

If you wish to enforce strictness globally, then just add require 'pl.strict' at the end of pl/init.lua, otherwise call it from your main program.

As from 1.1.0, this module provides a strict.module function which creates (or modifies) modules so that accessing an unknown function or field causes an error.

For example,

-- mymod.lua
local strict = require 'pl.strict'
local M = strict.module (...)

function M.answer ()
    return 42
end

return M

If you were to accidently type mymod.Answer(), then you would get a runtime error: "variable 'Answer' is not declared in 'mymod'".

This can be applied to existing modules. You may desire to have the same level of checking for the Lua standard libraries:

strict.make_all_strict(_G)

Thereafter a typo such as math.cosine will give you an explicit error, rather than merely returning a nil that will cause problems later.

What are function arguments in Penlight?

Many functions in Penlight themselves take function arguments, like map which applies a function to a list, element by element. You can use existing functions, like math.max, anonymous functions (like `function(x,y) return x > y end ), or operations by name (e.g '*' or '..'). The module pl.operator` exports all the standard Lua operations, like the Python module of the same name. Penlight allows these to be referred to by name, so operator.gt can be more concisely expressed as '>'.

Note that the map functions pass any extra arguments to the function, so we can have ls:filter('>',0), which is a shortcut for ls:filter(function(x) return x > 0 end).

Finally, pl.func supports placeholder expressions in the Boost lambda style, so that an anonymous function to multiply the two arguments can be expressed as \1*\2.

To use them directly, note that all function arguments in Penlight go through utils.function_arg. pl.func registers itself with this function, so that you can directly use placeholder expressions with standard methods:

> _1 = func._1
> = List{10,20,30}:map(_1+1)
{11,21,31}

Another option for short anonymous functions is provided by utils.string_lambda; this is invoked automatically:

> = List{10,20,30}:map '|x| x + 1'
{11,21,31}

Pros and Cons of Loopless Programming

The standard loops-and-ifs 'imperative' style of programming is dominant, and often seems to be the 'natural' way of telling a machine what to do. It is in fact very much how the machine does things, but we need to take a step back and find ways of expressing solutions in a higher-level way. For instance, applying a function to all elements of a list is a common operation:

local res = {}
for i = 1,#ls do
    res[i] = fun(ls[i])
end

This can be efficiently and succintly expressed as ls:map(fun). Not only is there less typing but the intention of the code is clearer. If readers of your code spend too much time trying to guess your intention by analyzing your loops, then you have failed to express yourself clearly. Similarly, ls:filter('>',0) will give you all the values in a list greater than zero. (Of course, if you don't feel like using List, or have non-list-like tables, then pl.tablex offers the same facilities. In fact, the List methods are implemented using tablex functions.)

A common observation is that loopless programming is less efficient, particularly in the way it uses memory. ls1:map2('*',ls2):reduce '+' will give you the dot product of two lists, but an unnecessary temporary list is created. But efficiency is relative to the actual situation, it may turn out to be fast enough, or may not appear in any crucial inner loops, etc.

Writing loops is 'error-prone and tedious', as Stroustrup says. But any half-decent editor can be taught to do much of that typing for you. The question should actually be: is it tedious to read loops? As with natural language, programmers tend to read chunks at a time. A for-loop causes no surprise, and probably little brain activity. One argument for loopless programming is the loops that you do write stand out more, and signal 'something different happening here'. It should not be an all-or-nothing thing, since most programs require a mixture of idioms that suit the problem. Some languages (like APL) do nearly everything with map and reduce operations on arrays, and so solutions can sometimes seem forced. Wisdom is knowing when a particular idiom makes a particular problem easy to solve and the solution easy to explain afterwards.

Generally useful functions.

The function printf discussed earlier is included in pl.utils because it makes properly formatted output easier. (There is an equivalent fprintf which also takes a file object parameter, just like the C function.)

Splitting a string using a delimiter is a fairly common operation, hence split.

Utility functions like is_type help with identifying what kind of animal you are dealing with. The Lua type function handles the basic types, but can't distinguish between different kinds of objects, which are all tables. So is_type handles both cases, like is_type(s,"string") and is_type(ls,List).

A common pattern when working with Lua varargs is capturing all the arguments in a table:

function t(...)
    local args = {...}
    ...
end

But this will bite you someday when nil is one of the arguments, since this will put a 'hole' in your table. In particular, #ls will only give you the size upto the nil value. Hence the need for table.pack - this is a new Lua 5.2 function which Penlight defines also for Lua 5.1.

function t(...)
    local args,n = table.pack(...)
    for i = 1,n do
      ...
    end
end

The 'memoize' pattern occurs when you have a function which is expensive to call, but will always return the same value subsequently. utils.memoize is given a function, and returns another function. This calls the function the first time, saves the value for that argument, and thereafter for that argument returns the saved value. This is a more flexible alternative to building a table of values upfront, since in general you won't know what values are needed.

sum = utils.memoize(function(n)
    local sum = 0
    for i = 1,n do sum = sum + i end
    return sum
end)
...
s = sum(1e8) --takes time!
...
s = sum(1e8) --returned saved value!

Penlight is fully compatible with Lua 5.1, 5.2 and LuaJIT 2. To ensure this, utils also defines the global Lua 5.2 load function as utils.load

  • the input (either a string or a function)
  • the source name used in debug information
  • the mode is a string that can have either or both of 'b' or 't', depending on whether the source is a binary chunk or text code (default is 'bt')
  • the environment for the compiled chunk

Using utils.load should reduce the need to call the deprecated function setfenv, and make your Lua 5.1 code 5.2-friendly.

The utils module exports getfenv and setfenv for Lua 5.2 as well, based on code by Sergey Rozhenko. Note that these functions can fail for functions which don't access any globals.

Application Support

app.parse_args is a simple command-line argument parser. If called without any arguments, it tries to use the global arg array. It returns the flags (options begining with '-') as a table of name/value pairs, and the arguments as an array. It knows about long GNU-style flag names, e.g. --value, and groups of short flags are understood, so that -ab is short for -a -b. The flags result would then look like {value=true,a=true,b=true}.

Flags may take values. The command-line --value=open -n10 would result in {value='open',n='10'}; generally you can use '=' or ':' to separate the flag from its value, except in the special case where a short flag is followed by an integer. Or you may specify upfront that some flags have associated values, and then the values will follow the flag.

> require 'pl'
> flags,args = app.parse_args({'-o','fred','-n10','fred.txt'},{o=true})
> pretty.dump(flags)
{o='fred',n='10'}

parse_args is not intelligent or psychic; it will not convert any flag values or arguments for you, or raise errors. For that, have a look at Lapp.

An application which consists of several files usually cannot use require to load files in the same directory as the main script. app.require_here() ensures that the Lua module path is modified so that files found locally are found first. In the examples directory, test-symbols.lua uses this function to ensure that it can find symbols.lua even if it is not run from this directory.

app.appfile will create a filename that your application can use to store its private data, based on the script name. For example, app.appfile "test.txt" from a script called testapp.lua produces the following file on my Windows machine:

C:\Documents and Settings\SJDonova\.testapp\test.txt

and the equivalent on my Linux machine:

/home/sdonovan/.testapp/test.txt

If .testapp does not exist, it will be created.

Penlight makes it convenient to save application data in Lua format. You can use pretty.dump(t,file) to write a Lua table in a human-readable form to a file, and pretty.read(file.read(file)) to generate the table again, using the pretty module.

Simplifying Object-Oriented Programming in Lua

Lua is similar to JavaScript in that the concept of class is not directly supported by the language. In fact, Lua has a very general mechanism for extending the behaviour of tables which makes it straightforward to implement classes. A table's behaviour is controlled by its metatable. If that metatable has a \\index function or table, this will handle looking up anything which is not found in the original table. A class is just a table with an __index key pointing to itself. Creating an object involves making a table and setting its metatable to the class; then when handling obj.fun, Lua first looks up fun in the table obj, and if not found it looks it up in the class. obj:fun(a) is just short for obj.fun(obj,a). So with the metatable mechanism and this bit of syntactic sugar, it is straightforward to implement classic object orientation.

-- animal.lua

class = require 'pl.class'

class.Animal()

function Animal:_init(name)
    self.name = name
end

function Animal:__tostring()
  return self.name..': '..self:speak()
end

class.Dog(Animal)

function Dog:speak()
  return 'bark'
end

class.Cat(Animal)

function Cat:_init(name,breed)
    self:super(name)  -- must init base!
    self.breed = breed
end

function Cat:speak()
  return 'meow'
end

class.Lion(Cat)

function Lion:speak()
  return 'roar'
end

fido = Dog('Fido')
felix = Cat('Felix','Tabby')
leo = Lion('Leo','African')

$ lua -i animal.lua
> = fido,felix,leo
Fido: bark      Felix: meow     Leo: roar
> = leo:is_a(Animal)
true
> = leo:is_a(Dog)
false
> = leo:is_a(Cat)
true

All Animal does is define \\tostring, which Lua will use whenever a string representation is needed of the object. In turn, this relies on speak, which is not defined. So it's what C++ people would call an abstract base class; the specific derived classes like Dog define speak. Please note that if derived classes have their own constructors, they must explicitly call the base constructor for their base class; this is conveniently available as the super method.

Note that (as always) there are multiple ways to implement OOP in Lua; this method uses the classic 'a class is the __index of its objects' but does 'fat inheritance'; methods from the base class are copied into the new class. The advantage of this is that you are not penalized for long inheritance chains, for the price of larger classes, but generally objects outnumber classes! (If not, something odd is going on with your design.)

All such objects will have a is_a method, which looks up the inheritance chain to find a match. Another form is class_of, which can be safely called on all objects, so instead of leo:is_a(Animal) one can say Animal:class_of(leo).

There are two ways to define a class, either class.Name() or Name = class(); both work identically, except that the first form will always put the class in the current environment (whether global or module); the second form provides more flexibility about where to store the class. The first form does name the class by setting the _name field, which can be useful in identifying the objects of this type later. This session illustrates the usefulness of having named classes, if no __tostring method is explicitly defined.

> class.Fred()
> a = Fred()
> = a
Fred: 00459330
> Alice = class()
> b = Alice()
> = b
table: 00459AE8
> Alice._name = 'Alice'
> = b
Alice: 00459AE8

So Alice = class(); Alice._name = 'Alice' is exactly the same as class.Alice().

This useful notation is borrowed from Hugo Etchegoyen's classlib which further extends this concept to allow for multiple inheritance. Notice that the more convenient form puts the class name in the current environment! That is, you may use it safely within modules using the old-fashioned module() or the new _ENV mechanism.

There is always more than one way of doing things in Lua; some may prefer this style for creating classes:

local class = require 'pl.class'

class.Named {
    _init = function(self,name)
        self.name = name
    end;

    __tostring = function(self)
        return 'boo '..self.name
    end;
}

b = Named 'dog'
print(b)
--> boo dog

Note that you have to explicitly declare self and end each function definition with a semi-colon or comma, since this is a Lua table. To inherit from a base class, set the special field _base to the class in this table.

Penlight provides a number of useful classes; there is List, which is a Lua clone of the standard Python list object, and Set which represents sets. There are three kinds of map defined: Map, MultiMap (where a key may have multiple values) and OrderedMap (where the order of insertion is remembered.). There is nothing special about these classes and you may inherit from them.

A powerful thing about dynamic languages is that you can redefine existing classes and functions, which is often called 'monkey patching' It's entertaining and convenient, but ultimately anti-social; you may modify List but then any other modules using this shared resource can no longer be sure about its behaviour. (This is why you must say stringx.import() explicitly if you want the extended string methods - it would be a bad default.) Lua is particularly open to modification but the community is not as tolerant of monkey-patching as the Ruby community, say. You may wish to add some new methods to List? Cool, but that's what subclassing is for.

class.Strings(List)

function Strings:my_method()
...
end

It's definitely more useful to define exactly how your objects behave in unknown conditions. All classes have a catch method you can use to set a handler for unknown lookups; the function you pass looks exactly like the __index metamethod.

Strings:catch(function(self,name)
    return function() error("no such method "..name,2) end
end)

In this case we're just customizing the error message, but creative things can be done. Consider this code from test-vector.lua:

Strings:catch(List.default_map_with(string))

ls = Strings{'one','two','three'}
asserteq(ls:upper(),{'ONE','TWO','THREE'})
asserteq(ls:sub(1,2),{'on','tw','th'})

So we've converted a unknown method invocation into a map using the function of that name found in string. So for a Vector (which is a specialization of List for numbers) it makes sense to make math the default map so that v:sin() makes sense.

Note that map operations return a object of the same type - this is often called covariance. So ls:upper() itself returns a Strings object.

This is not always what you want, but objects can always be cast to the desired type. (cast doesn't create a new object, but returns the object passed.)

local sizes = ls:map '#'
asserteq(sizes, {3,3,5})
asserteq(utils.type(sizes),'Strings')
asserteq(sizes:is_a(Strings),true)
sizes = Vector:cast(sizes)
asserteq(utils.type(sizes),'Vector')
asserteq(sizes+1,{4,4,6})

About utils.type: it can only return a string for a class type if that class does in fact have a _name field.

Properties are a useful object-oriented pattern. We wish to control access to a field, but don't wish to force the user of the class to say obj:get_field() etc. This excerpt from tests/test-class.lua shows how it is done:

local MyProps = class(class.properties)
local setted_a, got_b

function MyProps:_init ()
    self._a = 1
    self._b = 2
end

function MyProps:set_a (v)
    setted_a = true
    self._a = v
end

function MyProps:get_b ()
    got_b = true
    return self._b
end

local mp = MyProps()

mp.a = 10

asserteq(mp.a,10)
asserteq(mp.b,2)
asserteq(setted_a and got_b, true)

The convention is that the internal field name is prefixed with an underscore; when reading mp.a, first a check for an explicit getter get_a and then only look for _a. Simularly, writing mp.a causes the setter set_a to be used.

This is cool behaviour, but like much Lua metaprogramming, it is not free. Method lookup on such objects goes through \\index as before, but now \\index is a function which has to explicitly look up methods in the class, before doing any property indexing, which is not going to be as fast as field lookup. If however, your accessors actually do non-trivial things, then the extra overhead could be worth it.

This is not really intended for access control because external code can write to mp._a directly. It is possible to have this kind of control in Lua, but it again comes with run-time costs.

generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/02-arrays.md.html000066400000000000000000001565741416703176500206130ustar00rootroot00000000000000 Penlight Documentation

Tables and Arrays

Python-style Lists

One of the elegant things about Lua is that tables do the job of both lists and dicts (as called in Python) or vectors and maps, (as called in C++), and they do it efficiently. However, if we are dealing with 'tables with numerical indices' we may as well call them lists and look for operations which particularly make sense for lists. The Penlight List class was originally written by Nick Trout for Lua 5.0, and translated to 5.1 and extended by myself. It seemed that borrowing from Python was a good idea, and this eventually grew into Penlight.

Here is an example showing List in action; it redefines __tostring, so that it can print itself out more sensibly:

> List = require 'pl.List'  --> automatic with require 'pl' <---
> l = List()
> l:append(10)
> l:append(20)
> = l
{10,20}
> l:extend {30,40}
{10,20,30,40}
> l:insert(1,5)
{5,10,20,30,40}
> = l:pop()
40
> = l
{5,10,20,30}
> = l:index(30)
4
> = l:contains(30)
true
> = l:reverse()  ---> note: doesn't make a copy!
{30,20,10,5}

Although methods like sort and reverse operate in-place and change the list, they do return the original list. This makes it possible to do method chaining, like ls = ls:append(10):append(20):reverse():append(1). But (and this is an important but) no extra copy is made, so ls does not change identity. List objects (like tables) are mutable, unlike strings. If you want a copy of a list, then List(ls) will do the job, i.e. it acts like a copy constructor. However, if passed any other table, List will just set the metatable of the table and not make a copy.

A particular feature of Python lists is slicing. This is fully supported in this version of List, except we use 1-based indexing. So List.slice works rather like string.sub:

> l = List {10,20,30,40}
> = l:slice(1,1)  ---> note: creates a new list!
{10}
> = l:slice(2,2)
{20}
> = l:slice(2,3)
{20,30}
> = l:slice(2,-2)
{20,30}
> = l:slice_assign(2,2,{21,22,23})
{10,21,22,23,30,40}
> = l:chop(1,1)
{21,22,23,30,40}

Functions like slice_assign and chop modify the list; the first is equivalent to Pythonl[i1:i2] = seq and the second to del l[i1:i2].

List objects are ultimately just Lua 'list-like' tables, but they have extra operations defined on them, such as equality and concatention. For regular tables, equality is only true if the two tables are identical objects, whereas two lists are equal if they have the same contents, i.e. that l1[i]==l2[i] for all elements.

> l1 = List {1,2,3}
> l2 = List {1,2,3}
> = l1 == l2
true
> = l1..l2
{1,2,3,1,2,3}

The List constructor can be passed a function. If so, it's assumed that this is an iterator function that can be repeatedly called to generate a sequence. One such function is io.lines; the following short, intense little script counts the number of lines in standard input:

-- linecount.lua
require 'pl'
ls = List(io.lines())
print(#ls)

List.iterate captures what List considers a sequence. In particular, it can also iterate over all 'characters' in a string:

> for ch in List.iterate 'help' do io.write(ch,' ') end
h e l p >

Since the function iterate is used internally by the List constructor, strings can be made into lists of character strings very easily.

There are a number of operations that go beyond the standard Python methods. For instance, you can partition a list into a table of sublists using a function. In the simplest form, you use a predicate (a function returning a boolean value) to partition the list into two lists, one of elements matching and another of elements not matching. But you can use any function; if we use type then the keys will be the standard Lua type names.

> ls = List{1,2,3,4}
> ops = require 'pl.operator'
> ls:partition(function(x) return x > 2 end)
{false={1,2},true={3,4}}
> ls = List{'one',math.sin,List{1},10,20,List{1,2}}
> ls:partition(type)
{function={function: 00369110},string={one},number={10,20},table={{1},{1,2}}}

This is one List method which returns a table which is not a List. Bear in mind that you can always call a List method on a plain table argument, so List.partition(t,type) works as expected. But these functions will only operate on the array part of the table.

The 'nominal' type of the returned table is pl.Multimap, which describes a mapping between keys and multiple values. This does not mean that pl.Multimap is automatically loaded whenever you use partition (or List for that matter); this is one of the standard metatables which are only filled out when the appropriate module is loaded. This allows tables to be tagged appropriately without causing excessive coupling.

Stacks occur everywhere in computing. List supports stack-like operations; there is already pop (remove and return last value) and append acts like push (add a value to the end). push is provided as an alias for append, and the other stack operation (size) is simply the size operator #. Queues can also be implemented; you use pop to take values out of the queue, and put to insert a value at the begining.

You may derive classes from List, and since the list-returning methods are covariant, the result of slice etc will return lists of the derived type, not List. For instance, consider the specialization of a List type that contains numbers in tests/test-list.lua:

n1 = NA{10,20,30}
n2 = NA{1,2,3}
ns = n1 + 2*n2
asserteq(ns,{12,24,36})
min,max = ns:slice(1,2):minmax()
asserteq(T(min,max),T(12,24))
asserteq(n1:normalize():sum(),1,1e-8)

Map and Set classes

The Map class exposes what Python would call a 'dict' interface, and accesses the hash part of the table. The name 'Map' is used to emphasize the interface, not the implementation; it is an object which maps keys onto values; m['alice'] or the equivalent m.alice is the access operation. This class also provides explicit set and get methods, which are trivial for regular maps but get interesting when Map is subclassed. The other operation is update, which extends a map by copying the keys and values from another table, perhaps overwriting existing keys:

> Map = require 'pl.Map'
> m = Map{one=1,two=2}
> m:update {three=3,four=4,two=20}
> = m == M{one=1,two=20,three=3,four=4}
true

The method values returns a list of the values, and keys returns a list of the keys; there is no guarantee of order. getvalues is given a list of keys and returns a list of values associated with these keys:

> m = Map{one=1,two=2,three=3}
> = m:getvalues {'one','three'}
{1,3}
> = m:getvalues(m:keys()) == m:values()
true

When querying the value of a Map, it is best to use the get method:

> print(m:get 'one', m:get 'two')
1     2

The reason is that m[key] can be ambiguous; due to the current implementation, m["get"] will always succeed, because if a value is not present in the map, it will be looked up in the Map metatable, which contains a method get. There is currently no simple solution to this annoying restriction.

There are some useful classes which inherit from Map. An OrderedMap behaves like a Map but keeps its keys in order if you use its set method to add keys and values. Like all the 'container' classes in Penlight, it defines an iter method for iterating over its values; this will return the keys and values in the order of insertion; the keys and values methods likewise.

A MultiMap allows multiple values to be associated with a given key. So set (as before) takes a key and a value, but calling it with the same key and a different value does not overwrite but adds a new value. get (or using []) will return a list of values.

A Set can be seen as a special kind of Map, where all the values are true, the keys are the values, and the order is not important. So in this case Set.values is defined to return a list of the keys. Sets can display themselves, and the basic operations like union (+) and intersection (*) are defined.

> Set = require 'pl.Set'
> = Set{'one','two'} == Set{'two','one'}
true
> fruit = Set{'apple','banana','orange'}
> = fruit['banana']
true
> = fruit['hazelnut']
nil
> = fruit:values()
{apple,orange,banana}
> colours = Set{'red','orange','green','blue'}
> = fruit,colours
[apple,orange,banana]   [blue,green,orange,red]
> = fruit+colours
[blue,green,apple,red,orange,banana]
> = fruit*colours
[orange]

There are also the functions Set.difference and Set.symmetric_difference. The first answers the question 'what fruits are not colours?' and the second 'what are fruits and colours but not both?'

> = fruit - colours
[apple,banana]
> = fruit ^ colours
[blue,green,apple,red,banana]

Adding elements to a set is simply fruit['peach'] = true and removing is fruit['apple'] = nil . To make this simplicity work properly, the Set class has no methods - either you use the operator forms or explicitly use Set.intersect etc. In this way we avoid the ambiguity that plagues Map.

(See pl.Map and pl.Set)

Useful Operations on Tables

Some notes on terminology: Lua tables are usually list-like (like an array) or map-like (like an associative array or dict); they can of course have a list-like and a map-like part. Some of the table operations only make sense for list-like tables, and some only for map-like tables. (The usual Lua terminology is the array part and the hash part of the table, which reflects the actual implementation used; it is more accurate to say that a Lua table is an associative map which happens to be particularly efficient at acting like an array.)

The functions provided in table provide all the basic manipulations on Lua tables, but as we saw with the List class, it is useful to build higher-level operations on top of those functions. For instance, to copy a table involves this kind of loop:

local res = {}
for k,v in pairs(T) do
    res[k] = v
end
return res

The tablex module provides this as copy, which does a shallow copy of a table. There is also deepcopy which goes further than a simple loop in two ways; first, it also gives the copy the same metatable as the original (so it can copy objects like List above) and any nested tables will also be copied, to arbitrary depth. There is also icopy which operates on list-like tables, where you can set optionally set the start index of the source and destination as well. It ensures that any left-over elements will be deleted:

asserteq(icopy({1,2,3,4,5,6},{20,30}),{20,30})   -- start at 1
asserteq(icopy({1,2,3,4,5,6},{20,30},2),{1,20,30}) -- start at 2
asserteq(icopy({1,2,3,4,5,6},{20,30},2,2),{1,30}) -- start at 2, copy from 2

(This code from the tablex test module shows the use of pl.test.asserteq)

Whereas, move overwrites but does not delete the rest of the destination:

asserteq(move({1,2,3,4,5,6},{20,30}),{20,30,3,4,5,6})
asserteq(move({1,2,3,4,5,6},{20,30},2),{1,20,30,4,5,6})
asserteq(move({1,2,3,4,5,6},{20,30},2,2),{1,30,3,4,5,6})

(The difference is somewhat like that between C's strcpy and memmove.)

To summarize, use copy or deepcopy to make a copy of an arbitrary table. To copy into a map-like table, use update; to copy into a list-like table use icopy, and move if you are updating a range in the destination.

To complete this set of operations, there is insertvalues which works like table.insert except that one provides a table of values to be inserted, and removevalues which removes a range of values.

asserteq(insertvalues({1,2,3,4},2,{20,30}),{1,20,30,2,3,4})
asserteq(insertvalues({1,2},{3,4}),{1,2,3,4})

Another example:

> T = require 'pl.tablex'
> t = {10,20,30,40}
> = T.removevalues(t,2,3)
{10,40}
> = T.insertvalues(t,2,{20,30})
{10,20,30,40}

In a similar spirit to deepcopy, deepcompare will take two tables and return true only if they have exactly the same values and structure.

> t1 = {1,{2,3},4}
> t2 = deepcopy(t1)
> = t1 == t2
false
> = deepcompare(t1,t2)
true

find will return the index of a given value in a list-like table. Note that like string.find you can specify an index to start searching, so that all instances can be found. There is an optional fourth argument, which makes the search start at the end and go backwards, so we could define rfind like so:

function rfind(t,val,istart)
    return tablex.find(t,val,istart,true)
end

find does a linear search, so it can slow down code that depends on it. If efficiency is required for large tables, consider using an index map. index_map will return a table where the keys are the original values of the list, and the associated values are the indices. (It is almost exactly the representation needed for a set.)

> t = {'one','two','three'}
> = tablex.find(t,'two')
2
> = tablex.find(t,'four')
nil
> il = tablex.index_map(t)
> = il['two']
2
> = il.two
2

A version of index_map called makeset is also provided, where the values are just true. This is useful because two such sets can be compared for equality using deepcompare:

> = deepcompare(makeset {1,2,3},makeset {2,1,3})
true

Consider the problem of determining the new employees that have joined in a period. Assume we have two files of employee names:

(last-month.txt)
smith,john
brady,maureen
mongale,thabo

(this-month.txt)
smith,john
smit,johan
brady,maureen
mogale,thabo
van der Merwe,Piet

To find out differences, just make the employee lists into sets, like so:

require 'pl'

function read_employees(file)
  local ls = List(io.lines(file)) -- a list of employees
  return tablex.makeset(ls)
end

last = read_employees 'last-month.txt'
this = read_employees 'this-month.txt'

-- who is in this but not in last?
diff = tablex.difference(this,last)

-- in a set, the keys are the values...
for e in pairs(diff) do print(e) end

--  *output*
-- van der Merwe,Piet
-- smit,johan

The difference operation is easy to write and read:

for e in pairs(this) do
  if not last[e] then
    print(e)
  end
end

Using difference here is not that it is a tricky thing to code, it is that you are stating your intentions clearly to other readers of your code. (And naturally to your future self, in six months time.)

find_if will search a table using a function. The optional third argument is a value which will be passed as a second argument to the function. pl.operator provides the Lua operators conveniently wrapped as functions, so the basic comparison functions are available:

> ops = require 'pl.operator'
> = tablex.find_if({10,20,30,40},ops.gt,20)
3       true

Note that find_if will also return the actual value returned by the function, which of course is usually just true for a boolean function, but any value which is not nil and not false can be usefully passed back.

deepcompare does a thorough recursive comparison, but otherwise using the default equality operator. compare allows you to specify exactly what function to use when comparing two list-like tables, and compare_no_order is true if they contain exactly the same elements. Do note that the latter does not need an explicit comparison function - in this case the implementation is actually to compare the two sets, as above:

> = compare_no_order({1,2,3},{2,1,3})
true
> = compare_no_order({1,2,3},{2,1,3},'==')
true

(Note the special string '==' above; instead of saying ops.gt or ops.eq we can use the strings '>' or '==' respectively.)

sort and sortv return iterators that will iterate through the sorted elements of a table. sort iterates by sorted key order, and sortv iterates by sorted value order. For example, given a table with names and ages, it is trivial to iterate over the elements:

> t = {john=27,jane=31,mary=24}
> for name,age in tablex.sort(t) do print(name,age) end
jane    31
john    27
mary    24
> for name,age in tablex.sortv(t) do print(name,age) end
mary    24
john    27
jane    31

There are several ways to merge tables in PL. If they are list-like, then see the operations defined by pl.List, like concatenation. If they are map-like, then merge provides two basic operations. If the third arg is false, then the result only contains the keys that are in common between the two tables, and if true, then the result contains all the keys of both tables. These are in fact generalized set union and intersection operations:

> S1 = {john=27,jane=31,mary=24}
> S2 = {jane=31,jones=50}
> = tablex.merge(S1, S2, false)
{jane=31}
> = tablex.merge(S1, S2, true)
{mary=24,jane=31,john=27,jones=50}

When working with tables, you will often find yourself writing loops like in the first example. Loops are second nature to programmers, but they are often not the most elegant and self-describing way of expressing an operation. Consider the map function, which creates a new table by applying a function to each element of the original:

> = map(math.sin, {1,2,3,4})
{  0.84,  0.91,  0.14, -0.76}
> = map(function(x) return x*x end, {1,2,3,4})
{1,4,9,16}

map saves you from writing a loop, and the resulting code is often clearer, as well as being shorter. This is not to say that 'loops are bad' (although you will hear that from some extremists), just that it's good to capture standard patterns. Then the loops you do write will stand out and acquire more significance.

pairmap is interesting, because the function works with both the key and the value.

> t = {fred=10,bonzo=20,alice=4}
> = pairmap(function(k,v) return v end, t)
{4,10,20}
> = pairmap(function(k,v) return k end, t)
{'alice','fred','bonzo'}

(These are common enough operations that the first is defined as values and the second as keys.) If the function returns two values, then the second value is considered to be the new key:

> = pairmap(t,function(k,v) return v+10, k:upper() end)
{BONZO=30,FRED=20,ALICE=14}

map2 applies a function to two tables:

> map2(ops.add,{1,2},{10,20})
{11,22}
> map2('*',{1,2},{10,20})
{10,40}

The various map operations generate tables; reduce applies a function of two arguments over a table and returns the result as a scalar:

> reduce ('+', {1,2,3})
6
> reduce ('..', {'one','two','three'})
'onetwothree'

Finally, zip sews different tables together:

> = zip({1,2,3},{10,20,30})
{{1,10},{2,20},{3,30}}

Browsing through the documentation, you will find that tablex and List share methods. For instance, tablex.imap and List.map are basically the same function; they both operate over the array-part of the table and generate another table. This can also be expressed as a list comprehension C 'f(x) for x' (t) which makes the operation more explicit. So why are there different ways to do the same thing? The main reason is that not all tables are Lists: the expression ls:map('#') will return a list of the lengths of any elements of ls. A list is a thin wrapper around a table, provided by the metatable List. Sometimes you may wish to work with ordinary Lua tables; the List interface is not a compulsory way to use Penlight table operations.

Operations on two-dimensional tables

Two-dimensional tables are of course easy to represent in Lua, for instance {{1,2},{3,4}} where we store rows as subtables and index like so A[col][row]. This is the common representation used by matrix libraries like LuaMatrix. pl.array2d does not provide matrix operations, since that is the job for a specialized library, but rather provides generalizations of the higher-level operations provided by pl.tablex for one-dimensional arrays.

iter is a useful generalization of ipairs. (The extra parameter determines whether you want the indices as well.)

> a = {{1,2},{3,4}}
> for i,j,v in array2d.iter(a,true) do print(i,j,v) end
1       1       1
1       2       2
2       1       3
2       2       4

Note that you can always convert an arbitrary 2D array into a 'list of lists' with List(tablex.map(List,a))

map will apply a function over all elements (notice that extra arguments can be provided, so this operation is in effect function(x) return x-1 end)

> array2d.map('-',a,1)
{{0,1},{2,3}}

2D arrays are stored as an array of rows, but columns can be extracted:

> array2d.column(a,1)
{1,3}

There are three equivalents to tablex.reduce. You can either reduce along the rows (which is the most efficient) or reduce along the columns. Either one will give you a 1D array. And reduce2 will apply two operations: the first one reduces the rows, and the second reduces the result.

> array2d.reduce_rows('+',a)
{3,7}
> array2d.reduce_cols('+',a)
{4,6}
> -- same as tablex.reduce('*',array.reduce_rows('+',a))
> array2d.reduce2('*','+',a)
21    `

tablex.map2 applies an operation to two tables, giving another table. array2d.map2 does this for 2D arrays. Note that you have to provide the rank of the arrays involved, since it's hard to always correctly deduce this from the data:

> b = {{10,20},{30,40}}
> a = {{1,2},{3,4}}
> = array2d.map2('+',2,2,a,b)  -- two 2D arrays
{{11,22},{33,44}}
> = array2d.map2('+',1,2,{10,100},a)  -- 1D, 2D
{{11,102},{13,104}}
> = array2d.map2('*',2,1,a,{1,-1})  -- 2D, 1D
{{1,-2},{3,-4}}

Of course, you are not limited to simple arithmetic. Say we have a 2D array of strings, and wish to print it out with proper right justification. The first step is to create all the string lengths by mapping string.len over the array, the second is to reduce this along the columns using math.max to get maximum column widths, and last, apply stringx.rjust with these widths.

maxlens = reduce_cols(math.max,map('#',lines))
lines = map2(stringx.rjust,2,1,lines,maxlens)

There is product which returns the Cartesian product of two 1D arrays. The result is a 2D array formed from applying the function to all possible pairs from the two arrays.

> array2d.product('{}',{1,2},{'a','b'})
{{{1,'b'},{2,'a'}},{{1,'a'},{2,'b'}}}

There is a set of operations which work in-place on 2D arrays. You can swap_rows and swap_cols; the first really is a simple one-liner, but the idea here is to give the operation a name. remove_row and remove_col are generalizations of table.remove. Likewise, extract_rows and extract_cols are given arrays of indices and discard anything else. So, for instance, extract_cols(A,{2,4}) will leave just columns 2 and 4 in the array.

List.slice is often useful on 1D arrays; slice does the same thing, but is generally given a start (row,column) and a end (row,column).

> A = {{1,2,3},{4,5,6},{7,8,9}}
> B = slice(A,1,1,2,2)
> write(B)
 1 2
 4 5
> B = slice(A,2,2)
> write(B,nil,'%4.1f')
 5.0 6.0
 8.0 9.0

Here write is used to print out an array nicely; the second parameter is nil, which is the default (stdout) but can be any file object and the third parameter is an optional format (as used in string.format).

parse_range will take a spreadsheet range like 'A1:B2' or 'R1C1:R2C2' and return the range as four numbers, which can be passed to slice. The rule is that slice will return an array of the appropriate shape depending on the range; if a range represents a row or a column, the result is 1D, otherwise 2D.

This applies to iter as well, which can also optionally be given a range:

> for i,j,v in iter(A,true,2,2) do print(i,j,v) end
2       2       5
2       3       6
3       2       8
3       3       9

new will construct a new 2D array with the given dimensions. You provide an initial value for the elements, which is interpreted as a function if it's callable. With L being utils.string_lambda we then have the following way to make an identity matrix:

asserteq(
    array.new(3,3,L'|i,j| i==j and 1 or 0'),
    {{1,0,0},{0,1,0},{0,0,1}}
)

Please note that most functions in array2d are covariant, that is, they return an array of the same type as they receive. In particular, any objects created with data.new or matrix.new will remain data or matrix objects when reshaped or sliced, etc. Data objects have the array2d functions available as methods.

generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/03-strings.md.html000066400000000000000000000410341416703176500207640ustar00rootroot00000000000000 Penlight Documentation

Strings. Higher-level operations on strings.

Extra String Methods

These are convenient borrowings from Python, as described in 3.6.1 of the Python reference, but note that indices in Lua always begin at one. stringx defines functions like isalpha and isdigit, which return true if s is only composed of letters or digits respectively. startswith and endswith are convenient ways to find substrings. (endswith works as in Python 2.5, so that `f:endswith {'.bat','.exe','.cmd'}` will be true for any filename which ends with these extensions.) There are justify methods and whitespace trimming functions like strip.

> stringx.import()
> ('bonzo.dog'):endswith {'.dog','.cat'}
true
> ('bonzo.txt'):endswith {'.dog','.cat'}
false
> ('bonzo.cat'):endswith {'.dog','.cat'}
true
> (' stuff'):ljust(20,'+')
'++++++++++++++ stuff'
> ('  stuff '):lstrip()
'stuff '
> ('  stuff '):rstrip()
'  stuff'
> ('  stuff '):strip()
'stuff'
> for s in ('one\ntwo\nthree\n'):lines() do print(s) end
one
two
three

Most of these can be fairly easily implemented using the Lua string library, which is more general and powerful. But they are convenient operations to have easily at hand. Note that can be injected into the string table if you use stringx.import, but a simple alias like local stringx = require 'pl.stringx' is preferrable. This is the recommended practice when writing modules for consumption by other people, since it is bad manners to change the global state of the rest of the system. Magic may be used for convenience, but there is always a price.

String Templates

Another borrowing from Python, string templates allow you to substitute values looked up in a table:

local Template = require ('pl.text').Template
t = Template('${here} is the $answer')
print(t:substitute {here = 'Lua', answer = 'best'})
==>
Lua is the best

'$ variables' can optionally have curly braces; this form is useful if you are glueing text together to make variables, e.g ${prefix}_name_${postfix}. The substitute method will throw an error if a $ variable is not found in the table, and the safe_substitute method will not.

The Lua implementation has an extra method, indent_substitute which is very useful for inserting blocks of text, because it adjusts indentation. Consider this example:

-- testtemplate.lua
local Template = require ('pl.text').Template

t = Template [[
    for i = 1,#$t do
        $body
    end
]]

body = Template [[
local row = $t[i]
for j = 1,#row do
    fun(row[j])
end
]]

print(t:indent_substitute {body=body,t='tbl'})

And the output is:

for i = 1,#tbl do
    local row = tbl[i]
    for j = 1,#row do
        fun(row[j])
    end
end

indent_substitute can substitute templates, and in which case they themselves will be substituted using the given table. So in this case, $t was substituted twice.

pl.text also has a number of useful functions like dedent, which strips all the initial indentation from a multiline string. As in Python, this is useful for preprocessing multiline strings if you like indenting them with your code. The function wrap is passed a long string (a paragraph) and returns a list of lines that fit into a desired line width. As an extension, there is also indent for indenting multiline strings.

New in Penlight with the 0.9 series is text.format_operator. Calling this enables Python-style string formating using the modulo operator %:

> text.format_operator()
> = '%s[%d]' % {'dog',1}
dog[1]

So in its simplest form it saves the typing involved with string.format; it will also expand $ variables using named fields:

> = '$animal[$num]' % {animal='dog',num=1}
dog[1]

As with stringx.import you have to do this explicitly, since all strings share the same metatable. But in your own scripts you can feel free to do this.

Another Style of Template

A new module is template, which is a version of Rici Lake's Lua Preprocessor. This allows you to mix Lua code with your templates in a straightforward way. There are only two rules:

  • Lines begining with # are Lua
  • Otherwise, anything inside $() is a Lua expression.

So a template generating an HTML list would look like this:

<ul>
# for i,val in ipairs(T) do
<li>$(i) = $(val:upper())</li>
# end
</ul>

Assume the text is inside tmpl, then the template can be expanded using:

local template = require 'pl.template'
local my_env = {
  ipairs = ipairs,
  T = {'one','two','three'}
}
res = template.substitute(tmpl, my_env)

and we get

<ul>
<li>1 = ONE</li>
<li>2 = TWO</li>
<li>3 = THREE</li>
</ul>

There is a single function, template.substitute which is passed a template string and an environment table. This table may contain some special fields, like \_parent which can be set to a table representing a 'fallback' environment in case a symbol was not found. \_brackets is usually '()' and \_escape is usually '#' but it's sometimes necessary to redefine these if the defaults interfere with the target language - for instance, $(V) has another meaning in Make, and # means a preprocessor line in C/C++.

Finally, if something goes wrong, passing _debug will cause the intermediate Lua code to be dumped if there's a problem.

Here is a C code generation example; something that could easily be extended to be a minimal Lua extension skeleton generator.

local subst = require 'pl.template'.substitute

local templ = [[
#include <lua.h>
#include <lauxlib.h>
#include <lualib.h>

> for _,f in ipairs(mod) do
static int l_$(f.name) (lua_State *L) {

}
> end

static const luaL_reg $(mod.name)[] = {
> for _,f in ipairs(mod) do
    {"$(f.name)",l_$(f.name)},
> end
    {NULL,NULL}
};

int luaopen_$(mod.name) {
   luaL_register (L, "$(mod.name)", $(mod.name));
    return 1;
}
]]

print(subst(templ,{
    _escape = '>',
    ipairs = ipairs,
    mod = {
        name = 'baggins';
        {name='frodo'},
        {name='bilbo'}
    }
}))

File-style I/O on Strings

pl.stringio provides just three functions; stringio.open is passed a string, and returns a file-like object for reading. It supports a read method, which takes the same arguments as standard file objects:

> f = stringio.open 'first line\n10 20 30\n'
> = f:read()
first line
> = f:read('*n','*n','*n')
10    20    30

lines and seek are also supported.

stringio.lines is a useful short-cut for iterating over all the lines in a string.

stringio.create creates a writeable file-like object. You then use write to this stream, and finally extract the builded string using value. This 'string builder' pattern is useful for efficiently creating large strings.

generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/04-paths.md.html000066400000000000000000000373251416703176500204230ustar00rootroot00000000000000 Penlight Documentation

Paths and Directories

Working with Paths

Programs should not depend on quirks of your operating system. They will be harder to read, and need to be ported for other systems. The worst of course is hardcoding paths like 'c:\' in programs, and wondering why Vista complains so much. But even something like dir..'\'..file is a problem, since Unix can't understand backslashes in this way. dir..'/'..file is usually portable, but it's best to put this all into a simple function, path.join. If you consistently use path.join, then it's much easier to write cross-platform code, since it handles the directory separator for you.

pl.path provides the same functionality as Python's os.path module (11.1).

> p = 'c:\\bonzo\\DOG.txt'
> = path.normcase (p)  ---> only makes sense on Windows
c:\bonzo\dog.txt
> = path.splitext (p)
c:\bonzo\DOG    .txt
> = path.extension (p)
.txt
> = path.basename (p)
DOG.txt
> = path.exists(p)
false
> = path.join ('fred','alice.txt')
fred\alice.txt
> = path.exists 'pretty.lua'
true
> = path.getsize 'pretty.lua'
2125
> = path.isfile 'pretty.lua'
true
> = path.isdir 'pretty.lua'
false

It is very important for all programmers, not just on Unix, to only write to where they are allowed to write. path.expanduser will expand '~' (tilde) into the home directory. Depending on your OS, this will be a guaranteed place where you can create files:

> = path.expanduser '~/mydata.txt'
'C:\Documents and Settings\SJDonova/mydata.txt'

> = path.expanduser '~/mydata.txt'
/home/sdonovan/mydata.txt

Under Windows, os.tmpname returns a path which leads to your drive root full of temporary files. (And increasingly, you do not have access to this root folder.) This is corrected by path.tmpname, which uses the environment variable TMP:

> os.tmpname()  -- not a good place to put temporary files!
'\s25g.'
> path.tmpname()
'C:\DOCUME~1\SJDonova\LOCALS~1\Temp\s25g.1'

A useful extra function is pl.path.package_path, which will tell you the path of a particular Lua module. So on my system, package_path('pl.path') returns 'C:\Program Files\Lua\5.1\lualibs\pl\path.lua', and package_path('ifs') returns 'C:\Program Files\Lua\5.1\clibs\lfs.dll'. It is implemented in terms of package.searchpath, which is a new function in Lua 5.2 which has been implemented for Lua 5.1 in Penlight.

File Operations

pl.file is a new module that provides more sensible names for common file operations. For instance, file.read and file.write are aliases for utils.readfile and utils.writefile.

Smaller files can be efficiently read and written in one operation. file.read is passed a filename and returns the contents as a string, if successful; if not, then it returns nil and the actual error message. There is an optional boolean parameter if you want the file to be read in binary mode (this makes no difference on Unix but remains important with Windows.)

In previous versions of Penlight, utils.readfile would read standard input if the file was not specified, but this can lead to nasty bugs; use io.read '*a' to grab all of standard input.

Similarly, file.write takes a filename and a string which will be written to that file.

For example, this little script converts a file into upper case:

require 'pl'
assert(#arg == 2, 'supply two filenames')
text = assert(file.read(arg[1]))
assert(file.write(arg[2],text:upper()))

Copying files is suprisingly tricky. file.copy and file.move attempt to use the best implementation possible. On Windows, they link to the API functions CopyFile and MoveFile, but only if the alien package is installed (this is true for Lua for Windows.) Otherwise, the system copy command is used. This can be ugly when writing Windows GUI applications, because of the dreaded flashing black-box problem with launching processes.

Directory Operations

pl.dir provides some useful functions for working with directories. fnmatch will match a filename against a shell pattern, and filter will return any files in the supplied list which match the given pattern, which correspond to the functions in the Python fnmatch module. getdirectories will return all directories contained in a directory, and getfiles will return all files in a directory which match a shell pattern. These functions return the files as a table, unlike lfs.dir which returns an iterator.)

dir.makepath can create a full path, creating subdirectories as necessary; rmtree is the Nuclear Option of file deleting functions, since it will recursively clear out and delete all directories found begining at a path (there is a similar function with this name in the Python shutils module.)

> = dir.makepath 't\\temp\\bonzo'
> = path.isdir 't\\temp\\bonzo'
true
> = dir.rmtree 't'

dir.rmtree depends on dir.walk, which is a powerful tool for scanning a whole directory tree. Here is the implementation of dir.rmtree:

--- remove a whole directory tree.
-- @param path A directory path
function dir.rmtree(fullpath)
    for root,dirs,files in dir.walk(fullpath) do
        for i,f in ipairs(files) do
            os.remove(path.join(root,f))
        end
        lfs.rmdir(root)
    end
end

dir.clonetree clones directory trees. The first argument is a path that must exist, and the second path is the path to be cloned. (Note that this path cannot be inside the first path, since this leads to madness.) By default, it will then just recreate the directory structure. You can in addition provide a function, which will be applied for all files found.

-- make a copy of my libs folder
require 'pl'
p1 = [[d:\dev\lua\libs]]
p2 = [[D:\dev\lua\libs\..\tests]]
dir.clonetree(p1,p2,dir.copyfile)

A more sophisticated version, which only copies files which have been modified:

-- p1 and p2 as before, or from arg[1] and arg[2]
dir.clonetree(p1,p2,function(f1,f2)
  local res
  local t1,t2 = path.getmtime(f1),path.getmtime(f2)
  -- f2 might not exist, so be careful about t2
  if not t2 or t1 > t2 then
    res = dir.copyfile(f1,f2)
  end
  return res -- indicates successful operation
end)

dir.clonetree uses path.common_prefix. With p1 and p2 defined above, the common path is 'd:\dev\lua'. So 'd:\dev\lua\libs\testfunc.lua' is copied to 'd:\dev\lua\test\testfunc.lua', etc.

If you need to find the common path of list of files, then tablex.reduce will do the job:

> p3 = [[d:\dev]]
> = tablex.reduce(path.common_prefix,{p1,p2,p3})
'd:\dev'
generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/05-dates.md.html000066400000000000000000000335661416703176500204100ustar00rootroot00000000000000 Penlight Documentation

Date and Time

NOTE: the Date module is deprecated

Creating and Displaying Dates

The Date class provides a simplified way to work with date and time in Lua; it leans heavily on the functions os.date and os.time.

A Date object can be constructed from a table, just like with os.time. Methods are provided to get and set the various parts of the date.

> d = Date {year = 2011, month = 3, day = 2 }
> = d
2011-03-02 12:00:00
> = d:month(),d:year(),d:day()
3    2011    2
> d:month(4)
> = d
2011-04-02 12:00:00
> d:add {day=1}
> = d
2011-04-03 12:00:00

add takes a table containing one of the date table fields.

> = d:weekday_name()
Sun
> = d:last_day()
2011-04-30 12:00:00
> = d:month_name(true)
April

There is a default conversion to text for date objects, but Date.Format gives you full control of the format for both parsing and displaying dates:

> iso = Date.Format 'yyyy-mm-dd'
> d = iso:parse '2010-04-10'
> amer = Date.Format 'mm/dd/yyyy'
> = amer:tostring(d)
04/10/2010

With the 0.9.7 relase, the Date constructor has become more flexible. You may omit any of the 'year', 'month' or 'day' fields:

> = Date { year = 2008 }
2008-01-01 12:00:00
> = Date { month = 3 }
2011-03-01 12:00:00
> = Date { day = 20 }
2011-10-20 12:00:00
> = Date { hour = 14, min = 30 }
2011-10-13 14:30:00

If 'year' is omitted, then the current year is assumed, and likewise for 'month'.

To set the time on such a partial date, you can use the fact that the 'setter' methods return the date object and so you can 'chain' these methods.

> d = Date { day = 03 }
> = d:hour(18):min(30)
2011-10-03 18:30:00

Finally, Date also now accepts positional arguments:

> = Date(2011,10,3)
2011-10-03 12:00:00
> = Date(2011,10,3,18,30,23)
2011-10-03 18:30:23

Date.format has been extended. If you construct an instance without a pattern, then it will try to match against a set of known formats. This is useful for human-input dates since keeping to a strict format is not one of the strong points of users. It assumes that there will be a date, and then a date.

> df = Date.Format()
> = df:parse '5.30pm'
2011-10-13 17:30:00
> = df:parse '1730'
nil     day out of range: 1730 is not between 1 and 31
> = df:parse '17.30'
2011-10-13 17:30:00
> = df:parse 'mar'
2011-03-01 12:00:00
> = df:parse '3 March'
2011-03-03 12:00:00
> = df:parse '15 March'
2011-03-15 12:00:00
> = df:parse '15 March 2008'
2008-03-15 12:00:00
> = df:parse '15 March 2008 1.30pm'
2008-03-15 13:30:00
> = df:parse '2008-10-03 15:30:23'
2008-10-03 15:30:23

ISO date format is of course a good idea if you need to deal with users from different countries. Here is the default behaviour for 'short' dates:

> = df:parse '24/02/12'
2012-02-24 12:00:00

That's not what Americans expect! It's tricky to work out in a cross-platform way exactly what the expected format is, so there is an explicit flag:

> df:US_order(true)
> = df:parse '9/11/01'
2001-11-09 12:00:00
generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/06-data.md.html000066400000000000000000002143721416703176500202160ustar00rootroot00000000000000 Penlight Documentation

Data

Reading Data Files

The first thing to consider is this: do you actually need to write a custom file reader? And if the answer is yes, the next question is: can you write the reader in as clear a way as possible? Correctness, Robustness, and Speed; pick the first two and the third can be sorted out later, if necessary.

A common sort of data file is the configuration file format commonly used on Unix systems. This format is often called a property file in the Java world.

# Read timeout in seconds
read.timeout=10

# Write timeout in seconds
write.timeout=10

Here is a simple Lua implementation:

-- property file parsing with Lua string patterns
props = []
for line in io.lines() do
    if line:find('#',1,true) ~= 1 and not line:find('^%s*$') then
        local var,value = line:match('([^=]+)=(.*)')
        props[var] = value
    end
end

Very compact, but it suffers from a similar disease in equivalent Perl programs; it uses odd string patterns which are 'lexically noisy'. Noisy code like this slows the casual reader down. (For an even more direct way of doing this, see the next section, 'Reading Configuration Files')

Another implementation, using the Penlight libraries:

-- property file parsing with extended string functions
require 'pl'
stringx.import()
props = []
for line in io.lines() do
    if not line:startswith('#') and not line:isspace() then
        local var,value = line:splitv('=')
        props[var] = value
    end
end

This is more self-documenting; it is generally better to make the code express the intention, rather than having to scatter comments everywhere - comments are necessary, of course, but mostly to give the higher view of your intention that cannot be expressed in code. It is slightly slower, true, but in practice the speed of this script is determined by I/O, so further optimization is unnecessary.

Reading Unstructured Text Data

Text data is sometimes unstructured, for example a file containing words. The pl.input module has a number of functions which makes processing such files easier. For example, a script to count the number of words in standard input using import.words:

-- countwords.lua
require 'pl'
local k = 1
for w in input.words(io.stdin) do
    k = k + 1
end
print('count',k)

Or this script to calculate the average of a set of numbers using input.numbers:

-- average.lua
require 'pl'
local k = 1
local sum = 0
for n in input.numbers(io.stdin) do
    sum = sum + n
    k = k + 1
end
print('average',sum/k)

These scripts can be improved further by eliminating loops In the last case, there is a perfectly good function seq.sum which can already take a sequence of numbers and calculate these numbers for us:

-- average2.lua
require 'pl'
local total,n = seq.sum(input.numbers())
print('average',total/n)

A further simplification here is that if numbers or words are not passed an argument, they will grab their input from standard input. The first script can be rewritten:

-- countwords2.lua
require 'pl'
print('count',seq.count(input.words()))

A useful feature of a sequence generator like numbers is that it can read from a string source. Here is a script to calculate the sums of the numbers on each line in a file:

-- sums.lua
for line in io.lines() do
    print(seq.sum(input.numbers(line))
end

Reading Columnar Data

It is very common to find data in columnar form, either space or comma-separated, perhaps with an initial set of column headers. Here is a typical example:

EventID    Magnitude    LocationX    LocationY    LocationZ
981124001    2.0    18988.4    10047.1    4149.7
981125001    0.8    19104.0    9970.4    5088.7
981127003    0.5    19012.5    9946.9    3831.2
...

input.fields is designed to extract several columns, given some delimiter (default to whitespace). Here is a script to calculate the average X location of all the events:

-- avg-x.lua
require 'pl'
io.read() -- skip the header line
local sum,count = seq.sum(input.fields {3})
print(sum/count)

input.fields is passed either a field count, or a list of column indices, starting at one as usual. So in this case we're only interested in column 3. If you pass it a field count, then you get every field up to that count:

for id,mag,locX,locY,locZ in input.fields (5) do
....
end

input.fields by default tries to convert each field to a number. It will skip lines which clearly don't match the pattern, but will abort the script if there are any fields which cannot be converted to numbers.

The second parameter is a delimiter, by default spaces. ' ' is understood to mean 'any number of spaces', i.e. '%s+'. Any Lua string pattern can be used.

The third parameter is a data source, by default standard input (defined by input.create_getter.) It assumes that the data source has a read method which brings in the next line, i.e. it is a 'file-like' object. As a special case, a string will be split into its lines:

> for x,y in input.fields(2,' ','10 20\n30 40\n') do print(x,y) end
10      20
30      40

Note the default behaviour for bad fields, which is to show the offending line number:

> for x,y in input.fields(2,' ','10 20\n30 40x\n') do print(x,y) end
10      20
line 2: cannot convert '40x' to number

This behaviour of input.fields is appropriate for a script which you want to fail immediately with an appropriate user error message if conversion fails. The fourth optional parameter is an options table: {no_fail=true} means that conversion is attempted but if it fails it just returns the string, rather as AWK would operate. You are then responsible for checking the type of the returned field. {no_convert=true} switches off conversion altogether and all fields are returned as strings.

Sometimes it is useful to bring a whole dataset into memory, for operations such as extracting columns. Penlight provides a flexible reader specifically for reading this kind of data, using the data module. Given a file looking like this:

x,y
10,20
2,5
40,50

Then data.read will create a table like this, with each row represented by a sublist:

> t = data.read 'test.txt'
> pretty.dump(t)
{{10,20},{2,5},{40,50},fieldnames={'x','y'},delim=','}

You can now analyze this returned table using the supplied methods. For instance, the method column_by_name returns a table of all the values of that column.

-- testdata.lua
require 'pl'
d = data.read('fev.txt')
for _,name in ipairs(d.fieldnames) do
    local col = d:column_by_name(name)
    if type(col[1]) == 'number' then
        local total,n = seq.sum(col)
        utils.printf("Average for %s is %f\n",name,total/n)
    end
end

data.read tries to be clever when given data; by default it expects a first line of column names, unless any of them are numbers. It tries to deduce the column delimiter by looking at the first line. Sometimes it guesses wrong; these things can be specified explicitly. The second optional parameter is an options table: can override delim (a string pattern), fieldnames (a list or comma-separated string), specify no_convert (default is to convert), numfields (indices of columns known to be numbers, as a list) and thousands_dot (when the thousands separator in Excel CSV is '.')

A very powerful feature is a way to execute SQL-like queries on such data:

-- queries on tabular data
require 'pl'
local d = data.read('xyz.txt')
local q = d:select('x,y,z where x > 3 and z < 2 sort by y')
for x,y,z in q do
    print(x,y,z)
end

Please note that the format of queries is restricted to the following syntax:

FIELDLIST [ 'where' CONDITION ] [ 'sort by' FIELD [asc|desc]]

Any valid Lua code can appear in CONDITION; remember it is not SQL and you have to use == (this warning comes from experience.)

For this to work, field names must be Lua identifiers. So read will massage fieldnames so that all non-alphanumeric chars are replaced with underscores. However, the original_fieldnames field always contains the original un-massaged fieldnames.

read can handle standard CSV files fine, although doesn't try to be a full-blown CSV parser. With the csv=true option, it's possible to have double-quoted fields, which may contain commas; then trailing commas become significant as well.

Spreadsheet programs are not always the best tool to process such data, strange as this might seem to some people. This is a toy CSV file; to appreciate the problem, imagine thousands of rows and dozens of columns like this:

Department Name,Employee ID,Project,Hours Booked
sales,1231,overhead,4
sales,1255,overhead,3
engineering,1501,development,5
engineering,1501,maintenance,3
engineering,1433,maintenance,10

The task is to reduce the dataset to a relevant set of rows and columns, perhaps do some processing on row data, and write the result out to a new CSV file. The write_row method uses the delimiter to write the row to a file; Data.select_row is like Data.select, except it iterates over rows, not fields; this is necessary if we are dealing with a lot of columns!

names = {[1501]='don',[1433]='dilbert'}
keepcols = {'Employee_ID','Hours_Booked'}
t:write_row (outf,{'Employee','Hours_Booked'})
q = t:select_row {
    fields=keepcols,
    where=function(row) return row[1]=='engineering' end
}
for row in q do
    row[1] = names[row[1]]
    t:write_row(outf,row)
end

Data.select_row and Data.select can be passed a table specifying the query; a list of field names, a function defining the condition and an optional parameter sort_by. It isn't really necessary here, but if we had a more complicated row condition (such as belonging to a specified set) then it is not generally possible to express such a condition as a query string, without resorting to hackery such as global variables.

With 1.0.3, you can specify explicit conversion functions for selected columns. For instance, this is a log file with a Unix date stamp:

Time Message
1266840760 +# EE7C0600006F0D00C00F06010302054000000308010A00002B00407B00
1266840760 closure data 0.000000 1972 1972 0
1266840760 ++ 1266840760 EE 1
1266840760 +# EE7C0600006F0D00C00F06010302054000000408020A00002B00407B00
1266840764 closure data 0.000000 1972 1972 0

We would like the first column as an actual date object, so the convert field sets an explicit conversion for column 1. (Note that we have to explicitly convert the string to a number first.)

Date = require 'pl.Date'

function date_convert (ds)
    return Date(tonumber(ds))
end

d = data.read(f,{convert={[1]=date_convert},last_field_collect=true})

This gives us a two-column dataset, where the first column contains Date objects and the second column contains the rest of the line. Queries can then easily pick out events on a day of the week:

q = d:select "Time,Message where Time:weekday_name()=='Sun'"

Data does not have to come from files, nor does it necessarily come from the lab or the accounts department. On Linux, ps aux gives you a full listing of all processes running on your machine. It is straightforward to feed the output of this command into data.read and perform useful queries on it. Notice that non-identifier characters like '%' get converted into underscores:

require 'pl'
f = io.popen 'ps aux'
s = data.read (f,{last_field_collect=true})
f:close()
print(s.fieldnames)
print(s:column_by_name 'USER')
qs = 'COMMAND,_MEM where _MEM > 5 and USER=="steve"'
for name,mem in s:select(qs) do
    print(mem,name)
end

I've always been an admirer of the AWK programming language; with filter you can get Lua programs which are just as compact:

-- printxy.lua
require 'pl'
data.filter 'x,y where x > 3'

It is common enough to have data files without headers of field names. data.read makes a special exception for such files if all fields are numeric. Since there are no column names to use in query expressions, you can use AWK-like column indexes, e.g. '$1,$2 where $1 > 3'. I have a little executable script on my system called lf which looks like this:

#!/usr/bin/env lua
require 'pl.data'.filter(arg[1])

And it can be used generally as a filter command to extract columns from data. (The column specifications may be expressions or even constants.)

$ lf '$1,$5/10' < test.dat

(As with AWK, please note the single-quotes used in this command; this prevents the shell trying to expand the column indexes. If you are on Windows, then you must quote the expression in double-quotes so it is passed as one argument to your batch file.)

As a tutorial resource, have a look at test-data.lua in the PL tests directory for other examples of use, plus comments.

The data returned by read or constructed by Data.copy_select from a query is basically just an array of rows: {{1,2},{3,4}}. So you may use read to pull in any array-like dataset, and process with any function that expects such a implementation. In particular, the functions in array2d will work fine with this data. In fact, these functions are available as methods; e.g. array2d.flatten can be called directly like so to give us a one-dimensional list:

v = data.read('dat.txt'):flatten()

The data is also in exactly the right shape to be treated as matrices by LuaMatrix:

> matrix = require 'matrix'
> m = matrix(data.read 'mat.txt')
> = m
1       0.2     0.3
0.2     1       0.1
0.1     0.2     1
> = m^2  -- same as m*m
1.07    0.46    0.62
0.41    1.06    0.26
0.24    0.42    1.05

write will write matrices back to files for you.

Finally, for the curious, the global variable _DEBUG can be used to print out the actual iterator function which a query generates and dynamically compiles. By using code generation, we can get pretty much optimal performance out of arbitrary queries.

> lua -lpl -e "_DEBUG=true" -e "data.filter 'x,y where x > 4 sort by x'" < test.txt
return function (t)
        local i = 0
        local v
        local ls = {}
        for i,v in ipairs(t) do
            if v[1] > 4  then
                    ls[#ls+1] = v
            end
        end
        table.sort(ls,function(v1,v2)
            return v1[1] < v2[1]
        end)
        local n = #ls
        return function()
            i = i + 1
            v = ls[i]
            if i > n then return end
            return v[1],v[2]
        end
end

10,20
40,50

Reading Configuration Files

The config module provides a simple way to convert several kinds of configuration files into a Lua table. Consider the simple example:

# test.config
# Read timeout in seconds
read.timeout=10

# Write timeout in seconds
write.timeout=5

#acceptable ports
ports = 1002,1003,1004

This can be easily brought in using config.read and the result shown using pretty.write:

-- readconfig.lua
local config = require 'pl.config'
local pretty= require 'pl.pretty'

local t = config.read(arg[1])
print(pretty.write(t))

and the output of lua readconfig.lua test.config is:

{
  ports = {
    1002,
    1003,
    1004
  },
  write_timeout = 5,
  read_timeout = 10
}

That is, config.read will bring in all key/value pairs, ignore # comments, and ensure that the key names are proper Lua identifiers by replacing non-identifier characters with '_'. If the values are numbers, then they will be converted. (So the value of t.write_timeout is the number 5). In addition, any values which are separated by commas will be converted likewise into an array.

Any line can be continued with a backslash. So this will all be considered one line:

names=one,two,three, \
four,five,six,seven, \
eight,nine,ten

Windows-style INI files are also supported. The section structure of INI files translates naturally to nested tables in Lua:

; test.ini
[timeouts]
read=10 ; Read timeout in seconds
write=5 ; Write timeout in seconds
[portinfo]
ports = 1002,1003,1004

The output is:

{
  portinfo = {
    ports = {
      1002,
      1003,
      1004
    }
  },
  timeouts = {
    write = 5,
    read = 10
  }
}

You can now refer to the write timeout as t.timeouts.write.

As a final example of the flexibility of config.read, if passed this simple comma-delimited file

one,two,three
10,20,30
40,50,60
1,2,3

it will produce the following table:

{
  { "one", "two", "three" },
  { 10, 20, 30 },
  { 40, 50, 60  },
  { 1, 2, 3 }
}

config.read isn't designed to read all CSV files in general, but intended to support some Unix configuration files not structured as key-value pairs, such as '/etc/passwd'.

This function is intended to be a Swiss Army Knife of configuration readers, but it does have to make assumptions, and you may not like them. So there is an optional extra parameter which allows some control, which is table that may have the following fields:

{
   variablilize = true,
   convert_numbers = tonumber,
   trim_space = true,
   list_delim = ',',
   trim_quotes = true,
   ignore_assign = false,
   keysep = '=',
   smart = false,
}

variablilize is the option that converted write.timeout in the first example to the valid Lua identifier write_timeout. If convert_numbers is true, then an attempt is made to convert any string that starts like a number. You can specify your own function (say one that will convert a string like '5224 kb' into a number.)

trim_space ensures that there is no starting or trailing whitespace with values, and list_delim is the character that will be used to decide whether to split a value up into a list (it may be a Lua string pattern such as '%s+'.)

For instance, the password file in Unix is colon-delimited:

t = config.read('/etc/passwd',{list_delim=':'})

This produces the following output on my system (only last two lines shown):

{
  ...
  {
    "user",
    "x",
    "1000",
    "1000",
    "user,,,",
    "/home/user",
    "/bin/bash"
  },
  {
    "sdonovan",
    "x",
    "1001",
    "1001",
    "steve donovan,28,,",
    "/home/sdonovan",
    "/bin/bash"
  }
}

You can get this into a more sensible format, where the usernames are the keys, with this (the tablex.pairmap function must return value, key!)

t = tablex.pairmap(function(k,v) return v,v[1] end,t)

and you get:

{ ...
  sdonovan = {
    "sdonovan",
    "x",
    "1001",
    "1001",
    "steve donovan,28,,",
    "/home/sdonovan",
    "/bin/bash"
  }
...
}

Many common Unix configuration files can be read by tweaking these parameters. For /etc/fstab, the options {list_delim='%s+',ignore_assign=true} will correctly separate the columns. It's common to find 'KEY VALUE' assignments in files such as /etc/ssh/ssh_config; the options {keysep=' '} make config.read return a table where each KEY has a value VALUE.

Files in the Linux procfs usually use ':` as the field delimiter:

> t = config.read('/proc/meminfo',{keysep=':'})
> = t.MemFree
220140 kB

That result is a string, since tonumber doesn't like it, but defining the convert_numbers option as `function(s) return tonumber((s:gsub(' kB$',''))) end` will get the memory figures as actual numbers in the result. (The extra parentheses are necessary so that tonumber only gets the first result from gsub). From `tests/test-config.lua':

testconfig([[
MemTotal:        1024748 kB
MemFree:          220292 kB
]],
{ MemTotal = 1024748, MemFree = 220292 },
{
 keysep = ':',
 convert_numbers = function(s)
    s = s:gsub(' kB$','')
    return tonumber(s)
  end
 }
)

The smart option lets config.read make a reasonable guess for you; there are examples in tests/test-config.lua, but basically these common file formats (and those following the same pattern) can be processed directly in smart mode: 'etc/fstab', '/proc/XXXX/status', 'ssh_config' and 'pdatedb.conf'.

Please note that config.read can be passed a file-like object; if it's not a string and supports the read method, then that will be used. For instance, to read a configuration from a string, use stringio.open.

Lexical Scanning

Although Lua's string pattern matching is very powerful, there are times when something more powerful is needed. pl.lexer.scan provides lexical scanners which tokenize a string, classifying tokens into numbers, strings, etc.

> lua -lpl
Lua 5.1.4  Copyright (C) 1994-2008 Lua.org, PUC-Rio
> tok = lexer.scan 'alpha = sin(1.5)'
> = tok()
iden    alpha
> = tok()
=       =
> = tok()
iden    sin
> = tok()
(       (
> = tok()
number  1.5
> = tok()
)       )
> = tok()
(nil)

The scanner is a function, which is repeatedly called and returns the type and value of the token. Recognized basic types are 'iden','string','number', and 'space'. and everything else is represented by itself. Note that by default the scanner will skip any 'space' tokens.

'comment' and 'keyword' aren't applicable to the plain scanner, which is not language-specific, but a scanner which understands Lua is available. It recognizes the Lua keywords, and understands both short and long comments and strings.

> for t,v in lexer.lua 'for i=1,n do' do print(t,v) end
keyword for
iden    i
=       =
number  1
,       ,
iden    n
keyword do

A lexical scanner is useful where you have highly-structured data which is not nicely delimited by newlines. For example, here is a snippet of a in-house file format which it was my task to maintain:

points
    (818344.1,-20389.7,-0.1),(818337.9,-20389.3,-0.1),(818332.5,-20387.8,-0.1)
    ,(818327.4,-20388,-0.1),(818322,-20387.7,-0.1),(818316.3,-20388.6,-0.1)
    ,(818309.7,-20389.4,-0.1),(818303.5,-20390.6,-0.1),(818295.8,-20388.3,-0.1)
    ,(818290.5,-20386.9,-0.1),(818285.2,-20386.1,-0.1),(818279.3,-20383.6,-0.1)
    ,(818274,-20381.2,-0.1),(818274,-20380.7,-0.1);

Here is code to extract the points using pl.lexer:

-- assume 's' contains the text above...
local lexer = require 'pl.lexer'
local expecting = lexer.expecting
local append = table.insert

local tok = lexer.scan(s)

local points = {}
local t,v = tok() -- should be 'iden','points'

while t ~= ';' do
    c = {}
    expecting(tok,'(')
    c.x = expecting(tok,'number')
    expecting(tok,',')
    c.y = expecting(tok,'number')
    expecting(tok,',')
    c.z = expecting(tok,'number')
    expecting(tok,')')
    t,v = tok()  -- either ',' or ';'
    append(points,c)
end

The expecting function grabs the next token and if the type doesn't match, it throws an error. (pl.lexer, unlike other PL libraries, raises errors if something goes wrong, so you should wrap your code in pcall to catch the error gracefully.)

The scanners all have a second optional argument, which is a table which controls whether you want to exclude spaces and/or comments. The default for lexer.lua is {space=true,comments=true}. There is a third optional argument which determines how string and number tokens are to be processsed.

The ultimate highly-structured data is of course, program source. Here is a snippet from 'text-lexer.lua':

require 'pl'

lines = [[
for k,v in pairs(t) do
    if type(k) == 'number' then
        print(v) -- array-like case
    else
        print(k,v)
    end
end
]]

ls = List()
for tp,val in lexer.lua(lines,{space=true,comments=true}) do
    assert(tp ~= 'space' and tp ~= 'comment')
    if tp == 'keyword' then ls:append(val) end
end
test.asserteq(ls,List{'for','in','do','if','then','else','end','end'})

Here is a useful little utility that identifies all common global variables found in a lua module (ignoring those declared locally for the moment):

-- testglobal.lua
require 'pl'

local txt,err = utils.readfile(arg[1])
if not txt then return print(err) end

local globals = List()
for t,v in lexer.lua(txt) do
    if t == 'iden' and _G[v] then
        globals:append(v)
    end
end
pretty.dump(seq.count_map(globals))

Rather then dumping the whole list, with its duplicates, we pass it through seq.count_map which turns the list into a table where the keys are the values, and the associated values are the number of times those values occur in the sequence. Typical output looks like this:

{
  type = 2,
  pairs = 2,
  table = 2,
  print = 3,
  tostring = 2,
  require = 1,
  ipairs = 4
}

You could further pass this through tablex.keys to get a unique list of symbols. This can be useful when writing 'strict' Lua modules, where all global symbols must be defined as locals at the top of the file.

For a more detailed use of lexer.scan, please look at testxml.lua in the examples directory.

XML

New in the 0.9.7 release is some support for XML. This is a large topic, and Penlight does not provide a full XML stack, which is properly the task of a more specialized library.

Parsing and Pretty-Printing

The semi-standard XML parser in the Lua universe is lua-expat. In particular, it has a function called lxp.lom.parse which will parse XML into the Lua Object Model (LOM) format. However, it does not provide a way to convert this data back into XML text. xml.parse will use this function, if lua-expat is available, and otherwise switches back to a pure Lua parser originally written by Roberto Ierusalimschy.

The resulting document object knows how to render itself as a string, which is useful for debugging:

> d = xml.parse "<nodes><node id='1'>alice</node></nodes>"
> = d
<nodes><node id='1'>alice</node></nodes>
> pretty.dump (d)
{
  {
    "alice",
    attr = {
      "id",
      id = "1"
    },
    tag = "node"
  },
  attr = {
  },
  tag = "nodes"
}

Looking at the actual shape of the data reveals the structure of LOM:

  • every element has a tag field with its name
  • plus a attr field which is a table containing the attributes as fields, and also as an array. It is always present.
  • the children of the element are the array part of the element, so d[1] is the first child of d, etc.

It could be argued that having attributes also as the array part of attr is not essential (you cannot depend on attribute order in XML) but that's how it goes with this standard.

lua-expat is another soft dependency of Penlight; generally, the fallback parser is good enough for straightforward XML as is commonly found in configuration files, etc. doc.basic_parse is not intended to be a proper conforming parser (it's only sixty lines) but it handles simple kinds of documents that do not have comments or DTD directives. It is intelligent enough to ignore the <?xml directive and that is about it.

You can get pretty-printing by explicitly calling xml.tostring and passing it the initial indent and the per-element indent:

> = xml.tostring(d,'','  ')

<nodes>
  <node id='1'>alice</node>
</nodes>

There is a fourth argument which is the attribute indent:

> a = xml.parse "<frodo name='baggins' age='50' type='hobbit'/>"
> = xml.tostring(a,'','  ','  ')

<frodo
  type='hobbit'
  name='baggins'
  age='50'
/>

Parsing and Working with Configuration Files

It's common to find configurations expressed with XML these days. It's straightforward to 'walk' the LOM data and extract the data in the form you want:

require 'pl'

local config = [[
<config>
    <alpha>1.3</alpha>
    <beta>10</beta>
    <name>bozo</name>
</config>
]]
local d,err = xml.parse(config)

local t = {}
for item in d:childtags() do
    t[item.tag] = item[1]
end

pretty.dump(t)
--->
{
  beta = "10",
  alpha = "1.3",
  name = "bozo"
}

The only gotcha is that here we must use the Doc:childtags method, which will skip over any text elements.

A more involved example is this excerpt from serviceproviders.xml, which is usually found at /usr/share/mobile-broadband-provider-info/serviceproviders.xml on Debian/Ubuntu Linux systems.

d = xml.parse [[
<serviceproviders format="2.0">
...
<country code="za">
    <provider>
        <name>Cell-c</name>
        <gsm>
            <network-id mcc="655" mnc="07"/>
            <apn value="internet">
                <username>Cellcis</username>
                <dns>196.7.0.138</dns>
                <dns>196.7.142.132</dns>
            </apn>
        </gsm>
    </provider>
    <provider>
        <name>MTN</name>
        <gsm>
            <network-id mcc="655" mnc="10"/>
            <apn value="internet">
                <dns>196.11.240.241</dns>
                <dns>209.212.97.1</dns>
            </apn>
        </gsm>
    </provider>
    <provider>
        <name>Vodacom</name>
        <gsm>
            <network-id mcc="655" mnc="01"/>
            <apn value="internet">
                <dns>196.207.40.165</dns>
                <dns>196.43.46.190</dns>
            </apn>
            <apn value="unrestricted">
                <name>Unrestricted</name>
                <dns>196.207.32.69</dns>
                <dns>196.43.45.190</dns>
            </apn>
        </gsm>
    </provider>
    <provider>
        <name>Virgin Mobile</name>
        <gsm>
            <apn value="vdata">
                <dns>196.7.0.138</dns>
                <dns>196.7.142.132</dns>
            </apn>
        </gsm>
    </provider>
</country>
....
</serviceproviders>
]]

Getting the names of the providers per-country is straightforward:

local t = {}
for country in d:childtags() do
    local providers = {}
    t[country.attr.code] = providers
    for provider in country:childtags() do
        table.insert(providers,provider:child_with_name('name'):get_text())
    end
end

pretty.dump(t)
-->
{
  za = {
    "Cell-c",
    "MTN",
    "Vodacom",
    "Virgin Mobile"
  }
  ....
}

Generating XML with 'xmlification'

This feature is inspired by the htmlify function used by Orbit to simplify HTML generation, except that no function environment magic is used; the tags function returns a set of constructors for elements of the given tag names.

> nodes, node = xml.tags 'nodes, node'
> = node 'alice'
<node>alice</node>
> = nodes { node {id='1','alice'}}
<nodes><node id='1'>alice</node></nodes>

The flexibility of Lua tables is very useful here, since both the attributes and the children of an element can be encoded naturally. The argument to these tag constructors is either a single value (like a string) or a table where the attributes are the named keys and the children are the array values.

Generating XML using Templates

A template is a little XML document which contains dollar-variables. The subst method on a document is fed an array of tables containing values for these variables. Note how the parent tag name is specified:

> templ = xml.parse "<node id='$id'>$name</node>"
> = templ:subst {tag='nodes', {id=1,name='alice'},{id=2,name='john'}}
<nodes><node id='1'>alice</node><node id='2'>john</node></nodes>

Substitution is very related to filtering documents. One of the annoying things about XML is that it is a document markup language first, and a data language second. Standard parsers will assume you really care about all those extra text elements. Consider this fragment, which has been changed by a five-year old:

T = [[
  <weather>
    boops!
    <current_conditions>
      <condition data='$condition'/>
      <temp_c data='$temp'/>
      <bo>whoops!</bo>
    </current_conditions>
  </weather>
]]

Conformant parsers will give you text elements with the line feed after <current_conditions> although it makes handling the data more irritating.

local function parse (str)
    return xml.parse(str,false,true)
end

Second argument means 'string, not file' and third argument means use the built-in Lua parser (instead of LuaExpat if available) which by default is not interested in keeping such strings.

How to remove the string boops!? clone (also called filter when called as a method) copies a LOM document. It can be passed a filter function, which is applied to each string found. The powerful thing about this is that this function receives structural information - the parent node, and whether this was a tag name, a text element or a attribute name:

d = parse (T)
c = d:filter(function(s,kind,parent)
    print(stringx.strip(s),kind,parent and parent.tag or '?')
    if kind == '*TEXT' and #parent > 1 then return nil end
    return s
end)
--->
weather    *TAG    ?
boops!    *TEXT    weather
current_conditions    *TAG    weather
condition    *TAG    current_conditions
$condition    data    condition
temp_c    *TAG    current_conditions
$temp    data    temp_c
bo    *TAG    current_conditions
whoops!    *TEXT    bo

We can pull out 'boops' and not 'whoops' by discarding text elements which are not the single child of an element.

Extracting Data using Templates

Matching goes in the opposite direction. We have a document, and would like to extract values from it using a pattern.

A common use of this is parsing the XML result of API queries. The (undocumented and subsequently discontinued) Google Weather API is a good example. Grabbing the result of `http://www.google.com/ig/api?weather=Johannesburg,ZA" we get something like this, after pretty-printing:

<xml_api_reply version='1'>
  <weather module_id='0' tab_id='0' mobile_zipped='1' section='0' row='0'

mobile_row='0'>

<forecast_information>
  <city data='Johannesburg, Gauteng'/>
  <postal_code data='Johannesburg,ZA'/>
  <latitude_e6 data=''/>
  <longitude_e6 data=''/>
  <forecast_date data='2010-10-02'/>
  <current_date_time data='2010-10-02 18:30:00 +0000'/>
  <unit_system data='US'/>
</forecast_information>
<current_conditions>
  <condition data='Clear'/>
  <temp_f data='75'/>
  <temp_c data='24'/>
  <humidity data='Humidity: 19%'/>
  <icon data='/ig/images/weather/sunny.gif'/>
  <wind_condition data='Wind: NW at 7 mph'/>
</current_conditions>
<forecast_conditions>
  <day_of_week data='Sat'/>
  <low data='60'/>
  <high data='89'/>
  <icon data='/ig/images/weather/sunny.gif'/>
  <condition data='Clear'/>
</forecast_conditions>
....
/weather>
l_api_reply>

Assume that the above XML has been read into google. The idea is to write a pattern looking like a template, and use it to extract some values of interest:

t = [[
  <weather>
    <current_conditions>
      <condition data='$condition'/>
      <temp_c data='$temp'/>
    </current_conditions>
  </weather>
]]

local res, ret = google:match(t)
pretty.dump(res)

And the output is:

{
  condition = "Clear",
  temp = "24"
}

The match method can be passed a LOM document or some text, which will be parsed first.

But what if we need to extract values from repeated elements? Match templates may contain 'array matches' which are enclosed in '{{..}}':

<weather>
  {{<forecast_conditions>
    <day_of_week data='$day'/>
    <low data='$low'/>
    <high data='$high'/>
    <condition data='$condition'/>
  </forecast_conditions>}}
</weather>

And the match result is:

{
  {
    low = "60",
    high = "89",
    day = "Sat",
    condition = "Clear",
  },
  {
    low = "53",
    high = "86",
    day = "Sun",
    condition = "Clear",
  },
  {
    low = "57",
    high = "87",
    day = "Mon",
    condition = "Clear",
  },
  {
    low = "60",
    high = "84",
    day = "Tue",
    condition = "Clear",
  }
}

With this array of tables, you can use tablex or List to reshape into the desired form, if you choose. Just as with reading a Unix password file with config, you can make the array into a map of days to conditions using:

tablex.pairmap('|k,v| v,v.day',conditions)

(Here using the alternative string lambda option)

However, xml matches can shape the structure of the output. By replacing the day_of_week line of the template with <day_of_week data='$_'/> we get the same effect; $_ is a special symbol that means that this captured value (or simply capture) becomes the key.

Note that $NUMBER means a numerical index, so that $1 is the first element of the resulting array, and so forth. You can mix numbered and named captures, but it's strongly advised to make the numbered captures form a proper array sequence (everything from 1 to n inclusive). $0 has a special meaning; if it is the only capture ({[0]='foo'}) then the table is collapsed into 'foo'.

<weather>
  {{<forecast_conditions>
    <day_of_week data='$_'/>
    <low data='$1'/>
    <high data='$2'/>
    <condition data='$3'/>
  </forecast_conditions>}}
</weather>

Now the result is:

{
  Tue = {
    "60",
    "84",
    "Clear"
  },
  Sun = {
    "53",
    "86",
    "Clear"
  },
  Sat = {
    "60",
    "89",
    "Clear"
  },
  Mon = {
    "57",
    "87",
    "Clear"
  }
}

Applying matches to this config file poses another problem, because the actual tags matched are themselves meaningful.

<config>
    <alpha>1.3</alpha>
    <beta>10</beta>
    <name>bozo</name>
</config>

So there are tag 'wildcards' which are element names ending with a hyphen.

<config>
    {{<key->$value</key->}}
</config>

You will then get {{alpha='1.3'},...}. The most convenient format would be returned by this (note that _- behaves just like $_):

<config>
    {{<_->$0</_->}}
</config>

which would return {alpha='1.3',beta='10',name='bozo'}.

We could play this game endlessly, and encode ways of converting captures, but the scheme is complex enough, and it's easy to do the conversion later

local numbers = {alpha=true,beta=true}
for k,v in pairs(res) do
    if numbers[v] then res[k] = tonumber(v) end
end

HTML Parsing

HTML is an unusually degenerate form of XML, and Dennis Schridde has contributed a feature which makes parsing it easier. For instance, from the tests:

doc = xml.parsehtml [[
<BODY>
Hello dolly<br>
HTML is <b>slack</b><br>
</BODY>
]]

asserteq(xml.tostring(doc),[[
<body>
Hello dolly<br/>
HTML is <b>slack</b><br/></body>]])

That is, all tags are converted to lowercase, and empty HTML elements like br are properly closed; attributes do not need to be quoted.

Also, DOCTYPE directives and comments are skipped. For truly badly formed HTML, this is not the tool for you!

generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/07-functional.md.html000066400000000000000000001311701416703176500214420ustar00rootroot00000000000000 Penlight Documentation

Functional Programming

Sequences

A Lua iterator (in its simplest form) is a function which can be repeatedly called to return a set of one or more values. The for in statement understands these iterators, and loops until the function returns nil. There are standard sequence adapters for tables in Lua (ipairs and pairs), and io.lines returns an iterator over all the lines in a file. In the Penlight libraries, such iterators are also called sequences. A sequence of single values (say from io.lines) is called single-valued, whereas the sequence defined by pairs is double-valued.

pl.seq provides a number of useful iterators, and some functions which operate on sequences. At first sight this example looks like an attempt to write Python in Lua, (with the sequence being inclusive):

> for i in seq.range(1,4) do print(i) end
1
2
3
4

But range is actually equivalent to Python's xrange, since it generates a sequence, not a list. To get a list, use seq.copy(seq.range(1,10)), which takes any single-value sequence and makes a table from the result. seq.list is like ipairs except that it does not give you the index, just the value.

> for x in seq.list {1,2,3} do print(x) end
1
2
3

enum takes a sequence and turns it into a double-valued sequence consisting of a sequence number and the value, so enum(list(ls)) is actually equivalent to ipairs. A more interesting example prints out a file with line numbers:

for i,v in seq.enum(io.lines(fname)) do print(i..' '..v) end

Sequences can be combined, either by 'zipping' them or by concatenating them.

> for x,y in seq.zip(l1,l2) do print(x,y) end
10      1
20      2
30      3
> for x in seq.splice(l1,l2) do print(x) end
10
20
30
1
2
3

seq.printall is useful for printing out single-valued sequences, and provides some finer control over formating, such as a delimiter, the number of fields per line, and a format string to use (@see string.format)

> seq.printall(seq.random(10))
0.0012512588885159 0.56358531449324 0.19330423902097 ....
> seq.printall(seq.random(10), ',', 4, '%4.2f')
0.17,0.86,0.71,0.51
0.30,0.01,0.09,0.36
0.15,0.17,

map will apply a function to a sequence.

> seq.printall(seq.map(string.upper, {'one','two'}))
ONE TWO
> seq.printall(seq.map('+', {10,20,30}, 1))
11 21 31

filter will filter a sequence using a boolean function (often called a predicate). For instance, this code only prints lines in a file which are composed of digits:

for l in seq.filter(io.lines(file), stringx.isdigit) do print(l) end

The following returns a table consisting of all the positive values in the original table (equivalent to tablex.filter(ls, '>', 0))

ls = seq.copy(seq.filter(ls, '>', 0))

We're already encounted seq.sum when discussing input.numbers. This can also be expressed with seq.reduce:

> seq.reduce(function(x,y) return x + y end, seq.list{1,2,3,4})
10

seq.reduce applies a binary function in a recursive fashion, so that:

reduce(op,{1,2,3}) => op(1,reduce(op,{2,3}) => op(1,op(2,3))

it's now possible to easily generate other cumulative operations; the standard operations declared in pl.operator are useful here:

> ops = require 'pl.operator'
> -- can also say '*' instead of ops.mul
> = seq.reduce(ops.mul,input.numbers '1 2 3 4')
24

There are functions to extract statistics from a sequence of numbers:

> l1 = List {10,20,30}
> l2 = List {1,2,3}
> = seq.minmax(l1)
10      30
> = seq.sum(l1)
60      3

It is common to get sequences where values are repeated, say the words in a file. count_map will take such a sequence and count the values, returning a table where the keys are the unique values, and the value associated with each key is the number of times they occurred:

> t = seq.count_map {'one','fred','two','one','two','two'}
> = t
{one=2,fred=1,two=3}

This will also work on numerical sequences, but you cannot expect the result to be a proper list, i.e. having no 'holes'. Instead, you always need to use pairs to iterate over the result - note that there is a hole at index 5:

> t = seq.count_map {1,2,4,2,2,3,4,2,6}
> for k,v in pairs(t) do print(k,v) end
1       1
2       4
3       1
4       2
6       1

unique uses count_map to return a list of the unique values, that is, just the keys of the resulting table.

last turns a single-valued sequence into a double-valued sequence with the current value and the last value:

> for current,last in seq.last {10,20,30,40} do print (current,last) end
20      10
30      20
40      30

This makes it easy to do things like identify repeated lines in a file, or construct differences between values. filter can handle double-valued sequences as well, so one could filter such a sequence to only return cases where the current value is less than the last value by using operator.lt or just '<'. This code then copies the resulting code into a table.

> ls = {10,9,10,3}
> = seq.copy(seq.filter(seq.last(s),'<'))
{9,3}

Sequence Wrappers

The functions in pl.seq cover the common patterns when dealing with sequences, but chaining these functions together can lead to ugly code. Consider the last example of the previous section; seq is repeated three times and the resulting expression has to be read right-to-left. The first issue can be helped by local aliases, so that the expression becomes copy(filter(last(s),'<')) but the second issue refers to the somewhat unnatural order of functional application. We tend to prefer reading operations from left to right, which is one reason why object-oriented notation has become popular. Sequence adapters allow this expression to be written like so:

seq(s):last():filter('<'):copy()

With this notation, the operation becomes a chain of method calls running from left to right.

'Sequence' is not a basic Lua type, they are generally functions or callable objects. The expression seq(s) wraps a sequence in a sequence wrapper, which is an object which understands all the functions in pl.seq as methods. This object then explicitly represents sequences.

As a special case, the constructor (which is when you call the table seq) will make a wrapper for a plain list-like table. Here we apply the length operator to a sequence of strings, and print them out.

> seq{'one','tw','t'} :map '#' :printall()
3 2 1

As a convenience, there is a function seq.lines which behaves just like io.lines except it wraps the result as an explicit sequence type. This takes the first 10 lines from standard input, makes it uppercase, turns it into a sequence with a count and the value, glues these together with the concatenation operator, and finally prints out the sequence delimited by a newline.

seq.lines():take(10):upper():enum():map('..'):printall '\n'

Note the method upper, which is not a seq function. if an unknown method is called, sequence wrappers apply that method to all the values in the sequence (this is implicit use of mapmethod)

It is straightforward to create custom sequences that can be used in this way. On Unix, /dev/random gives you an endless sequence of random bytes, so we use take to limit the sequence, and then map to scale the result into the desired range. The key step is to use seq to wrap the iterator function:

-- random.lua
local seq = require 'pl.seq'

function dev_random()
    local f = io.open('/dev/random')
    local byte = string.byte
    return seq(function()
        -- read two bytes into a string and convert into a 16-bit number
        local s = f:read(2)
        return byte(s,1) + 256*byte(s,2)
    end)
end

-- print 10 random numbers from 0 to 1 !
dev_random():take(10):map('%',100):map('/',100):printall ','

Another Linux one-liner depends on the /proc filesystem and makes a list of all the currently running processes:

pids = seq(lfs.dir '/proc'):filter(stringx.isdigit):map(tonumber):copy()

This version of Penlight has an experimental feature which relies on the fact that all Lua types can have metatables, including functions. This makes implicit sequence wrapping possible:

> seq.import()
> seq.random(5):printall(',',5,'%4.1f')
 0.0, 0.1, 0.4, 0.1, 0.2

This avoids the awkward seq(seq.random(5)) construction. Or the iterator can come from somewhere else completely:

> ('one two three'):gfind('%a+'):printall(',')
one,two,three,

After seq.import, it is no longer necessary to explicitly wrap sequence functions.

But there is a price to pay for this convenience. Every function is affected, so that any function can be used, appropriate or not:

> math.sin:printall()
..seq.lua:287: bad argument #1 to '(for generator)' (number expected, got nil)
> a = tostring
> = a:find(' ')
function: 0042C920

What function is returned? It's almost certain to be something that makes no sense in the current context. So implicit sequences may make certain kinds of programming mistakes harder to catch - they are best used for interactive exploration and small scripts.

List Comprehensions

List comprehensions are a compact way to create tables by specifying their elements. In Python, you can say this:

ls = [x for x in range(5)]  # == [0,1,2,3,4]

In Lua, using pl.comprehension:

> C = require('pl.comprehension').new()
> = C ('x for x=1,10') ()
{1,2,3,4,5,6,7,8,9,10}

C is a function which compiles a list comprehension string into a function. In this case, the function has no arguments. The parentheses are redundant for a function taking a string argument, so this works as well:

> = C 'x^2 for x=1,4' ()
{1,4,9,16}
> = C '{x,x^2} for x=1,4' ()
{{1,1},{2,4},{3,9},{4,16}}

Note that the expression can be any function of the variable x!

The basic syntax so far is <expr> for <set>, where <set> can be anything that the Lua for statement understands. <set> can also just be the variable, in which case the values will come from the argument of the comprehension. Here I'm emphasizing that a comprehension is a function which can take a list argument:

> = C '2*x for x' {1,2,3}
{2,4,6}
> dbl = C '2*x for x'
> = dbl {10,20,30}
{20,40,60}

Here is a somewhat more explicit way of saying the same thing; _1 is a placeholder refering to the first argument passed to the comprehension.

> = C '2*x for _,x in pairs(_1)' {10,20,30}
{20,40,60}
> = C '_1(x) for x'(tostring,{1,2,3,4})
{'1','2','3','4'}

This extended syntax is useful when you wish to collect the result of some iterator, such as io.lines. This comprehension creates a function which creates a table of all the lines in a file:

> f = io.open('array.lua')
> lines = C 'line for line in _1:lines()' (f)
> = #lines
118

There are a number of functions that may be applied to the result of a comprehension:

> = C 'min(x for x)' {1,44,0}
0
> = C 'max(x for x)' {1,44,0}
44
> = C 'sum(x for x)' {1,44,0}
45

(These are equivalent to a reduce operation on a list.)

After the for part, there may be a condition, which filters the output. This comprehension collects the even numbers from a list:

> = C 'x for x if x % 2 == 0' {1,2,3,4,5}
{2,4}

There may be a number of for parts:

> = C '{x,y} for x = 1,2 for y = 1,2' ()
{{1,1},{1,2},{2,1},{2,2}}
> = C '{x,y} for x for y' ({1,2},{10,20})
{{1,10},{1,20},{2,10},{2,20}}

These comprehensions are useful when dealing with functions of more than one variable, and are not so easily achieved with the other Penlight functional forms.

Creating Functions from Functions

Lua functions may be treated like any other value, although of course you cannot multiply or add them. One operation that makes sense is function composition, which chains function calls (so (f * g)(x) is f(g(x)).)

> func = require 'pl.func'
> printf = func.compose(io.write,string.format)
> printf("hello %s\n",'world')
hello world
true

Many functions require you to pass a function as an argument, say to apply to all values of a sequence or as a callback. Often useful functions have the wrong number of arguments. So there is a need to construct a function of one argument from one of two arguments, binding the extra argument to a given value.

partial application takes a function of n arguments and returns a function of n-1 arguments where the first argument is bound to some value:

> p2 = func.bind1(print,'start>')
> p2('hello',2)
start>  hello   2
> ops = require 'pl.operator'
> = tablex.filter({1,-2,10,-1,2},bind1(ops.gt,0))
{-2,-1}
> tablex.filter({1,-2,10,-1,2},bind1(ops.le,0))
{1,10,2}

The last example unfortunately reads backwards, because bind1 alway binds the first argument! Also unfortunately, in my youth I confused 'currying' with 'partial application', so the old name for bind1 is curry - this alias still exists.

This is a specialized form of function argument binding. Here is another way to say the print example:

> p2 = func.bind(print,'start>',func._1,func._2)
> p2('hello',2)
start>  hello   2

where _1 and _2 are placeholder variables, corresponding to the first and second argument respectively.

Having func all over the place is distracting, so it's useful to pull all of pl.func into the local context. Here is the filter example, this time the right way around:

> utils.import 'pl.func'
> tablex.filter({1,-2,10,-1,2},bind(ops.gt, _1, 0))
{1,10,2}

tablex.merge does a general merge of two tables. This example shows the usefulness of binding the last argument of a function.

> S1 = {john=27, jane=31, mary=24}
> S2 = {jane=31, jones=50}
> intersection = bind(tablex.merge, _1, _2, false)
> union = bind(tablex.merge, _1, _2, true)
> = intersection(S1,S2)
{jane=31}
> = union(S1,S2)
{mary=24,jane=31,john=27,jones=50}

When using bind with print, we got a function of precisely two arguments, whereas we really want our function to use varargs like print. This is the role of _0:

> _DEBUG = true
> p = bind(print,'start>', _0)
return function (fn,_v1)
    return function(...) return fn(_v1,...) end
end

> p(1,2,3,4,5)
start>  1       2       3       4       5

I've turned on the global _DEBUG flag, so that the function generated is printed out. It is actually a function which generates the required function; the first call binds the value of _v1 to 'start>'.

Placeholder Expressions

A common pattern in Penlight is a function which applies another function to all elements in a table or a sequence, such as tablex.map or seq.filter. Lua does anonymous functions well, although they can be a bit tedious to type:

> = tablex.map(function(x) return x*x end, {1,2,3,4})
{1,4,9,16}

pl.func allows you to define placeholder expressions, which can cut down on the typing required, and also make your intent clearer. First, we bring contents of pl.func into our context, and then supply an expression using placeholder variables, such as _1,_2,etc. (C++ programmers will recognize this from the Boost libraries.)

> utils.import 'pl.func'
> = tablex.map(_1*_1, {1,2,3,4})
{1,4,9,16}

Functions of up to 5 arguments can be generated.

> = tablex.map2(_1+_2,{1,2,3}, {10,20,30})
{11,22,33}

These expressions can use arbitrary functions, altho they must first be registered with the functional library. func.register brings in a single function, and func.import brings in a whole table of functions, such as math.

> sin = register(math.sin)
> = tablex.map(sin(_1), {1,2,3,4})
{0.8414709848079,0.90929742682568,0.14112000805987,-0.75680249530793}
> import 'math'
> = tablex.map(cos(2*_1),{1,2,3,4})
{-0.41614683654714,-0.65364362086361,0.96017028665037,-0.14550003380861}

A common operation is calling a method of a set of objects:

> = tablex.map(_1:sub(1,1), {'one','four','x'})
{'o','f','x'}

There are some restrictions on what operators can be used in PEs. For instance, because the __len metamethod cannot be overriden by plain Lua tables, we need to define a special function to express `#_1':

> = tablex.map(Len(_1), {'one','four','x'})
{3,4,1}

Likewise for comparison operators, which cannot be overloaded for different types, and thus also have to be expressed as a special function:

> = tablex.filter(Gt(_1,0), {1,-1,2,4,-3})
{1,2,4}

It is useful to express the fact that a function returns multiple values. For instance, tablex.pairmap expects a function that will be called with the key and the value, and returns the new value and the key, in that order.

> = pairmap(Args(_2,_1:upper()),{fred=1,alice=2})
{ALICE=2,FRED=1}

PEs cannot contain nil values, since PE function arguments are represented as an array. Instead, a special value called Nil is provided. So say _1:f(Nil,1) instead of _1:f(nil,1).

A placeholder expression cannot be automatically used as a Lua function. The technical reason is that the call operator must be overloaded to construct function calls like _1(1). If you want to force a PE to return a function, use func.I.

> = tablex.map(_1(10),{I(2*_1),I(_1*_1),I(_1+2)})
{20,100,12}

Here we make a table of functions taking a single argument, and then call them all with a value of 10.

The essential idea with PEs is to 'quote' an expression so that it is not immediately evaluated, but instead turned into a function that can be applied later to some arguments. The basic mechanism is to wrap values and placeholders so that the usual Lua operators have the effect of building up an expression tree. (It turns out that you can do symbolic algebra using PEs, see symbols.lua in the examples directory, and its test runner testsym.lua, which demonstrates symbolic differentiation.)

The rule is that if any operator has a PE operand, the result will be quoted. Sometimes we need to quote things explicitly. For instance, say we want to pass a function to a filter that must return true if the element value is in a set. set[_1] is the obvious expression, but it does not give the desired result, since it evaluates directly, giving nil. Indexing works differently than a binary operation like addition (set+_1 is properly quoted) so there is a need for an explicit quoting or wrapping operation. This is the job of the _ function; the PE in this case should be _(set)[_1]. This works for functions as well, as a convenient alternative to registering functions: _(math.sin)(_1). This is equivalent to using the `lines' method:

for line in I(_(f):read()) do print(line) end

Now this will work for any 'file-like' object which which has a read method returning the next line. If you had a LuaSocket client which was being 'pushed' by lines sent from a server, then _(s):receive '*l' would create an iterator for accepting input. These forms can be convenient for adapting your data flow so that it can be passed to the sequence functions in `pl.seq'.

Placeholder expressions can be mixed with sequence wrapper expressions. lexer.lua will give us a double-valued sequence of tokens, where the first value is a type, and the second is a value. We filter out only the values where the type is 'iden', extract the actual value using map, get the unique values and finally copy to a list.

> str = 'for i=1,10 do for j = 1,10 do print(i,j) end end'
> = seq(lexer.lua(str)):filter('==','iden'):map(_2):unique():copy()
{i,print,j}

This is a particularly intense line (and I don't always suggest making everything a one-liner!); the key is the behaviour of map, which will take both values of the sequence, so _2 returns the value part. (Since filter here takes extra arguments, it only operates on the type values.)

There are some performance considerations to using placeholder expressions. Instantiating a PE requires constructing and compiling a function, which is not such a fast operation. So to get best performance, factor out PEs from loops like this;

local fn = I(_1:f() + _2:g())
for i = 1,n do
    res[i] = tablex.map2(fn,first[i],second[i])
end
generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/08-additional.md.html000066400000000000000000001067411416703176500214170ustar00rootroot00000000000000 Penlight Documentation

Additional Libraries

Libraries in this section are no longer considered to be part of the Penlight core, but still provide specialized functionality when needed.

Simple Input Patterns

Lua string pattern matching is very powerful, and usually you will not need a traditional regular expression library. Even so, sometimes Lua code ends up looking like Perl, which happens because string patterns are not always the easiest things to read, especially for the casual reader. Here is a program which needs to understand three distinct date formats:

-- parsing dates using Lua string patterns
months={Jan=1,Feb=2,Mar=3,Apr=4,May=5,Jun=6,
Jul=7,Aug=8,Sep=9,Oct=10,Nov=11,Dec=12}

function check_and_process(d,m,y)
    d = tonumber(d)
    m = tonumber(m)
    y = tonumber(y)
    ....
end

for line in f:lines() do
    -- ordinary (English) date format
    local d,m,y = line:match('(%d+)/(%d+)/(%d+)')
    if d then
        check_and_process(d,m,y)
    else -- ISO date??
        y,m,d = line:match('(%d+)%-(%d+)%-(%d+)')
        if y then
            check_and_process(d,m,y)
        else -- <day> <month-name> <year>?
            d,mm,y = line:match('%(d+)%s+(%a+)%s+(%d+)')
            m = months[mm]
            check_and_process(d,m,y)
        end
    end
end

These aren't particularly difficult patterns, but already typical issues are appearing, such as having to escape '-'. Also, string.match returns its captures, so that we're forced to use a slightly awkward nested if-statement.

Verification issues will further cloud the picture, since regular expression people try to enforce constraints (like year cannot be more than four digits) using regular expressions, on the usual grounds that you shouldn't stop using a hammer when you are enjoying yourself.

pl.sip provides a simple, intuitive way to detect patterns in strings and extract relevant parts.

> sip = require 'pl.sip'
> dump = require('pl.pretty').dump
> res = {}
> c = sip.compile 'ref=$S{file}:$d{line}'
> = c('ref=hello.c:10',res)
true
> dump(res)
{
  line = 10,
  file = "hello.c"
}
> = c('ref=long name, no line',res)
false

sip.compile creates a pattern matcher function, which takes a string and a table as arguments. If the string matches the pattern, then true is returned and the table is populated according to the captures within the pattern.

Here is another version of the date parser:

-- using SIP patterns
function check(t)
    check_and_process(t.day,t.month,t.year)
end

shortdate = sip.compile('$d{day}/$d{month}/$d{year}')
longdate = sip.compile('$d{day} $v{mon} $d{year}')
isodate = sip.compile('$d{year}-$d{month}-$d{day}')

for line in f:lines() do
    local res = {}
    if shortdate(str,res) then
        check(res)
    elseif isodate(str,res) then
        check(res)
    elseif longdate(str,res) then
        res.month = months[res.mon]
        check(res)
    end
end

SIP captures start with '$', then a one-character type, and then an optional variable name in curly braces.

Type      Meaning
v         identifier
i         possibly signed integer
f         floating-point number
r         rest of line
q         quoted string (quoted using either ' or ")
p         a path name
(         anything inside balanced parentheses
[         anything inside balanced brackets
{         anything inside balanced curly brackets
<         anything inside balanced angle brackets

If a type is not one of the above, then it's assumed to be one of the standard Lua character classes, and will match one or more repetitions of that class. Any spaces you leave in your pattern will match any number of spaces, including zero, unless the spaces are between two identifier characters or patterns matching them; in that case, at least one space will be matched.

SIP captures (like $v{mon}) do not have to be named. You can use just $v, but you have to be consistent; if a pattern contains unnamed captures, then all captures must be unnamed. In this case, the result table is a simple list of values.

sip.match is a useful shortcut if you want to compile and match in one call, without saving the compiled pattern. It caches the result, so it is not much slower than explicitly using sip.compile.

> sip.match('($q{first},$q{second})','("john","smith")',res)
true
> res
{second='smith',first='john'}
> res = {}
> sip.match('($q,$q)','("jan","smit")',res)  -- unnamed captures
true
> res
{'jan','smit'}
> sip.match('($q,$q)','("jan", "smit")',res)
false   ---> oops! Can't handle extra space!
> sip.match('( $q , $q )','("jan", "smit")',res)
true

As a general rule, allow for whitespace in your patterns.

Finally, putting a '$' at the end of a pattern means 'capture the rest of the line, starting at the first non-space'. It is a shortcut for '$r{rest}', or just '$r' if no named captures are used.

> sip.match('( $q , $q ) $','("jan", "smit") and a string',res)
true
> res
{'jan','smit','and a string'}
> res = {}
> sip.match('( $q{first} , $q{last} ) $','("jan", "smit") and a string',res)
true
> res
{first='jan',rest='and a string',last='smit'}

Command-line Programs with Lapp

pl.lapp is a small and focused Lua module which aims to make standard command-line parsing easier and intuitive. It implements the standard GNU style, i.e. short flags with one letter start with '-', and there may be an additional long flag which starts with '--'. Generally options which take an argument expect to find it as the next parameter (e.g. 'gcc test.c -o test') but single short options taking a value can dispense with the space (e.g. 'head -n4 test.c' or gcc -I/usr/include/lua/5.1 ...)

As far as possible, Lapp will convert parameters into their equivalent Lua types, i.e. convert numbers and convert filenames into file objects. If any conversion fails, or a required parameter is missing, an error will be issued and the usage text will be written out. So there are two necessary tasks, supplying the flag and option names and associating them with a type.

For any non-trivial script, even for personal consumption, it's necessary to supply usage text. The novelty of Lapp is that it starts from that point and defines a loose format for usage strings which can specify the names and types of the parameters.

An example will make this clearer:

-- scale.lua
  lapp = require 'pl.lapp'
  local args = lapp [[
  Does some calculations
    -o,--offset (default 0.0)  Offset to add to scaled number
    -s,--scale  (number)  Scaling factor
    <number> (number)  Number to be scaled
  ]]

  print(args.offset + args.scale * args.number)

Here is a command-line session using this script:

$ lua scale.lua
scale.lua:missing required parameter: scale

Does some calculations
 -o,--offset (default 0.0)  Offset to add to scaled number
 -s,--scale  (number)  Scaling factor
  <number> (number )  Number to be scaled

$ lua scale.lua -s 2.2 10
22

$ lua scale.lua -s 2.2 x10
scale.lua:unable to convert to number: x10

....(usage as before)

There are two kinds of lines in Lapp usage strings which are meaningful; option and parameter lines. An option line gives the short option, optionally followed by the corresponding long option. A type specifier in parentheses may follow. Similarly, a parameter line starts with '', followed by a type specifier.

Type specifiers usually start with a type name: one of 'boolean', 'string','number','file-in' or 'file-out'. You may leave this out, but then must say 'default' followed by a value. If a flag or parameter has a default, it is not required and is set to the default. The actual type is deduced from this value (number, string, file or boolean) if not provided directly. 'Deduce' is a fancy word for 'guess' and it can be wrong, e.g '(default 1)' will always be a number. You can say '(string default 1)' to override the guess. There are file values for the predefined console streams: stdin, stdout, stderr.

The boolean type is the default for flags. Not providing the type specifier is equivalent to '(boolean default false)`. If the flag is meant to be 'turned off' then either the full '(boolean default true)` or the shortcut '(default true)' will work.

An alternative to default is optional:

local lapp = require 'pl.lapp'
local args = lapp [[
   --cmd (optional string) Command to run.
]]

if args.cmd then
  os.execute(args.cmd)
end

Here we're implying that cmd need not be specified (just as with default) but if not present, then args.cmd is nil, which will always test false.

The rest of the line is ignored and can be used for explanatory text.

This script shows the relation between the specified parameter names and the fields in the output table.

-- simple.lua
local args = require ('pl.lapp') [[
Various flags and option types
  -p          A simple optional flag, defaults to false
  -q,--quiet  A simple flag with long name
  -o  (string)  A required option with argument
  -s  (default 'save') Optional string with default 'save' (single quotes ignored)
  -n  (default 1) Optional numerical flag with default 1
  -b  (string default 1)  Optional string flag with default '1' (type explicit)
  <input> (default stdin)  Optional input file parameter, reads from stdin
]]

for k,v in pairs(args) do
    print(k,v)
end

I've just dumped out all values of the args table; note that args.quiet has become true, because it's specified; args.p defaults to false. If there is a long name for an option, that will be used in preference as a field name. A type or default specifier is not necessary for simple flags, since the default type is boolean.

$ simple -o test -q simple.lua
p       false
input   file (781C1BD8)
quiet   true
o       test
input_name      simple.lua
D:\dev\lua\lapp>simple -o test simple.lua one two three
1       one
2       two
3       three
p       false
quiet   false
input   file (781C1BD8)
o       test
input_name      simple.lua

The parameter input has been set to an open read-only file object - we know it must be a read-only file since that is the type of the default value. The field input_name is automatically generated, since it's often useful to have access to the original filename.

Notice that any extra parameters supplied will be put in the result table with integer indices, i.e. args[i] where i goes from 1 to #args.

Files don't really have to be closed explicitly for short scripts with a quick well-defined mission, since the result of garbage-collecting file objects is to close them.

Enforcing a Range and Enumerations

The type specifier can also be of the form '(' MIN '..' MAX ')' or a set of strings separated by '|'.

local lapp = require 'pl.lapp'
local args = lapp [[
    Setting ranges
    <x> (1..10)  A number from 1 to 10
    <y> (-5..1e6) Bigger range
    <z> (slow|medium|fast)
]]

print(args.x,args.y)

Here the meaning of ranges is that the value is greater or equal to MIN and less or equal to MAX. An 'enum' is a string that can only have values from a specified set.

Custom Types

There is no builti-in way to force a parameter to be a whole number, but you may define a custom type that does this:

lapp = require ('pl.lapp')

lapp.add_type('integer','number',
    function(x)
        lapp.assert(math.ceil(x) == x, 'not an integer!')
    end
)

local args =  lapp [[
    <ival> (integer) Process PID
]]

print(args.ival)

lapp.add_type takes three parameters, a type name, a converter and a constraint function. The constraint function is expected to throw an assertion if some condition is not true; we use lapp.assert because it fails in the standard way for a command-line script. The converter argument can either be a type name known to Lapp, or a function which takes a string and generates a value.

Here's a useful custom type that allows dates to be input as pl.Date values:

local df = Date.Format()

lapp.add_type('date',
    function(s)
        local d,e = df:parse(s)
        lapp.assert(d,e)
        return d
    end
)

'varargs' Parameter Arrays

lapp = require 'pl.lapp'
local args = lapp [[
Summing numbers
    <numbers...> (number) A list of numbers to be summed
]]

local sum = 0
for i,x in ipairs(args.numbers) do
    sum = sum + x
end
print ('sum is '..sum)

The parameter number has a trailing '...', which indicates that this parameter is a 'varargs' parameter. It must be the last parameter, and args.number will be an array.

Consider this implementation of the head utility from Mac OS X:

-- implements a BSD-style head
-- (see http://www.manpagez.com/man/1/head/osx-10.3.php)

lapp = require ('pl.lapp')

local args = lapp [[
Print the first few lines of specified files
   -n         (default 10)    Number of lines to print
   <files...> (default stdin) Files to print
]]

-- by default, lapp converts file arguments to an actual Lua file object.
-- But the actual filename is always available as <file>_name.
-- In this case, 'files' is a varargs array, so that 'files_name' is
-- also an array.
local nline = args.n
local nfile = #args.files
for i = 1,nfile do
    local file = args.files[i]
    if nfile > 1 then
        print('==> '..args.files_name[i]..' <==')
    end
    local n = 0
    for line in file:lines() do
        print(line)
        n = n + 1
        if n == nline then break end
    end
end

Note how we have access to all the filenames, because the auto-generated field files_name is also an array!

(This is probably not a very considerate script, since Lapp will open all the files provided, and only close them at the end of the script. See the xhead.lua example for another implementation.)

Flags and options may also be declared as vararg arrays, and can occur anywhere. If there is both a short and long form, then the trailing "..." must happen after the long form, for example "-x,--network... (string)...",

Bear in mind that short options can be combined (like 'tar -xzf'), so it's perfectly legal to have '-vvv'. But normally the value of args.v is just a simple true value.

local args = require ('pl.lapp') [[
   -v...  Verbosity level; can be -v, -vv or -vvv
]]
vlevel = not args.v[1] and 0 or #args.v
print(vlevel)

The vlevel assigment is a bit of Lua voodoo, so consider the cases:

* No -v flag, v is just { false }
* One -v flags, v is { true }
* Two -v flags, v is { true, true }
* Three -v flags, v is { true, true, true }

Defining a Parameter Callback

If a script implements lapp.callback, then Lapp will call it after each argument is parsed. The callback is passed the parameter name, the raw unparsed value, and the result table. It is called immediately after assignment of the value, so the corresponding field is available.

lapp = require ('pl.lapp')

function lapp.callback(parm,arg,args)
    print('+',parm,arg)
end

local args = lapp [[
Testing parameter handling
    -p               Plain flag (defaults to false)
    -q,--quiet       Plain flag with GNU-style optional long name
    -o  (string)     Required string option
    -n  (number)     Required number option
    -s (default 1.0) Option that takes a number, but will default
    <start> (number) Required number argument
    <input> (default stdin)  A parameter which is an input file
    <output> (default stdout) One that is an output file
]]
print 'args'
for k,v in pairs(args) do
    print(k,v)
end

This produces the following output:

$ args -o name -n 2 10 args.lua
+       o       name
+       n       2
+       start   10
+       input   args.lua
args
p       false
s       1
input_name      args.lua
quiet   false
output  file (781C1B98)
start   10
input   file (781C1BD8)
o       name
n       2

Callbacks are needed when you want to take action immediately on parsing an argument.

Slack Mode

If you'd like to use a multi-letter 'short' parameter you need to set the lapp.slack variable to true.

In the following example we also see how default false and default true flags can be used and how to overwrite the default -h help flag (--help still works fine) - this applies to non-slack mode as well.

-- Parsing the command line ----------------------------------------------------
-- test.lua
local lapp = require 'pl.lapp'
local pretty = require 'pl.pretty'
lapp.slack = true
local args = lapp [[
Does some calculations
   -v, --video              (string)             Specify input video
   -w, --width              (default 256)        Width of the video
   -h, --height             (default 144)        Height of the video
   -t, --time               (default 10)         Seconds of video to process
   -sk,--seek               (default 0)          Seek number of seconds
   -f1,--flag1                                   A false flag
   -f2,--flag2                                   A false flag
   -f3,--flag3              (default true)       A true flag
   -f4,--flag4              (default true)       A true flag
]]

pretty.dump(args)

And here we can see the output of test.lua:

$> lua test.lua -v abc --time 40 -h 20 -sk 15 --flag1 -f3
---->
{
  width = 256,
  flag1 = true,
  flag3 = false,
  seek = 15,
  flag2 = false,
  video = abc,
  time = 40,
  height = 20,
  flag4 = true
}

Simple Test Framework

pl.test was originally developed for the sole purpose of testing Penlight itself, but you may find it useful for your own applications. (There are many other options.)

Most of the goodness is in test.asserteq. It uses tablex.deepcompare on its two arguments, and by default quits the test application with a non-zero exit code, and an informative message printed to stderr:

local test = require 'pl.test'

test.asserteq({10,20,30},{10,20,30.1})

--~ test-test.lua:3: assertion failed
--~ got:    {
--~  [1] = 10,
--~  [2] = 20,
--~  [3] = 30
--~ }
--~ needed:    {
--~  [1] = 10,
--~  [2] = 20,
--~  [3] = 30.1
--~ }
--~ these values were not equal

This covers most cases but it's also useful to compare strings using string.match

-- must start with bonzo the dog
test.assertmatch ('bonzo the dog is here','^bonzo the dog')
-- must end with an integer
test.assertmatch ('hello 42','%d+$')

Since Lua errors are usually strings, this matching strategy is used to test 'exceptions':

test.assertraise(function()
    local t = nil
    print(t.bonzo)
end,'nil value')

(Some care is needed to match the essential part of the thrown error if you care for portability, since in Lua 5.2 the exact error is "attempt to index local 't' (a nil value)" and in Lua 5.3 the error is "attempt to index a nil value (local 't')")

There is an extra optional argument to these test functions, which is helpful when writing test helper functions. There you want to highlight the failed line, not the actual call to asserteq or assertmatch - line 33 here is the call to is_iden

function is_iden(str)
    test.assertmatch(str,'^[%a_][%w_]*$',1)
end

is_iden 'alpha_dog'
is_iden '$dollars'

--~ test-test.lua:33: assertion failed
--~ got:    "$dollars"
--~ needed:    "^[%a_][%w_]*$"
--~ these strings did not match

Useful Lua functions often return multiple values, and test.tuple is a convenient way to capture these values, whether they contain nils or not.

T = test.tuple

--- common error pattern
function failing()
    return nil,'failed'
end

test.asserteq(T(failing()),T(nil,'failed'))
generated by LDoc 1.4.6
Penlight-1.12.0/docs/manual/09-discussion.md.html000066400000000000000000000266051416703176500214730ustar00rootroot00000000000000 Penlight Documentation

Technical Choices

Modularity and Granularity

In an ideal world, a program should only load the libraries it needs. Penlight is intended to work in situations where an extra 100Kb of bytecode could be a problem. It is straightforward but tedious to load exactly what you need:

local data = require 'pl.data'
local List = require 'pl.List'
local array2d = require 'pl.array2d'
local seq = require 'pl.seq'
local utils = require 'pl.utils'

This is the style that I follow in Penlight itself, so that modules don't mess with the global environment; also, stringx.import() is not used because it will update the global string table.

But require 'pl' is more convenient in scripts; the question is how to ensure that one doesn't load the whole kitchen sink as the price of convenience. The strategy is to only load modules when they are referenced. In 'init.lua' (which is loaded by require 'pl') a metatable is attached to the global table with an __index metamethod. Any unknown name is looked up in the list of modules, and if found, we require it and make that module globally available. So when tablex.deepcompare is encountered, looking up tablex causes 'pl.tablex' to be required. .

Modifying the behaviour of the global table has consequences. For instance, there is the famous module strict which comes with Lua itself (perhaps the only standard Lua module written in Lua itself) which also does this modification so that global variiables must be defined before use. So the implementation in 'init.lua' allows for a 'not found' hook, which 'pl.strict.lua' uses. Other libraries may install their own metatables for _G, but Penlight will now forward any unknown name to the __index defined by the original metatable.

But the strategy is worth the effort: the old 'kitchen sink' 'init.lua' would pull in about 260K of bytecode, whereas now typical programs use about 100K less, and short scripts even better - for instance, if they were only needing functionality in utils.

There are some functions which mark their output table with a special metatable, when it seems particularly appropriate. For instance, tablex.makeset creates a Set, and seq.copy creates a List. But this does not automatically result in the loading of pl.Set and pl.List; only if you try to access any of these methods. In 'utils.lua', there is an exported table called stdmt:

stdmt = { List = {}, Map = {}, Set = {}, MultiMap = {} }

If you go through 'init.lua', then these plain little 'identity' tables get an __index metamethod which forces the loading of the full functionality. Here is the code from 'list.lua' which starts the ball rolling for lists:

List = utils.stdmt.List
List.__index = List
List._name = "List"
List._class = List

The 'load-on-demand' strategy helps to modularize the library. Especially for more casual use, require 'pl' is a good compromise between convenience and modularity.

In this current version, I have generally reduced the amount of trickery involved. Previously, Map was defined in pl.class; now it is sensibly defined in pl.Map; pl.class only contains the basic class mechanism (and returns that function.) For consistency, List is returned directly by require 'pl.List' (note the uppercase 'L'), Also, the amount of module dependencies in the non-core libraries like pl.config have been reduced.

Defining what is Callable

'utils.lua' exports function_arg which is used extensively throughout Penlight. It defines what is meant by 'callable'. Obviously true functions are immediately passed back. But what about strings? The first option is that it represents an operator in 'operator.lua', so that '<' is just an alias for operator.lt.

We then check whether there is a function factory defined for the metatable of the value.

(It is true that strings can be made callable, but in practice this turns out to be a cute but dubious idea, since all strings share the same metatable. A common programming error is to pass the wrong kind of object to a function, and it's better to get a nice clean 'attempting to call a string' message rather than some obscure trace from the bowels of your library.)

The other module that registers a function factory is pl.func. Placeholder expressions cannot be directly calleable, and so need to be instantiated and cached in as efficient way as possible.

(An inconsistency is that utils.is_callable does not do this thorough check.)

generated by LDoc 1.4.6
Penlight-1.12.0/docs_topics/000077500000000000000000000000001416703176500156705ustar00rootroot00000000000000Penlight-1.12.0/docs_topics/01-introduction.md000066400000000000000000000622311416703176500211550ustar00rootroot00000000000000## Introduction ### Purpose It is often said of Lua that it does not include batteries. That is because the goal of Lua is to produce a lean expressive language that will be used on all sorts of machines, (some of which don't even have hierarchical filesystems). The Lua language is the equivalent of an operating system kernel; the creators of Lua do not see it as their responsibility to create a full software ecosystem around the language. That is the role of the community. A principle of software design is to recognize common patterns and reuse them. If you find yourself writing things like `io.write(string.format('the answer is %d ',42))` more than a number of times then it becomes useful just to define a function `printf`. This is good, not just because repeated code is harder to maintain, but because such code is easier to read, once people understand your libraries. Penlight captures many such code patterns, so that the intent of your code becomes clearer. For instance, a Lua idiom to copy a table is `{unpack(t)}`, but this will only work for 'small' tables (for a given value of 'small') so it is not very robust. Also, the intent is not clear. So `tablex.deepcopy` is provided, which will also copy nested tables and and associated metatables, so it can be used to clone complex objects. The default error handling policy follows that of the Lua standard libraries: if a argument is the wrong type, then an error will be thrown, but otherwise we return `nil,message` if there is a problem. There are some exceptions; functions like `input.fields` default to shutting down the program immediately with a useful message. This is more appropriate behaviour for a _script_ than providing a stack trace. (However, this default can be changed.) The lexer functions always throw errors, to simplify coding, and so should be wrapped in `pcall`. If you are used to Python conventions, please note that all indices consistently start at 1. The Lua function `table.foreach` has been deprecated in favour of the `for in` statement, but such an operation becomes particularly useful with the higher-order function support in Penlight. Note that `tablex.foreach` reverses the order, so that the function is passed the value and then the key. Although perverse, this matches the intended use better. The only important external dependence of Penlight is [LuaFileSystem](http://keplerproject.github.com/luafilesystem/manual.html) (`lfs`), and if you want `dir.copyfile` to work cleanly on Windows, you will need either [alien](http://alien.luaforge.net/) or be using [LuaJIT](http://luajit.org) as well. (The fallback is to call the equivalent shell commands.) ### To Inject or not to Inject? It was realized a long time ago that large programs needed a way to keep names distinct by putting them into tables (Lua), namespaces (C++) or modules (Python). It is obviously impossible to run a company where everyone is called 'Bruce', except in Monty Python skits. These 'namespace clashes' are more of a problem in a simple language like Lua than in C++, because C++ does more complicated lookup over 'injected namespaces'. However, in a small group of friends, 'Bruce' is usually unique, so in particular situations it's useful to drop the formality and not use last names. It depends entirely on what kind of program you are writing, whether it is a ten line script or a ten thousand line program. So the Penlight library provides the formal way and the informal way, without imposing any preference. You can do it formally like: local utils = require 'pl.utils' utils.printf("%s\n","hello, world!") or informally like: require 'pl' utils.printf("%s\n","That feels better") `require 'pl'` makes all the separate Penlight modules available, without needing to require them each individually. Generally, the formal way is better when writing modules, since then there are no global side-effects and the dependencies of your module are made explicit. Andrew Starks has contributed another way, which balances nicely between the formal need to keep the global table uncluttered and the informal need for convenience. `require'pl.import_into'` returns a function, which accepts a table for injecting Penlight into, or if no table is given, it passes back a new one. local pl = require'pl.import_into'() The table `pl` is a 'lazy table' which loads modules as needed, so we can then use `pl.utils.printf` and so forth, without an explicit `require' or harming any globals. If you are using `_ENV` with Lua 5.2 to define modules, then here is a way to make Penlight available within a module: local _ENV,M = require 'pl.import_into' () function answer () -- all the Penlight modules are available! return pretty.write(utils.split '10 20 30', '') end return M The default is to put Penlight into `\_ENV`, which has the unintended effect of making it available from the module (much as `module(...,package.seeall)` does). To satisfy both convenience and safety, you may pass `true` to this function, and then the _module_ `M` is not the same as `\_ENV`, but only contains the exported functions. Otherwise, Penlight will _not_ bring in functions into the global table, or clobber standard tables like 'io'. require('pl') will bring tables like 'utils','tablex',etc into the global table _if they are used_. This 'load-on-demand' strategy ensures that the whole kitchen sink is not loaded up front, so this method is as efficient as explicitly loading required modules. You have an option to bring the `pl.stringx` methods into the standard string table. All strings have a metatable that allows for automatic lookup in `string`, so we can say `s:upper()`. Importing `stringx` allows for its functions to also be called as methods: `s:strip()`,etc: require 'pl' stringx.import() or, more explicitly: require('pl.stringx').import() A more delicate operation is importing tables into the local environment. This is convenient when the context makes the meaning of a name very clear: > require 'pl' > utils.import(math) > = sin(1.2) 0.93203908596723 `utils.import` can also be passed a module name as a string, which is first required and then imported. If used in a module, `import` will bring the symbols into the module context. Keeping the global scope simple is very necessary with dynamic languages. Using global variables in a big program is always asking for trouble, especially since you do not have the spell-checking provided by a compiler. The `pl.strict` module enforces a simple rule: globals must be 'declared'. This means that they must be assigned before use; assigning to `nil` is sufficient. > require 'pl.strict' > print(x) stdin:1: variable 'x' is not declared > x = nil > print(x) nil The `strict` module provided by Penlight is compatible with the 'load-on-demand' scheme used by `require 'pl`. `strict` also disallows assignment to global variables, except in the main program. Generally, modules have no business messing with global scope; if you must do it, then use a call to `rawset`. Similarly, if you have to check for the existence of a global, use `rawget`. If you wish to enforce strictness globally, then just add `require 'pl.strict'` at the end of `pl/init.lua`, otherwise call it from your main program. As from 1.1.0, this module provides a `strict.module` function which creates (or modifies) modules so that accessing an unknown function or field causes an error. For example, -- mymod.lua local strict = require 'pl.strict' local M = strict.module (...) function M.answer () return 42 end return M If you were to accidently type `mymod.Answer()`, then you would get a runtime error: "variable 'Answer' is not declared in 'mymod'". This can be applied to existing modules. You may desire to have the same level of checking for the Lua standard libraries: strict.make_all_strict(_G) Thereafter a typo such as `math.cosine` will give you an explicit error, rather than merely returning a `nil` that will cause problems later. ### What are function arguments in Penlight? Many functions in Penlight themselves take function arguments, like `map` which applies a function to a list, element by element. You can use existing functions, like `math.max`, anonymous functions (like `function(x,y) return x > y end` ), or operations by name (e.g '*' or '..'). The module `pl.operator` exports all the standard Lua operations, like the Python module of the same name. Penlight allows these to be referred to by name, so `operator.gt` can be more concisely expressed as '>'. Note that the `map` functions pass any extra arguments to the function, so we can have `ls:filter('>',0)`, which is a shortcut for `ls:filter(function(x) return x > 0 end)`. Finally, `pl.func` supports _placeholder expressions_ in the Boost lambda style, so that an anonymous function to multiply the two arguments can be expressed as `\_1*\_2`. To use them directly, note that _all_ function arguments in Penlight go through `utils.function_arg`. `pl.func` registers itself with this function, so that you can directly use placeholder expressions with standard methods: > _1 = func._1 > = List{10,20,30}:map(_1+1) {11,21,31} Another option for short anonymous functions is provided by `utils.string_lambda`; this is invoked automatically: > = List{10,20,30}:map '|x| x + 1' {11,21,31} ### Pros and Cons of Loopless Programming The standard loops-and-ifs 'imperative' style of programming is dominant, and often seems to be the 'natural' way of telling a machine what to do. It is in fact very much how the machine does things, but we need to take a step back and find ways of expressing solutions in a higher-level way. For instance, applying a function to all elements of a list is a common operation: local res = {} for i = 1,#ls do res[i] = fun(ls[i]) end This can be efficiently and succintly expressed as `ls:map(fun)`. Not only is there less typing but the intention of the code is clearer. If readers of your code spend too much time trying to guess your intention by analyzing your loops, then you have failed to express yourself clearly. Similarly, `ls:filter('>',0)` will give you all the values in a list greater than zero. (Of course, if you don't feel like using `List`, or have non-list-like tables, then `pl.tablex` offers the same facilities. In fact, the `List` methods are implemented using `tablex` functions.) A common observation is that loopless programming is less efficient, particularly in the way it uses memory. `ls1:map2('*',ls2):reduce '+'` will give you the dot product of two lists, but an unnecessary temporary list is created. But efficiency is relative to the actual situation, it may turn out to be _fast enough_, or may not appear in any crucial inner loops, etc. Writing loops is 'error-prone and tedious', as Stroustrup says. But any half-decent editor can be taught to do much of that typing for you. The question should actually be: is it tedious to _read_ loops? As with natural language, programmers tend to read chunks at a time. A for-loop causes no surprise, and probably little brain activity. One argument for loopless programming is the loops that you _do_ write stand out more, and signal 'something different happening here'. It should not be an all-or-nothing thing, since most programs require a mixture of idioms that suit the problem. Some languages (like APL) do nearly everything with map and reduce operations on arrays, and so solutions can sometimes seem forced. Wisdom is knowing when a particular idiom makes a particular problem easy to _solve_ and the solution easy to _explain_ afterwards. ### Generally useful functions. The function `printf` discussed earlier is included in `pl.utils` because it makes properly formatted output easier. (There is an equivalent `fprintf` which also takes a file object parameter, just like the C function.) Splitting a string using a delimiter is a fairly common operation, hence `split`. Utility functions like `is_type` help with identifying what kind of animal you are dealing with. The Lua `type` function handles the basic types, but can't distinguish between different kinds of objects, which are all tables. So `is_type` handles both cases, like `is_type(s,"string")` and `is_type(ls,List)`. A common pattern when working with Lua varargs is capturing all the arguments in a table: function t(...) local args = {...} ... end But this will bite you someday when `nil` is one of the arguments, since this will put a 'hole' in your table. In particular, `#ls` will only give you the size upto the `nil` value. Hence the need for `table.pack` - this is a new Lua 5.2 function which Penlight defines also for Lua 5.1. function t(...) local args,n = table.pack(...) for i = 1,n do ... end end The 'memoize' pattern occurs when you have a function which is expensive to call, but will always return the same value subsequently. `utils.memoize` is given a function, and returns another function. This calls the function the first time, saves the value for that argument, and thereafter for that argument returns the saved value. This is a more flexible alternative to building a table of values upfront, since in general you won't know what values are needed. sum = utils.memoize(function(n) local sum = 0 for i = 1,n do sum = sum + i end return sum end) ... s = sum(1e8) --takes time! ... s = sum(1e8) --returned saved value! Penlight is fully compatible with Lua 5.1, 5.2 and LuaJIT 2. To ensure this, `utils` also defines the global Lua 5.2 [load](http://www.lua.org/work/doc/manual.html#pdf-load) function as `utils.load` * the input (either a string or a function) * the source name used in debug information * the mode is a string that can have either or both of 'b' or 't', depending on whether the source is a binary chunk or text code (default is 'bt') * the environment for the compiled chunk Using `utils.load` should reduce the need to call the deprecated function `setfenv`, and make your Lua 5.1 code 5.2-friendly. The `utils` module exports `getfenv` and `setfenv` for Lua 5.2 as well, based on code by Sergey Rozhenko. Note that these functions can fail for functions which don't access any globals. ### Application Support `app.parse_args` is a simple command-line argument parser. If called without any arguments, it tries to use the global `arg` array. It returns the _flags_ (options begining with '-') as a table of name/value pairs, and the _arguments_ as an array. It knows about long GNU-style flag names, e.g. `--value`, and groups of short flags are understood, so that `-ab` is short for `-a -b`. The flags result would then look like `{value=true,a=true,b=true}`. Flags may take values. The command-line `--value=open -n10` would result in `{value='open',n='10'}`; generally you can use '=' or ':' to separate the flag from its value, except in the special case where a short flag is followed by an integer. Or you may specify upfront that some flags have associated values, and then the values will follow the flag. > require 'pl' > flags,args = app.parse_args({'-o','fred','-n10','fred.txt'},{o=true}) > pretty.dump(flags) {o='fred',n='10'} `parse_args` is not intelligent or psychic; it will not convert any flag values or arguments for you, or raise errors. For that, have a look at @{08-additional.md.Command_line_Programs_with_Lapp|Lapp}. An application which consists of several files usually cannot use `require` to load files in the same directory as the main script. `app.require_here()` ensures that the Lua module path is modified so that files found locally are found first. In the `examples` directory, `test-symbols.lua` uses this function to ensure that it can find `symbols.lua` even if it is not run from this directory. `app.appfile` will create a filename that your application can use to store its private data, based on the script name. For example, `app.appfile "test.txt"` from a script called `testapp.lua` produces the following file on my Windows machine: @plain C:\Documents and Settings\SJDonova\.testapp\test.txt and the equivalent on my Linux machine: @plain /home/sdonovan/.testapp/test.txt If `.testapp` does not exist, it will be created. Penlight makes it convenient to save application data in Lua format. You can use `pretty.dump(t,file)` to write a Lua table in a human-readable form to a file, and `pretty.read(file.read(file))` to generate the table again, using the `pretty` module. ### Simplifying Object-Oriented Programming in Lua Lua is similar to JavaScript in that the concept of class is not directly supported by the language. In fact, Lua has a very general mechanism for extending the behaviour of tables which makes it straightforward to implement classes. A table's behaviour is controlled by its metatable. If that metatable has a `\_\_index` function or table, this will handle looking up anything which is not found in the original table. A class is just a table with an `__index` key pointing to itself. Creating an object involves making a table and setting its metatable to the class; then when handling `obj.fun`, Lua first looks up `fun` in the table `obj`, and if not found it looks it up in the class. `obj:fun(a)` is just short for `obj.fun(obj,a)`. So with the metatable mechanism and this bit of syntactic sugar, it is straightforward to implement classic object orientation. -- animal.lua class = require 'pl.class' class.Animal() function Animal:_init(name) self.name = name end function Animal:__tostring() return self.name..': '..self:speak() end class.Dog(Animal) function Dog:speak() return 'bark' end class.Cat(Animal) function Cat:_init(name,breed) self:super(name) -- must init base! self.breed = breed end function Cat:speak() return 'meow' end class.Lion(Cat) function Lion:speak() return 'roar' end fido = Dog('Fido') felix = Cat('Felix','Tabby') leo = Lion('Leo','African') $ lua -i animal.lua > = fido,felix,leo Fido: bark Felix: meow Leo: roar > = leo:is_a(Animal) true > = leo:is_a(Dog) false > = leo:is_a(Cat) true All Animal does is define `\_\_tostring`, which Lua will use whenever a string representation is needed of the object. In turn, this relies on `speak`, which is not defined. So it's what C++ people would call an abstract base class; the specific derived classes like Dog define `speak`. Please note that _if_ derived classes have their own constructors, they must explicitly call the base constructor for their base class; this is conveniently available as the `super` method. Note that (as always) there are multiple ways to implement OOP in Lua; this method uses the classic 'a class is the __index of its objects' but does 'fat inheritance'; methods from the base class are copied into the new class. The advantage of this is that you are not penalized for long inheritance chains, for the price of larger classes, but generally objects outnumber classes! (If not, something odd is going on with your design.) All such objects will have a `is_a` method, which looks up the inheritance chain to find a match. Another form is `class_of`, which can be safely called on all objects, so instead of `leo:is_a(Animal)` one can say `Animal:class_of(leo)`. There are two ways to define a class, either `class.Name()` or `Name = class()`; both work identically, except that the first form will always put the class in the current environment (whether global or module); the second form provides more flexibility about where to store the class. The first form does _name_ the class by setting the `_name` field, which can be useful in identifying the objects of this type later. This session illustrates the usefulness of having named classes, if no `__tostring` method is explicitly defined. > class.Fred() > a = Fred() > = a Fred: 00459330 > Alice = class() > b = Alice() > = b table: 00459AE8 > Alice._name = 'Alice' > = b Alice: 00459AE8 So `Alice = class(); Alice._name = 'Alice'` is exactly the same as `class.Alice()`. This useful notation is borrowed from Hugo Etchegoyen's [classlib](http://lua-users.org/wiki/MultipleInheritanceClasses) which further extends this concept to allow for multiple inheritance. Notice that the more convenient form puts the class name in the _current environment_! That is, you may use it safely within modules using the old-fashioned `module()` or the new `_ENV` mechanism. There is always more than one way of doing things in Lua; some may prefer this style for creating classes: local class = require 'pl.class' class.Named { _init = function(self,name) self.name = name end; __tostring = function(self) return 'boo '..self.name end; } b = Named 'dog' print(b) --> boo dog Note that you have to explicitly declare `self` and end each function definition with a semi-colon or comma, since this is a Lua table. To inherit from a base class, set the special field `_base` to the class in this table. Penlight provides a number of useful classes; there is `List`, which is a Lua clone of the standard Python list object, and `Set` which represents sets. There are three kinds of _map_ defined: `Map`, `MultiMap` (where a key may have multiple values) and `OrderedMap` (where the order of insertion is remembered.). There is nothing special about these classes and you may inherit from them. A powerful thing about dynamic languages is that you can redefine existing classes and functions, which is often called 'monkey patching' It's entertaining and convenient, but ultimately anti-social; you may modify `List` but then any other modules using this _shared_ resource can no longer be sure about its behaviour. (This is why you must say `stringx.import()` explicitly if you want the extended string methods - it would be a bad default.) Lua is particularly open to modification but the community is not as tolerant of monkey-patching as the Ruby community, say. You may wish to add some new methods to `List`? Cool, but that's what subclassing is for. class.Strings(List) function Strings:my_method() ... end It's definitely more useful to define exactly how your objects behave in _unknown_ conditions. All classes have a `catch` method you can use to set a handler for unknown lookups; the function you pass looks exactly like the `__index` metamethod. Strings:catch(function(self,name) return function() error("no such method "..name,2) end end) In this case we're just customizing the error message, but creative things can be done. Consider this code from `test-vector.lua`: Strings:catch(List.default_map_with(string)) ls = Strings{'one','two','three'} asserteq(ls:upper(),{'ONE','TWO','THREE'}) asserteq(ls:sub(1,2),{'on','tw','th'}) So we've converted a unknown method invocation into a map using the function of that name found in `string`. So for a `Vector` (which is a specialization of `List` for numbers) it makes sense to make `math` the default map so that `v:sin()` makes sense. Note that `map` operations return a object of the same type - this is often called _covariance_. So `ls:upper()` itself returns a `Strings` object. This is not _always_ what you want, but objects can always be cast to the desired type. (`cast` doesn't create a new object, but returns the object passed.) local sizes = ls:map '#' asserteq(sizes, {3,3,5}) asserteq(utils.type(sizes),'Strings') asserteq(sizes:is_a(Strings),true) sizes = Vector:cast(sizes) asserteq(utils.type(sizes),'Vector') asserteq(sizes+1,{4,4,6}) About `utils.type`: it can only return a string for a class type if that class does in fact have a `_name` field. _Properties_ are a useful object-oriented pattern. We wish to control access to a field, but don't wish to force the user of the class to say `obj:get_field()` etc. This excerpt from `tests/test-class.lua` shows how it is done: local MyProps = class(class.properties) local setted_a, got_b function MyProps:_init () self._a = 1 self._b = 2 end function MyProps:set_a (v) setted_a = true self._a = v end function MyProps:get_b () got_b = true return self._b end local mp = MyProps() mp.a = 10 asserteq(mp.a,10) asserteq(mp.b,2) asserteq(setted_a and got_b, true) The convention is that the internal field name is prefixed with an underscore; when reading `mp.a`, first a check for an explicit _getter_ `get_a` and then only look for `_a`. Simularly, writing `mp.a` causes the _setter_ `set_a` to be used. This is cool behaviour, but like much Lua metaprogramming, it is not free. Method lookup on such objects goes through `\_\_index` as before, but now `\_\_index` is a function which has to explicitly look up methods in the class, before doing any property indexing, which is not going to be as fast as field lookup. If however, your accessors actually do non-trivial things, then the extra overhead could be worth it. This is not really intended for _access control_ because external code can write to `mp._a` directly. It is possible to have this kind of control in Lua, but it again comes with run-time costs. Penlight-1.12.0/docs_topics/02-arrays.md000066400000000000000000000610151416703176500177350ustar00rootroot00000000000000## Tables and Arrays ### Python-style Lists One of the elegant things about Lua is that tables do the job of both lists and dicts (as called in Python) or vectors and maps, (as called in C++), and they do it efficiently. However, if we are dealing with 'tables with numerical indices' we may as well call them lists and look for operations which particularly make sense for lists. The Penlight `List` class was originally written by Nick Trout for Lua 5.0, and translated to 5.1 and extended by myself. It seemed that borrowing from Python was a good idea, and this eventually grew into Penlight. Here is an example showing `List` in action; it redefines `__tostring`, so that it can print itself out more sensibly: > List = require 'pl.List' --> automatic with require 'pl' <--- > l = List() > l:append(10) > l:append(20) > = l {10,20} > l:extend {30,40} {10,20,30,40} > l:insert(1,5) {5,10,20,30,40} > = l:pop() 40 > = l {5,10,20,30} > = l:index(30) 4 > = l:contains(30) true > = l:reverse() ---> note: doesn't make a copy! {30,20,10,5} Although methods like `sort` and `reverse` operate in-place and change the list, they do return the original list. This makes it possible to do _method chaining_, like `ls = ls:append(10):append(20):reverse():append(1)`. But (and this is an important but) no extra copy is made, so `ls` does not change identity. `List` objects (like tables) are _mutable_, unlike strings. If you want a copy of a list, then `List(ls)` will do the job, i.e. it acts like a copy constructor. However, if passed any other table, `List` will just set the metatable of the table and _not_ make a copy. A particular feature of Python lists is _slicing_. This is fully supported in this version of `List`, except we use 1-based indexing. So `List.slice` works rather like `string.sub`: > l = List {10,20,30,40} > = l:slice(1,1) ---> note: creates a new list! {10} > = l:slice(2,2) {20} > = l:slice(2,3) {20,30} > = l:slice(2,-2) {20,30} > = l:slice_assign(2,2,{21,22,23}) {10,21,22,23,30,40} > = l:chop(1,1) {21,22,23,30,40} Functions like `slice_assign` and `chop` modify the list; the first is equivalent to Python`l[i1:i2] = seq` and the second to `del l[i1:i2]`. List objects are ultimately just Lua 'list-like' tables, but they have extra operations defined on them, such as equality and concatention. For regular tables, equality is only true if the two tables are _identical objects_, whereas two lists are equal if they have the same contents, i.e. that `l1[i]==l2[i]` for all elements. > l1 = List {1,2,3} > l2 = List {1,2,3} > = l1 == l2 true > = l1..l2 {1,2,3,1,2,3} The `List` constructor can be passed a function. If so, it's assumed that this is an iterator function that can be repeatedly called to generate a sequence. One such function is `io.lines`; the following short, intense little script counts the number of lines in standard input: -- linecount.lua require 'pl' ls = List(io.lines()) print(#ls) `List.iterate` captures what `List` considers a sequence. In particular, it can also iterate over all 'characters' in a string: > for ch in List.iterate 'help' do io.write(ch,' ') end h e l p > Since the function `iterate` is used internally by the `List` constructor, strings can be made into lists of character strings very easily. There are a number of operations that go beyond the standard Python methods. For instance, you can _partition_ a list into a table of sublists using a function. In the simplest form, you use a predicate (a function returning a boolean value) to partition the list into two lists, one of elements matching and another of elements not matching. But you can use any function; if we use `type` then the keys will be the standard Lua type names. > ls = List{1,2,3,4} > ops = require 'pl.operator' > ls:partition(function(x) return x > 2 end) {false={1,2},true={3,4}} > ls = List{'one',math.sin,List{1},10,20,List{1,2}} > ls:partition(type) {function={function: 00369110},string={one},number={10,20},table={{1},{1,2}}} This is one `List` method which returns a table which is not a `List`. Bear in mind that you can always call a `List` method on a plain table argument, so `List.partition(t,type)` works as expected. But these functions will only operate on the array part of the table. The 'nominal' type of the returned table is `pl.Multimap`, which describes a mapping between keys and multiple values. This does not mean that `pl.Multimap` is automatically loaded whenever you use `partition` (or `List` for that matter); this is one of the standard metatables which are only filled out when the appropriate module is loaded. This allows tables to be tagged appropriately without causing excessive coupling. Stacks occur everywhere in computing. `List` supports stack-like operations; there is already `pop` (remove and return last value) and `append` acts like `push` (add a value to the end). `push` is provided as an alias for `append`, and the other stack operation (size) is simply the size operator `#`. Queues can also be implemented; you use `pop` to take values out of the queue, and `put` to insert a value at the begining. You may derive classes from `List`, and since the list-returning methods are covariant, the result of `slice` etc will return lists of the derived type, not `List`. For instance, consider the specialization of a `List` type that contains numbers in `tests/test-list.lua`: n1 = NA{10,20,30} n2 = NA{1,2,3} ns = n1 + 2*n2 asserteq(ns,{12,24,36}) min,max = ns:slice(1,2):minmax() asserteq(T(min,max),T(12,24)) asserteq(n1:normalize():sum(),1,1e-8) ### Map and Set classes The `Map` class exposes what Python would call a 'dict' interface, and accesses the hash part of the table. The name 'Map' is used to emphasize the interface, not the implementation; it is an object which maps keys onto values; `m['alice']` or the equivalent `m.alice` is the access operation. This class also provides explicit `set` and `get` methods, which are trivial for regular maps but get interesting when `Map` is subclassed. The other operation is `update`, which extends a map by copying the keys and values from another table, perhaps overwriting existing keys: > Map = require 'pl.Map' > m = Map{one=1,two=2} > m:update {three=3,four=4,two=20} > = m == M{one=1,two=20,three=3,four=4} true The method `values` returns a list of the values, and `keys` returns a list of the keys; there is no guarantee of order. `getvalues` is given a list of keys and returns a list of values associated with these keys: > m = Map{one=1,two=2,three=3} > = m:getvalues {'one','three'} {1,3} > = m:getvalues(m:keys()) == m:values() true When querying the value of a `Map`, it is best to use the `get` method: > print(m:get 'one', m:get 'two') 1 2 The reason is that `m[key]` can be ambiguous; due to the current implementation, `m["get"]` will always succeed, because if a value is not present in the map, it will be looked up in the `Map` metatable, which contains a method `get`. There is currently no simple solution to this annoying restriction. There are some useful classes which inherit from `Map`. An `OrderedMap` behaves like a `Map` but keeps its keys in order if you use its `set` method to add keys and values. Like all the 'container' classes in Penlight, it defines an `iter` method for iterating over its values; this will return the keys and values in the order of insertion; the `keys` and `values` methods likewise. A `MultiMap` allows multiple values to be associated with a given key. So `set` (as before) takes a key and a value, but calling it with the same key and a different value does not overwrite but adds a new value. `get` (or using `[]`) will return a list of values. A `Set` can be seen as a special kind of `Map`, where all the values are `true`, the keys are the values, and the order is not important. So in this case `Set.values` is defined to return a list of the keys. Sets can display themselves, and the basic operations like `union` (`+`) and `intersection` (`*`) are defined. > Set = require 'pl.Set' > = Set{'one','two'} == Set{'two','one'} true > fruit = Set{'apple','banana','orange'} > = fruit['banana'] true > = fruit['hazelnut'] nil > = fruit:values() {apple,orange,banana} > colours = Set{'red','orange','green','blue'} > = fruit,colours [apple,orange,banana] [blue,green,orange,red] > = fruit+colours [blue,green,apple,red,orange,banana] > = fruit*colours [orange] There are also the functions `Set.difference` and `Set.symmetric_difference`. The first answers the question 'what fruits are not colours?' and the second 'what are fruits and colours but not both?' > = fruit - colours [apple,banana] > = fruit ^ colours [blue,green,apple,red,banana] Adding elements to a set is simply `fruit['peach'] = true` and removing is `fruit['apple'] = nil` . To make this simplicity work properly, the `Set` class has no methods - either you use the operator forms or explicitly use `Set.intersect` etc. In this way we avoid the ambiguity that plagues `Map`. (See `pl.Map` and `pl.Set`) ### Useful Operations on Tables @lookup pl.tablex Some notes on terminology: Lua tables are usually _list-like_ (like an array) or _map-like_ (like an associative array or dict); they can of course have a list-like and a map-like part. Some of the table operations only make sense for list-like tables, and some only for map-like tables. (The usual Lua terminology is the array part and the hash part of the table, which reflects the actual implementation used; it is more accurate to say that a Lua table is an associative map which happens to be particularly efficient at acting like an array.) The functions provided in `table` provide all the basic manipulations on Lua tables, but as we saw with the `List` class, it is useful to build higher-level operations on top of those functions. For instance, to copy a table involves this kind of loop: local res = {} for k,v in pairs(T) do res[k] = v end return res The `tablex` module provides this as `copy`, which does a _shallow_ copy of a table. There is also `deepcopy` which goes further than a simple loop in two ways; first, it also gives the copy the same metatable as the original (so it can copy objects like `List` above) and any nested tables will also be copied, to arbitrary depth. There is also `icopy` which operates on list-like tables, where you can set optionally set the start index of the source and destination as well. It ensures that any left-over elements will be deleted: asserteq(icopy({1,2,3,4,5,6},{20,30}),{20,30}) -- start at 1 asserteq(icopy({1,2,3,4,5,6},{20,30},2),{1,20,30}) -- start at 2 asserteq(icopy({1,2,3,4,5,6},{20,30},2,2),{1,30}) -- start at 2, copy from 2 (This code from the `tablex` test module shows the use of `pl.test.asserteq`) Whereas, `move` overwrites but does not delete the rest of the destination: asserteq(move({1,2,3,4,5,6},{20,30}),{20,30,3,4,5,6}) asserteq(move({1,2,3,4,5,6},{20,30},2),{1,20,30,4,5,6}) asserteq(move({1,2,3,4,5,6},{20,30},2,2),{1,30,3,4,5,6}) (The difference is somewhat like that between C's `strcpy` and `memmove`.) To summarize, use `copy` or `deepcopy` to make a copy of an arbitrary table. To copy into a map-like table, use `update`; to copy into a list-like table use `icopy`, and `move` if you are updating a range in the destination. To complete this set of operations, there is `insertvalues` which works like `table.insert` except that one provides a table of values to be inserted, and `removevalues` which removes a range of values. asserteq(insertvalues({1,2,3,4},2,{20,30}),{1,20,30,2,3,4}) asserteq(insertvalues({1,2},{3,4}),{1,2,3,4}) Another example: > T = require 'pl.tablex' > t = {10,20,30,40} > = T.removevalues(t,2,3) {10,40} > = T.insertvalues(t,2,{20,30}) {10,20,30,40} In a similar spirit to `deepcopy`, `deepcompare` will take two tables and return true only if they have exactly the same values and structure. > t1 = {1,{2,3},4} > t2 = deepcopy(t1) > = t1 == t2 false > = deepcompare(t1,t2) true `find` will return the index of a given value in a list-like table. Note that like `string.find` you can specify an index to start searching, so that all instances can be found. There is an optional fourth argument, which makes the search start at the end and go backwards, so we could define `rfind` like so: function rfind(t,val,istart) return tablex.find(t,val,istart,true) end `find` does a linear search, so it can slow down code that depends on it. If efficiency is required for large tables, consider using an _index map_. `index_map` will return a table where the keys are the original values of the list, and the associated values are the indices. (It is almost exactly the representation needed for a _set_.) > t = {'one','two','three'} > = tablex.find(t,'two') 2 > = tablex.find(t,'four') nil > il = tablex.index_map(t) > = il['two'] 2 > = il.two 2 A version of `index_map` called `makeset` is also provided, where the values are just `true`. This is useful because two such sets can be compared for equality using `deepcompare`: > = deepcompare(makeset {1,2,3},makeset {2,1,3}) true Consider the problem of determining the new employees that have joined in a period. Assume we have two files of employee names: (last-month.txt) smith,john brady,maureen mongale,thabo (this-month.txt) smith,john smit,johan brady,maureen mogale,thabo van der Merwe,Piet To find out differences, just make the employee lists into sets, like so: require 'pl' function read_employees(file) local ls = List(io.lines(file)) -- a list of employees return tablex.makeset(ls) end last = read_employees 'last-month.txt' this = read_employees 'this-month.txt' -- who is in this but not in last? diff = tablex.difference(this,last) -- in a set, the keys are the values... for e in pairs(diff) do print(e) end -- *output* -- van der Merwe,Piet -- smit,johan The `difference` operation is easy to write and read: for e in pairs(this) do if not last[e] then print(e) end end Using `difference` here is not that it is a tricky thing to code, it is that you are stating your intentions clearly to other readers of your code. (And naturally to your future self, in six months time.) `find_if` will search a table using a function. The optional third argument is a value which will be passed as a second argument to the function. `pl.operator` provides the Lua operators conveniently wrapped as functions, so the basic comparison functions are available: > ops = require 'pl.operator' > = tablex.find_if({10,20,30,40},ops.gt,20) 3 true Note that `find_if` will also return the _actual value_ returned by the function, which of course is usually just `true` for a boolean function, but any value which is not `nil` and not `false` can be usefully passed back. `deepcompare` does a thorough recursive comparison, but otherwise using the default equality operator. `compare` allows you to specify exactly what function to use when comparing two list-like tables, and `compare_no_order` is true if they contain exactly the same elements. Do note that the latter does not need an explicit comparison function - in this case the implementation is actually to compare the two sets, as above: > = compare_no_order({1,2,3},{2,1,3}) true > = compare_no_order({1,2,3},{2,1,3},'==') true (Note the special string '==' above; instead of saying `ops.gt` or `ops.eq` we can use the strings '>' or '==' respectively.) `sort` and `sortv` return iterators that will iterate through the sorted elements of a table. `sort` iterates by sorted key order, and `sortv` iterates by sorted value order. For example, given a table with names and ages, it is trivial to iterate over the elements: > t = {john=27,jane=31,mary=24} > for name,age in tablex.sort(t) do print(name,age) end jane 31 john 27 mary 24 > for name,age in tablex.sortv(t) do print(name,age) end mary 24 john 27 jane 31 There are several ways to merge tables in PL. If they are list-like, then see the operations defined by `pl.List`, like concatenation. If they are map-like, then `merge` provides two basic operations. If the third arg is false, then the result only contains the keys that are in common between the two tables, and if true, then the result contains all the keys of both tables. These are in fact generalized set union and intersection operations: > S1 = {john=27,jane=31,mary=24} > S2 = {jane=31,jones=50} > = tablex.merge(S1, S2, false) {jane=31} > = tablex.merge(S1, S2, true) {mary=24,jane=31,john=27,jones=50} When working with tables, you will often find yourself writing loops like in the first example. Loops are second nature to programmers, but they are often not the most elegant and self-describing way of expressing an operation. Consider the `map` function, which creates a new table by applying a function to each element of the original: > = map(math.sin, {1,2,3,4}) { 0.84, 0.91, 0.14, -0.76} > = map(function(x) return x*x end, {1,2,3,4}) {1,4,9,16} `map` saves you from writing a loop, and the resulting code is often clearer, as well as being shorter. This is not to say that 'loops are bad' (although you will hear that from some extremists), just that it's good to capture standard patterns. Then the loops you do write will stand out and acquire more significance. `pairmap` is interesting, because the function works with both the key and the value. > t = {fred=10,bonzo=20,alice=4} > = pairmap(function(k,v) return v end, t) {4,10,20} > = pairmap(function(k,v) return k end, t) {'alice','fred','bonzo'} (These are common enough operations that the first is defined as `values` and the second as `keys`.) If the function returns two values, then the _second_ value is considered to be the new key: > = pairmap(t,function(k,v) return v+10, k:upper() end) {BONZO=30,FRED=20,ALICE=14} `map2` applies a function to two tables: > map2(ops.add,{1,2},{10,20}) {11,22} > map2('*',{1,2},{10,20}) {10,40} The various map operations generate tables; `reduce` applies a function of two arguments over a table and returns the result as a scalar: > reduce ('+', {1,2,3}) 6 > reduce ('..', {'one','two','three'}) 'onetwothree' Finally, `zip` sews different tables together: > = zip({1,2,3},{10,20,30}) {{1,10},{2,20},{3,30}} Browsing through the documentation, you will find that `tablex` and `List` share methods. For instance, `tablex.imap` and `List.map` are basically the same function; they both operate over the array-part of the table and generate another table. This can also be expressed as a _list comprehension_ `C 'f(x) for x' (t)` which makes the operation more explicit. So why are there different ways to do the same thing? The main reason is that not all tables are Lists: the expression `ls:map('#')` will return a _list_ of the lengths of any elements of `ls`. A list is a thin wrapper around a table, provided by the metatable `List`. Sometimes you may wish to work with ordinary Lua tables; the `List` interface is not a compulsory way to use Penlight table operations. ### Operations on two-dimensional tables @lookup pl.array2d Two-dimensional tables are of course easy to represent in Lua, for instance `{{1,2},{3,4}}` where we store rows as subtables and index like so `A[col][row]`. This is the common representation used by matrix libraries like [LuaMatrix](http://lua-users.org/wiki/LuaMatrix). `pl.array2d` does not provide matrix operations, since that is the job for a specialized library, but rather provides generalizations of the higher-level operations provided by `pl.tablex` for one-dimensional arrays. `iter` is a useful generalization of `ipairs`. (The extra parameter determines whether you want the indices as well.) > a = {{1,2},{3,4}} > for i,j,v in array2d.iter(a,true) do print(i,j,v) end 1 1 1 1 2 2 2 1 3 2 2 4 Note that you can always convert an arbitrary 2D array into a 'list of lists' with `List(tablex.map(List,a))` `map` will apply a function over all elements (notice that extra arguments can be provided, so this operation is in effect `function(x) return x-1 end`) > array2d.map('-',a,1) {{0,1},{2,3}} 2D arrays are stored as an array of rows, but columns can be extracted: > array2d.column(a,1) {1,3} There are three equivalents to `tablex.reduce`. You can either reduce along the rows (which is the most efficient) or reduce along the columns. Either one will give you a 1D array. And `reduce2` will apply two operations: the first one reduces the rows, and the second reduces the result. > array2d.reduce_rows('+',a) {3,7} > array2d.reduce_cols('+',a) {4,6} > -- same as tablex.reduce('*',array.reduce_rows('+',a)) > array2d.reduce2('*','+',a) 21 ` `tablex.map2` applies an operation to two tables, giving another table. `array2d.map2` does this for 2D arrays. Note that you have to provide the _rank_ of the arrays involved, since it's hard to always correctly deduce this from the data: > b = {{10,20},{30,40}} > a = {{1,2},{3,4}} > = array2d.map2('+',2,2,a,b) -- two 2D arrays {{11,22},{33,44}} > = array2d.map2('+',1,2,{10,100},a) -- 1D, 2D {{11,102},{13,104}} > = array2d.map2('*',2,1,a,{1,-1}) -- 2D, 1D {{1,-2},{3,-4}} Of course, you are not limited to simple arithmetic. Say we have a 2D array of strings, and wish to print it out with proper right justification. The first step is to create all the string lengths by mapping `string.len` over the array, the second is to reduce this along the columns using `math.max` to get maximum column widths, and last, apply `stringx.rjust` with these widths. maxlens = reduce_cols(math.max,map('#',lines)) lines = map2(stringx.rjust,2,1,lines,maxlens) There is `product` which returns the _Cartesian product_ of two 1D arrays. The result is a 2D array formed from applying the function to all possible pairs from the two arrays. > array2d.product('{}',{1,2},{'a','b'}) {{{1,'b'},{2,'a'}},{{1,'a'},{2,'b'}}} There is a set of operations which work in-place on 2D arrays. You can `swap_rows` and `swap_cols`; the first really is a simple one-liner, but the idea here is to give the operation a name. `remove_row` and `remove_col` are generalizations of `table.remove`. Likewise, `extract_rows` and `extract_cols` are given arrays of indices and discard anything else. So, for instance, `extract_cols(A,{2,4})` will leave just columns 2 and 4 in the array. `List.slice` is often useful on 1D arrays; `slice` does the same thing, but is generally given a start (row,column) and a end (row,column). > A = {{1,2,3},{4,5,6},{7,8,9}} > B = slice(A,1,1,2,2) > write(B) 1 2 4 5 > B = slice(A,2,2) > write(B,nil,'%4.1f') 5.0 6.0 8.0 9.0 Here `write` is used to print out an array nicely; the second parameter is `nil`, which is the default (stdout) but can be any file object and the third parameter is an optional format (as used in `string.format`). `parse_range` will take a spreadsheet range like 'A1:B2' or 'R1C1:R2C2' and return the range as four numbers, which can be passed to `slice`. The rule is that `slice` will return an array of the appropriate shape depending on the range; if a range represents a row or a column, the result is 1D, otherwise 2D. This applies to `iter` as well, which can also optionally be given a range: > for i,j,v in iter(A,true,2,2) do print(i,j,v) end 2 2 5 2 3 6 3 2 8 3 3 9 `new` will construct a new 2D array with the given dimensions. You provide an initial value for the elements, which is interpreted as a function if it's callable. With `L` being `utils.string_lambda` we then have the following way to make an _identity matrix_: asserteq( array.new(3,3,L'|i,j| i==j and 1 or 0'), {{1,0,0},{0,1,0},{0,0,1}} ) Please note that most functions in `array2d` are _covariant_, that is, they return an array of the same type as they receive. In particular, any objects created with `data.new` or `matrix.new` will remain data or matrix objects when reshaped or sliced, etc. Data objects have the `array2d` functions available as methods. Penlight-1.12.0/docs_topics/03-strings.md000066400000000000000000000157741416703176500201410ustar00rootroot00000000000000## Strings. Higher-level operations on strings. ### Extra String Methods @lookup pl.stringx These are convenient borrowings from Python, as described in 3.6.1 of the Python reference, but note that indices in Lua always begin at one. `stringx` defines functions like `isalpha` and `isdigit`, which return `true` if s is only composed of letters or digits respectively. `startswith` and `endswith` are convenient ways to find substrings. (`endswith` works as in Python 2.5, so that `f:endswith {'.bat','.exe','.cmd'}` will be true for any filename which ends with these extensions.) There are justify methods and whitespace trimming functions like `strip`. > stringx.import() > ('bonzo.dog'):endswith {'.dog','.cat'} true > ('bonzo.txt'):endswith {'.dog','.cat'} false > ('bonzo.cat'):endswith {'.dog','.cat'} true > (' stuff'):ljust(20,'+') '++++++++++++++ stuff' > (' stuff '):lstrip() 'stuff ' > (' stuff '):rstrip() ' stuff' > (' stuff '):strip() 'stuff' > for s in ('one\ntwo\nthree\n'):lines() do print(s) end one two three Most of these can be fairly easily implemented using the Lua string library, which is more general and powerful. But they are convenient operations to have easily at hand. Note that can be injected into the `string` table if you use `stringx.import`, but a simple alias like `local stringx = require 'pl.stringx'` is preferrable. This is the recommended practice when writing modules for consumption by other people, since it is bad manners to change the global state of the rest of the system. Magic may be used for convenience, but there is always a price. ### String Templates @lookup pl.text Another borrowing from Python, string templates allow you to substitute values looked up in a table: local Template = require ('pl.text').Template t = Template('${here} is the $answer') print(t:substitute {here = 'Lua', answer = 'best'}) ==> Lua is the best '$ variables' can optionally have curly braces; this form is useful if you are glueing text together to make variables, e.g `${prefix}_name_${postfix}`. The `substitute` method will throw an error if a $ variable is not found in the table, and the `safe_substitute` method will not. The Lua implementation has an extra method, `indent_substitute` which is very useful for inserting blocks of text, because it adjusts indentation. Consider this example: -- testtemplate.lua local Template = require ('pl.text').Template t = Template [[ for i = 1,#$t do $body end ]] body = Template [[ local row = $t[i] for j = 1,#row do fun(row[j]) end ]] print(t:indent_substitute {body=body,t='tbl'}) And the output is: for i = 1,#tbl do local row = tbl[i] for j = 1,#row do fun(row[j]) end end `indent_substitute` can substitute templates, and in which case they themselves will be substituted using the given table. So in this case, `$t` was substituted twice. `pl.text` also has a number of useful functions like `dedent`, which strips all the initial indentation from a multiline string. As in Python, this is useful for preprocessing multiline strings if you like indenting them with your code. The function `wrap` is passed a long string (a _paragraph_) and returns a list of lines that fit into a desired line width. As an extension, there is also `indent` for indenting multiline strings. New in Penlight with the 0.9 series is `text.format_operator`. Calling this enables Python-style string formating using the modulo operator `%`: > text.format_operator() > = '%s[%d]' % {'dog',1} dog[1] So in its simplest form it saves the typing involved with `string.format`; it will also expand `$` variables using named fields: > = '$animal[$num]' % {animal='dog',num=1} dog[1] As with `stringx.import` you have to do this explicitly, since all strings share the same metatable. But in your own scripts you can feel free to do this. ### Another Style of Template A new module is `template`, which is a version of Rici Lake's [Lua Preprocessor](http://lua-users.org/wiki/SlightlyLessSimpleLuaPreprocessor). This allows you to mix Lua code with your templates in a straightforward way. There are only two rules: - Lines begining with `#` are Lua - Otherwise, anything inside `$()` is a Lua expression. So a template generating an HTML list would look like this: Assume the text is inside `tmpl`, then the template can be expanded using: local template = require 'pl.template' local my_env = { ipairs = ipairs, T = {'one','two','three'} } res = template.substitute(tmpl, my_env) and we get There is a single function, `template.substitute` which is passed a template string and an environment table. This table may contain some special fields, like `\_parent` which can be set to a table representing a 'fallback' environment in case a symbol was not found. `\_brackets` is usually '()' and `\_escape` is usually '#' but it's sometimes necessary to redefine these if the defaults interfere with the target language - for instance, `$(V)` has another meaning in Make, and `#` means a preprocessor line in C/C++. Finally, if something goes wrong, passing `_debug` will cause the intermediate Lua code to be dumped if there's a problem. Here is a C code generation example; something that could easily be extended to be a minimal Lua extension skeleton generator. local subst = require 'pl.template'.substitute local templ = [[ #include #include #include > for _,f in ipairs(mod) do static int l_$(f.name) (lua_State *L) { } > end static const luaL_reg $(mod.name)[] = { > for _,f in ipairs(mod) do {"$(f.name)",l_$(f.name)}, > end {NULL,NULL} }; int luaopen_$(mod.name) { luaL_register (L, "$(mod.name)", $(mod.name)); return 1; } ]] print(subst(templ,{ _escape = '>', ipairs = ipairs, mod = { name = 'baggins'; {name='frodo'}, {name='bilbo'} } })) ### File-style I/O on Strings `pl.stringio` provides just three functions; `stringio.open` is passed a string, and returns a file-like object for reading. It supports a `read` method, which takes the same arguments as standard file objects: > f = stringio.open 'first line\n10 20 30\n' > = f:read() first line > = f:read('*n','*n','*n') 10 20 30 `lines` and `seek` are also supported. `stringio.lines` is a useful short-cut for iterating over all the lines in a string. `stringio.create` creates a writeable file-like object. You then use `write` to this stream, and finally extract the builded string using `value`. This 'string builder' pattern is useful for efficiently creating large strings. Penlight-1.12.0/docs_topics/04-paths.md000066400000000000000000000151051416703176500175540ustar00rootroot00000000000000## Paths and Directories ### Working with Paths Programs should not depend on quirks of your operating system. They will be harder to read, and need to be ported for other systems. The worst of course is hardcoding paths like 'c:\\' in programs, and wondering why Vista complains so much. But even something like `dir..'\\'..file` is a problem, since Unix can't understand backslashes in this way. `dir..'/'..file` is _usually_ portable, but it's best to put this all into a simple function, `path.join`. If you consistently use `path.join`, then it's much easier to write cross-platform code, since it handles the directory separator for you. `pl.path` provides the same functionality as Python's `os.path` module (11.1). > p = 'c:\\bonzo\\DOG.txt' > = path.normcase (p) ---> only makes sense on Windows c:\bonzo\dog.txt > = path.splitext (p) c:\bonzo\DOG .txt > = path.extension (p) .txt > = path.basename (p) DOG.txt > = path.exists(p) false > = path.join ('fred','alice.txt') fred\alice.txt > = path.exists 'pretty.lua' true > = path.getsize 'pretty.lua' 2125 > = path.isfile 'pretty.lua' true > = path.isdir 'pretty.lua' false It is very important for all programmers, not just on Unix, to only write to where they are allowed to write. `path.expanduser` will expand '~' (tilde) into the home directory. Depending on your OS, this will be a guaranteed place where you can create files: > = path.expanduser '~/mydata.txt' 'C:\Documents and Settings\SJDonova/mydata.txt' > = path.expanduser '~/mydata.txt' /home/sdonovan/mydata.txt Under Windows, `os.tmpname` returns a path which leads to your drive root full of temporary files. (And increasingly, you do not have access to this root folder.) This is corrected by `path.tmpname`, which uses the environment variable TMP: > os.tmpname() -- not a good place to put temporary files! '\s25g.' > path.tmpname() 'C:\DOCUME~1\SJDonova\LOCALS~1\Temp\s25g.1' A useful extra function is `pl.path.package_path`, which will tell you the path of a particular Lua module. So on my system, `package_path('pl.path')` returns 'C:\Program Files\Lua\5.1\lualibs\pl\path.lua', and `package_path('ifs')` returns 'C:\Program Files\Lua\5.1\clibs\lfs.dll'. It is implemented in terms of `package.searchpath`, which is a new function in Lua 5.2 which has been implemented for Lua 5.1 in Penlight. ### File Operations `pl.file` is a new module that provides more sensible names for common file operations. For instance, `file.read` and `file.write` are aliases for `utils.readfile` and `utils.writefile`. Smaller files can be efficiently read and written in one operation. `file.read` is passed a filename and returns the contents as a string, if successful; if not, then it returns `nil` and the actual error message. There is an optional boolean parameter if you want the file to be read in binary mode (this makes no difference on Unix but remains important with Windows.) In previous versions of Penlight, `utils.readfile` would read standard input if the file was not specified, but this can lead to nasty bugs; use `io.read '*a'` to grab all of standard input. Similarly, `file.write` takes a filename and a string which will be written to that file. For example, this little script converts a file into upper case: require 'pl' assert(#arg == 2, 'supply two filenames') text = assert(file.read(arg[1])) assert(file.write(arg[2],text:upper())) Copying files is suprisingly tricky. `file.copy` and `file.move` attempt to use the best implementation possible. On Windows, they link to the API functions `CopyFile` and `MoveFile`, but only if the `alien` package is installed (this is true for Lua for Windows.) Otherwise, the system copy command is used. This can be ugly when writing Windows GUI applications, because of the dreaded flashing black-box problem with launching processes. ### Directory Operations `pl.dir` provides some useful functions for working with directories. `fnmatch` will match a filename against a shell pattern, and `filter` will return any files in the supplied list which match the given pattern, which correspond to the functions in the Python `fnmatch` module. `getdirectories` will return all directories contained in a directory, and `getfiles` will return all files in a directory which match a shell pattern. These functions return the files as a table, unlike `lfs.dir` which returns an iterator.) `dir.makepath` can create a full path, creating subdirectories as necessary; `rmtree` is the Nuclear Option of file deleting functions, since it will recursively clear out and delete all directories found begining at a path (there is a similar function with this name in the Python `shutils` module.) > = dir.makepath 't\\temp\\bonzo' > = path.isdir 't\\temp\\bonzo' true > = dir.rmtree 't' `dir.rmtree` depends on `dir.walk`, which is a powerful tool for scanning a whole directory tree. Here is the implementation of `dir.rmtree`: --- remove a whole directory tree. -- @param path A directory path function dir.rmtree(fullpath) for root,dirs,files in dir.walk(fullpath) do for i,f in ipairs(files) do os.remove(path.join(root,f)) end lfs.rmdir(root) end end `dir.clonetree` clones directory trees. The first argument is a path that must exist, and the second path is the path to be cloned. (Note that this path cannot be _inside_ the first path, since this leads to madness.) By default, it will then just recreate the directory structure. You can in addition provide a function, which will be applied for all files found. -- make a copy of my libs folder require 'pl' p1 = [[d:\dev\lua\libs]] p2 = [[D:\dev\lua\libs\..\tests]] dir.clonetree(p1,p2,dir.copyfile) A more sophisticated version, which only copies files which have been modified: -- p1 and p2 as before, or from arg[1] and arg[2] dir.clonetree(p1,p2,function(f1,f2) local res local t1,t2 = path.getmtime(f1),path.getmtime(f2) -- f2 might not exist, so be careful about t2 if not t2 or t1 > t2 then res = dir.copyfile(f1,f2) end return res -- indicates successful operation end) `dir.clonetree` uses `path.common_prefix`. With `p1` and `p2` defined above, the common path is 'd:\dev\lua'. So 'd:\dev\lua\libs\testfunc.lua' is copied to 'd:\dev\lua\test\testfunc.lua', etc. If you need to find the common path of list of files, then `tablex.reduce` will do the job: > p3 = [[d:\dev]] > = tablex.reduce(path.common_prefix,{p1,p2,p3}) 'd:\dev' Penlight-1.12.0/docs_topics/05-dates.md000066400000000000000000000062461416703176500175440ustar00rootroot00000000000000## Date and Time NOTE: the Date module is deprecated ### Creating and Displaying Dates The `Date` class provides a simplified way to work with [date and time](http://www.lua.org/pil/22.1.html) in Lua; it leans heavily on the functions `os.date` and `os.time`. A `Date` object can be constructed from a table, just like with `os.time`. Methods are provided to get and set the various parts of the date. > d = Date {year = 2011, month = 3, day = 2 } > = d 2011-03-02 12:00:00 > = d:month(),d:year(),d:day() 3 2011 2 > d:month(4) > = d 2011-04-02 12:00:00 > d:add {day=1} > = d 2011-04-03 12:00:00 `add` takes a table containing one of the date table fields. > = d:weekday_name() Sun > = d:last_day() 2011-04-30 12:00:00 > = d:month_name(true) April There is a default conversion to text for date objects, but `Date.Format` gives you full control of the format for both parsing and displaying dates: > iso = Date.Format 'yyyy-mm-dd' > d = iso:parse '2010-04-10' > amer = Date.Format 'mm/dd/yyyy' > = amer:tostring(d) 04/10/2010 With the 0.9.7 relase, the `Date` constructor has become more flexible. You may omit any of the 'year', 'month' or 'day' fields: > = Date { year = 2008 } 2008-01-01 12:00:00 > = Date { month = 3 } 2011-03-01 12:00:00 > = Date { day = 20 } 2011-10-20 12:00:00 > = Date { hour = 14, min = 30 } 2011-10-13 14:30:00 If 'year' is omitted, then the current year is assumed, and likewise for 'month'. To set the time on such a partial date, you can use the fact that the 'setter' methods return the date object and so you can 'chain' these methods. > d = Date { day = 03 } > = d:hour(18):min(30) 2011-10-03 18:30:00 Finally, `Date` also now accepts positional arguments: > = Date(2011,10,3) 2011-10-03 12:00:00 > = Date(2011,10,3,18,30,23) 2011-10-03 18:30:23 `Date.format` has been extended. If you construct an instance without a pattern, then it will try to match against a set of known formats. This is useful for human-input dates since keeping to a strict format is not one of the strong points of users. It assumes that there will be a date, and then a date. > df = Date.Format() > = df:parse '5.30pm' 2011-10-13 17:30:00 > = df:parse '1730' nil day out of range: 1730 is not between 1 and 31 > = df:parse '17.30' 2011-10-13 17:30:00 > = df:parse 'mar' 2011-03-01 12:00:00 > = df:parse '3 March' 2011-03-03 12:00:00 > = df:parse '15 March' 2011-03-15 12:00:00 > = df:parse '15 March 2008' 2008-03-15 12:00:00 > = df:parse '15 March 2008 1.30pm' 2008-03-15 13:30:00 > = df:parse '2008-10-03 15:30:23' 2008-10-03 15:30:23 ISO date format is of course a good idea if you need to deal with users from different countries. Here is the default behaviour for 'short' dates: > = df:parse '24/02/12' 2012-02-24 12:00:00 That's not what Americans expect! It's tricky to work out in a cross-platform way exactly what the expected format is, so there is an explicit flag: > df:US_order(true) > = df:parse '9/11/01' 2001-11-09 12:00:00 Penlight-1.12.0/docs_topics/06-data.md000066400000000000000000001172251416703176500173560ustar00rootroot00000000000000## Data ### Reading Data Files The first thing to consider is this: do you actually need to write a custom file reader? And if the answer is yes, the next question is: can you write the reader in as clear a way as possible? Correctness, Robustness, and Speed; pick the first two and the third can be sorted out later, _if necessary_. A common sort of data file is the configuration file format commonly used on Unix systems. This format is often called a _property_ file in the Java world. # Read timeout in seconds read.timeout=10 # Write timeout in seconds write.timeout=10 Here is a simple Lua implementation: -- property file parsing with Lua string patterns props = [] for line in io.lines() do if line:find('#',1,true) ~= 1 and not line:find('^%s*$') then local var,value = line:match('([^=]+)=(.*)') props[var] = value end end Very compact, but it suffers from a similar disease in equivalent Perl programs; it uses odd string patterns which are 'lexically noisy'. Noisy code like this slows the casual reader down. (For an even more direct way of doing this, see the next section, 'Reading Configuration Files') Another implementation, using the Penlight libraries: -- property file parsing with extended string functions require 'pl' stringx.import() props = [] for line in io.lines() do if not line:startswith('#') and not line:isspace() then local var,value = line:splitv('=') props[var] = value end end This is more self-documenting; it is generally better to make the code express the _intention_, rather than having to scatter comments everywhere - comments are necessary, of course, but mostly to give the higher view of your intention that cannot be expressed in code. It is slightly slower, true, but in practice the speed of this script is determined by I/O, so further optimization is unnecessary. ### Reading Unstructured Text Data Text data is sometimes unstructured, for example a file containing words. The `pl.input` module has a number of functions which makes processing such files easier. For example, a script to count the number of words in standard input using `import.words`: -- countwords.lua require 'pl' local k = 1 for w in input.words(io.stdin) do k = k + 1 end print('count',k) Or this script to calculate the average of a set of numbers using `input.numbers`: -- average.lua require 'pl' local k = 1 local sum = 0 for n in input.numbers(io.stdin) do sum = sum + n k = k + 1 end print('average',sum/k) These scripts can be improved further by _eliminating loops_ In the last case, there is a perfectly good function `seq.sum` which can already take a sequence of numbers and calculate these numbers for us: -- average2.lua require 'pl' local total,n = seq.sum(input.numbers()) print('average',total/n) A further simplification here is that if `numbers` or `words` are not passed an argument, they will grab their input from standard input. The first script can be rewritten: -- countwords2.lua require 'pl' print('count',seq.count(input.words())) A useful feature of a sequence generator like `numbers` is that it can read from a string source. Here is a script to calculate the sums of the numbers on each line in a file: -- sums.lua for line in io.lines() do print(seq.sum(input.numbers(line)) end ### Reading Columnar Data It is very common to find data in columnar form, either space or comma-separated, perhaps with an initial set of column headers. Here is a typical example: EventID Magnitude LocationX LocationY LocationZ 981124001 2.0 18988.4 10047.1 4149.7 981125001 0.8 19104.0 9970.4 5088.7 981127003 0.5 19012.5 9946.9 3831.2 ... `input.fields` is designed to extract several columns, given some delimiter (default to whitespace). Here is a script to calculate the average X location of all the events: -- avg-x.lua require 'pl' io.read() -- skip the header line local sum,count = seq.sum(input.fields {3}) print(sum/count) `input.fields` is passed either a field count, or a list of column indices, starting at one as usual. So in this case we're only interested in column 3. If you pass it a field count, then you get every field up to that count: for id,mag,locX,locY,locZ in input.fields (5) do .... end `input.fields` by default tries to convert each field to a number. It will skip lines which clearly don't match the pattern, but will abort the script if there are any fields which cannot be converted to numbers. The second parameter is a delimiter, by default spaces. ' ' is understood to mean 'any number of spaces', i.e. '%s+'. Any Lua string pattern can be used. The third parameter is a _data source_, by default standard input (defined by `input.create_getter`.) It assumes that the data source has a `read` method which brings in the next line, i.e. it is a 'file-like' object. As a special case, a string will be split into its lines: > for x,y in input.fields(2,' ','10 20\n30 40\n') do print(x,y) end 10 20 30 40 Note the default behaviour for bad fields, which is to show the offending line number: > for x,y in input.fields(2,' ','10 20\n30 40x\n') do print(x,y) end 10 20 line 2: cannot convert '40x' to number This behaviour of `input.fields` is appropriate for a script which you want to fail immediately with an appropriate _user_ error message if conversion fails. The fourth optional parameter is an options table: `{no_fail=true}` means that conversion is attempted but if it fails it just returns the string, rather as AWK would operate. You are then responsible for checking the type of the returned field. `{no_convert=true}` switches off conversion altogether and all fields are returned as strings. @lookup pl.data Sometimes it is useful to bring a whole dataset into memory, for operations such as extracting columns. Penlight provides a flexible reader specifically for reading this kind of data, using the `data` module. Given a file looking like this: x,y 10,20 2,5 40,50 Then `data.read` will create a table like this, with each row represented by a sublist: > t = data.read 'test.txt' > pretty.dump(t) {{10,20},{2,5},{40,50},fieldnames={'x','y'},delim=','} You can now analyze this returned table using the supplied methods. For instance, the method `column_by_name` returns a table of all the values of that column. -- testdata.lua require 'pl' d = data.read('fev.txt') for _,name in ipairs(d.fieldnames) do local col = d:column_by_name(name) if type(col[1]) == 'number' then local total,n = seq.sum(col) utils.printf("Average for %s is %f\n",name,total/n) end end `data.read` tries to be clever when given data; by default it expects a first line of column names, unless any of them are numbers. It tries to deduce the column delimiter by looking at the first line. Sometimes it guesses wrong; these things can be specified explicitly. The second optional parameter is an options table: can override `delim` (a string pattern), `fieldnames` (a list or comma-separated string), specify `no_convert` (default is to convert), numfields (indices of columns known to be numbers, as a list) and `thousands_dot` (when the thousands separator in Excel CSV is '.') A very powerful feature is a way to execute SQL-like queries on such data: -- queries on tabular data require 'pl' local d = data.read('xyz.txt') local q = d:select('x,y,z where x > 3 and z < 2 sort by y') for x,y,z in q do print(x,y,z) end Please note that the format of queries is restricted to the following syntax: FIELDLIST [ 'where' CONDITION ] [ 'sort by' FIELD [asc|desc]] Any valid Lua code can appear in `CONDITION`; remember it is _not_ SQL and you have to use `==` (this warning comes from experience.) For this to work, _field names must be Lua identifiers_. So `read` will massage fieldnames so that all non-alphanumeric chars are replaced with underscores. However, the `original_fieldnames` field always contains the original un-massaged fieldnames. `read` can handle standard CSV files fine, although doesn't try to be a full-blown CSV parser. With the `csv=true` option, it's possible to have double-quoted fields, which may contain commas; then trailing commas become significant as well. Spreadsheet programs are not always the best tool to process such data, strange as this might seem to some people. This is a toy CSV file; to appreciate the problem, imagine thousands of rows and dozens of columns like this: Department Name,Employee ID,Project,Hours Booked sales,1231,overhead,4 sales,1255,overhead,3 engineering,1501,development,5 engineering,1501,maintenance,3 engineering,1433,maintenance,10 The task is to reduce the dataset to a relevant set of rows and columns, perhaps do some processing on row data, and write the result out to a new CSV file. The `write_row` method uses the delimiter to write the row to a file; `Data.select_row` is like `Data.select`, except it iterates over _rows_, not fields; this is necessary if we are dealing with a lot of columns! names = {[1501]='don',[1433]='dilbert'} keepcols = {'Employee_ID','Hours_Booked'} t:write_row (outf,{'Employee','Hours_Booked'}) q = t:select_row { fields=keepcols, where=function(row) return row[1]=='engineering' end } for row in q do row[1] = names[row[1]] t:write_row(outf,row) end `Data.select_row` and `Data.select` can be passed a table specifying the query; a list of field names, a function defining the condition and an optional parameter `sort_by`. It isn't really necessary here, but if we had a more complicated row condition (such as belonging to a specified set) then it is not generally possible to express such a condition as a query string, without resorting to hackery such as global variables. With 1.0.3, you can specify explicit conversion functions for selected columns. For instance, this is a log file with a Unix date stamp: Time Message 1266840760 +# EE7C0600006F0D00C00F06010302054000000308010A00002B00407B00 1266840760 closure data 0.000000 1972 1972 0 1266840760 ++ 1266840760 EE 1 1266840760 +# EE7C0600006F0D00C00F06010302054000000408020A00002B00407B00 1266840764 closure data 0.000000 1972 1972 0 We would like the first column as an actual date object, so the `convert` field sets an explicit conversion for column 1. (Note that we have to explicitly convert the string to a number first.) Date = require 'pl.Date' function date_convert (ds) return Date(tonumber(ds)) end d = data.read(f,{convert={[1]=date_convert},last_field_collect=true}) This gives us a two-column dataset, where the first column contains `Date` objects and the second column contains the rest of the line. Queries can then easily pick out events on a day of the week: q = d:select "Time,Message where Time:weekday_name()=='Sun'" Data does not have to come from files, nor does it necessarily come from the lab or the accounts department. On Linux, `ps aux` gives you a full listing of all processes running on your machine. It is straightforward to feed the output of this command into `data.read` and perform useful queries on it. Notice that non-identifier characters like '%' get converted into underscores: require 'pl' f = io.popen 'ps aux' s = data.read (f,{last_field_collect=true}) f:close() print(s.fieldnames) print(s:column_by_name 'USER') qs = 'COMMAND,_MEM where _MEM > 5 and USER=="steve"' for name,mem in s:select(qs) do print(mem,name) end I've always been an admirer of the AWK programming language; with `filter` you can get Lua programs which are just as compact: -- printxy.lua require 'pl' data.filter 'x,y where x > 3' It is common enough to have data files without headers of field names. `data.read` makes a special exception for such files if all fields are numeric. Since there are no column names to use in query expressions, you can use AWK-like column indexes, e.g. '$1,$2 where $1 > 3'. I have a little executable script on my system called `lf` which looks like this: #!/usr/bin/env lua require 'pl.data'.filter(arg[1]) And it can be used generally as a filter command to extract columns from data. (The column specifications may be expressions or even constants.) $ lf '$1,$5/10' < test.dat (As with AWK, please note the single-quotes used in this command; this prevents the shell trying to expand the column indexes. If you are on Windows, then you must quote the expression in double-quotes so it is passed as one argument to your batch file.) As a tutorial resource, have a look at `test-data.lua` in the PL tests directory for other examples of use, plus comments. The data returned by `read` or constructed by `Data.copy_select` from a query is basically just an array of rows: `{{1,2},{3,4}}`. So you may use `read` to pull in any array-like dataset, and process with any function that expects such a implementation. In particular, the functions in `array2d` will work fine with this data. In fact, these functions are available as methods; e.g. `array2d.flatten` can be called directly like so to give us a one-dimensional list: v = data.read('dat.txt'):flatten() The data is also in exactly the right shape to be treated as matrices by [LuaMatrix](http://lua-users.org/wiki/LuaMatrix): > matrix = require 'matrix' > m = matrix(data.read 'mat.txt') > = m 1 0.2 0.3 0.2 1 0.1 0.1 0.2 1 > = m^2 -- same as m*m 1.07 0.46 0.62 0.41 1.06 0.26 0.24 0.42 1.05 `write` will write matrices back to files for you. Finally, for the curious, the global variable `_DEBUG` can be used to print out the actual iterator function which a query generates and dynamically compiles. By using code generation, we can get pretty much optimal performance out of arbitrary queries. > lua -lpl -e "_DEBUG=true" -e "data.filter 'x,y where x > 4 sort by x'" < test.txt return function (t) local i = 0 local v local ls = {} for i,v in ipairs(t) do if v[1] > 4 then ls[#ls+1] = v end end table.sort(ls,function(v1,v2) return v1[1] < v2[1] end) local n = #ls return function() i = i + 1 v = ls[i] if i > n then return end return v[1],v[2] end end 10,20 40,50 ### Reading Configuration Files The `config` module provides a simple way to convert several kinds of configuration files into a Lua table. Consider the simple example: # test.config # Read timeout in seconds read.timeout=10 # Write timeout in seconds write.timeout=5 #acceptable ports ports = 1002,1003,1004 This can be easily brought in using `config.read` and the result shown using `pretty.write`: -- readconfig.lua local config = require 'pl.config' local pretty= require 'pl.pretty' local t = config.read(arg[1]) print(pretty.write(t)) and the output of `lua readconfig.lua test.config` is: { ports = { 1002, 1003, 1004 }, write_timeout = 5, read_timeout = 10 } That is, `config.read` will bring in all key/value pairs, ignore # comments, and ensure that the key names are proper Lua identifiers by replacing non-identifier characters with '_'. If the values are numbers, then they will be converted. (So the value of `t.write_timeout` is the number 5). In addition, any values which are separated by commas will be converted likewise into an array. Any line can be continued with a backslash. So this will all be considered one line: names=one,two,three, \ four,five,six,seven, \ eight,nine,ten Windows-style INI files are also supported. The section structure of INI files translates naturally to nested tables in Lua: ; test.ini [timeouts] read=10 ; Read timeout in seconds write=5 ; Write timeout in seconds [portinfo] ports = 1002,1003,1004 The output is: { portinfo = { ports = { 1002, 1003, 1004 } }, timeouts = { write = 5, read = 10 } } You can now refer to the write timeout as `t.timeouts.write`. As a final example of the flexibility of `config.read`, if passed this simple comma-delimited file one,two,three 10,20,30 40,50,60 1,2,3 it will produce the following table: { { "one", "two", "three" }, { 10, 20, 30 }, { 40, 50, 60 }, { 1, 2, 3 } } `config.read` isn't designed to read all CSV files in general, but intended to support some Unix configuration files not structured as key-value pairs, such as '/etc/passwd'. This function is intended to be a Swiss Army Knife of configuration readers, but it does have to make assumptions, and you may not like them. So there is an optional extra parameter which allows some control, which is table that may have the following fields: { variablilize = true, convert_numbers = tonumber, trim_space = true, list_delim = ',', trim_quotes = true, ignore_assign = false, keysep = '=', smart = false, } `variablilize` is the option that converted `write.timeout` in the first example to the valid Lua identifier `write_timeout`. If `convert_numbers` is true, then an attempt is made to convert any string that starts like a number. You can specify your own function (say one that will convert a string like '5224 kb' into a number.) `trim_space` ensures that there is no starting or trailing whitespace with values, and `list_delim` is the character that will be used to decide whether to split a value up into a list (it may be a Lua string pattern such as '%s+'.) For instance, the password file in Unix is colon-delimited: t = config.read('/etc/passwd',{list_delim=':'}) This produces the following output on my system (only last two lines shown): { ... { "user", "x", "1000", "1000", "user,,,", "/home/user", "/bin/bash" }, { "sdonovan", "x", "1001", "1001", "steve donovan,28,,", "/home/sdonovan", "/bin/bash" } } You can get this into a more sensible format, where the usernames are the keys, with this (the `tablex.pairmap` function must return value, key!) t = tablex.pairmap(function(k,v) return v,v[1] end,t) and you get: { ... sdonovan = { "sdonovan", "x", "1001", "1001", "steve donovan,28,,", "/home/sdonovan", "/bin/bash" } ... } Many common Unix configuration files can be read by tweaking these parameters. For `/etc/fstab`, the options `{list_delim='%s+',ignore_assign=true}` will correctly separate the columns. It's common to find 'KEY VALUE' assignments in files such as `/etc/ssh/ssh_config`; the options `{keysep=' '}` make `config.read` return a table where each KEY has a value VALUE. Files in the Linux `procfs` usually use ':` as the field delimiter: > t = config.read('/proc/meminfo',{keysep=':'}) > = t.MemFree 220140 kB That result is a string, since `tonumber` doesn't like it, but defining the `convert_numbers` option as `function(s) return tonumber((s:gsub(' kB$',''))) end` will get the memory figures as actual numbers in the result. (The extra parentheses are necessary so that `tonumber` only gets the first result from `gsub`). From `tests/test-config.lua': testconfig([[ MemTotal: 1024748 kB MemFree: 220292 kB ]], { MemTotal = 1024748, MemFree = 220292 }, { keysep = ':', convert_numbers = function(s) s = s:gsub(' kB$','') return tonumber(s) end } ) The `smart` option lets `config.read` make a reasonable guess for you; there are examples in `tests/test-config.lua`, but basically these common file formats (and those following the same pattern) can be processed directly in smart mode: 'etc/fstab', '/proc/XXXX/status', 'ssh_config' and 'pdatedb.conf'. Please note that `config.read` can be passed a _file-like object_; if it's not a string and supports the `read` method, then that will be used. For instance, to read a configuration from a string, use `stringio.open`. ### Lexical Scanning Although Lua's string pattern matching is very powerful, there are times when something more powerful is needed. `pl.lexer.scan` provides lexical scanners which _tokenize_ a string, classifying tokens into numbers, strings, etc. > lua -lpl Lua 5.1.4 Copyright (C) 1994-2008 Lua.org, PUC-Rio > tok = lexer.scan 'alpha = sin(1.5)' > = tok() iden alpha > = tok() = = > = tok() iden sin > = tok() ( ( > = tok() number 1.5 > = tok() ) ) > = tok() (nil) The scanner is a function, which is repeatedly called and returns the _type_ and _value_ of the token. Recognized basic types are 'iden','string','number', and 'space'. and everything else is represented by itself. Note that by default the scanner will skip any 'space' tokens. 'comment' and 'keyword' aren't applicable to the plain scanner, which is not language-specific, but a scanner which understands Lua is available. It recognizes the Lua keywords, and understands both short and long comments and strings. > for t,v in lexer.lua 'for i=1,n do' do print(t,v) end keyword for iden i = = number 1 , , iden n keyword do A lexical scanner is useful where you have highly-structured data which is not nicely delimited by newlines. For example, here is a snippet of a in-house file format which it was my task to maintain: points (818344.1,-20389.7,-0.1),(818337.9,-20389.3,-0.1),(818332.5,-20387.8,-0.1) ,(818327.4,-20388,-0.1),(818322,-20387.7,-0.1),(818316.3,-20388.6,-0.1) ,(818309.7,-20389.4,-0.1),(818303.5,-20390.6,-0.1),(818295.8,-20388.3,-0.1) ,(818290.5,-20386.9,-0.1),(818285.2,-20386.1,-0.1),(818279.3,-20383.6,-0.1) ,(818274,-20381.2,-0.1),(818274,-20380.7,-0.1); Here is code to extract the points using `pl.lexer`: -- assume 's' contains the text above... local lexer = require 'pl.lexer' local expecting = lexer.expecting local append = table.insert local tok = lexer.scan(s) local points = {} local t,v = tok() -- should be 'iden','points' while t ~= ';' do c = {} expecting(tok,'(') c.x = expecting(tok,'number') expecting(tok,',') c.y = expecting(tok,'number') expecting(tok,',') c.z = expecting(tok,'number') expecting(tok,')') t,v = tok() -- either ',' or ';' append(points,c) end The `expecting` function grabs the next token and if the type doesn't match, it throws an error. (`pl.lexer`, unlike other PL libraries, raises errors if something goes wrong, so you should wrap your code in `pcall` to catch the error gracefully.) The scanners all have a second optional argument, which is a table which controls whether you want to exclude spaces and/or comments. The default for `lexer.lua` is `{space=true,comments=true}`. There is a third optional argument which determines how string and number tokens are to be processsed. The ultimate highly-structured data is of course, program source. Here is a snippet from 'text-lexer.lua': require 'pl' lines = [[ for k,v in pairs(t) do if type(k) == 'number' then print(v) -- array-like case else print(k,v) end end ]] ls = List() for tp,val in lexer.lua(lines,{space=true,comments=true}) do assert(tp ~= 'space' and tp ~= 'comment') if tp == 'keyword' then ls:append(val) end end test.asserteq(ls,List{'for','in','do','if','then','else','end','end'}) Here is a useful little utility that identifies all common global variables found in a lua module (ignoring those declared locally for the moment): -- testglobal.lua require 'pl' local txt,err = utils.readfile(arg[1]) if not txt then return print(err) end local globals = List() for t,v in lexer.lua(txt) do if t == 'iden' and _G[v] then globals:append(v) end end pretty.dump(seq.count_map(globals)) Rather then dumping the whole list, with its duplicates, we pass it through `seq.count_map` which turns the list into a table where the keys are the values, and the associated values are the number of times those values occur in the sequence. Typical output looks like this: { type = 2, pairs = 2, table = 2, print = 3, tostring = 2, require = 1, ipairs = 4 } You could further pass this through `tablex.keys` to get a unique list of symbols. This can be useful when writing 'strict' Lua modules, where all global symbols must be defined as locals at the top of the file. For a more detailed use of `lexer.scan`, please look at `testxml.lua` in the examples directory. ### XML New in the 0.9.7 release is some support for XML. This is a large topic, and Penlight does not provide a full XML stack, which is properly the task of a more specialized library. #### Parsing and Pretty-Printing The semi-standard XML parser in the Lua universe is [lua-expat](http://matthewwild.co.uk/projects/luaexpat/). In particular, it has a function called `lxp.lom.parse` which will parse XML into the Lua Object Model (LOM) format. However, it does not provide a way to convert this data back into XML text. `xml.parse` will use this function, _if_ `lua-expat` is available, and otherwise switches back to a pure Lua parser originally written by Roberto Ierusalimschy. The resulting document object knows how to render itself as a string, which is useful for debugging: > d = xml.parse "alice" > = d alice > pretty.dump (d) { { "alice", attr = { "id", id = "1" }, tag = "node" }, attr = { }, tag = "nodes" } Looking at the actual shape of the data reveals the structure of LOM: * every element has a `tag` field with its name * plus a `attr` field which is a table containing the attributes as fields, and also as an array. It is always present. * the children of the element are the array part of the element, so `d[1]` is the first child of `d`, etc. It could be argued that having attributes also as the array part of `attr` is not essential (you cannot depend on attribute order in XML) but that's how it goes with this standard. `lua-expat` is another _soft dependency_ of Penlight; generally, the fallback parser is good enough for straightforward XML as is commonly found in configuration files, etc. `doc.basic_parse` is not intended to be a proper conforming parser (it's only sixty lines) but it handles simple kinds of documents that do not have comments or DTD directives. It is intelligent enough to ignore the ` = xml.tostring(d,'',' ') alice There is a fourth argument which is the _attribute indent_: > a = xml.parse "" > = xml.tostring(a,'',' ',' ') #### Parsing and Working with Configuration Files It's common to find configurations expressed with XML these days. It's straightforward to 'walk' the [LOM](http://matthewwild.co.uk/projects/luaexpat/lom.html) data and extract the data in the form you want: require 'pl' local config = [[ 1.3 10 bozo ]] local d,err = xml.parse(config) local t = {} for item in d:childtags() do t[item.tag] = item[1] end pretty.dump(t) ---> { beta = "10", alpha = "1.3", name = "bozo" } The only gotcha is that here we must use the `Doc:childtags` method, which will skip over any text elements. A more involved example is this excerpt from `serviceproviders.xml`, which is usually found at `/usr/share/mobile-broadband-provider-info/serviceproviders.xml` on Debian/Ubuntu Linux systems. d = xml.parse [[ ... Cell-c Cellcis 196.7.0.138 196.7.142.132 MTN 196.11.240.241 209.212.97.1 Vodacom 196.207.40.165 196.43.46.190 Unrestricted 196.207.32.69 196.43.45.190 Virgin Mobile 196.7.0.138 196.7.142.132 .... ]] Getting the names of the providers per-country is straightforward: local t = {} for country in d:childtags() do local providers = {} t[country.attr.code] = providers for provider in country:childtags() do table.insert(providers,provider:child_with_name('name'):get_text()) end end pretty.dump(t) --> { za = { "Cell-c", "MTN", "Vodacom", "Virgin Mobile" } .... } #### Generating XML with 'xmlification' This feature is inspired by the `htmlify` function used by [Orbit](http://keplerproject.github.com/orbit/) to simplify HTML generation, except that no function environment magic is used; the `tags` function returns a set of _constructors_ for elements of the given tag names. > nodes, node = xml.tags 'nodes, node' > = node 'alice' alice > = nodes { node {id='1','alice'}} alice The flexibility of Lua tables is very useful here, since both the attributes and the children of an element can be encoded naturally. The argument to these tag constructors is either a single value (like a string) or a table where the attributes are the named keys and the children are the array values. #### Generating XML using Templates A template is a little XML document which contains dollar-variables. The `subst` method on a document is fed an array of tables containing values for these variables. Note how the parent tag name is specified: > templ = xml.parse "$name" > = templ:subst {tag='nodes', {id=1,name='alice'},{id=2,name='john'}} alicejohn Substitution is very related to _filtering_ documents. One of the annoying things about XML is that it is a document markup language first, and a data language second. Standard parsers will assume you really care about all those extra text elements. Consider this fragment, which has been changed by a five-year old: T = [[ boops! whoops! ]] Conformant parsers will give you text elements with the line feed after `` although it makes handling the data more irritating. local function parse (str) return xml.parse(str,false,true) end Second argument means 'string, not file' and third argument means use the built-in Lua parser (instead of LuaExpat if available) which _by default_ is not interested in keeping such strings. How to remove the string `boops!`? `clone` (also called `filter` when called as a method) copies a LOM document. It can be passed a filter function, which is applied to each string found. The powerful thing about this is that this function receives structural information - the parent node, and whether this was a tag name, a text element or a attribute name: d = parse (T) c = d:filter(function(s,kind,parent) print(stringx.strip(s),kind,parent and parent.tag or '?') if kind == '*TEXT' and #parent > 1 then return nil end return s end) ---> weather *TAG ? boops! *TEXT weather current_conditions *TAG weather condition *TAG current_conditions $condition data condition temp_c *TAG current_conditions $temp data temp_c bo *TAG current_conditions whoops! *TEXT bo We can pull out 'boops' and not 'whoops' by discarding text elements which are not the single child of an element. #### Extracting Data using Templates Matching goes in the opposite direction. We have a document, and would like to extract values from it using a pattern. A common use of this is parsing the XML result of API queries. The [(undocumented and subsequently discontinued) Google Weather API](http://blog.programmableweb.com/2010/02/08/googles-secret-weather-api/) is a good example. Grabbing the result of `http://www.google.com/ig/api?weather=Johannesburg,ZA" we get something like this, after pretty-printing: .... Assume that the above XML has been read into `google`. The idea is to write a pattern looking like a template, and use it to extract some values of interest: t = [[ ]] local res, ret = google:match(t) pretty.dump(res) And the output is: { condition = "Clear", temp = "24" } The `match` method can be passed a LOM document or some text, which will be parsed first. But what if we need to extract values from repeated elements? Match templates may contain 'array matches' which are enclosed in '{{..}}': {{ }} And the match result is: { { low = "60", high = "89", day = "Sat", condition = "Clear", }, { low = "53", high = "86", day = "Sun", condition = "Clear", }, { low = "57", high = "87", day = "Mon", condition = "Clear", }, { low = "60", high = "84", day = "Tue", condition = "Clear", } } With this array of tables, you can use `tablex` or `List` to reshape into the desired form, if you choose. Just as with reading a Unix password file with `config`, you can make the array into a map of days to conditions using: `tablex.pairmap`('|k,v| v,v.day',conditions) (Here using the alternative string lambda option) However, xml matches can shape the structure of the output. By replacing the `day_of_week` line of the template with `` we get the same effect; `$_` is a special symbol that means that this captured value (or simply _capture_) becomes the key. Note that `$NUMBER` means a numerical index, so that `$1` is the first element of the resulting array, and so forth. You can mix numbered and named captures, but it's strongly advised to make the numbered captures form a proper array sequence (everything from `1` to `n` inclusive). `$0` has a special meaning; if it is the only capture (`{[0]='foo'}`) then the table is collapsed into 'foo'. {{ }} Now the result is: { Tue = { "60", "84", "Clear" }, Sun = { "53", "86", "Clear" }, Sat = { "60", "89", "Clear" }, Mon = { "57", "87", "Clear" } } Applying matches to this config file poses another problem, because the actual tags matched are themselves meaningful. 1.3 10 bozo So there are tag 'wildcards' which are element names ending with a hyphen. {{$value}} You will then get `{{alpha='1.3'},...}`. The most convenient format would be returned by this (note that `_-` behaves just like `$_`): {{<_->$0}} which would return `{alpha='1.3',beta='10',name='bozo'}`. We could play this game endlessly, and encode ways of converting captures, but the scheme is complex enough, and it's easy to do the conversion later local numbers = {alpha=true,beta=true} for k,v in pairs(res) do if numbers[v] then res[k] = tonumber(v) end end #### HTML Parsing HTML is an unusually degenerate form of XML, and Dennis Schridde has contributed a feature which makes parsing it easier. For instance, from the tests: doc = xml.parsehtml [[ Hello dolly
HTML is slack
]] asserteq(xml.tostring(doc),[[ Hello dolly
HTML is slack
]]) That is, all tags are converted to lowercase, and empty HTML elements like `br` are properly closed; attributes do not need to be quoted. Also, DOCTYPE directives and comments are skipped. For truly badly formed HTML, this is not the tool for you! Penlight-1.12.0/docs_topics/07-functional.md000066400000000000000000000502331416703176500206030ustar00rootroot00000000000000## Functional Programming ### Sequences @lookup pl.seq A Lua iterator (in its simplest form) is a function which can be repeatedly called to return a set of one or more values. The `for in` statement understands these iterators, and loops until the function returns `nil`. There are standard sequence adapters for tables in Lua (`ipairs` and `pairs`), and `io.lines` returns an iterator over all the lines in a file. In the Penlight libraries, such iterators are also called _sequences_. A sequence of single values (say from `io.lines`) is called _single-valued_, whereas the sequence defined by `pairs` is _double-valued_. `pl.seq` provides a number of useful iterators, and some functions which operate on sequences. At first sight this example looks like an attempt to write Python in Lua, (with the sequence being inclusive): > for i in seq.range(1,4) do print(i) end 1 2 3 4 But `range` is actually equivalent to Python's `xrange`, since it generates a sequence, not a list. To get a list, use `seq.copy(seq.range(1,10))`, which takes any single-value sequence and makes a table from the result. `seq.list` is like `ipairs` except that it does not give you the index, just the value. > for x in seq.list {1,2,3} do print(x) end 1 2 3 `enum` takes a sequence and turns it into a double-valued sequence consisting of a sequence number and the value, so `enum(list(ls))` is actually equivalent to `ipairs`. A more interesting example prints out a file with line numbers: for i,v in seq.enum(io.lines(fname)) do print(i..' '..v) end Sequences can be _combined_, either by 'zipping' them or by concatenating them. > for x,y in seq.zip(l1,l2) do print(x,y) end 10 1 20 2 30 3 > for x in seq.splice(l1,l2) do print(x) end 10 20 30 1 2 3 `seq.printall` is useful for printing out single-valued sequences, and provides some finer control over formating, such as a delimiter, the number of fields per line, and a format string to use (@see string.format) > seq.printall(seq.random(10)) 0.0012512588885159 0.56358531449324 0.19330423902097 .... > seq.printall(seq.random(10), ',', 4, '%4.2f') 0.17,0.86,0.71,0.51 0.30,0.01,0.09,0.36 0.15,0.17, `map` will apply a function to a sequence. > seq.printall(seq.map(string.upper, {'one','two'})) ONE TWO > seq.printall(seq.map('+', {10,20,30}, 1)) 11 21 31 `filter` will filter a sequence using a boolean function (often called a _predicate_). For instance, this code only prints lines in a file which are composed of digits: for l in seq.filter(io.lines(file), stringx.isdigit) do print(l) end The following returns a table consisting of all the positive values in the original table (equivalent to `tablex.filter(ls, '>', 0)`) ls = seq.copy(seq.filter(ls, '>', 0)) We're already encounted `seq.sum` when discussing `input.numbers`. This can also be expressed with `seq.reduce`: > seq.reduce(function(x,y) return x + y end, seq.list{1,2,3,4}) 10 `seq.reduce` applies a binary function in a recursive fashion, so that: reduce(op,{1,2,3}) => op(1,reduce(op,{2,3}) => op(1,op(2,3)) it's now possible to easily generate other cumulative operations; the standard operations declared in `pl.operator` are useful here: > ops = require 'pl.operator' > -- can also say '*' instead of ops.mul > = seq.reduce(ops.mul,input.numbers '1 2 3 4') 24 There are functions to extract statistics from a sequence of numbers: > l1 = List {10,20,30} > l2 = List {1,2,3} > = seq.minmax(l1) 10 30 > = seq.sum(l1) 60 3 It is common to get sequences where values are repeated, say the words in a file. `count_map` will take such a sequence and count the values, returning a table where the _keys_ are the unique values, and the value associated with each key is the number of times they occurred: > t = seq.count_map {'one','fred','two','one','two','two'} > = t {one=2,fred=1,two=3} This will also work on numerical sequences, but you cannot expect the result to be a proper list, i.e. having no 'holes'. Instead, you always need to use `pairs` to iterate over the result - note that there is a hole at index 5: > t = seq.count_map {1,2,4,2,2,3,4,2,6} > for k,v in pairs(t) do print(k,v) end 1 1 2 4 3 1 4 2 6 1 `unique` uses `count_map` to return a list of the unique values, that is, just the keys of the resulting table. `last` turns a single-valued sequence into a double-valued sequence with the current value and the last value: > for current,last in seq.last {10,20,30,40} do print (current,last) end 20 10 30 20 40 30 This makes it easy to do things like identify repeated lines in a file, or construct differences between values. `filter` can handle double-valued sequences as well, so one could filter such a sequence to only return cases where the current value is less than the last value by using `operator.lt` or just '<'. This code then copies the resulting code into a table. > ls = {10,9,10,3} > = seq.copy(seq.filter(seq.last(s),'<')) {9,3} ### Sequence Wrappers The functions in `pl.seq` cover the common patterns when dealing with sequences, but chaining these functions together can lead to ugly code. Consider the last example of the previous section; `seq` is repeated three times and the resulting expression has to be read right-to-left. The first issue can be helped by local aliases, so that the expression becomes `copy(filter(last(s),'<'))` but the second issue refers to the somewhat unnatural order of functional application. We tend to prefer reading operations from left to right, which is one reason why object-oriented notation has become popular. Sequence adapters allow this expression to be written like so: seq(s):last():filter('<'):copy() With this notation, the operation becomes a chain of method calls running from left to right. 'Sequence' is not a basic Lua type, they are generally functions or callable objects. The expression `seq(s)` wraps a sequence in a _sequence wrapper_, which is an object which understands all the functions in `pl.seq` as methods. This object then explicitly represents sequences. As a special case, the constructor (which is when you call the table `seq`) will make a wrapper for a plain list-like table. Here we apply the length operator to a sequence of strings, and print them out. > seq{'one','tw','t'} :map '#' :printall() 3 2 1 As a convenience, there is a function `seq.lines` which behaves just like `io.lines` except it wraps the result as an explicit sequence type. This takes the first 10 lines from standard input, makes it uppercase, turns it into a sequence with a count and the value, glues these together with the concatenation operator, and finally prints out the sequence delimited by a newline. seq.lines():take(10):upper():enum():map('..'):printall '\n' Note the method `upper`, which is not a `seq` function. if an unknown method is called, sequence wrappers apply that method to all the values in the sequence (this is implicit use of `mapmethod`) It is straightforward to create custom sequences that can be used in this way. On Unix, `/dev/random` gives you an _endless_ sequence of random bytes, so we use `take` to limit the sequence, and then `map` to scale the result into the desired range. The key step is to use `seq` to wrap the iterator function: -- random.lua local seq = require 'pl.seq' function dev_random() local f = io.open('/dev/random') local byte = string.byte return seq(function() -- read two bytes into a string and convert into a 16-bit number local s = f:read(2) return byte(s,1) + 256*byte(s,2) end) end -- print 10 random numbers from 0 to 1 ! dev_random():take(10):map('%',100):map('/',100):printall ',' Another Linux one-liner depends on the `/proc` filesystem and makes a list of all the currently running processes: pids = seq(lfs.dir '/proc'):filter(stringx.isdigit):map(tonumber):copy() This version of Penlight has an experimental feature which relies on the fact that _all_ Lua types can have metatables, including functions. This makes _implicit sequence wrapping_ possible: > seq.import() > seq.random(5):printall(',',5,'%4.1f') 0.0, 0.1, 0.4, 0.1, 0.2 This avoids the awkward `seq(seq.random(5))` construction. Or the iterator can come from somewhere else completely: > ('one two three'):gfind('%a+'):printall(',') one,two,three, After `seq.import`, it is no longer necessary to explicitly wrap sequence functions. But there is a price to pay for this convenience. _Every_ function is affected, so that any function can be used, appropriate or not: > math.sin:printall() ..seq.lua:287: bad argument #1 to '(for generator)' (number expected, got nil) > a = tostring > = a:find(' ') function: 0042C920 What function is returned? It's almost certain to be something that makes no sense in the current context. So implicit sequences may make certain kinds of programming mistakes harder to catch - they are best used for interactive exploration and small scripts.
### List Comprehensions List comprehensions are a compact way to create tables by specifying their elements. In Python, you can say this: ls = [x for x in range(5)] # == [0,1,2,3,4] In Lua, using `pl.comprehension`: > C = require('pl.comprehension').new() > = C ('x for x=1,10') () {1,2,3,4,5,6,7,8,9,10} `C` is a function which compiles a list comprehension _string_ into a _function_. In this case, the function has no arguments. The parentheses are redundant for a function taking a string argument, so this works as well: > = C 'x^2 for x=1,4' () {1,4,9,16} > = C '{x,x^2} for x=1,4' () {{1,1},{2,4},{3,9},{4,16}} Note that the expression can be _any_ function of the variable `x`! The basic syntax so far is ` for `, where `` can be anything that the Lua `for` statement understands. `` can also just be the variable, in which case the values will come from the _argument_ of the comprehension. Here I'm emphasizing that a comprehension is a function which can take a list argument: > = C '2*x for x' {1,2,3} {2,4,6} > dbl = C '2*x for x' > = dbl {10,20,30} {20,40,60} Here is a somewhat more explicit way of saying the same thing; `_1` is a _placeholder_ refering to the _first_ argument passed to the comprehension. > = C '2*x for _,x in pairs(_1)' {10,20,30} {20,40,60} > = C '_1(x) for x'(tostring,{1,2,3,4}) {'1','2','3','4'} This extended syntax is useful when you wish to collect the result of some iterator, such as `io.lines`. This comprehension creates a function which creates a table of all the lines in a file: > f = io.open('array.lua') > lines = C 'line for line in _1:lines()' (f) > = #lines 118 There are a number of functions that may be applied to the result of a comprehension: > = C 'min(x for x)' {1,44,0} 0 > = C 'max(x for x)' {1,44,0} 44 > = C 'sum(x for x)' {1,44,0} 45 (These are equivalent to a reduce operation on a list.) After the `for` part, there may be a condition, which filters the output. This comprehension collects the even numbers from a list: > = C 'x for x if x % 2 == 0' {1,2,3,4,5} {2,4} There may be a number of `for` parts: > = C '{x,y} for x = 1,2 for y = 1,2' () {{1,1},{1,2},{2,1},{2,2}} > = C '{x,y} for x for y' ({1,2},{10,20}) {{1,10},{1,20},{2,10},{2,20}} These comprehensions are useful when dealing with functions of more than one variable, and are not so easily achieved with the other Penlight functional forms. ### Creating Functions from Functions @lookup pl.func Lua functions may be treated like any other value, although of course you cannot multiply or add them. One operation that makes sense is _function composition_, which chains function calls (so `(f * g)(x)` is `f(g(x))`.) > func = require 'pl.func' > printf = func.compose(io.write,string.format) > printf("hello %s\n",'world') hello world true Many functions require you to pass a function as an argument, say to apply to all values of a sequence or as a callback. Often useful functions have the wrong number of arguments. So there is a need to construct a function of one argument from one of two arguments, _binding_ the extra argument to a given value. _partial application_ takes a function of n arguments and returns a function of n-1 arguments where the first argument is bound to some value: > p2 = func.bind1(print,'start>') > p2('hello',2) start> hello 2 > ops = require 'pl.operator' > = tablex.filter({1,-2,10,-1,2},bind1(ops.gt,0)) {-2,-1} > tablex.filter({1,-2,10,-1,2},bind1(ops.le,0)) {1,10,2} The last example unfortunately reads backwards, because `bind1` alway binds the first argument! Also unfortunately, in my youth I confused 'currying' with 'partial application', so the old name for `bind1` is `curry` - this alias still exists. This is a specialized form of function argument binding. Here is another way to say the `print` example: > p2 = func.bind(print,'start>',func._1,func._2) > p2('hello',2) start> hello 2 where `_1` and `_2` are _placeholder variables_, corresponding to the first and second argument respectively. Having `func` all over the place is distracting, so it's useful to pull all of `pl.func` into the local context. Here is the filter example, this time the right way around: > utils.import 'pl.func' > tablex.filter({1,-2,10,-1,2},bind(ops.gt, _1, 0)) {1,10,2} `tablex.merge` does a general merge of two tables. This example shows the usefulness of binding the last argument of a function. > S1 = {john=27, jane=31, mary=24} > S2 = {jane=31, jones=50} > intersection = bind(tablex.merge, _1, _2, false) > union = bind(tablex.merge, _1, _2, true) > = intersection(S1,S2) {jane=31} > = union(S1,S2) {mary=24,jane=31,john=27,jones=50} When using `bind` with `print`, we got a function of precisely two arguments, whereas we really want our function to use varargs like `print`. This is the role of `_0`: > _DEBUG = true > p = bind(print,'start>', _0) return function (fn,_v1) return function(...) return fn(_v1,...) end end > p(1,2,3,4,5) start> 1 2 3 4 5 I've turned on the global `_DEBUG` flag, so that the function generated is printed out. It is actually a function which _generates_ the required function; the first call _binds the value_ of `_v1` to 'start>'. ### Placeholder Expressions A common pattern in Penlight is a function which applies another function to all elements in a table or a sequence, such as `tablex.map` or `seq.filter`. Lua does anonymous functions well, although they can be a bit tedious to type: > = tablex.map(function(x) return x*x end, {1,2,3,4}) {1,4,9,16} `pl.func` allows you to define _placeholder expressions_, which can cut down on the typing required, and also make your intent clearer. First, we bring contents of `pl.func` into our context, and then supply an expression using placeholder variables, such as `_1`,`_2`,etc. (C++ programmers will recognize this from the Boost libraries.) > utils.import 'pl.func' > = tablex.map(_1*_1, {1,2,3,4}) {1,4,9,16} Functions of up to 5 arguments can be generated. > = tablex.map2(_1+_2,{1,2,3}, {10,20,30}) {11,22,33} These expressions can use arbitrary functions, altho they must first be registered with the functional library. `func.register` brings in a single function, and `func.import` brings in a whole table of functions, such as `math`. > sin = register(math.sin) > = tablex.map(sin(_1), {1,2,3,4}) {0.8414709848079,0.90929742682568,0.14112000805987,-0.75680249530793} > import 'math' > = tablex.map(cos(2*_1),{1,2,3,4}) {-0.41614683654714,-0.65364362086361,0.96017028665037,-0.14550003380861} A common operation is calling a method of a set of objects: > = tablex.map(_1:sub(1,1), {'one','four','x'}) {'o','f','x'} There are some restrictions on what operators can be used in PEs. For instance, because the `__len` metamethod cannot be overriden by plain Lua tables, we need to define a special function to express `#_1': > = tablex.map(Len(_1), {'one','four','x'}) {3,4,1} Likewise for comparison operators, which cannot be overloaded for _different_ types, and thus also have to be expressed as a special function: > = tablex.filter(Gt(_1,0), {1,-1,2,4,-3}) {1,2,4} It is useful to express the fact that a function returns multiple values. For instance, `tablex.pairmap` expects a function that will be called with the key and the value, and returns the new value and the key, in that order. > = pairmap(Args(_2,_1:upper()),{fred=1,alice=2}) {ALICE=2,FRED=1} PEs cannot contain `nil` values, since PE function arguments are represented as an array. Instead, a special value called `Nil` is provided. So say `_1:f(Nil,1)` instead of `_1:f(nil,1)`. A placeholder expression cannot be automatically used as a Lua function. The technical reason is that the call operator must be overloaded to construct function calls like `_1(1)`. If you want to force a PE to return a function, use `func.I`. > = tablex.map(_1(10),{I(2*_1),I(_1*_1),I(_1+2)}) {20,100,12} Here we make a table of functions taking a single argument, and then call them all with a value of 10. The essential idea with PEs is to 'quote' an expression so that it is not immediately evaluated, but instead turned into a function that can be applied later to some arguments. The basic mechanism is to wrap values and placeholders so that the usual Lua operators have the effect of building up an _expression tree_. (It turns out that you can do _symbolic algebra_ using PEs, see `symbols.lua` in the examples directory, and its test runner `testsym.lua`, which demonstrates symbolic differentiation.) The rule is that if any operator has a PE operand, the result will be quoted. Sometimes we need to quote things explicitly. For instance, say we want to pass a function to a filter that must return true if the element value is in a set. `set[_1]` is the obvious expression, but it does not give the desired result, since it evaluates directly, giving `nil`. Indexing works differently than a binary operation like addition (set+_1 _is_ properly quoted) so there is a need for an explicit quoting or wrapping operation. This is the job of the `_` function; the PE in this case should be `_(set)[_1]`. This works for functions as well, as a convenient alternative to registering functions: `_(math.sin)(_1)`. This is equivalent to using the `lines' method: for line in I(_(f):read()) do print(line) end Now this will work for _any_ 'file-like' object which which has a `read` method returning the next line. If you had a LuaSocket client which was being 'pushed' by lines sent from a server, then `_(s):receive '*l'` would create an iterator for accepting input. These forms can be convenient for adapting your data flow so that it can be passed to the sequence functions in `pl.seq'. Placeholder expressions can be mixed with sequence wrapper expressions. `lexer.lua` will give us a double-valued sequence of tokens, where the first value is a type, and the second is a value. We filter out only the values where the type is 'iden', extract the actual value using `map`, get the unique values and finally copy to a list. > str = 'for i=1,10 do for j = 1,10 do print(i,j) end end' > = seq(lexer.lua(str)):filter('==','iden'):map(_2):unique():copy() {i,print,j} This is a particularly intense line (and I don't always suggest making everything a one-liner!); the key is the behaviour of `map`, which will take both values of the sequence, so `_2` returns the value part. (Since `filter` here takes extra arguments, it only operates on the type values.) There are some performance considerations to using placeholder expressions. Instantiating a PE requires constructing and compiling a function, which is not such a fast operation. So to get best performance, factor out PEs from loops like this; local fn = I(_1:f() + _2:g()) for i = 1,n do res[i] = tablex.map2(fn,first[i],second[i]) end Penlight-1.12.0/docs_topics/08-additional.md000066400000000000000000000504011416703176500205470ustar00rootroot00000000000000## Additional Libraries Libraries in this section are no longer considered to be part of the Penlight core, but still provide specialized functionality when needed. ### Simple Input Patterns Lua string pattern matching is very powerful, and usually you will not need a traditional regular expression library. Even so, sometimes Lua code ends up looking like Perl, which happens because string patterns are not always the easiest things to read, especially for the casual reader. Here is a program which needs to understand three distinct date formats: -- parsing dates using Lua string patterns months={Jan=1,Feb=2,Mar=3,Apr=4,May=5,Jun=6, Jul=7,Aug=8,Sep=9,Oct=10,Nov=11,Dec=12} function check_and_process(d,m,y) d = tonumber(d) m = tonumber(m) y = tonumber(y) .... end for line in f:lines() do -- ordinary (English) date format local d,m,y = line:match('(%d+)/(%d+)/(%d+)') if d then check_and_process(d,m,y) else -- ISO date?? y,m,d = line:match('(%d+)%-(%d+)%-(%d+)') if y then check_and_process(d,m,y) else -- ? d,mm,y = line:match('%(d+)%s+(%a+)%s+(%d+)') m = months[mm] check_and_process(d,m,y) end end end These aren't particularly difficult patterns, but already typical issues are appearing, such as having to escape '-'. Also, `string.match` returns its captures, so that we're forced to use a slightly awkward nested if-statement. Verification issues will further cloud the picture, since regular expression people try to enforce constraints (like year cannot be more than four digits) using regular expressions, on the usual grounds that you shouldn't stop using a hammer when you are enjoying yourself. `pl.sip` provides a simple, intuitive way to detect patterns in strings and extract relevant parts. > sip = require 'pl.sip' > dump = require('pl.pretty').dump > res = {} > c = sip.compile 'ref=$S{file}:$d{line}' > = c('ref=hello.c:10',res) true > dump(res) { line = 10, file = "hello.c" } > = c('ref=long name, no line',res) false `sip.compile` creates a pattern matcher function, which takes a string and a table as arguments. If the string matches the pattern, then `true` is returned and the table is populated according to the captures within the pattern. Here is another version of the date parser: -- using SIP patterns function check(t) check_and_process(t.day,t.month,t.year) end shortdate = sip.compile('$d{day}/$d{month}/$d{year}') longdate = sip.compile('$d{day} $v{mon} $d{year}') isodate = sip.compile('$d{year}-$d{month}-$d{day}') for line in f:lines() do local res = {} if shortdate(str,res) then check(res) elseif isodate(str,res) then check(res) elseif longdate(str,res) then res.month = months[res.mon] check(res) end end SIP captures start with '$', then a one-character type, and then an optional variable name in curly braces. Type Meaning v identifier i possibly signed integer f floating-point number r rest of line q quoted string (quoted using either ' or ") p a path name ( anything inside balanced parentheses [ anything inside balanced brackets { anything inside balanced curly brackets < anything inside balanced angle brackets If a type is not one of the above, then it's assumed to be one of the standard Lua character classes, and will match one or more repetitions of that class. Any spaces you leave in your pattern will match any number of spaces, including zero, unless the spaces are between two identifier characters or patterns matching them; in that case, at least one space will be matched. SIP captures (like `$v{mon}`) do not have to be named. You can use just `$v`, but you have to be consistent; if a pattern contains unnamed captures, then all captures must be unnamed. In this case, the result table is a simple list of values. `sip.match` is a useful shortcut if you want to compile and match in one call, without saving the compiled pattern. It caches the result, so it is not much slower than explicitly using `sip.compile`. > sip.match('($q{first},$q{second})','("john","smith")',res) true > res {second='smith',first='john'} > res = {} > sip.match('($q,$q)','("jan","smit")',res) -- unnamed captures true > res {'jan','smit'} > sip.match('($q,$q)','("jan", "smit")',res) false ---> oops! Can't handle extra space! > sip.match('( $q , $q )','("jan", "smit")',res) true As a general rule, allow for whitespace in your patterns. Finally, putting a '$' at the end of a pattern means 'capture the rest of the line, starting at the first non-space'. It is a shortcut for '$r{rest}', or just '$r' if no named captures are used. > sip.match('( $q , $q ) $','("jan", "smit") and a string',res) true > res {'jan','smit','and a string'} > res = {} > sip.match('( $q{first} , $q{last} ) $','("jan", "smit") and a string',res) true > res {first='jan',rest='and a string',last='smit'} ### Command-line Programs with Lapp `pl.lapp` is a small and focused Lua module which aims to make standard command-line parsing easier and intuitive. It implements the standard GNU style, i.e. short flags with one letter start with '-', and there may be an additional long flag which starts with '--'. Generally options which take an argument expect to find it as the next parameter (e.g. 'gcc test.c -o test') but single short options taking a value can dispense with the space (e.g. 'head -n4 test.c' or `gcc -I/usr/include/lua/5.1 ...`) As far as possible, Lapp will convert parameters into their equivalent Lua types, i.e. convert numbers and convert filenames into file objects. If any conversion fails, or a required parameter is missing, an error will be issued and the usage text will be written out. So there are two necessary tasks, supplying the flag and option names and associating them with a type. For any non-trivial script, even for personal consumption, it's necessary to supply usage text. The novelty of Lapp is that it starts from that point and defines a loose format for usage strings which can specify the names and types of the parameters. An example will make this clearer: -- scale.lua lapp = require 'pl.lapp' local args = lapp [[ Does some calculations -o,--offset (default 0.0) Offset to add to scaled number -s,--scale (number) Scaling factor (number) Number to be scaled ]] print(args.offset + args.scale * args.number) Here is a command-line session using this script: $ lua scale.lua scale.lua:missing required parameter: scale Does some calculations -o,--offset (default 0.0) Offset to add to scaled number -s,--scale (number) Scaling factor (number ) Number to be scaled $ lua scale.lua -s 2.2 10 22 $ lua scale.lua -s 2.2 x10 scale.lua:unable to convert to number: x10 ....(usage as before) There are two kinds of lines in Lapp usage strings which are meaningful; option and parameter lines. An option line gives the short option, optionally followed by the corresponding long option. A type specifier in parentheses may follow. Similarly, a parameter line starts with '', followed by a type specifier. Type specifiers usually start with a type name: one of 'boolean', 'string','number','file-in' or 'file-out'. You may leave this out, but then _must_ say 'default' followed by a value. If a flag or parameter has a default, it is not _required_ and is set to the default. The actual type is deduced from this value (number, string, file or boolean) if not provided directly. 'Deduce' is a fancy word for 'guess' and it can be wrong, e.g '(default 1)' will always be a number. You can say '(string default 1)' to override the guess. There are file values for the predefined console streams: stdin, stdout, stderr. The boolean type is the default for flags. Not providing the type specifier is equivalent to '(boolean default false)`. If the flag is meant to be 'turned off' then either the full '(boolean default true)` or the shortcut '(default true)' will work. An alternative to `default` is `optional`: local lapp = require 'pl.lapp' local args = lapp [[ --cmd (optional string) Command to run. ]] if args.cmd then os.execute(args.cmd) end Here we're implying that `cmd` need not be specified (just as with `default`) but if not present, then `args.cmd` is `nil`, which will always test false. The rest of the line is ignored and can be used for explanatory text. This script shows the relation between the specified parameter names and the fields in the output table. -- simple.lua local args = require ('pl.lapp') [[ Various flags and option types -p A simple optional flag, defaults to false -q,--quiet A simple flag with long name -o (string) A required option with argument -s (default 'save') Optional string with default 'save' (single quotes ignored) -n (default 1) Optional numerical flag with default 1 -b (string default 1) Optional string flag with default '1' (type explicit) (default stdin) Optional input file parameter, reads from stdin ]] for k,v in pairs(args) do print(k,v) end I've just dumped out all values of the args table; note that args.quiet has become true, because it's specified; args.p defaults to false. If there is a long name for an option, that will be used in preference as a field name. A type or default specifier is not necessary for simple flags, since the default type is boolean. $ simple -o test -q simple.lua p false input file (781C1BD8) quiet true o test input_name simple.lua D:\dev\lua\lapp>simple -o test simple.lua one two three 1 one 2 two 3 three p false quiet false input file (781C1BD8) o test input_name simple.lua The parameter input has been set to an open read-only file object - we know it must be a read-only file since that is the type of the default value. The field input_name is automatically generated, since it's often useful to have access to the original filename. Notice that any extra parameters supplied will be put in the result table with integer indices, i.e. args[i] where i goes from 1 to #args. Files don't really have to be closed explicitly for short scripts with a quick well-defined mission, since the result of garbage-collecting file objects is to close them. #### Enforcing a Range and Enumerations The type specifier can also be of the form '(' MIN '..' MAX ')' or a set of strings separated by '|'. local lapp = require 'pl.lapp' local args = lapp [[ Setting ranges (1..10) A number from 1 to 10 (-5..1e6) Bigger range (slow|medium|fast) ]] print(args.x,args.y) Here the meaning of ranges is that the value is greater or equal to MIN and less or equal to MAX. An 'enum' is a _string_ that can only have values from a specified set. #### Custom Types There is no builti-in way to force a parameter to be a whole number, but you may define a custom type that does this: lapp = require ('pl.lapp') lapp.add_type('integer','number', function(x) lapp.assert(math.ceil(x) == x, 'not an integer!') end ) local args = lapp [[ (integer) Process PID ]] print(args.ival) `lapp.add_type` takes three parameters, a type name, a converter and a constraint function. The constraint function is expected to throw an assertion if some condition is not true; we use `lapp.assert` because it fails in the standard way for a command-line script. The converter argument can either be a type name known to Lapp, or a function which takes a string and generates a value. Here's a useful custom type that allows dates to be input as @{pl.Date} values: local df = Date.Format() lapp.add_type('date', function(s) local d,e = df:parse(s) lapp.assert(d,e) return d end ) #### 'varargs' Parameter Arrays lapp = require 'pl.lapp' local args = lapp [[ Summing numbers (number) A list of numbers to be summed ]] local sum = 0 for i,x in ipairs(args.numbers) do sum = sum + x end print ('sum is '..sum) The parameter number has a trailing '...', which indicates that this parameter is a 'varargs' parameter. It must be the last parameter, and args.number will be an array. Consider this implementation of the head utility from Mac OS X: -- implements a BSD-style head -- (see http://www.manpagez.com/man/1/head/osx-10.3.php) lapp = require ('pl.lapp') local args = lapp [[ Print the first few lines of specified files -n (default 10) Number of lines to print (default stdin) Files to print ]] -- by default, lapp converts file arguments to an actual Lua file object. -- But the actual filename is always available as _name. -- In this case, 'files' is a varargs array, so that 'files_name' is -- also an array. local nline = args.n local nfile = #args.files for i = 1,nfile do local file = args.files[i] if nfile > 1 then print('==> '..args.files_name[i]..' <==') end local n = 0 for line in file:lines() do print(line) n = n + 1 if n == nline then break end end end Note how we have access to all the filenames, because the auto-generated field `files_name` is also an array! (This is probably not a very considerate script, since Lapp will open all the files provided, and only close them at the end of the script. See the `xhead.lua` example for another implementation.) Flags and options may also be declared as vararg arrays, and can occur anywhere. If there is both a short and long form, then the trailing "..." must happen after the long form, for example "-x,--network... (string)...", Bear in mind that short options can be combined (like 'tar -xzf'), so it's perfectly legal to have '-vvv'. But normally the value of args.v is just a simple `true` value. local args = require ('pl.lapp') [[ -v... Verbosity level; can be -v, -vv or -vvv ]] vlevel = not args.v[1] and 0 or #args.v print(vlevel) The vlevel assigment is a bit of Lua voodoo, so consider the cases: * No -v flag, v is just { false } * One -v flags, v is { true } * Two -v flags, v is { true, true } * Three -v flags, v is { true, true, true } #### Defining a Parameter Callback If a script implements `lapp.callback`, then Lapp will call it after each argument is parsed. The callback is passed the parameter name, the raw unparsed value, and the result table. It is called immediately after assignment of the value, so the corresponding field is available. lapp = require ('pl.lapp') function lapp.callback(parm,arg,args) print('+',parm,arg) end local args = lapp [[ Testing parameter handling -p Plain flag (defaults to false) -q,--quiet Plain flag with GNU-style optional long name -o (string) Required string option -n (number) Required number option -s (default 1.0) Option that takes a number, but will default (number) Required number argument (default stdin) A parameter which is an input file (default stdout) One that is an output file ]] print 'args' for k,v in pairs(args) do print(k,v) end This produces the following output: $ args -o name -n 2 10 args.lua + o name + n 2 + start 10 + input args.lua args p false s 1 input_name args.lua quiet false output file (781C1B98) start 10 input file (781C1BD8) o name n 2 Callbacks are needed when you want to take action immediately on parsing an argument. #### Slack Mode If you'd like to use a multi-letter 'short' parameter you need to set the `lapp.slack` variable to `true`. In the following example we also see how default `false` and default `true` flags can be used and how to overwrite the default `-h` help flag (`--help` still works fine) - this applies to non-slack mode as well. -- Parsing the command line ---------------------------------------------------- -- test.lua local lapp = require 'pl.lapp' local pretty = require 'pl.pretty' lapp.slack = true local args = lapp [[ Does some calculations -v, --video (string) Specify input video -w, --width (default 256) Width of the video -h, --height (default 144) Height of the video -t, --time (default 10) Seconds of video to process -sk,--seek (default 0) Seek number of seconds -f1,--flag1 A false flag -f2,--flag2 A false flag -f3,--flag3 (default true) A true flag -f4,--flag4 (default true) A true flag ]] pretty.dump(args) And here we can see the output of `test.lua`: $> lua test.lua -v abc --time 40 -h 20 -sk 15 --flag1 -f3 ----> { width = 256, flag1 = true, flag3 = false, seek = 15, flag2 = false, video = abc, time = 40, height = 20, flag4 = true } ### Simple Test Framework `pl.test` was originally developed for the sole purpose of testing Penlight itself, but you may find it useful for your own applications. ([There are many other options](http://lua-users.org/wiki/UnitTesting).) Most of the goodness is in `test.asserteq`. It uses `tablex.deepcompare` on its two arguments, and by default quits the test application with a non-zero exit code, and an informative message printed to stderr: local test = require 'pl.test' test.asserteq({10,20,30},{10,20,30.1}) --~ test-test.lua:3: assertion failed --~ got: { --~ [1] = 10, --~ [2] = 20, --~ [3] = 30 --~ } --~ needed: { --~ [1] = 10, --~ [2] = 20, --~ [3] = 30.1 --~ } --~ these values were not equal This covers most cases but it's also useful to compare strings using `string.match` -- must start with bonzo the dog test.assertmatch ('bonzo the dog is here','^bonzo the dog') -- must end with an integer test.assertmatch ('hello 42','%d+$') Since Lua errors are usually strings, this matching strategy is used to test 'exceptions': test.assertraise(function() local t = nil print(t.bonzo) end,'nil value') (Some care is needed to match the essential part of the thrown error if you care for portability, since in Lua 5.2 the exact error is "attempt to index local 't' (a nil value)" and in Lua 5.3 the error is "attempt to index a nil value (local 't')") There is an extra optional argument to these test functions, which is helpful when writing test helper functions. There you want to highlight the failed line, not the actual call to `asserteq` or `assertmatch` - line 33 here is the call to `is_iden` function is_iden(str) test.assertmatch(str,'^[%a_][%w_]*$',1) end is_iden 'alpha_dog' is_iden '$dollars' --~ test-test.lua:33: assertion failed --~ got: "$dollars" --~ needed: "^[%a_][%w_]*$" --~ these strings did not match Useful Lua functions often return multiple values, and `test.tuple` is a convenient way to capture these values, whether they contain nils or not. T = test.tuple --- common error pattern function failing() return nil,'failed' end test.asserteq(T(failing()),T(nil,'failed')) Penlight-1.12.0/docs_topics/09-discussion.md000066400000000000000000000106571416703176500206340ustar00rootroot00000000000000## Technical Choices ### Modularity and Granularity In an ideal world, a program should only load the libraries it needs. Penlight is intended to work in situations where an extra 100Kb of bytecode could be a problem. It is straightforward but tedious to load exactly what you need: local data = require 'pl.data' local List = require 'pl.List' local array2d = require 'pl.array2d' local seq = require 'pl.seq' local utils = require 'pl.utils' This is the style that I follow in Penlight itself, so that modules don't mess with the global environment; also, `stringx.import()` is not used because it will update the global `string` table. But `require 'pl'` is more convenient in scripts; the question is how to ensure that one doesn't load the whole kitchen sink as the price of convenience. The strategy is to only load modules when they are referenced. In 'init.lua' (which is loaded by `require 'pl'`) a metatable is attached to the global table with an `__index` metamethod. Any unknown name is looked up in the list of modules, and if found, we require it and make that module globally available. So when `tablex.deepcompare` is encountered, looking up `tablex` causes 'pl.tablex' to be required. . Modifying the behaviour of the global table has consequences. For instance, there is the famous module `strict` which comes with Lua itself (perhaps the only standard Lua module written in Lua itself) which also does this modification so that global variiables must be defined before use. So the implementation in 'init.lua' allows for a 'not found' hook, which 'pl.strict.lua' uses. Other libraries may install their own metatables for `_G`, but Penlight will now forward any unknown name to the `__index` defined by the original metatable. But the strategy is worth the effort: the old 'kitchen sink' 'init.lua' would pull in about 260K of bytecode, whereas now typical programs use about 100K less, and short scripts even better - for instance, if they were only needing functionality in `utils`. There are some functions which mark their output table with a special metatable, when it seems particularly appropriate. For instance, `tablex.makeset` creates a `Set`, and `seq.copy` creates a `List`. But this does not automatically result in the loading of `pl.Set` and `pl.List`; only if you try to access any of these methods. In 'utils.lua', there is an exported table called `stdmt`: stdmt = { List = {}, Map = {}, Set = {}, MultiMap = {} } If you go through 'init.lua', then these plain little 'identity' tables get an `__index` metamethod which forces the loading of the full functionality. Here is the code from 'list.lua' which starts the ball rolling for lists: List = utils.stdmt.List List.__index = List List._name = "List" List._class = List The 'load-on-demand' strategy helps to modularize the library. Especially for more casual use, `require 'pl'` is a good compromise between convenience and modularity. In this current version, I have generally reduced the amount of trickery involved. Previously, `Map` was defined in `pl.class`; now it is sensibly defined in `pl.Map`; `pl.class` only contains the basic class mechanism (and returns that function.) For consistency, `List` is returned directly by `require 'pl.List'` (note the uppercase 'L'), Also, the amount of module dependencies in the non-core libraries like `pl.config` have been reduced. ### Defining what is Callable 'utils.lua' exports `function_arg` which is used extensively throughout Penlight. It defines what is meant by 'callable'. Obviously true functions are immediately passed back. But what about strings? The first option is that it represents an operator in 'operator.lua', so that '<' is just an alias for `operator.lt`. We then check whether there is a _function factory_ defined for the metatable of the value. (It is true that strings can be made callable, but in practice this turns out to be a cute but dubious idea, since _all_ strings share the same metatable. A common programming error is to pass the wrong kind of object to a function, and it's better to get a nice clean 'attempting to call a string' message rather than some obscure trace from the bowels of your library.) The other module that registers a function factory is `pl.func`. Placeholder expressions cannot be directly calleable, and so need to be instantiated and cached in as efficient way as possible. (An inconsistency is that `utils.is_callable` does not do this thorough check.) Penlight-1.12.0/examples/000077500000000000000000000000001416703176500151755ustar00rootroot00000000000000Penlight-1.12.0/examples/seesubst.lua000066400000000000000000000024711416703176500175410ustar00rootroot00000000000000-- shows how replacing '@see module' in the Markdown documentation -- can be done more elegantly using PL. -- We either have something like 'pl.config' (a module reference) -- or 'pl.seq.map' (a function reference); these cases must be distinguished -- and a Markdown link generated pointing to the LuaDoc file. local sip = require 'pl.sip' local stringx = require 'pl.stringx' local res = {} local s = [[ (@see pl.bonzo.dog) remember about @see pl.bonzo ]] local _gsub_patterns = {} local function gsub (s,pat,subst,start) local fpat = _gsub_patterns[pat] if not fpat then -- use SIP to generate a proper string pattern. -- the _whole thing_ is a capture, to get the whole match -- and the unnamed capture. fpat = '('..sip.create_pattern(pat)..')' _gsub_patterns[pat] = fpat end return s:gsub(fpat,subst,start) end local mod = sip.compile '$v.$v' local fun = sip.compile '$v.$v.$v' for line in stringx.lines(s) do line = gsub(line,'@see $p',function(see,path) if fun(path,res) or mod(path,res) then local ret = ('[see %s](%s.%s.html'):format(path,res[1],res[2]) if res[3] then return ret..'#'..res[3]..')' else return ret..')' end end end) print(line) end Penlight-1.12.0/examples/sipscan.lua000066400000000000000000000014521416703176500173420ustar00rootroot00000000000000-- another SIP example, shows how an awkward log file format -- can be parsed. It also prints out the actual Lua string -- pattern generated: -- SYNC%s*%[([+%-%d]%d*)%]%s*([+%-%d]%d*)%s*([+%-%d]%d*) local sip = require 'pl.sip' local stringx = require 'pl.stringx' local s = [[ SYNC [1] 0 547 (14679 sec) SYNC [2] 0 555 (14679 sec) SYNC [3] 0 563 (14679 sec) SYNC [4] 0 571 (14679 sec) SYNC [5] -1 580 (14679 sec) SYNC [6] 0 587 (14679 sec) ]] local first = true local expected local res = {} local pat = 'SYNC [$i{seq}] $i{diff} $i{val}' print(sip.create_pattern(pat)) local match = sip.compile(pat) for line in stringx.lines(s) do if match(line,res) then if first then expected = res.val first = false end print(res.val,expected - res.val) expected = expected + 8 end end Penlight-1.12.0/examples/symbols.lua000066400000000000000000000145661416703176500174040ustar00rootroot00000000000000require 'pl' utils.import 'pl.func' local ops = require 'pl.operator' local List = require 'pl.List' local append,concat = table.insert,table.concat local compare,find_if,compare_no_order,imap,reduce,count_map = tablex.compare,tablex.find_if,tablex.compare_no_order,tablex.imap,tablex.reduce,tablex.count_map local unpack = table.unpack function bindval (self,val) rawset(self,'value',val) end local optable = ops.optable function sexpr (e) if isPE(e) then if e.op ~= 'X' then local args = tablex.imap(sexpr,e) return '('..e.op..' '..table.concat(args,' ')..')' else return e.repr end else return tostring(e) end end psexpr = compose(print,sexpr) function equals (e1,e2) local p1,p2 = isPE(e1),isPE(e2) if p1 ~= p2 then return false end -- different kinds of animals! if p1 and p2 then -- both PEs -- operators must be the same if e1.op ~= e2.op then return false end -- PHs are equal if their representations are equal if e1.op == 'X' then return e1.repr == e2.repr -- commutative operators elseif e1.op == '+' or e1.op == '*' then return compare_no_order(e1,e2,equals) else -- arguments must be the same return compare(e1,e2,equals) end else -- fall back on simple equality for non PEs return e1 == e2 end end -- run down an unbalanced operator chain (like a+b+c) and return the arguments {a,b,c} function tcollect (op,e,ls) if isPE(e) and e.op == op then for i = 1,#e do tcollect(op,e[i],ls) end else ls:append(e) return end end function rcollect (e) local res = List() tcollect(e.op,e,res) return res end -- balance ensures that +/* chains are collected together, operates in-place. -- thus (+(+ a b) c) or (+ a (+ b c)) becomes (+ a b c), order immaterial function balance (e) if isPE(e) and e.op ~= 'X' then local op,args = e.op if op == '+' or op == '*' then args = rcollect(e) else args = imap(balance,e) end for i = 1,#args do e[i] = args[i] end end return e end -- fold constants in an expression function fold (e) if isPE(e) then if e.op == 'X' then -- there could be _bound values_! local val = rawget(e,'value') return val and val or e else local op = e.op local addmul = op == '*' or op == '+' -- first fold all arguments local args = imap(fold,e) if not addmul and not find_if(args,isPE) then -- no placeholders in these args, we can fold the expression. local opfn = optable[op] if opfn then return opfn(unpack(args)) else return '?' end elseif addmul then -- enforce a few rules for + and * -- split the args into two classes, PE args and non-PE args. local classes = List.partition(args,isPE) local pe,npe = classes[true],classes[false] if npe then -- there's at least one non PE argument -- so fold them if #npe == 1 then npe = npe[1] else npe = npe:reduce(optable[op]) end -- if the result is a constant, return it if not pe then return npe end -- either (* 1 x) => x or (* 1 x y ...) => (* x y ...) if op == '*' then if npe == 0 then return 0 elseif npe == 1 then -- identity if #pe == 1 then return pe[1] else npe = nil end end else -- special cases for + if npe == 0 then -- identity if #pe == 1 then return pe[1] else npe = nil end end end end -- build up the final arguments local res = {} if npe then append(res,npe) end for val,count in pairs(count_map(pe,equals)) do if count > 1 then if op == '*' then val = val ^ count else val = val * count end end append(res,val) end if #res == 1 then return res[1] end return PE{op=op,unpack(res)} elseif op == '^' then if args[2] == 1 then return args[1] end -- identity if args[2] == 0 then return 1 end end return PE{op=op,unpack(args)} end else return e end end function expand (e) if isPE(e) and e.op == '*' and isPE(e[2]) and e[2].op == '+' then local a,b = e[1],e[2] return expand(b[1]*a) + expand(b[2]*a) else return e end end function isnumber (x) return type(x) == 'number' end -- does this PE contain a reference to x? function references (e,x) if isPE(e) then if e.op == 'X' then return x.repr == e.repr else return find_if(e,references,x) end else return false end end local function muli (args) return PE{op='*',unpack(args)} end local function addi (args) return PE{op='+',unpack(args)} end function diff (e,x) if isPE(e) and references(e,x) then local op = e.op if op == 'X' then return 1 else local a,b = e[1],e[2] if op == '+' then -- differentiation is linear local args = imap(diff,e,x) return balance(addi(args)) elseif op == '*' then -- product rule local res,d,ee = {} for i = 1,#e do d = fold(diff(e[i],x)) if d ~= 0 then ee = {unpack(e)} ee[i] = d append(res,balance(muli(ee))) end end if #res > 1 then return addi(res) else return res[1] end elseif op == '^' and isnumber(b) then -- power rule return b*x^(b-1) end end else return 0 end end Penlight-1.12.0/examples/test-cmp.lua000066400000000000000000000001711416703176500174330ustar00rootroot00000000000000local A = require 'pl.tablex' print(A.compare_no_order({1,2,3},{2,1,3})) print(A.compare_no_order({1,2,3},{2,1,3},'==')) Penlight-1.12.0/examples/test-listcallbacks.lua000066400000000000000000000005211416703176500214660ustar00rootroot00000000000000-- demonstrates how to use a list of callbacks local List = require 'pl.List' local utils = require 'pl.utils' local actions = List() local L = utils.string_lambda actions:append(function() print 'hello' end) actions:append(L '|| print "yay"') -- '()' is a shortcut for operator.call or function(x) return x() end actions:foreach '()' Penlight-1.12.0/examples/test-pretty.lua000066400000000000000000000003401416703176500202010ustar00rootroot00000000000000local pretty = require 'pl.pretty' local tb = { 'one','two','three',{1,2,3}, alpha=1,beta=2,gamma=3,['&']=true,[0]=false, _fred = {true,true}, s = [[ hello dolly you're so fine ]] } print(pretty.write(tb)) Penlight-1.12.0/examples/test-symbols.lua000066400000000000000000000022631416703176500203500ustar00rootroot00000000000000require 'pl' -- force us to look in the script's directory when requiring... app.require_here() require 'symbols' local MT = getmetatable(_1) add = MT.__add mul = MT.__mul pow = MT.__pow function testeq (e1,e2) if not equals(e1,e2) then print ('Not equal',repr(e1),repr(e2)) end end sin = register(math.sin,'sin') f = register(function(x,y,z) end) --[[ testeq (_1,_1) testeq (_1+_2,_1+_2) testeq (_1 + 3*_2,_1 + 3*_2) testeq (_2+_1,_1+_2) testeq (sin(_1),sin(_1)) testeq (1+f(10,20,'ok'),f(10,20,'ok')+1) --]] function testexpand (e) print(repr(fold(expand(e)))) --fold end --[[ testexpand (a*(a+1)) testexpand ((x+2)*(b+1)) ]]-- function testfold (e) print(repr(fold(e))) end a,b,c,x,y = Var 'a,b,c,x,y' --~ testfold(_1 + _2) --~ testfold(add(10,20)) --~ testfold(add(mul(2,_1),mul(3,_2))) --[[ testfold(sin(a)) e = a^(b+2) testfold(e) bindval(b,1) testfold(e) bindval(a,2) testfold(e) bindval(a) bindval(b) ]] function testdiff (e) balance(e) e = diff(e,x) balance(e) print('+ ',e) e = fold(e) print('- ',e) end testdiff(x^2+1) testdiff(3*x^2) testdiff(x^2 + 2*x^3) testdiff(x^2 + 2*a*x^3 + x^4) testdiff(2*a*x^3) testdiff(x*x*x) Penlight-1.12.0/examples/testclone.lua000066400000000000000000000015021416703176500176760ustar00rootroot00000000000000--cloning a directory tree. local lfs = require 'lfs' local path = require 'pl.path' local dir = require 'pl.dir' local p1 = [[examples]] local p2 = [[copy/of/examples]] if not path.isfile 'examples/testclone.lua' then return print 'please run this in the penlight folder (below examples)' end -- make a copy of the examples folder dir.clonetree(p1,p2,dir.copyfile) assert(path.isdir 'copy') print '---' local t = os.time() print(lfs.touch('examples/testclone.lua',t,t+10)) -- this should only update this file dir.clonetree(p1,p2, function(f1,f2) local t1 = path.getmtime(f1) local t2 = path.getmtime(f2) --print(f1,t1,f2,t2) if t1 > t2 then dir.copyfile(f1,f2) print(f1,f2,t1,t2) end return true end) -- and get rid of the whole copy directory, with subdirs dir.rmtree 'copy' assert(not path.exists 'copy') Penlight-1.12.0/examples/testconfig.lua000066400000000000000000000016151416703176500200500ustar00rootroot00000000000000local stringio = require 'pl.stringio' local config = require 'pl.config' local function dump(t,indent) if type(t) == 'table' then io.write(indent,'{\n') local newindent = indent..' ' for k,v in pairs(t) do io.write(newindent,k,'=') dump(v,indent) io.write('\n') end io.write(newindent,'},\n') else io.write(indent,t,'(',type(t),')') end end local function testconfig(test) local f = stringio.open(test) local c = config.read(f) f:close() dump(c,' ') print '-----' end testconfig [[ ; comment 2 (an ini file) [section!] bonzo.dog=20,30 config_parm=here we go again depth = 2 [another] felix="cat" ]] testconfig [[ # this is a more Unix-y config file fred = 1 alice = 2 home = /bonzo/dog/etc ]] testconfig [[ # this is just a set of comma-separated values 1000,444,222 44,555,224 ]] Penlight-1.12.0/examples/testglobal.lua000066400000000000000000000015011416703176500200350ustar00rootroot00000000000000-- very simple lexer program which looks at all identifiers in a Lua -- file and checks whether they're in the global namespace. -- At the end, we dump out the result of count_map, which will give us -- unique identifiers with their usage count. -- (an example of a program which itself needs to be careful about what -- goes into the global namespace) local utils = require 'pl.utils' local file = require 'pl.file' local lexer = require 'pl.lexer' local List = require 'pl.List' local pretty = require 'pl.pretty' local seq = require 'pl.seq' local path = require 'pl.path' utils.on_error 'quit' local txt = file.read(arg[1] or path.normpath('examples/testglobal.lua')) local globals = List() for t,v in lexer.lua(txt) do if t == 'iden' and rawget(_G,v) then globals:append(v) end end pretty.dump(seq.count_map(globals)) Penlight-1.12.0/examples/testinputfields.lua000066400000000000000000000005111416703176500211230ustar00rootroot00000000000000local input = require 'pl.input' local sum = 0.0 local count = 0 local text = [[ 981124001 2.0 18988.4 10047.1 4149.7 981125001 0.8 19104.0 9970.4 5088.7 981127003 0.5 19012.5 9946.9 3831.2 ]] for id,magn,x in input.fields(3,' ',text) do sum = sum + x count = count + 1 end print('average x coord is ',sum/count) Penlight-1.12.0/examples/testinputfields2.lua000066400000000000000000000004221416703176500212060ustar00rootroot00000000000000local input = require 'pl.input' local seq = require 'pl.seq' local text = [[ 981124001 2.0 18988.4 10047.1 4149.7 981125001 0.8 19104.0 9970.4 5088.7 981127003 0.5 19012.5 9946.9 3831.2 ]] local sum,count = seq.sum(input.fields ({3},' ',text)) print(sum/count) Penlight-1.12.0/examples/testxml.lua000066400000000000000000000042301416703176500173770ustar00rootroot00000000000000-- an example showing 'pl.lexer' doing some serious work. -- The resulting Lua table is in the same LOM format used by luaexpat. -- This is (clearly) not a professional XML parser, so don't use it -- on your homework! local lexer = require 'pl.lexer' local pretty = require 'pl.pretty' local append = table.insert local skipws,expecting = lexer.skipws,lexer.expecting local function parse_element (tok,tag) local tbl,t,v,attrib tbl = {} tbl.tag = tag -- LOM 'tag' is the element tag t,v = skipws(tok) while v ~= '/' and v ~= '>' do if t ~= 'iden' then error('expecting attribute identifier') end attrib = v expecting(tok,'=') v = expecting(tok,'string') -- LOM: 'attr' subtable contains attrib/value pairs and an ordered list of attribs if not tbl.attr then tbl.attr = {} end tbl.attr[attrib] = v append(tbl.attr,attrib) t,v = skipws(tok) end if v == '/' then expecting(tok,'>') return tbl end -- pick up element data t,v = tok() while true do if t == '<' then t,v = skipws(tok) if t == '/' then -- element end tag t,v = tok() if t == '>' then return tbl end if t == 'iden' and v == tag then if tok() == '>' then return tbl end end error('expecting end tag '..tag) else append(tbl,parse_element(tok,v)) -- LOM: child elements added to table t,v = skipws(tok) end else append(tbl,v) -- LOM: text added to table t,v = skipws(tok) end end end local function parse_xml (tok) local t = skipws(tok) local v while t == '<' do t,v = tok() if t == '?' or t == '!' then -- skip meta stuff and commentary repeat t = tok() until t == '>' t = expecting(tok,'<') else return parse_element(tok,v) end end end local s = [[ ]] local tok = lexer.scan(s,nil,{space=false},{string=true}) local res = parse_xml(tok) print(pretty.write(res)) Penlight-1.12.0/examples/which.lua000066400000000000000000000014401416703176500170010ustar00rootroot00000000000000-- a simple implementation of the which command. This looks for -- the given file on the path. On windows, it will assume an extension -- of .exe if no extension is given. local List = require 'pl.List' local path = require 'pl.path' local app = require 'pl.app' local pathl = List.split(os.getenv 'PATH',path.dirsep) local function which (file) local res = pathl:map(path.join,file) res = res:filter(path.exists) if res then return res[1] end end local _,lua = app.lua() local file = arg[1] or lua -- i.e. location of lua executable local try if not file then return print 'must provide a filename' end if path.extension(file) == '' and path.is_windows then try = which(file..'.exe') else try = which(file) end if try then print(try) else print 'cannot find on path' end Penlight-1.12.0/ldoc.ltp000066400000000000000000000211021416703176500150150ustar00rootroot00000000000000 $(ldoc.title) # if ldoc.custom_css then -- add custom CSS file if configured. # end
# local no_spaces = ldoc.no_spaces # local use_li = ldoc.use_li # local display_name = ldoc.display_name # local iter = ldoc.modules.iter # local function M(txt,item) return ldoc.markup(txt,item,ldoc.plain) end # local nowrap = ldoc.wrap and '' or 'nowrap'
# if ldoc.body then -- verbatim HTML as contents; 'non-code' entries $(ldoc.body) # elseif module then -- module documentation

$(ldoc.module_typename(module)) $(module.name)

$(M(module.summary,module))

$(M(module.description,module))

# if module.tags.include then $(M(ldoc.include_file(module.tags.include))) # end # if module.see then # local li,il = use_li(module.see)

See also:

    # for see in iter(module.see) do $(li)$(see.label)$(il) # end -- for
# end -- if see # if module.usage then # local li,il = use_li(module.usage)

Usage:

    # for usage in iter(module.usage) do $(li)
    $(ldoc.escape(usage))
    $(il) # end -- for
# end -- if usage # if module.info then

Info:

    # for tag, value in module.info:iter() do
  • $(tag): $(M(value,module))
  • # end
# end -- if module.info # if not ldoc.no_summary then # -- bang out the tables of item types for this module (e.g Functions, Tables, etc) # for kind,items in module.kinds() do

$(kind)

# for item in items() do # end -- for items
$(display_name(item)) $(M(item.summary,item))
#end -- for kinds

#end -- if not no_summary # --- currently works for both Functions and Tables. The params field either contains # --- function parameters or table fields. # local show_return = not ldoc.no_return_or_parms # local show_parms = show_return # for kind, items in module.kinds() do # local kitem = module.kinds:get_item(kind) # local has_description = kitem and ldoc.descript(kitem) ~= ""

$(kind)

$(M(module.kinds:get_section_description(kind),nil)) # if kitem then # if has_description then
$(M(ldoc.descript(kitem),kitem))
# end # if kitem.usage then

Usage:

$(ldoc.prettify(kitem.usage[1]))
# end # end
# for item in items() do
$(display_name(item)) # if ldoc.prettify_files and ldoc.is_file_prettified[item.module.file.filename] then line $(item.lineno) # end
$(M(ldoc.descript(item),item)) # if ldoc.custom_tags then # for custom in iter(ldoc.custom_tags) do # local tag = item.tags[custom[1]] # if tag and not custom.hidden then # local li,il = use_li(tag)

$(custom.title or custom[1]):

    # for value in iter(tag) do $(li)$(custom.format and custom.format(value) or M(value))$(il) # end -- for # end -- if tag
# end -- iter tags # end # if show_parms and item.params and #item.params > 0 then # local subnames = module.kinds:type_of(item).subnames # if subnames then

$(subnames):

# end
    # for parm in iter(item.params) do # local param,sublist = item:subparam(parm) # if sublist then
  • $(sublist)$(M(item.params.map[sublist],item))
      # end # for p in iter(param) do # local name,tp,def = item:display_name_of(p), ldoc.typename(item:type_of_param(p)), item:default_of_param(p)
    • $(name) # if tp ~= '' then $(tp) # end $(M(item.params.map[p],item)) # if def == true then (optional) # elseif def then (default $(def)) # end # if item:readonly(p) then readonly # end
    • # end # if sublist then
    # end # end -- for
# end -- if params # if show_return and item.retgroups then local groups = item.retgroups

Returns:

# for i,group in ldoc.ipairs(groups) do local li,il = use_li(group)
    # for r in group:iter() do local type, ctypes = item:return_type(r); local rt = ldoc.typename(type) $(li) # if rt ~= '' then $(rt) # end $(M(r.text,item))$(il) # if ctypes then
      # for c in ctypes:iter() do
    • $(c.name) $(ldoc.typename(c.type)) $(M(c.comment,item))
    • # end
    # end -- if ctypes # end -- for r
# if i < #groups then

Or

# end # end -- for group # end -- if returns # if show_return and item.raise then

Raises:

$(M(item.raise,item)) # end # if item.see then # local li,il = use_li(item.see)

See also:

    # for see in iter(item.see) do $(li)$(see.label)$(il) # end -- for
# end -- if see # if item.usage then # local li,il = use_li(item.usage)

Usage:

    # for usage in iter(item.usage) do $(li)
    $(ldoc.prettify(usage))
    $(il) # end -- for
# end -- if usage
# end -- for items
# end -- for kinds # else -- if module; project-level contents # if ldoc.description then

$(M(ldoc.description,nil))

# end # if ldoc.full_description then

$(M(ldoc.full_description,nil))

# end # for kind, mods in ldoc.kinds() do

$(kind)

# kind = kind:lower() # for m in mods() do # end -- for modules
$(m.name) $(M(ldoc.strip_header(m.summary),m))
# end -- for kinds # end -- if module
generated by LDoc $(ldoc.version)
Penlight-1.12.0/lua/000077500000000000000000000000001416703176500141405ustar00rootroot00000000000000Penlight-1.12.0/lua/pl/000077500000000000000000000000001416703176500145535ustar00rootroot00000000000000Penlight-1.12.0/lua/pl/Date.lua000066400000000000000000000441171416703176500161420ustar00rootroot00000000000000--- Date and Date Format classes. -- See @{05-dates.md|the Guide}. -- -- NOTE: the date module is deprecated! see -- https://github.com/lunarmodules/Penlight/issues/285 -- -- Dependencies: `pl.class`, `pl.stringx`, `pl.utils` -- @classmod pl.Date -- @pragma nostrip local class = require 'pl.class' local os_time, os_date = os.time, os.date local stringx = require 'pl.stringx' local utils = require 'pl.utils' local assert_arg,assert_string = utils.assert_arg,utils.assert_string utils.raise_deprecation { source = "Penlight " .. utils._VERSION, message = "the 'Date' module is deprecated, see https://github.com/lunarmodules/Penlight/issues/285", version_removed = "2.0.0", version_deprecated = "1.9.2", } local Date = class() Date.Format = class() --- Date constructor. -- @param t this can be either -- -- * `nil` or empty - use current date and time -- * number - seconds since epoch (as returned by `os.time`). Resulting time is UTC -- * `Date` - make a copy of this date -- * table - table containing year, month, etc as for `os.time`. You may leave out year, month or day, -- in which case current values will be used. -- * year (will be followed by month, day etc) -- -- @param ... true if Universal Coordinated Time, or two to five numbers: month,day,hour,min,sec -- @function Date function Date:_init(t,...) local time local nargs = select('#',...) if nargs > 2 then local extra = {...} local year = t t = { year = year, month = extra[1], day = extra[2], hour = extra[3], min = extra[4], sec = extra[5] } end if nargs == 1 then self.utc = select(1,...) == true end if t == nil or t == 'utc' then time = os_time() self.utc = t == 'utc' elseif type(t) == 'number' then time = t if self.utc == nil then self.utc = true end elseif type(t) == 'table' then if getmetatable(t) == Date then -- copy ctor time = t.time self.utc = t.utc else if not (t.year and t.month) then local lt = os_date('*t') if not t.year and not t.month and not t.day then t.year = lt.year t.month = lt.month t.day = lt.day else t.year = t.year or lt.year t.month = t.month or (t.day and lt.month or 1) t.day = t.day or 1 end end t.day = t.day or 1 time = os_time(t) end else error("bad type for Date constructor: "..type(t),2) end self:set(time) end --- set the current time of this Date object. -- @int t seconds since epoch function Date:set(t) self.time = t if self.utc then self.tab = os_date('!*t',t) else self.tab = os_date('*t',t) end end --- get the time zone offset from UTC. -- @int ts seconds ahead of UTC function Date.tzone (ts) if ts == nil then ts = os_time() elseif type(ts) == "table" then if getmetatable(ts) == Date then ts = ts.time else ts = Date(ts).time end end local utc = os_date('!*t',ts) local lcl = os_date('*t',ts) lcl.isdst = false return os.difftime(os_time(lcl), os_time(utc)) end --- convert this date to UTC. function Date:toUTC () local ndate = Date(self) if not self.utc then ndate.utc = true ndate:set(ndate.time) end return ndate end --- convert this UTC date to local. function Date:toLocal () local ndate = Date(self) if self.utc then ndate.utc = false ndate:set(ndate.time) --~ ndate:add { sec = Date.tzone(self) } end return ndate end --- set the year. -- @int y Four-digit year -- @class function -- @name Date:year --- set the month. -- @int m month -- @class function -- @name Date:month --- set the day. -- @int d day -- @class function -- @name Date:day --- set the hour. -- @int h hour -- @class function -- @name Date:hour --- set the minutes. -- @int min minutes -- @class function -- @name Date:min --- set the seconds. -- @int sec seconds -- @class function -- @name Date:sec --- set the day of year. -- @class function -- @int yday day of year -- @name Date:yday --- get the year. -- @int y Four-digit year -- @class function -- @name Date:year --- get the month. -- @class function -- @name Date:month --- get the day. -- @class function -- @name Date:day --- get the hour. -- @class function -- @name Date:hour --- get the minutes. -- @class function -- @name Date:min --- get the seconds. -- @class function -- @name Date:sec --- get the day of year. -- @class function -- @name Date:yday for _,c in ipairs{'year','month','day','hour','min','sec','yday'} do Date[c] = function(self,val) if val then assert_arg(1,val,"number") self.tab[c] = val self:set(os_time(self.tab)) return self else return self.tab[c] end end end --- name of day of week. -- @bool full abbreviated if true, full otherwise. -- @ret string name function Date:weekday_name(full) return os_date(full and '%A' or '%a',self.time) end --- name of month. -- @int full abbreviated if true, full otherwise. -- @ret string name function Date:month_name(full) return os_date(full and '%B' or '%b',self.time) end --- is this day on a weekend?. function Date:is_weekend() return self.tab.wday == 1 or self.tab.wday == 7 end --- add to a date object. -- @param t a table containing one of the following keys and a value: -- one of `year`,`month`,`day`,`hour`,`min`,`sec` -- @return this date function Date:add(t) local old_dst = self.tab.isdst local key,val = next(t) self.tab[key] = self.tab[key] + val self:set(os_time(self.tab)) if old_dst ~= self.tab.isdst then self.tab.hour = self.tab.hour - (old_dst and 1 or -1) self:set(os_time(self.tab)) end return self end --- last day of the month. -- @return int day function Date:last_day() local d = 28 local m = self.tab.month while self.tab.month == m do d = d + 1 self:add{day=1} end self:add{day=-1} return self end --- difference between two Date objects. -- @tparam Date other Date object -- @treturn Date.Interval object function Date:diff(other) local dt = self.time - other.time if dt < 0 then error("date difference is negative!",2) end return Date.Interval(dt) end --- long numerical ISO data format version of this date. function Date:__tostring() local fmt = '%Y-%m-%dT%H:%M:%S' if self.utc then fmt = "!"..fmt end local t = os_date(fmt,self.time) if self.utc then return t .. 'Z' else local offs = self:tzone() if offs == 0 then return t .. 'Z' end local sign = offs > 0 and '+' or '-' local h = math.ceil(offs/3600) local m = (offs % 3600)/60 if m == 0 then return t .. ('%s%02d'):format(sign,h) else return t .. ('%s%02d:%02d'):format(sign,h,m) end end end --- equality between Date objects. function Date:__eq(other) return self.time == other.time end --- ordering between Date objects. function Date:__lt(other) return self.time < other.time end --- difference between Date objects. -- @function Date:__sub Date.__sub = Date.diff --- add a date and an interval. -- @param other either a `Date.Interval` object or a table such as -- passed to `Date:add` function Date:__add(other) local nd = Date(self) if Date.Interval:class_of(other) then other = {sec=other.time} end nd:add(other) return nd end Date.Interval = class(Date) ---- Date.Interval constructor -- @int t an interval in seconds -- @function Date.Interval function Date.Interval:_init(t) self:set(t) end function Date.Interval:set(t) self.time = t self.tab = os_date('!*t',self.time) end local function ess(n) if n > 1 then return 's ' else return ' ' end end --- If it's an interval then the format is '2 hours 29 sec' etc. function Date.Interval:__tostring() local t, res = self.tab, '' local y,m,d = t.year - 1970, t.month - 1, t.day - 1 if y > 0 then res = res .. y .. ' year'..ess(y) end if m > 0 then res = res .. m .. ' month'..ess(m) end if d > 0 then res = res .. d .. ' day'..ess(d) end if y == 0 and m == 0 then local h = t.hour if h > 0 then res = res .. h .. ' hour'..ess(h) end if t.min > 0 then res = res .. t.min .. ' min ' end if t.sec > 0 then res = res .. t.sec .. ' sec ' end end if res == '' then res = 'zero' end return res end ------------ Date.Format class: parsing and renderinig dates ------------ -- short field names, explicit os.date names, and a mask for allowed field repeats local formats = { d = {'day',{true,true}}, y = {'year',{false,true,false,true}}, m = {'month',{true,true}}, H = {'hour',{true,true}}, M = {'min',{true,true}}, S = {'sec',{true,true}}, } --- Date.Format constructor. -- @string fmt. A string where the following fields are significant: -- -- * d day (either d or dd) -- * y year (either yy or yyy) -- * m month (either m or mm) -- * H hour (either H or HH) -- * M minute (either M or MM) -- * S second (either S or SS) -- -- Alternatively, if fmt is nil then this returns a flexible date parser -- that tries various date/time schemes in turn: -- -- * [ISO 8601](http://en.wikipedia.org/wiki/ISO_8601), like `2010-05-10 12:35:23Z` or `2008-10-03T14:30+02` -- * times like 15:30 or 8.05pm (assumed to be today's date) -- * dates like 28/10/02 (European order!) or 5 Feb 2012 -- * month name like march or Mar (case-insensitive, first 3 letters); here the -- day will be 1 and the year this current year -- -- A date in format 3 can be optionally followed by a time in format 2. -- Please see test-date.lua in the tests folder for more examples. -- @usage df = Date.Format("yyyy-mm-dd HH:MM:SS") -- @class function -- @name Date.Format function Date.Format:_init(fmt) if not fmt then self.fmt = '%Y-%m-%d %H:%M:%S' self.outf = self.fmt self.plain = true return end local append = table.insert local D,PLUS,OPENP,CLOSEP = '\001','\002','\003','\004' local vars,used = {},{} local patt,outf = {},{} local i = 1 while i < #fmt do local ch = fmt:sub(i,i) local df = formats[ch] if df then if used[ch] then error("field appeared twice: "..ch,4) end used[ch] = true -- this field may be repeated local _,inext = fmt:find(ch..'+',i+1) local cnt = not _ and 1 or inext-i+1 if not df[2][cnt] then error("wrong number of fields: "..ch,4) end -- single chars mean 'accept more than one digit' local p = cnt==1 and (D..PLUS) or (D):rep(cnt) append(patt,OPENP..p..CLOSEP) append(vars,ch) if ch == 'y' then append(outf,cnt==2 and '%y' or '%Y') else append(outf,'%'..ch) end i = i + cnt else append(patt,ch) append(outf,ch) i = i + 1 end end -- escape any magic characters fmt = utils.escape(table.concat(patt)) -- fmt = table.concat(patt):gsub('[%-%.%+%[%]%(%)%$%^%%%?%*]','%%%1') -- replace markers with their magic equivalents fmt = fmt:gsub(D,'%%d'):gsub(PLUS,'+'):gsub(OPENP,'('):gsub(CLOSEP,')') self.fmt = fmt self.outf = table.concat(outf) self.vars = vars end local parse_date --- parse a string into a Date object. -- @string str a date string -- @return date object function Date.Format:parse(str) assert_string(1,str) if self.plain then return parse_date(str,self.us) end local res = {str:match(self.fmt)} if #res==0 then return nil, 'cannot parse '..str end local tab = {} for i,v in ipairs(self.vars) do local name = formats[v][1] -- e.g. 'y' becomes 'year' tab[name] = tonumber(res[i]) end -- os.date() requires these fields; if not present, we assume -- that the time set is for the current day. if not (tab.year and tab.month and tab.day) then local today = Date() tab.year = tab.year or today:year() tab.month = tab.month or today:month() tab.day = tab.day or today:day() end local Y = tab.year if Y < 100 then -- classic Y2K pivot tab.year = Y + (Y < 35 and 2000 or 1999) elseif not Y then tab.year = 1970 end return Date(tab) end --- convert a Date object into a string. -- @param d a date object, or a time value as returned by @{os.time} -- @return string function Date.Format:tostring(d) local tm local fmt = self.outf if type(d) == 'number' then tm = d else tm = d.time if d.utc then fmt = '!'..fmt end end return os_date(fmt,tm) end --- force US order in dates like 9/11/2001 function Date.Format:US_order(yesno) self.us = yesno end --local months = {jan=1,feb=2,mar=3,apr=4,may=5,jun=6,jul=7,aug=8,sep=9,oct=10,nov=11,dec=12} local months local parse_date_unsafe local function create_months() local ld, day1 = parse_date_unsafe '2000-12-31', {day=1} months = {} for i = 1,12 do ld = ld:last_day() ld:add(day1) local mon = ld:month_name():lower() months [mon] = i end end --[[ Allowed patterns: - [day] [monthname] [year] [time] - [day]/[month][/year] [time] ]] local function looks_like_a_month(w) return w:match '^%a+,*$' ~= nil end local is_number = stringx.isdigit local function tonum(s,l1,l2,kind) kind = kind or '' local n = tonumber(s) if not n then error(("%snot a number: '%s'"):format(kind,s)) end if n < l1 or n > l2 then error(("%s out of range: %s is not between %d and %d"):format(kind,s,l1,l2)) end return n end local function parse_iso_end(p,ns,sec) -- may be fractional part of seconds local _,nfrac,secfrac = p:find('^%.%d+',ns+1) if secfrac then sec = sec .. secfrac p = p:sub(nfrac+1) else p = p:sub(ns+1) end -- ISO 8601 dates may end in Z (for UTC) or [+-][isotime] -- (we're working with the date as lower case, hence 'z') if p:match 'z$' then -- we're UTC! return sec, {h=0,m=0} end p = p:gsub(':','') -- turn 00:30 to 0030 local _,_,sign,offs = p:find('^([%+%-])(%d+)') if not sign then return sec, nil end -- not UTC if #offs == 2 then offs = offs .. '00' end -- 01 to 0100 local tz = { h = tonumber(offs:sub(1,2)), m = tonumber(offs:sub(3,4)) } if sign == '-' then tz.h = -tz.h; tz.m = -tz.m end return sec, tz end function parse_date_unsafe (s,US) s = s:gsub('T',' ') -- ISO 8601 local parts = stringx.split(s:lower()) local i,p = 1,parts[1] local function nextp() i = i + 1; p = parts[i] end local year,min,hour,sec,apm local tz local _,nxt,day, month = p:find '^(%d+)/(%d+)' if day then -- swop for US case if US then day, month = month, day end _,_,year = p:find('^/(%d+)',nxt+1) nextp() else -- ISO year,month,day = p:match('^(%d+)%-(%d+)%-(%d+)') if year then nextp() end end if p and not year and is_number(p) then -- has to be date if #p < 4 then day = p nextp() else -- unless it looks like a 24-hour time year = true end end if p and looks_like_a_month(p) then -- date followed by month p = p:sub(1,3) if not months then create_months() end local mon = months[p] if mon then month = mon else error("not a month: " .. p) end nextp() end if p and not year and is_number(p) then year = p nextp() end if p then -- time is hh:mm[:ss], hhmm[ss] or H.M[am|pm] _,nxt,hour,min = p:find '^(%d+):(%d+)' local ns if nxt then -- are there seconds? _,ns,sec = p:find ('^:(%d+)',nxt+1) --if ns then sec,tz = parse_iso_end(p,ns or nxt,sec) --end else -- might be h.m _,ns,hour,min = p:find '^(%d+)%.(%d+)' if ns then apm = p:match '[ap]m$' else -- or hhmm[ss] local hourmin _,nxt,hourmin = p:find ('^(%d+)') if nxt then hour = hourmin:sub(1,2) min = hourmin:sub(3,4) sec = hourmin:sub(5,6) if #sec == 0 then sec = nil end sec,tz = parse_iso_end(p,nxt,sec) end end end end local today if year == true then year = nil end if not (year and month and day) then today = Date() end day = day and tonum(day,1,31,'day') or (month and 1 or today:day()) month = month and tonum(month,1,12,'month') or today:month() year = year and tonumber(year) or today:year() if year < 100 then -- two-digit year pivot around year < 2035 year = year + (year < 35 and 2000 or 1900) end hour = hour and tonum(hour,0,apm and 12 or 24,'hour') or 12 if apm == 'pm' then hour = hour + 12 end min = min and tonum(min,0,59) or 0 sec = sec and tonum(sec,0,60) or 0 --60 used to indicate leap second local res = Date {year = year, month = month, day = day, hour = hour, min = min, sec = sec} if tz then -- ISO 8601 UTC time local corrected = false if tz.h ~= 0 then res:add {hour = -tz.h}; corrected = true end if tz.m ~= 0 then res:add {min = -tz.m}; corrected = true end res.utc = true -- we're in UTC, so let's go local... if corrected then res = res:toLocal() end-- we're UTC! end return res end function parse_date (s) local ok, d = pcall(parse_date_unsafe,s) if not ok then -- error d = d:gsub('.-:%d+: ','') return nil, d else return d end end return Date Penlight-1.12.0/lua/pl/List.lua000066400000000000000000000366601416703176500162040ustar00rootroot00000000000000--- Python-style list class. -- -- **Please Note**: methods that change the list will return the list. -- This is to allow for method chaining, but please note that `ls = ls:sort()` -- does not mean that a new copy of the list is made. In-place (mutable) methods -- are marked as returning 'the list' in this documentation. -- -- See the Guide for further @{02-arrays.md.Python_style_Lists|discussion} -- -- See http://www.python.org/doc/current/tut/tut.html, section 5.1 -- -- **Note**: The comments before some of the functions are from the Python docs -- and contain Python code. -- -- Written for Lua version Nick Trout 4.0; Redone for Lua 5.1, Steve Donovan. -- -- Dependencies: `pl.utils`, `pl.tablex`, `pl.class` -- @classmod pl.List -- @pragma nostrip local tinsert,tremove,concat,tsort = table.insert,table.remove,table.concat,table.sort local setmetatable, getmetatable,type,tostring,string = setmetatable,getmetatable,type,tostring,string local tablex = require 'pl.tablex' local filter,imap,imap2,reduce,transform,tremovevalues = tablex.filter,tablex.imap,tablex.imap2,tablex.reduce,tablex.transform,tablex.removevalues local tsub = tablex.sub local utils = require 'pl.utils' local class = require 'pl.class' local array_tostring,split,assert_arg,function_arg = utils.array_tostring,utils.split,utils.assert_arg,utils.function_arg local normalize_slice = tablex._normalize_slice -- metatable for our list and map objects has already been defined.. local Multimap = utils.stdmt.MultiMap local List = utils.stdmt.List local iter class(nil,nil,List) -- we want the result to be _covariant_, i.e. t must have type of obj if possible local function makelist (t,obj) local klass = List if obj then klass = getmetatable(obj) end return setmetatable(t,klass) end local function simple_table(t) return type(t) == 'table' and not getmetatable(t) and #t > 0 end function List._create (src) if simple_table(src) then return src end end function List:_init (src) if self == src then return end -- existing table used as self! if src then for v in iter(src) do tinsert(self,v) end end end --- Create a new list. Can optionally pass a table; -- passing another instance of List will cause a copy to be created; -- this will return a plain table with an appropriate metatable. -- we pass anything which isn't a simple table to iterate() to work out -- an appropriate iterator -- @see List.iterate -- @param[opt] t An optional list-like table -- @return a new List -- @usage ls = List(); ls = List {1,2,3,4} -- @function List.new List.new = List --- Make a copy of an existing list. -- The difference from a plain 'copy constructor' is that this returns -- the actual List subtype. function List:clone() local ls = makelist({},self) ls:extend(self) return ls end --- Add an item to the end of the list. -- @param i An item -- @return the list function List:append(i) tinsert(self,i) return self end List.push = tinsert --- Extend the list by appending all the items in the given list. -- equivalent to 'a[len(a):] = L'. -- @tparam List L Another List -- @return the list function List:extend(L) assert_arg(1,L,'table') for i = 1,#L do tinsert(self,L[i]) end return self end --- Insert an item at a given position. i is the index of the -- element before which to insert. -- @int i index of element before whichh to insert -- @param x A data item -- @return the list function List:insert(i, x) assert_arg(1,i,'number') tinsert(self,i,x) return self end --- Insert an item at the begining of the list. -- @param x a data item -- @return the list function List:put (x) return self:insert(1,x) end --- Remove an element given its index. -- (equivalent of Python's del s[i]) -- @int i the index -- @return the list function List:remove (i) assert_arg(1,i,'number') tremove(self,i) return self end --- Remove the first item from the list whose value is given. -- (This is called 'remove' in Python; renamed to avoid confusion -- with table.remove) -- Return nil if there is no such item. -- @param x A data value -- @return the list function List:remove_value(x) for i=1,#self do if self[i]==x then tremove(self,i) return self end end return self end --- Remove the item at the given position in the list, and return it. -- If no index is specified, a:pop() returns the last item in the list. -- The item is also removed from the list. -- @int[opt] i An index -- @return the item function List:pop(i) if not i then i = #self end assert_arg(1,i,'number') return tremove(self,i) end List.get = List.pop --- Return the index in the list of the first item whose value is given. -- Return nil if there is no such item. -- @function List:index -- @param x A data value -- @int[opt=1] idx where to start search -- @return the index, or nil if not found. local tfind = tablex.find List.index = tfind --- Does this list contain the value? -- @param x A data value -- @return true or false function List:contains(x) return tfind(self,x) and true or false end --- Return the number of times value appears in the list. -- @param x A data value -- @return number of times x appears function List:count(x) local cnt=0 for i=1,#self do if self[i]==x then cnt=cnt+1 end end return cnt end --- Sort the items of the list, in place. -- @func[opt='<'] cmp an optional comparison function -- @return the list function List:sort(cmp) if cmp then cmp = function_arg(1,cmp) end tsort(self,cmp) return self end --- Return a sorted copy of this list. -- @func[opt='<'] cmp an optional comparison function -- @return a new list function List:sorted(cmp) return List(self):sort(cmp) end --- Reverse the elements of the list, in place. -- @return the list function List:reverse() local t = self local n = #t for i = 1,n/2 do t[i],t[n] = t[n],t[i] n = n - 1 end return self end --- Return the minimum and the maximum value of the list. -- @return minimum value -- @return maximum value function List:minmax() local vmin,vmax = 1e70,-1e70 for i = 1,#self do local v = self[i] if v < vmin then vmin = v end if v > vmax then vmax = v end end return vmin,vmax end --- Emulate list slicing. like 'list[first:last]' in Python. -- If first or last are negative then they are relative to the end of the list -- eg. slice(-2) gives last 2 entries in a list, and -- slice(-4,-2) gives from -4th to -2nd -- @param first An index -- @param last An index -- @return a new List function List:slice(first,last) return tsub(self,first,last) end --- Empty the list. -- @return the list function List:clear() for i=1,#self do tremove(self) end return self end local eps = 1.0e-10 --- Emulate Python's range(x) function. -- Include it in List table for tidiness -- @int start A number -- @int[opt] finish A number greater than start; if absent, -- then start is 1 and finish is start -- @int[opt=1] incr an increment (may be less than 1) -- @return a List from start .. finish -- @usage List.range(0,3) == List{0,1,2,3} -- @usage List.range(4) = List{1,2,3,4} -- @usage List.range(5,1,-1) == List{5,4,3,2,1} function List.range(start,finish,incr) if not finish then finish = start start = 1 end if incr then assert_arg(3,incr,'number') if math.ceil(incr) ~= incr then finish = finish + eps end else incr = 1 end assert_arg(1,start,'number') assert_arg(2,finish,'number') local t = List() for i=start,finish,incr do tinsert(t,i) end return t end --- list:len() is the same as #list. function List:len() return #self end -- Extended operations -- --- Remove a subrange of elements. -- equivalent to 'del s[i1:i2]' in Python. -- @int i1 start of range -- @int i2 end of range -- @return the list function List:chop(i1,i2) return tremovevalues(self,i1,i2) end --- Insert a sublist into a list -- equivalent to 's[idx:idx] = list' in Python -- @int idx index -- @tparam List list list to insert -- @return the list -- @usage l = List{10,20}; l:splice(2,{21,22}); assert(l == List{10,21,22,20}) function List:splice(idx,list) assert_arg(1,idx,'number') idx = idx - 1 local i = 1 for v in iter(list) do tinsert(self,i+idx,v) i = i + 1 end return self end --- General slice assignment s[i1:i2] = seq. -- @int i1 start index -- @int i2 end index -- @tparam List seq a list -- @return the list function List:slice_assign(i1,i2,seq) assert_arg(1,i1,'number') assert_arg(1,i2,'number') i1,i2 = normalize_slice(self,i1,i2) if i2 >= i1 then self:chop(i1,i2) end self:splice(i1,seq) return self end --- Concatenation operator. -- @within metamethods -- @tparam List L another List -- @return a new list consisting of the list with the elements of the new list appended function List:__concat(L) assert_arg(1,L,'table') local ls = self:clone() ls:extend(L) return ls end --- Equality operator ==. True iff all elements of two lists are equal. -- @within metamethods -- @tparam List L another List -- @return true or false function List:__eq(L) if #self ~= #L then return false end for i = 1,#self do if self[i] ~= L[i] then return false end end return true end --- Join the elements of a list using a delimiter. -- This method uses tostring on all elements. -- @string[opt=''] delim a delimiter string, can be empty. -- @return a string function List:join (delim) delim = delim or '' assert_arg(1,delim,'string') return concat(array_tostring(self),delim) end --- Join a list of strings.
-- Uses `table.concat` directly. -- @function List:concat -- @string[opt=''] delim a delimiter -- @return a string List.concat = concat local function tostring_q(val) local s = tostring(val) if type(val) == 'string' then s = '"'..s..'"' end return s end --- How our list should be rendered as a string. Uses join(). -- @within metamethods -- @see List:join function List:__tostring() return '{'..self:join(',',tostring_q)..'}' end --- Call the function on each element of the list. -- @func fun a function or callable object -- @param ... optional values to pass to function function List:foreach (fun,...) fun = function_arg(1,fun) for i = 1,#self do fun(self[i],...) end end local function lookup_fun (obj,name) local f = obj[name] if not f then error(type(obj).." does not have method "..name,3) end return f end --- Call the named method on each element of the list. -- @string name the method name -- @param ... optional values to pass to function function List:foreachm (name,...) for i = 1,#self do local obj = self[i] local f = lookup_fun(obj,name) f(obj,...) end end --- Create a list of all elements which match a function. -- @func fun a boolean function -- @param[opt] arg optional argument to be passed as second argument of the predicate -- @return a new filtered list. function List:filter (fun,arg) return makelist(filter(self,fun,arg),self) end --- Split a string using a delimiter. -- @string s the string -- @string[opt] delim the delimiter (default spaces) -- @return a List of strings -- @see pl.utils.split function List.split (s,delim) assert_arg(1,s,'string') return makelist(split(s,delim)) end --- Apply a function to all elements. -- Any extra arguments will be passed to the function. -- @func fun a function of at least one argument -- @param ... arbitrary extra arguments. -- @return a new list: {f(x) for x in self} -- @usage List{'one','two'}:map(string.upper) == {'ONE','TWO'} -- @see pl.tablex.imap function List:map (fun,...) return makelist(imap(fun,self,...),self) end --- Apply a function to all elements, in-place. -- Any extra arguments are passed to the function. -- @func fun A function that takes at least one argument -- @param ... arbitrary extra arguments. -- @return the list. function List:transform (fun,...) transform(fun,self,...) return self end --- Apply a function to elements of two lists. -- Any extra arguments will be passed to the function -- @func fun a function of at least two arguments -- @tparam List ls another list -- @param ... arbitrary extra arguments. -- @return a new list: {f(x,y) for x in self, for x in arg1} -- @see pl.tablex.imap2 function List:map2 (fun,ls,...) return makelist(imap2(fun,self,ls,...),self) end --- apply a named method to all elements. -- Any extra arguments will be passed to the method. -- @string name name of method -- @param ... extra arguments -- @return a new list of the results -- @see pl.seq.mapmethod function List:mapm (name,...) local res = {} for i = 1,#self do local val = self[i] local fn = lookup_fun(val,name) res[i] = fn(val,...) end return makelist(res,self) end local function composite_call (method,f) return function(self,...) return self[method](self,f,...) end end function List.default_map_with(T) return function(self,name) local m if T then local f = lookup_fun(T,name) m = composite_call('map',f) else m = composite_call('mapn',name) end getmetatable(self)[name] = m -- and cache.. return m end end List.default_map = List.default_map_with --- 'reduce' a list using a binary function. -- @func fun a function of two arguments -- @return result of the function -- @see pl.tablex.reduce function List:reduce (fun) return reduce(fun,self) end --- Partition a list using a classifier function. -- The function may return nil, but this will be converted to the string key ''. -- @func fun a function of at least one argument -- @param ... will also be passed to the function -- @treturn MultiMap a table where the keys are the returned values, and the values are Lists -- of values where the function returned that key. -- @see pl.MultiMap function List:partition (fun,...) fun = function_arg(1,fun) local res = {} for i = 1,#self do local val = self[i] local klass = fun(val,...) if klass == nil then klass = '' end if not res[klass] then res[klass] = List() end res[klass]:append(val) end return setmetatable(res,Multimap) end --- return an iterator over all values. function List:iter () return iter(self) end --- Create an iterator over a seqence. -- This captures the Python concept of 'sequence'. -- For tables, iterates over all values with integer indices. -- @param seq a sequence; a string (over characters), a table, a file object (over lines) or an iterator function -- @usage for x in iterate {1,10,22,55} do io.write(x,',') end ==> 1,10,22,55 -- @usage for ch in iterate 'help' do do io.write(ch,' ') end ==> h e l p function List.iterate(seq) if type(seq) == 'string' then local idx = 0 local n = #seq local sub = string.sub return function () idx = idx + 1 if idx > n then return nil else return sub(seq,idx,idx) end end elseif type(seq) == 'table' then local idx = 0 local n = #seq return function() idx = idx + 1 if idx > n then return nil else return seq[idx] end end elseif type(seq) == 'function' then return seq elseif type(seq) == 'userdata' and io.type(seq) == 'file' then return seq:lines() end end iter = List.iterate return List Penlight-1.12.0/lua/pl/Map.lua000066400000000000000000000053421416703176500157770ustar00rootroot00000000000000--- A Map class. -- -- > Map = require 'pl.Map' -- > m = Map{one=1,two=2} -- > m:update {three=3,four=4,two=20} -- > = m == M{one=1,two=20,three=3,four=4} -- true -- -- Dependencies: `pl.utils`, `pl.class`, `pl.tablex`, `pl.pretty` -- @classmod pl.Map local tablex = require 'pl.tablex' local utils = require 'pl.utils' local stdmt = utils.stdmt local deepcompare = tablex.deepcompare local pretty_write = require 'pl.pretty' . write local Map = stdmt.Map local Set = stdmt.Set local class = require 'pl.class' -- the Map class --------------------- class(nil,nil,Map) function Map:_init (t) local mt = getmetatable(t) if mt == Set or mt == Map then self:update(t) else return t -- otherwise assumed to be a map-like table end end local function makelist(t) return setmetatable(t, require('pl.List')) end --- list of keys. Map.keys = tablex.keys --- list of values. Map.values = tablex.values --- return an iterator over all key-value pairs. function Map:iter () return pairs(self) end --- return a List of all key-value pairs, sorted by the keys. function Map:items() local ls = makelist(tablex.pairmap (function (k,v) return makelist {k,v} end, self)) ls:sort(function(t1,t2) return t1[1] < t2[1] end) return ls end --- set a value in the map if it doesn't exist yet. -- @param key the key -- @param default value to set -- @return the value stored in the map (existing value, or the new value) function Map:setdefault(key, default) local val = self[key] if val ~= nil then return val end self:set(key,default) return default end --- size of map. -- note: this is a relatively expensive operation! -- @class function -- @name Map:len Map.len = tablex.size --- put a value into the map. -- This will remove the key if the value is `nil` -- @param key the key -- @param val the value function Map:set (key,val) self[key] = val end --- get a value from the map. -- @param key the key -- @return the value, or nil if not found. function Map:get (key) return rawget(self,key) end local index_by = tablex.index_by --- get a list of values indexed by a list of keys. -- @param keys a list-like table of keys -- @return a new list function Map:getvalues (keys) return makelist(index_by(self,keys)) end --- update the map using key/value pairs from another table. -- @tab table -- @function Map:update Map.update = tablex.update --- equality between maps. -- @within metamethods -- @tparam Map m another map. function Map:__eq (m) -- note we explicitly ask deepcompare _not_ to use __eq! return deepcompare(self,m,true) end --- string representation of a map. -- @within metamethods function Map:__tostring () return pretty_write(self,'') end return Map Penlight-1.12.0/lua/pl/MultiMap.lua000066400000000000000000000023061416703176500170070ustar00rootroot00000000000000--- MultiMap, a Map which has multiple values per key. -- -- Dependencies: `pl.utils`, `pl.class`, `pl.List`, `pl.Map` -- @classmod pl.MultiMap local utils = require 'pl.utils' local class = require 'pl.class' local List = require 'pl.List' local Map = require 'pl.Map' -- MultiMap is a standard MT local MultiMap = utils.stdmt.MultiMap class(Map,nil,MultiMap) MultiMap._name = 'MultiMap' function MultiMap:_init (t) if not t then return end self:update(t) end --- update a MultiMap using a table. -- @param t either a Multimap or a map-like table. -- @return the map function MultiMap:update (t) utils.assert_arg(1,t,'table') if Map:class_of(t) then for k,v in pairs(t) do self[k] = List() self[k]:append(v) end else for k,v in pairs(t) do self[k] = List(v) end end end --- add a new value to a key. Setting a nil value removes the key. -- @param key the key -- @param val the value -- @return the map function MultiMap:set (key,val) if val == nil then self[key] = nil else if not self[key] then self[key] = List() end self[key]:append(val) end end return MultiMap Penlight-1.12.0/lua/pl/OrderedMap.lua000066400000000000000000000106321416703176500173020ustar00rootroot00000000000000--- OrderedMap, a map which preserves ordering. -- -- Derived from `pl.Map`. -- -- Dependencies: `pl.utils`, `pl.tablex`, `pl.class`, `pl.List`, `pl.Map` -- @classmod pl.OrderedMap local tablex = require 'pl.tablex' local utils = require 'pl.utils' local List = require 'pl.List' local index_by,tsort,concat = tablex.index_by,table.sort,table.concat local class = require 'pl.class' local Map = require 'pl.Map' local OrderedMap = class(Map) OrderedMap._name = 'OrderedMap' local rawset = rawset --- construct an OrderedMap. -- Will throw an error if the argument is bad. -- @param t optional initialization table, same as for @{OrderedMap:update} function OrderedMap:_init (t) rawset(self,'_keys',List()) if t then local map,err = self:update(t) if not map then error(err,2) end end end local assert_arg,raise = utils.assert_arg,utils.raise --- update an OrderedMap using a table. -- If the table is itself an OrderedMap, then its entries will be appended. -- if it s a table of the form `{{key1=val1},{key2=val2},...}` these will be appended. -- -- Otherwise, it is assumed to be a map-like table, and order of extra entries is arbitrary. -- @tab t a table. -- @return the map, or nil in case of error -- @return the error message function OrderedMap:update (t) assert_arg(1,t,'table') if OrderedMap:class_of(t) then for k,v in t:iter() do self:set(k,v) end elseif #t > 0 then -- an array must contain {key=val} tables if type(t[1]) == 'table' then for _,pair in ipairs(t) do local key,value = next(pair) if not key then return raise 'empty pair initialization table' end self:set(key,value) end else return raise 'cannot use an array to initialize an OrderedMap' end else for k,v in pairs(t) do self:set(k,v) end end return self end --- set the key's value. This key will be appended at the end of the map. -- -- If the value is nil, then the key is removed. -- @param key the key -- @param val the value -- @return the map function OrderedMap:set (key,val) if rawget(self, key) == nil and val ~= nil then -- new key self._keys:append(key) -- we keep in order rawset(self,key,val) -- don't want to provoke __newindex! else -- existing key-value pair if val == nil then self._keys:remove_value(key) rawset(self,key,nil) else self[key] = val end end return self end OrderedMap.__newindex = OrderedMap.set --- insert a key/value pair before a given position. -- Note: if the map already contains the key, then this effectively -- moves the item to the new position by first removing at the old position. -- Has no effect if the key does not exist and val is nil -- @int pos a position starting at 1 -- @param key the key -- @param val the value; if nil use the old value function OrderedMap:insert (pos,key,val) local oldval = self[key] val = val or oldval if oldval then self._keys:remove_value(key) end if val then self._keys:insert(pos,key) rawset(self,key,val) end return self end --- return the keys in order. -- (Not a copy!) -- @return List function OrderedMap:keys () return self._keys end --- return the values in order. -- this is relatively expensive. -- @return List function OrderedMap:values () return List(index_by(self,self._keys)) end --- sort the keys. -- @func cmp a comparison function as for @{table.sort} -- @return the map function OrderedMap:sort (cmp) tsort(self._keys,cmp) return self end --- iterate over key-value pairs in order. function OrderedMap:iter () local i = 0 local keys = self._keys local idx return function() i = i + 1 if i > #keys then return nil end idx = keys[i] return idx,self[idx] end end --- iterate over an ordered map (5.2). -- @within metamethods -- @function OrderedMap:__pairs OrderedMap.__pairs = OrderedMap.iter --- string representation of an ordered map. -- @within metamethods function OrderedMap:__tostring () local res = {} for i,v in ipairs(self._keys) do local val = self[v] local vs = tostring(val) if type(val) ~= 'number' then vs = '"'..vs..'"' end res[i] = tostring(v)..'='..vs end return '{'..concat(res,',')..'}' end return OrderedMap Penlight-1.12.0/lua/pl/Set.lua000066400000000000000000000123571416703176500160210ustar00rootroot00000000000000--- A Set class. -- -- > Set = require 'pl.Set' -- > = Set{'one','two'} == Set{'two','one'} -- true -- > fruit = Set{'apple','banana','orange'} -- > = fruit['banana'] -- true -- > = fruit['hazelnut'] -- nil -- > colours = Set{'red','orange','green','blue'} -- > = fruit,colours -- [apple,orange,banana] [blue,green,orange,red] -- > = fruit+colours -- [blue,green,apple,red,orange,banana] -- [orange] -- > more_fruits = fruit + 'apricot' -- > = fruit*colours -- > = more_fruits, fruit -- [banana,apricot,apple,orange] [banana,apple,orange] -- -- Dependencies: `pl.utils`, `pl.tablex`, `pl.class`, `pl.Map`, (`pl.List` if __tostring is used) -- @classmod pl.Set local tablex = require 'pl.tablex' local utils = require 'pl.utils' local array_tostring, concat = utils.array_tostring, table.concat local merge,difference = tablex.merge,tablex.difference local Map = require 'pl.Map' local class = require 'pl.class' local stdmt = utils.stdmt local Set = stdmt.Set -- the Set class -------------------- class(Map,nil,Set) -- note: Set has _no_ methods! Set.__index = nil local function makeset (t) return setmetatable(t,Set) end --- create a set.
-- @param t may be a Set, Map or list-like table. -- @class function -- @name Set function Set:_init (t) t = t or {} local mt = getmetatable(t) if mt == Set or mt == Map then for k in pairs(t) do self[k] = true end else for _,v in ipairs(t) do self[v] = true end end end --- string representation of a set. -- @within metamethods function Set:__tostring () return '['..concat(array_tostring(Set.values(self)),',')..']' end --- get a list of the values in a set. -- @param self a Set -- @function Set.values -- @return a list Set.values = Map.keys --- map a function over the values of a set. -- @param self a Set -- @param fn a function -- @param ... extra arguments to pass to the function. -- @return a new set function Set.map (self,fn,...) fn = utils.function_arg(1,fn) local res = {} for k in pairs(self) do res[fn(k,...)] = true end return makeset(res) end --- union of two sets (also +). -- @param self a Set -- @param set another set -- @return a new set function Set.union (self,set) return merge(self,set,true) end --- modifies '+' operator to allow addition of non-Set elements --- Preserves +/- semantics - does not modify first argument. local function setadd(self,other) local mt = getmetatable(other) if mt == Set or mt == Map then return Set.union(self,other) else local new = Set(self) new[other] = true return new end end --- union of sets. -- @within metamethods -- @function Set.__add Set.__add = setadd --- intersection of two sets (also *). -- @param self a Set -- @param set another set -- @return a new set -- @usage -- > s = Set{10,20,30} -- > t = Set{20,30,40} -- > = t -- [20,30,40] -- > = Set.intersection(s,t) -- [30,20] -- > = s*t -- [30,20] function Set.intersection (self,set) return merge(self,set,false) end --- intersection of sets. -- @within metamethods -- @function Set.__mul Set.__mul = Set.intersection --- new set with elements in the set that are not in the other (also -). -- @param self a Set -- @param set another set -- @return a new set function Set.difference (self,set) return difference(self,set,false) end --- modifies "-" operator to remove non-Set values from set. --- Preserves +/- semantics - does not modify first argument. local function setminus (self,other) local mt = getmetatable(other) if mt == Set or mt == Map then return Set.difference(self,other) else local new = Set(self) new[other] = nil return new end end --- difference of sets. -- @within metamethods -- @function Set.__sub Set.__sub = setminus -- a new set with elements in _either_ the set _or_ other but not both (also ^). -- @param self a Set -- @param set another set -- @return a new set function Set.symmetric_difference (self,set) return difference(self,set,true) end --- symmetric difference of sets. -- @within metamethods -- @function Set.__pow Set.__pow = Set.symmetric_difference --- is the first set a subset of the second (also <)?. -- @param self a Set -- @param set another set -- @return true or false function Set.issubset (self,set) for k in pairs(self) do if not set[k] then return false end end return true end --- first set subset of second? -- @within metamethods -- @function Set.__lt Set.__lt = Set.issubset --- is the set empty?. -- @param self a Set -- @return true or false function Set.isempty (self) return next(self) == nil end --- are the sets disjoint? (no elements in common). -- Uses naive definition, i.e. that intersection is empty -- @param s1 a Set -- @param s2 another set -- @return true or false function Set.isdisjoint (s1,s2) return Set.isempty(Set.intersection(s1,s2)) end --- size of this set (also # for 5.2). -- @param s a Set -- @return size -- @function Set.len Set.len = tablex.size --- cardinality of set (5.2). -- @within metamethods -- @function Set.__len Set.__len = Set.len --- equality between sets. -- @within metamethods function Set.__eq (s1,s2) return Set.issubset(s1,s2) and Set.issubset(s2,s1) end return Set Penlight-1.12.0/lua/pl/app.lua000066400000000000000000000245331416703176500160450ustar00rootroot00000000000000--- Application support functions. -- See @{01-introduction.md.Application_Support|the Guide} -- -- Dependencies: `pl.utils`, `pl.path` -- @module pl.app local io,package,require = _G.io, _G.package, _G.require local utils = require 'pl.utils' local path = require 'pl.path' local app = {} --- return the name of the current script running. -- The name will be the name as passed on the command line -- @return string filename function app.script_name() if _G.arg and _G.arg[0] then return _G.arg[0] end return utils.raise("No script name found") end --- prefixes the current script's path to the Lua module path. -- Applies to both the source and the binary module paths. It makes it easy for -- the main file of a multi-file program to access its modules in the same directory. -- `base` allows these modules to be put in a specified subdirectory, to allow for -- cleaner deployment and resolve potential conflicts between a script name and its -- library directory. -- -- Note: the path is prefixed, so it is searched first when requiring modules. -- @string base optional base directory (absolute, or relative path). -- @treturn string the current script's path with a trailing slash function app.require_here (base) local p = path.dirname(app.script_name()) if not path.isabs(p) then p = path.join(path.currentdir(),p) end if p:sub(-1,-1) ~= path.sep then p = p..path.sep end if base then if path.is_windows then base = base:gsub('/','\\') end if path.isabs(base) then p = base .. path.sep else p = p..base..path.sep end end local so_ext = path.is_windows and 'dll' or 'so' local lsep = package.path:find '^;' and '' or ';' local csep = package.cpath:find '^;' and '' or ';' package.path = ('%s?.lua;%s?%sinit.lua%s%s'):format(p,p,path.sep,lsep,package.path) package.cpath = ('%s?.%s%s%s'):format(p,so_ext,csep,package.cpath) return p end --- return a suitable path for files private to this application. -- These will look like '~/.SNAME/file', with '~' as with expanduser and -- SNAME is the name of the script without .lua extension. -- If the directory does not exist, it will be created. -- @string file a filename (w/out path) -- @return a full pathname, or nil -- @return cannot create directory error -- @usage -- -- when run from a script called 'testapp' (on Windows): -- local app = require 'pl.app' -- print(app.appfile 'test.txt') -- -- C:\Documents and Settings\steve\.testapp\test.txt function app.appfile(file) local sfullname, err = app.script_name() if not sfullname then return utils.raise(err) end local sname = path.basename(sfullname) local name = path.splitext(sname) local dir = path.join(path.expanduser('~'),'.'..name) if not path.isdir(dir) then local ret = path.mkdir(dir) if not ret then return utils.raise('cannot create '..dir) end end return path.join(dir,file) end --- return string indicating operating system. -- @return 'Windows','OSX' or whatever uname returns (e.g. 'Linux') function app.platform() if path.is_windows then return 'Windows' else local f = io.popen('uname') local res = f:read() if res == 'Darwin' then res = 'OSX' end f:close() return res end end --- return the full command-line used to invoke this script. -- It will not include the scriptname itself, see `app.script_name`. -- @return command-line -- @return name of Lua program used -- @usage -- -- execute: lua -lluacov -e 'print(_VERSION)' myscript.lua -- -- -- myscript.lua -- print(require("pl.app").lua()) --> "lua -lluacov -e 'print(_VERSION)'", "lua" function app.lua() local args = _G.arg if not args then return utils.raise "not in a main program" end local cmd = {} local i = -1 while true do table.insert(cmd, 1, args[i]) if not args[i-1] then return utils.quote_arg(cmd), args[i] end i = i - 1 end end --- parse command-line arguments into flags and parameters. -- Understands GNU-style command-line flags; short (`-f`) and long (`--flag`). -- -- These may be given a value with either '=' or ':' (`-k:2`,`--alpha=3.2`,`-n2`), -- a number value can be given without a space. If the flag is marked -- as having a value, then a space-separated value is also accepted (`-i hello`), -- see the `flags_with_values` argument). -- -- Multiple short args can be combined like so: ( `-abcd`). -- -- When specifying the `flags_valid` parameter, its contents can also contain -- aliasses, to convert short/long flags to the same output name. See the -- example below. -- -- Note: if a flag is repeated, the last value wins. -- @tparam {string} args an array of strings (default is the global `arg`) -- @tab flags_with_values any flags that take values, either list or hash -- table e.g. `{ out=true }` or `{ "out" }`. -- @tab flags_valid (optional) flags that are valid, either list or hashtable. -- If not given, everything -- will be accepted(everything in `flags_with_values` will automatically be allowed) -- @return a table of flags (flag=value pairs) -- @return an array of parameters -- @raise if args is nil, then the global `args` must be available! -- @usage -- -- Simple form: -- local flags, params = app.parse_args(nil, -- { "hello", "world" }, -- list of flags taking values -- { "l", "a", "b"}) -- list of allowed flags (value ones will be added) -- -- -- More complex example using aliasses: -- local valid = { -- long = "l", -- if 'l' is specified, it is reported as 'long' -- new = { "n", "old" }, -- here both 'n' and 'old' will go into 'new' -- } -- local values = { -- "value", -- will automatically be added to the allowed set of flags -- "new", -- will mark 'n' and 'old' as requiring a value as well -- } -- local flags, params = app.parse_args(nil, values, valid) -- -- -- command: myapp.lua -l --old:hello --value world param1 param2 -- -- will yield: -- flags = { -- long = true, -- input from 'l' -- new = "hello", -- input from 'old' -- value = "world", -- allowed because it was in 'values', note: space separated! -- } -- params = { -- [1] = "param1" -- [2] = "param2" -- } function app.parse_args (args,flags_with_values, flags_valid) if not args then args = _G.arg if not args then utils.raise "Not in a main program: 'arg' not found" end end local with_values = {} for k,v in pairs(flags_with_values or {}) do if type(k) == "number" then k = v end with_values[k] = true end local valid if not flags_valid then -- if no allowed flags provided, we create a table that always returns -- the keyname, no matter what you look up valid = setmetatable({},{ __index = function(_, key) return key end }) else valid = {} for k,aliasses in pairs(flags_valid) do if type(k) == "number" then -- array/list entry k = aliasses end if type(aliasses) == "string" then -- single alias aliasses = { aliasses } end if type(aliasses) == "table" then -- list of aliasses -- it's the alternate name, so add the proper mappings for i, alias in ipairs(aliasses) do valid[alias] = k end end valid[k] = k end do local new_with_values = {} -- needed to prevent "invalid key to 'next'" error for k,v in pairs(with_values) do if not valid[k] then valid[k] = k -- add the with_value entry as a valid one new_with_values[k] = true else new_with_values[valid[k]] = true --set, but by its alias end end with_values = new_with_values end end -- now check that all flags with values are reported as such under all -- of their aliasses for k, main_alias in pairs(valid) do if with_values[main_alias] then with_values[k] = true end end local _args = {} local flags = {} local i = 1 while i <= #args do local a = args[i] local v = a:match('^-(.+)') local is_long if not v then -- we have a parameter _args[#_args+1] = a else -- it's a flag if v:find '^-' then is_long = true v = v:sub(2) end if with_values[v] then if i == #args or args[i+1]:find '^-' then return utils.raise ("no value for '"..v.."'") end flags[valid[v]] = args[i+1] i = i + 1 else -- a value can also be indicated with = or : local var,val = utils.splitv (v,'[=:]', false, 2) var = var or v val = val or true if not is_long then if #var > 1 then if var:find '.%d+' then -- short flag, number value val = var:sub(2) var = var:sub(1,1) else -- multiple short flags for i = 1,#var do local f = var:sub(i,i) if not valid[f] then return utils.raise("unknown flag '"..f.."'") else f = valid[f] end flags[f] = true end val = nil -- prevents use of var as a flag below end else -- single short flag (can have value, defaults to true) val = val or true end end if val then if not valid[var] then return utils.raise("unknown flag '"..var.."'") else var = valid[var] end flags[var] = val end end end i = i + 1 end return flags,_args end return app Penlight-1.12.0/lua/pl/array2d.lua000066400000000000000000000402671416703176500166330ustar00rootroot00000000000000--- Operations on two-dimensional arrays. -- See @{02-arrays.md.Operations_on_two_dimensional_tables|The Guide} -- -- The size of the arrays is determined by using the length operator `#` hence -- the module is not `nil` safe, and the usual precautions apply. -- -- Note: all functions taking `i1,j1,i2,j2` as arguments will normalize the -- arguments using `default_range`. -- -- Dependencies: `pl.utils`, `pl.tablex`, `pl.types` -- @module pl.array2d local tonumber,tostring,io,ipairs,string,table = _G.tonumber,_G.tostring,_G.io,_G.ipairs,_G.string,_G.table local setmetatable,getmetatable = setmetatable,getmetatable local tablex = require 'pl.tablex' local utils = require 'pl.utils' local types = require 'pl.types' local imap,tmap,reduce,keys,tmap2,tset,index_by = tablex.imap,tablex.map,tablex.reduce,tablex.keys,tablex.map2,tablex.set,tablex.index_by local remove = table.remove local splitv,fprintf,assert_arg = utils.splitv,utils.fprintf,utils.assert_arg local byte = string.byte local stdout = io.stdout local min = math.min local array2d = {} local function obj (int,out) local mt = getmetatable(int) if mt then setmetatable(out,mt) end return out end local function makelist (res) return setmetatable(res, require('pl.List')) end --- return the row and column size. -- Size is calculated using the Lua length operator #, so usual precautions -- regarding `nil` values apply. -- @array2d a a 2d array -- @treturn int number of rows (`#a`) -- @treturn int number of cols (`#a[1]`) function array2d.size (a) assert_arg(1,a,'table') return #a,#a[1] end do local function index (t,k) return t[k] end --- extract a column from the 2D array. -- @array2d a 2d array -- @param j column index -- @return 1d array function array2d.column (a,j) assert_arg(1,a,'table') return makelist(imap(index,a,j)) end end local column = array2d.column --- extract a row from the 2D array. -- Added in line with `column`, for read-only purposes directly -- accessing a[i] is more performant. -- @array2d a 2d array -- @param i row index -- @return 1d array (copy of the row) function array2d.row(a,i) assert_arg(1,a,'table') local row = a[i] local r = {} for n,v in ipairs(row) do r[n] = v end return makelist(r) end --- map a function over a 2D array -- @func f a function of at least one argument -- @array2d a 2d array -- @param arg an optional extra argument to be passed to the function. -- @return 2d array function array2d.map (f,a,arg) assert_arg(2,a,'table') f = utils.function_arg(1,f) return obj(a,imap(function(row) return imap(f,row,arg) end, a)) end --- reduce the rows using a function. -- @func f a binary function -- @array2d a 2d array -- @return 1d array -- @see pl.tablex.reduce function array2d.reduce_rows (f,a) assert_arg(1,a,'table') return tmap(function(row) return reduce(f,row) end, a) end --- reduce the columns using a function. -- @func f a binary function -- @array2d a 2d array -- @return 1d array -- @see pl.tablex.reduce function array2d.reduce_cols (f,a) assert_arg(1,a,'table') return tmap(function(c) return reduce(f,column(a,c)) end, keys(a[1])) end --- reduce a 2D array into a scalar, using two operations. -- @func opc operation to reduce the final result -- @func opr operation to reduce the rows -- @param a 2D array function array2d.reduce2 (opc,opr,a) assert_arg(3,a,'table') local tmp = array2d.reduce_rows(opr,a) return reduce(opc,tmp) end --- map a function over two arrays. -- They can be both or either 2D arrays -- @func f function of at least two arguments -- @int ad order of first array (`1` if `a` is a list/array, `2` if it is a 2d array) -- @int bd order of second array (`1` if `b` is a list/array, `2` if it is a 2d array) -- @tab a 1d or 2d array -- @tab b 1d or 2d array -- @param arg optional extra argument to pass to function -- @return 2D array, unless both arrays are 1D function array2d.map2 (f,ad,bd,a,b,arg) assert_arg(1,a,'table') assert_arg(2,b,'table') f = utils.function_arg(1,f) if ad == 1 and bd == 2 then return imap(function(row) return tmap2(f,a,row,arg) end, b) elseif ad == 2 and bd == 1 then return imap(function(row) return tmap2(f,row,b,arg) end, a) elseif ad == 1 and bd == 1 then return tmap2(f,a,b) elseif ad == 2 and bd == 2 then return tmap2(function(rowa,rowb) return tmap2(f,rowa,rowb,arg) end, a,b) end end --- cartesian product of two 1d arrays. -- @func f a function of 2 arguments -- @array t1 a 1d table -- @array t2 a 1d table -- @return 2d table -- @usage product('..',{1,2},{'a','b'}) == {{'1a','2a'},{'1b','2b'}} function array2d.product (f,t1,t2) f = utils.function_arg(1,f) assert_arg(2,t1,'table') assert_arg(3,t2,'table') local res = {} for i,v in ipairs(t2) do res[i] = tmap(f,t1,v) end return res end --- flatten a 2D array. -- (this goes over columns first.) -- @array2d t 2d table -- @return a 1d table -- @usage flatten {{1,2},{3,4},{5,6}} == {1,2,3,4,5,6} function array2d.flatten (t) local res = {} local k = 1 local rows, cols = array2d.size(t) for r = 1, rows do local row = t[r] for c = 1, cols do res[k] = row[c] k = k + 1 end end return makelist(res) end --- reshape a 2D array. Reshape the aray by specifying a new nr of rows. -- @array2d t 2d array -- @int nrows new number of rows -- @bool co use column-order (Fortran-style) (default false) -- @return a new 2d array function array2d.reshape (t,nrows,co) local nr,nc = array2d.size(t) local ncols = nr*nc / nrows local res = {} local ir,ic = 1,1 for i = 1,nrows do local row = {} for j = 1,ncols do row[j] = t[ir][ic] if not co then ic = ic + 1 if ic > nc then ir = ir + 1 ic = 1 end else ir = ir + 1 if ir > nr then ic = ic + 1 ir = 1 end end end res[i] = row end return obj(t,res) end --- transpose a 2D array. -- @array2d t 2d array -- @return a new 2d array function array2d.transpose(t) assert_arg(1,t,'table') local _, c = array2d.size(t) return array2d.reshape(t,c,true) end --- swap two rows of an array. -- @array2d t a 2d array -- @int i1 a row index -- @int i2 a row index -- @return t (same, modified 2d array) function array2d.swap_rows (t,i1,i2) assert_arg(1,t,'table') t[i1],t[i2] = t[i2],t[i1] return t end --- swap two columns of an array. -- @array2d t a 2d array -- @int j1 a column index -- @int j2 a column index -- @return t (same, modified 2d array) function array2d.swap_cols (t,j1,j2) assert_arg(1,t,'table') for _, row in ipairs(t) do row[j1],row[j2] = row[j2],row[j1] end return t end --- extract the specified rows. -- @array2d t 2d array -- @tparam {int} ridx a table of row indices -- @return a new 2d array with the extracted rows function array2d.extract_rows (t,ridx) return obj(t,index_by(t,ridx)) end --- extract the specified columns. -- @array2d t 2d array -- @tparam {int} cidx a table of column indices -- @return a new 2d array with the extracted colums function array2d.extract_cols (t,cidx) assert_arg(1,t,'table') local res = {} for i = 1,#t do res[i] = index_by(t[i],cidx) end return obj(t,res) end --- remove a row from an array. -- @function array2d.remove_row -- @array2d t a 2d array -- @int i a row index array2d.remove_row = remove --- remove a column from an array. -- @array2d t a 2d array -- @int j a column index function array2d.remove_col (t,j) assert_arg(1,t,'table') for i = 1,#t do remove(t[i],j) end end do local function _parse (s) local r, c = s:match 'R(%d+)C(%d+)' if r then r,c = tonumber(r),tonumber(c) return r,c end c,r = s:match '(%a+)(%d+)' if c then local cv = 0 for i = 1, #c do cv = cv * 26 + byte(c:sub(i,i)) - byte 'A' + 1 end return tonumber(r), cv end error('bad cell specifier: '..s) end --- parse a spreadsheet range or cell. -- The range/cell can be specified either as 'A1:B2' or 'R1C1:R2C2' or for -- single cells as 'A1' or 'R1C1'. -- @string s a range (case insensitive). -- @treturn int start row -- @treturn int start col -- @treturn int end row (or `nil` if the range was a single cell) -- @treturn int end col (or `nil` if the range was a single cell) function array2d.parse_range (s) assert_arg(1,s,'string') s = s:upper() if s:find ':' then local start,finish = splitv(s,':') local i1,j1 = _parse(start) local i2,j2 = _parse(finish) return i1,j1,i2,j2 else -- single value local i,j = _parse(s) return i,j end end end --- get a slice of a 2D array. -- Same as `slice`. -- @see slice function array2d.range (...) return array2d.slice(...) end local default_range do local function norm_value(v, max) if not v then return v end if v < 0 then v = max + v + 1 end if v < 1 then v = 1 end if v > max then v = max end return v end --- normalizes coordinates to valid positive entries and defaults. -- Negative indices will be counted from the end, too low, or too high -- will be limited by the array sizes. -- @array2d t a 2D array -- @tparam[opt=1] int|string i1 start row or spreadsheet range passed to `parse_range` -- @tparam[opt=1] int j1 start col -- @tparam[opt=N] int i2 end row -- @tparam[opt=M] int j2 end col -- @see parse_range -- @return i1, j1, i2, j2 function array2d.default_range (t,i1,j1,i2,j2) if (type(i1) == 'string') and not (j1 or i2 or j2) then i1, j1, i2, j2 = array2d.parse_range(i1) end local nr, nc = array2d.size(t) i1 = norm_value(i1 or 1, nr) j1 = norm_value(j1 or 1, nc) i2 = norm_value(i2 or nr, nr) j2 = norm_value(j2 or nc, nc) return i1,j1,i2,j2 end default_range = array2d.default_range end --- get a slice of a 2D array. Note that if the specified range has -- a 1D result, the rank of the result will be 1. -- @array2d t a 2D array -- @tparam[opt=1] int|string i1 start row or spreadsheet range passed to `parse_range` -- @tparam[opt=1] int j1 start col -- @tparam[opt=N] int i2 end row -- @tparam[opt=M] int j2 end col -- @see parse_range -- @return an array, 2D in general but 1D in special cases. function array2d.slice (t,i1,j1,i2,j2) assert_arg(1,t,'table') i1,j1,i2,j2 = default_range(t,i1,j1,i2,j2) local res = {} for i = i1,i2 do local val local row = t[i] if j1 == j2 then val = row[j1] else val = {} for j = j1,j2 do val[#val+1] = row[j] end end res[#res+1] = val end if i1 == i2 then res = res[1] end return obj(t,res) end --- set a specified range of an array to a value. -- @array2d t a 2D array -- @param value the value (may be a function, called as `val(i,j)`) -- @tparam[opt=1] int|string i1 start row or spreadsheet range passed to `parse_range` -- @tparam[opt=1] int j1 start col -- @tparam[opt=N] int i2 end row -- @tparam[opt=M] int j2 end col -- @see parse_range -- @see tablex.set function array2d.set (t,value,i1,j1,i2,j2) i1,j1,i2,j2 = default_range(t,i1,j1,i2,j2) local i = i1 if types.is_callable(value) then local old_f = value value = function(j) return old_f(i,j) end end while i <= i2 do tset(t[i],value,j1,j2) i = i + 1 end end --- write a 2D array to a file. -- @array2d t a 2D array -- @param f a file object (default stdout) -- @string fmt a format string (default is just to use tostring) -- @tparam[opt=1] int|string i1 start row or spreadsheet range passed to `parse_range` -- @tparam[opt=1] int j1 start col -- @tparam[opt=N] int i2 end row -- @tparam[opt=M] int j2 end col -- @see parse_range function array2d.write (t,f,fmt,i1,j1,i2,j2) assert_arg(1,t,'table') f = f or stdout local rowop if fmt then rowop = function(row,j) fprintf(f,fmt,row[j]) end else rowop = function(row,j) f:write(tostring(row[j]),' ') end end local function newline() f:write '\n' end array2d.forall(t,rowop,newline,i1,j1,i2,j2) end --- perform an operation for all values in a 2D array. -- @array2d t 2D array -- @func row_op function to call on each value; `row_op(row,j)` -- @func end_row_op function to call at end of each row; `end_row_op(i)` -- @tparam[opt=1] int|string i1 start row or spreadsheet range passed to `parse_range` -- @tparam[opt=1] int j1 start col -- @tparam[opt=N] int i2 end row -- @tparam[opt=M] int j2 end col -- @see parse_range function array2d.forall (t,row_op,end_row_op,i1,j1,i2,j2) assert_arg(1,t,'table') i1,j1,i2,j2 = default_range(t,i1,j1,i2,j2) for i = i1,i2 do local row = t[i] for j = j1,j2 do row_op(row,j) end if end_row_op then end_row_op(i) end end end ---- move a block from the destination to the source. -- @array2d dest a 2D array -- @int di start row in dest -- @int dj start col in dest -- @array2d src a 2D array -- @tparam[opt=1] int|string i1 start row or spreadsheet range passed to `parse_range` -- @tparam[opt=1] int j1 start col -- @tparam[opt=N] int i2 end row -- @tparam[opt=M] int j2 end col -- @see parse_range function array2d.move (dest,di,dj,src,i1,j1,i2,j2) assert_arg(1,dest,'table') assert_arg(4,src,'table') i1,j1,i2,j2 = default_range(src,i1,j1,i2,j2) local nr,nc = array2d.size(dest) i2, j2 = min(nr,i2), min(nc,j2) --i1, j1 = max(1,i1), max(1,j1) dj = dj - 1 for i = i1,i2 do local drow, srow = dest[i+di-1], src[i] for j = j1,j2 do drow[j+dj] = srow[j] end end end --- iterate over all elements in a 2D array, with optional indices. -- @array2d a 2D array -- @bool indices with indices (default false) -- @tparam[opt=1] int|string i1 start row or spreadsheet range passed to `parse_range` -- @tparam[opt=1] int j1 start col -- @tparam[opt=N] int i2 end row -- @tparam[opt=M] int j2 end col -- @see parse_range -- @return either `value` or `i,j,value` depending on the value of `indices` function array2d.iter(a,indices,i1,j1,i2,j2) assert_arg(1,a,'table') i1,j1,i2,j2 = default_range(a,i1,j1,i2,j2) local i,j = i1,j1-1 local row = a[i] return function() j = j + 1 if j > j2 then j = j1 i = i + 1 row = a[i] if i > i2 then return nil end end if indices then return i,j,row[j] else return row[j] end end end --- iterate over all columns. -- @array2d a a 2D array -- @return column, column-index function array2d.columns(a) assert_arg(1,a,'table') local n = #a[1] local i = 0 return function() i = i + 1 if i > n then return nil end return column(a,i), i end end --- iterate over all rows. -- Returns a copy of the row, for read-only purposes directly iterating -- is more performant; `ipairs(a)` -- @array2d a a 2D array -- @return row, row-index function array2d.rows(a) assert_arg(1,a,'table') local n = #a local i = 0 return function() i = i + 1 if i > n then return nil end return array2d.row(a,i), i end end --- new array of specified dimensions -- @int rows number of rows -- @int cols number of cols -- @param val initial value; if it's a function then use `val(i,j)` -- @return new 2d array function array2d.new(rows,cols,val) local res = {} local fun = types.is_callable(val) for i = 1,rows do local row = {} if fun then for j = 1,cols do row[j] = val(i,j) end else for j = 1,cols do row[j] = val end end res[i] = row end return res end return array2d Penlight-1.12.0/lua/pl/class.lua000066400000000000000000000202531416703176500163650ustar00rootroot00000000000000--- Provides a reuseable and convenient framework for creating classes in Lua. -- Two possible notations: -- -- B = class(A) -- class.B(A) -- -- The latter form creates a named class within the current environment. Note -- that this implicitly brings in `pl.utils` as a dependency. -- -- See the Guide for further @{01-introduction.md.Simplifying_Object_Oriented_Programming_in_Lua|discussion} -- @module pl.class local error, getmetatable, io, pairs, rawget, rawset, setmetatable, tostring, type = _G.error, _G.getmetatable, _G.io, _G.pairs, _G.rawget, _G.rawset, _G.setmetatable, _G.tostring, _G.type local compat -- this trickery is necessary to prevent the inheritance of 'super' and -- the resulting recursive call problems. local function call_ctor (c,obj,...) local init = rawget(c,'_init') local parent_with_init = rawget(c,'_parent_with_init') if parent_with_init then if not init then -- inheriting an init init = rawget(parent_with_init, '_init') parent_with_init = rawget(parent_with_init, '_parent_with_init') end if parent_with_init then -- super() points to one above whereever _init came from rawset(obj,'super',function(obj,...) call_ctor(parent_with_init,obj,...) end) end else -- Without this, calling super() where none exists will sometimes loop and stack overflow rawset(obj,'super',nil) end local res = init(obj,...) if parent_with_init then -- If this execution of call_ctor set a super, unset it rawset(obj,'super',nil) end return res end --- initializes an __instance__ upon creation. -- @function class:_init -- @param ... parameters passed to the constructor -- @usage local Cat = class() -- function Cat:_init(name) -- --self:super(name) -- call the ancestor initializer if needed -- self.name = name -- end -- -- local pussycat = Cat("pussycat") -- print(pussycat.name) --> pussycat --- checks whether an __instance__ is derived from some class. -- Works the other way around as `class_of`. It has two ways of using; -- 1) call with a class to check against, 2) call without params. -- @function instance:is_a -- @param some_class class to check against, or `nil` to return the class -- @return `true` if `instance` is derived from `some_class`, or if `some_class == nil` then -- it returns the class table of the instance -- @usage local pussycat = Lion() -- assuming Lion derives from Cat -- if pussycat:is_a(Cat) then -- -- it's true, it is a Lion, but also a Cat -- end -- -- if pussycat:is_a() == Lion then -- -- It's true -- end local function is_a(self,klass) if klass == nil then -- no class provided, so return the class this instance is derived from return getmetatable(self) end local m = getmetatable(self) if not m then return false end --*can't be an object! while m do if m == klass then return true end m = rawget(m,'_base') end return false end --- checks whether an __instance__ is derived from some class. -- Works the other way around as `is_a`. -- @function some_class:class_of -- @param some_instance instance to check against -- @return `true` if `some_instance` is derived from `some_class` -- @usage local pussycat = Lion() -- assuming Lion derives from Cat -- if Cat:class_of(pussycat) then -- -- it's true -- end local function class_of(klass,obj) if type(klass) ~= 'table' or not rawget(klass,'is_a') then return false end return klass.is_a(obj,klass) end --- cast an object to another class. -- It is not clever (or safe!) so use carefully. -- @param some_instance the object to be changed -- @function some_class:cast local function cast (klass, obj) return setmetatable(obj,klass) end local function _class_tostring (obj) local mt = obj._class local name = rawget(mt,'_name') setmetatable(obj,nil) local str = tostring(obj) setmetatable(obj,mt) if name then str = name ..str:gsub('table','') end return str end local function tupdate(td,ts,dont_override) for k,v in pairs(ts) do if not dont_override or td[k] == nil then td[k] = v end end end local function _class(base,c_arg,c) -- the class `c` will be the metatable for all its objects, -- and they will look up their methods in it. local mt = {} -- a metatable for the class to support __call and _handler -- can define class by passing it a plain table of methods local plain = type(base) == 'table' and not getmetatable(base) if plain then c = base base = c._base else c = c or {} end if type(base) == 'table' then -- our new class is a shallow copy of the base class! -- but be careful not to wipe out any methods we have been given at this point! tupdate(c,base,plain) c._base = base -- inherit the 'not found' handler, if present if rawget(c,'_handler') then mt.__index = c._handler end elseif base ~= nil then error("must derive from a table type",3) end c.__index = c setmetatable(c,mt) if not plain then if base and rawget(base,'_init') then c._parent_with_init = base end -- For super and inherited init c._init = nil end if base and rawget(base,'_class_init') then base._class_init(c,c_arg) end -- expose a ctor which can be called by () mt.__call = function(class_tbl,...) local obj if rawget(c,'_create') then obj = c._create(...) end if not obj then obj = {} end setmetatable(obj,c) if rawget(c,'_init') or rawget(c,'_parent_with_init') then -- constructor exists local res = call_ctor(c,obj,...) if res then -- _if_ a ctor returns a value, it becomes the object... obj = res setmetatable(obj,c) end end if base and rawget(base,'_post_init') then base._post_init(obj) end return obj end -- Call Class.catch to set a handler for methods/properties not found in the class! c.catch = function(self, handler) if type(self) == "function" then -- called using . instead of : handler = self end c._handler = handler mt.__index = handler end c.is_a = is_a c.class_of = class_of c.cast = cast c._class = c if not rawget(c,'__tostring') then c.__tostring = _class_tostring end return c end --- create a new class, derived from a given base class. -- Supporting two class creation syntaxes: -- either `Name = class(base)` or `class.Name(base)`. -- The first form returns the class directly and does not set its `_name`. -- The second form creates a variable `Name` in the current environment set -- to the class, and also sets `_name`. -- @function class -- @param base optional base class -- @param c_arg optional parameter to class constructor -- @param c optional table to be used as class local class class = setmetatable({},{ __call = function(fun,...) return _class(...) end, __index = function(tbl,key) if key == 'class' then io.stderr:write('require("pl.class").class is deprecated. Use require("pl.class")\n') return class end compat = compat or require 'pl.compat' local env = compat.getfenv(2) return function(...) local c = _class(...) c._name = key rawset(env,key,c) return c end end }) class.properties = class() function class.properties._class_init(klass) klass.__index = function(t,key) -- normal class lookup! local v = klass[key] if v then return v end -- is it a getter? v = rawget(klass,'get_'..key) if v then return v(t) end -- is it a field? return rawget(t,'_'..key) end klass.__newindex = function (t,key,value) -- if there's a setter, use that, otherwise directly set table local p = 'set_'..key local setter = klass[p] if setter then setter(t,value) else rawset(t,key,value) end end end return class Penlight-1.12.0/lua/pl/compat.lua000066400000000000000000000211541416703176500165440ustar00rootroot00000000000000---------------- --- Lua 5.1/5.2/5.3 compatibility. -- Injects `table.pack`, `table.unpack`, and `package.searchpath` in the global -- environment, to make sure they are available for Lua 5.1 and LuaJIT. -- -- All other functions are exported as usual in the returned module table. -- -- NOTE: everything in this module is also available in `pl.utils`. -- @module pl.compat local compat = {} --- boolean flag this is Lua 5.1 (or LuaJIT). -- @field lua51 compat.lua51 = _VERSION == 'Lua 5.1' --- boolean flag this is LuaJIT. -- @field jit compat.jit = (tostring(assert):match('builtin') ~= nil) --- boolean flag this is LuaJIT with 5.2 compatibility compiled in. -- @field jit52 if compat.jit then -- 'goto' is a keyword when 52 compatibility is enabled in LuaJit compat.jit52 = not loadstring("local goto = 1") end --- the directory separator character for the current platform. -- @field dir_separator compat.dir_separator = _G.package.config:sub(1,1) --- boolean flag this is a Windows platform. -- @field is_windows compat.is_windows = compat.dir_separator == '\\' --- execute a shell command, in a compatible and platform independent way. -- This is a compatibility function that returns the same for Lua 5.1 and -- Lua 5.2+. -- -- NOTE: Windows systems can use signed 32bit integer exitcodes. Posix systems -- only use exitcodes 0-255, anything else is undefined. -- -- NOTE2: In Lua 5.2 and 5.3 a Windows exitcode of -1 would not properly be -- returned, this function will return it properly for all versions. -- @param cmd a shell command -- @return true if successful -- @return actual return code function compat.execute(cmd) local res1,res2,res3 = os.execute(cmd) if res2 == "No error" and res3 == 0 and compat.is_windows then -- os.execute bug in Lua 5.2/5.3 not reporting -1 properly on Windows -- this was fixed in 5.4 res3 = -1 end if compat.lua51 and not compat.jit52 then if compat.is_windows then return res1==0,res1 else res1 = res1 > 255 and res1 / 256 or res1 return res1==0,res1 end else if compat.is_windows then return res3==0,res3 else return not not res1,res3 end end end ---------------- -- Load Lua code as a text or binary chunk (in a Lua 5.2 compatible way). -- @param ld code string or loader -- @param[opt] source name of chunk for errors -- @param[opt] mode 'b', 't' or 'bt' -- @param[opt] env environment to load the chunk in -- @function compat.load --------------- -- Get environment of a function (in a Lua 5.1 compatible way). -- Not 100% compatible, so with Lua 5.2 it may return nil for a function with no -- global references! -- Based on code by [Sergey Rozhenko](http://lua-users.org/lists/lua-l/2010-06/msg00313.html) -- @param f a function or a call stack reference -- @function compat.getfenv --------------- -- Set environment of a function (in a Lua 5.1 compatible way). -- @param f a function or a call stack reference -- @param env a table that becomes the new environment of `f` -- @function compat.setfenv if compat.lua51 then -- define Lua 5.2 style load() if not compat.jit then -- but LuaJIT's load _is_ compatible local lua51_load = load function compat.load(str,src,mode,env) local chunk,err if type(str) == 'string' then if str:byte(1) == 27 and not (mode or 'bt'):find 'b' then return nil,"attempt to load a binary chunk" end chunk,err = loadstring(str,src) else chunk,err = lua51_load(str,src) end if chunk and env then setfenv(chunk,env) end return chunk,err end else compat.load = load end compat.setfenv, compat.getfenv = setfenv, getfenv else compat.load = load -- setfenv/getfenv replacements for Lua 5.2 -- by Sergey Rozhenko -- http://lua-users.org/lists/lua-l/2010-06/msg00313.html -- Roberto Ierusalimschy notes that it is possible for getfenv to return nil -- in the case of a function with no globals: -- http://lua-users.org/lists/lua-l/2010-06/msg00315.html function compat.setfenv(f, t) f = (type(f) == 'function' and f or debug.getinfo(f + 1, 'f').func) local name local up = 0 repeat up = up + 1 name = debug.getupvalue(f, up) until name == '_ENV' or name == nil if name then debug.upvaluejoin(f, up, function() return name end, 1) -- use unique upvalue debug.setupvalue(f, up, t) end if f ~= 0 then return f end end function compat.getfenv(f) local f = f or 0 f = (type(f) == 'function' and f or debug.getinfo(f + 1, 'f').func) local name, val local up = 0 repeat up = up + 1 name, val = debug.getupvalue(f, up) until name == '_ENV' or name == nil return val end end --- Global exported functions (for Lua 5.1 & LuaJIT) -- @section lua52 --- pack an argument list into a table. -- @param ... any arguments -- @return a table with field n set to the length -- @function table.pack if not table.pack then function table.pack (...) -- luacheck: ignore return {n=select('#',...); ...} end end --- unpack a table and return the elements. -- -- NOTE: this version does NOT honor the n field, and hence it is not nil-safe. -- See `utils.unpack` for a version that is nil-safe. -- @param t table to unpack -- @param[opt] i index from which to start unpacking, defaults to 1 -- @param[opt] j index of the last element to unpack, defaults to #t -- @return multiple return values from the table -- @function table.unpack -- @see utils.unpack if not table.unpack then table.unpack = unpack -- luacheck: ignore end --- return the full path where a file name would be matched. -- This function was introduced in Lua 5.2, so this compatibility version -- will be injected in Lua 5.1 engines. -- @string name file name, possibly dotted -- @string path a path-template in the same form as package.path or package.cpath -- @string[opt] sep template separate character to be replaced by path separator. Default: "." -- @string[opt] rep the path separator to use, defaults to system separator. Default; "/" on Unixes, "\" on Windows. -- @see path.package_path -- @function package.searchpath -- @return on success: path of the file -- @return on failure: nil, error string listing paths tried if not package.searchpath then function package.searchpath (name,path,sep,rep) -- luacheck: ignore if type(name) ~= "string" then error(("bad argument #1 to 'searchpath' (string expected, got %s)"):format(type(path)), 2) end if type(path) ~= "string" then error(("bad argument #2 to 'searchpath' (string expected, got %s)"):format(type(path)), 2) end if sep ~= nil and type(sep) ~= "string" then error(("bad argument #3 to 'searchpath' (string expected, got %s)"):format(type(path)), 2) end if rep ~= nil and type(rep) ~= "string" then error(("bad argument #4 to 'searchpath' (string expected, got %s)"):format(type(path)), 2) end sep = sep or "." rep = rep or compat.dir_separator do local s, e = name:find(sep, nil, true) while s do name = name:sub(1, s-1) .. rep .. name:sub(e+1, -1) s, e = name:find(sep, s + #rep + 1, true) end end local tried = {} for m in path:gmatch('[^;]+') do local nm = m:gsub('?', name) tried[#tried+1] = nm local f = io.open(nm,'r') if f then f:close(); return nm end end return nil, "\tno file '" .. table.concat(tried, "'\n\tno file '") .. "'" end end --- Global exported functions (for Lua < 5.4) -- @section lua54 --- raise a warning message. -- This functions mimics the `warn` function added in Lua 5.4. -- @function warn -- @param ... any arguments if not warn then -- luacheck: ignore local enabled = false function warn(arg1, ...) -- luacheck: ignore if type(arg1) == "string" and arg1:sub(1, 1) == "@" then -- control message if arg1 == "@on" then enabled = true return end if arg1 == "@off" then enabled = false return end return -- ignore unknown control messages end if enabled then io.stderr:write("Lua warning: ", arg1, ...) io.stderr:write("\n") end end end return compat Penlight-1.12.0/lua/pl/comprehension.lua000066400000000000000000000215461416703176500201370ustar00rootroot00000000000000--- List comprehensions implemented in Lua. -- -- See the [wiki page](http://lua-users.org/wiki/ListComprehensions) -- -- local C= require 'pl.comprehension' . new() -- -- C ('x for x=1,10') () -- ==> {1,2,3,4,5,6,7,8,9,10} -- C 'x^2 for x=1,4' () -- ==> {1,4,9,16} -- C '{x,x^2} for x=1,4' () -- ==> {{1,1},{2,4},{3,9},{4,16}} -- C '2*x for x' {1,2,3} -- ==> {2,4,6} -- dbl = C '2*x for x' -- dbl {10,20,30} -- ==> {20,40,60} -- C 'x for x if x % 2 == 0' {1,2,3,4,5} -- ==> {2,4} -- C '{x,y} for x = 1,2 for y = 1,2' () -- ==> {{1,1},{1,2},{2,1},{2,2}} -- C '{x,y} for x for y' ({1,2},{10,20}) -- ==> {{1,10},{1,20},{2,10},{2,20}} -- assert(C 'sum(x^2 for x)' {2,3,4} == 2^2+3^2+4^2) -- -- (c) 2008 David Manura. Licensed under the same terms as Lua (MIT license). -- -- Dependencies: `pl.utils`, `pl.luabalanced` -- -- See @{07-functional.md.List_Comprehensions|the Guide} -- @module pl.comprehension local utils = require 'pl.utils' local status,lb = pcall(require, "pl.luabalanced") if not status then lb = require 'luabalanced' end local math_max = math.max local table_concat = table.concat -- fold operations -- http://en.wikipedia.org/wiki/Fold_(higher-order_function) local ops = { list = {init=' {} ', accum=' __result[#__result+1] = (%s) '}, table = {init=' {} ', accum=' local __k, __v = %s __result[__k] = __v '}, sum = {init=' 0 ', accum=' __result = __result + (%s) '}, min = {init=' nil ', accum=' local __tmp = %s ' .. ' if __result then if __tmp < __result then ' .. '__result = __tmp end else __result = __tmp end '}, max = {init=' nil ', accum=' local __tmp = %s ' .. ' if __result then if __tmp > __result then ' .. '__result = __tmp end else __result = __tmp end '}, } -- Parses comprehension string expr. -- Returns output expression list string, array of for types -- ('=', 'in' or nil) , array of input variable name -- strings , array of input variable value strings -- , array of predicate expression strings , -- operation name string , and number of placeholder -- parameters . -- -- The is equivalent to the mathematical set-builder notation: -- -- { | in , } -- -- @usage "x^2 for x" -- array values -- @usage "x^2 for x=1,10,2" -- numeric for -- @usage "k^v for k,v in pairs(_1)" -- iterator for -- @usage "(x+y)^2 for x for y if x > y" -- nested -- local function parse_comprehension(expr) local pos = 1 -- extract opname (if exists) local opname local tok, post = expr:match('^%s*([%a_][%w_]*)%s*%(()', pos) local pose = #expr + 1 if tok then local tok2, posb = lb.match_bracketed(expr, post-1) assert(tok2, 'syntax error') if expr:match('^%s*$', posb) then opname = tok pose = posb - 1 pos = post end end opname = opname or "list" -- extract out expression list local out; out, pos = lb.match_explist(expr, pos) assert(out, "syntax error: missing expression list") out = table_concat(out, ', ') -- extract "for" clauses local fortypes = {} local invarlists = {} local invallists = {} while 1 do local post = expr:match('^%s*for%s+()', pos) if not post then break end pos = post -- extract input vars local iv; iv, pos = lb.match_namelist(expr, pos) assert(#iv > 0, 'syntax error: zero variables') for _,ident in ipairs(iv) do assert(not ident:match'^__', "identifier " .. ident .. " may not contain __ prefix") end invarlists[#invarlists+1] = iv -- extract '=' or 'in' (optional) local fortype, post = expr:match('^(=)%s*()', pos) if not fortype then fortype, post = expr:match('^(in)%s+()', pos) end if fortype then pos = post -- extract input value range local il; il, pos = lb.match_explist(expr, pos) assert(#il > 0, 'syntax error: zero expressions') assert(fortype ~= '=' or #il == 2 or #il == 3, 'syntax error: numeric for requires 2 or three expressions') fortypes[#invarlists] = fortype invallists[#invarlists] = il else fortypes[#invarlists] = false invallists[#invarlists] = false end end assert(#invarlists > 0, 'syntax error: missing "for" clause') -- extract "if" clauses local preds = {} while 1 do local post = expr:match('^%s*if%s+()', pos) if not post then break end pos = post local pred; pred, pos = lb.match_expression(expr, pos) assert(pred, 'syntax error: predicated expression not found') preds[#preds+1] = pred end -- extract number of parameter variables (name matching "_%d+") local stmp = ''; lb.gsub(expr, function(u, sin) -- strip comments/strings if u == 'e' then stmp = stmp .. ' ' .. sin .. ' ' end end) local max_param = 0; stmp:gsub('[%a_][%w_]*', function(s) local s = s:match('^_(%d+)$') if s then max_param = math_max(max_param, tonumber(s)) end end) if pos ~= pose then assert(false, "syntax error: unrecognized " .. expr:sub(pos)) end --DEBUG: --print('----\n', string.format("%q", expr), string.format("%q", out), opname) --for k,v in ipairs(invarlists) do print(k,v, invallists[k]) end --for k,v in ipairs(preds) do print(k,v) end return out, fortypes, invarlists, invallists, preds, opname, max_param end -- Create Lua code string representing comprehension. -- Arguments are in the form returned by parse_comprehension. local function code_comprehension( out, fortypes, invarlists, invallists, preds, opname, max_param ) local op = assert(ops[opname]) local code = op.accum:gsub('%%s', out) for i=#preds,1,-1 do local pred = preds[i] code = ' if ' .. pred .. ' then ' .. code .. ' end ' end for i=#invarlists,1,-1 do if not fortypes[i] then local arrayname = '__in' .. i local idx = '__idx' .. i code = ' for ' .. idx .. ' = 1, #' .. arrayname .. ' do ' .. ' local ' .. invarlists[i][1] .. ' = ' .. arrayname .. '['..idx..'] ' .. code .. ' end ' else code = ' for ' .. table_concat(invarlists[i], ', ') .. ' ' .. fortypes[i] .. ' ' .. table_concat(invallists[i], ', ') .. ' do ' .. code .. ' end ' end end code = ' local __result = ( ' .. op.init .. ' ) ' .. code return code end -- Convert code string represented by code_comprehension -- into Lua function. Also must pass ninputs = #invarlists, -- max_param, and invallists (from parse_comprehension). -- Uses environment env. local function wrap_comprehension(code, ninputs, max_param, invallists, env) assert(ninputs > 0) local ts = {} for i=1,max_param do ts[#ts+1] = '_' .. i end for i=1,ninputs do if not invallists[i] then local name = '__in' .. i ts[#ts+1] = name end end if #ts > 0 then code = ' local ' .. table_concat(ts, ', ') .. ' = ... ' .. code end code = code .. ' return __result ' --print('DEBUG:', code) local f, err = utils.load(code,'tmp','t',env) if not f then assert(false, err .. ' with generated code ' .. code) end return f end -- Build Lua function from comprehension string. -- Uses environment env. local function build_comprehension(expr, env) local out, fortypes, invarlists, invallists, preds, opname, max_param = parse_comprehension(expr) local code = code_comprehension( out, fortypes, invarlists, invallists, preds, opname, max_param) local f = wrap_comprehension(code, #invarlists, max_param, invallists, env) return f end -- Creates new comprehension cache. -- Any list comprehension function created are set to the environment -- env (defaults to caller of new). local function new(env) -- Note: using a single global comprehension cache would have had -- security implications (e.g. retrieving cached functions created -- in other environments). -- The cache lookup function could have instead been written to retrieve -- the caller's environment, lookup up the cache private to that -- environment, and then looked up the function in that cache. -- That would avoid the need for this call to -- explicitly manage caches; however, that might also have an undue -- performance penalty. if not env then env = utils.getfenv(2) end local mt = {} local cache = setmetatable({}, mt) -- Index operator builds, caches, and returns Lua function -- corresponding to comprehension expression string. -- -- Example: f = comprehension['x^2 for x'] -- function mt:__index(expr) local f = build_comprehension(expr, env) self[expr] = f -- cache return f end -- Convenience syntax. -- Allows comprehension 'x^2 for x' instead of comprehension['x^2 for x']. mt.__call = mt.__index cache.new = new return cache end local comprehension = {} comprehension.new = new return comprehension Penlight-1.12.0/lua/pl/config.lua000066400000000000000000000151161416703176500165270ustar00rootroot00000000000000--- Reads configuration files into a Lua table. -- Understands INI files, classic Unix config files, and simple -- delimited columns of values. See @{06-data.md.Reading_Configuration_Files|the Guide} -- -- # test.config -- # Read timeout in seconds -- read.timeout=10 -- # Write timeout in seconds -- write.timeout=5 -- #acceptable ports -- ports = 1002,1003,1004 -- -- -- readconfig.lua -- local config = require 'config' -- local t = config.read 'test.config' -- print(pretty.write(t)) -- -- ### output ##### -- { -- ports = { -- 1002, -- 1003, -- 1004 -- }, -- write_timeout = 5, -- read_timeout = 10 -- } -- -- @module pl.config local type,tonumber,ipairs,io, table = _G.type,_G.tonumber,_G.ipairs,_G.io,_G.table local function split(s,re) local res = {} local t_insert = table.insert re = '[^'..re..']+' for k in s:gmatch(re) do t_insert(res,k) end return res end local function strip(s) return s:gsub('^%s+',''):gsub('%s+$','') end local function strip_quotes (s) return s:gsub("['\"](.*)['\"]",'%1') end local config = {} --- like io.lines(), but allows for lines to be continued with '\'. -- @param file a file-like object (anything where read() returns the next line) or a filename. -- Defaults to stardard input. -- @return an iterator over the lines, or nil -- @return error 'not a file-like object' or 'file is nil' function config.lines(file) local f,openf,err local line = '' if type(file) == 'string' then f,err = io.open(file,'r') if not f then return nil,err end openf = true else f = file or io.stdin if not file.read then return nil, 'not a file-like object' end end if not f then return nil, 'file is nil' end return function() local l = f:read() while l do -- only for non-blank lines that don't begin with either ';' or '#' if l:match '%S' and not l:match '^%s*[;#]' then -- does the line end with '\'? local i = l:find '\\%s*$' if i then -- if so, line = line..l:sub(1,i-1) elseif line == '' then return l else l = line..l line = '' return l end end l = f:read() end if openf then f:close() end end end --- read a configuration file into a table -- @param file either a file-like object or a string, which must be a filename -- @tab[opt] cnfg a configuration table that may contain these fields: -- -- * `smart` try to deduce what kind of config file we have (default false) -- * `variabilize` make names into valid Lua identifiers (default true) -- * `convert_numbers` try to convert values into numbers (default true) -- * `trim_space` ensure that there is no starting or trailing whitespace with values (default true) -- * `trim_quotes` remove quotes from strings (default false) -- * `list_delim` delimiter to use when separating columns (default ',') -- * `keysep` separator between key and value pairs (default '=') -- -- @return a table containing items, or `nil` -- @return error message (same as @{config.lines} function config.read(file,cnfg) local auto local iter,err = config.lines(file) if not iter then return nil,err end local line = iter() cnfg = cnfg or {} if cnfg.smart then auto = true if line:match '^[^=]+=' then cnfg.keysep = '=' elseif line:match '^[^:]+:' then cnfg.keysep = ':' cnfg.list_delim = ':' elseif line:match '^%S+%s+' then cnfg.keysep = ' ' -- more than two columns assume that it's a space-delimited list -- cf /etc/fstab with /etc/ssh/ssh_config if line:match '^%S+%s+%S+%s+%S+' then cnfg.list_delim = ' ' end cnfg.variabilize = false end end local function check_cnfg (var,def) local val = cnfg[var] if val == nil then return def else return val end end local initial_digits = '^[%d%+%-]' local t = {} local top_t = t local variabilize = check_cnfg ('variabilize',true) local list_delim = check_cnfg('list_delim',',') local convert_numbers = check_cnfg('convert_numbers',true) local convert_boolean = check_cnfg('convert_boolean',false) local trim_space = check_cnfg('trim_space',true) local trim_quotes = check_cnfg('trim_quotes',false) local ignore_assign = check_cnfg('ignore_assign',false) local keysep = check_cnfg('keysep','=') local keypat = keysep == ' ' and '%s+' or '%s*'..keysep..'%s*' if list_delim == ' ' then list_delim = '%s+' end local function process_name(key) if variabilize then key = key:gsub('[^%w]','_') end return key end local function process_value(value) if list_delim and value:find(list_delim) then value = split(value,list_delim) for i,v in ipairs(value) do value[i] = process_value(v) end elseif convert_numbers and value:find(initial_digits) then local val = tonumber(value) if not val and value:match ' kB$' then value = value:gsub(' kB','') val = tonumber(value) end if val then value = val end elseif convert_boolean and value == 'true' then return true elseif convert_boolean and value == 'false' then return false end if type(value) == 'string' then if trim_space then value = strip(value) end if not trim_quotes and auto and value:match '^"' then trim_quotes = true end if trim_quotes then value = strip_quotes(value) end end return value end while line do if line:find('^%[') then -- section! local section = process_name(line:match('%[([^%]]+)%]')) t = top_t t[section] = {} t = t[section] else line = line:gsub('^%s*','') local i1,i2 = line:find(keypat) if i1 and not ignore_assign then -- key,value assignment local key = process_name(line:sub(1,i1-1)) local value = process_value(line:sub(i2+1)) t[key] = value else -- a plain list of values... t[#t+1] = process_value(line) end end line = iter() end return top_t end return config Penlight-1.12.0/lua/pl/data.lua000066400000000000000000000474131416703176500162000ustar00rootroot00000000000000--- Reading and querying simple tabular data. -- -- data.read 'test.txt' -- ==> {{10,20},{2,5},{40,50},fieldnames={'x','y'},delim=','} -- -- Provides a way of creating basic SQL-like queries. -- -- require 'pl' -- local d = data.read('xyz.txt') -- local q = d:select('x,y,z where x > 3 and z < 2 sort by y') -- for x,y,z in q do -- print(x,y,z) -- end -- -- See @{06-data.md.Reading_Columnar_Data|the Guide} -- -- Dependencies: `pl.utils`, `pl.array2d` (fallback methods) -- @module pl.data local utils = require 'pl.utils' local _DEBUG = rawget(_G,'_DEBUG') local patterns,function_arg,usplit,array_tostring = utils.patterns,utils.function_arg,utils.split,utils.array_tostring local append,concat = table.insert,table.concat local gsub = string.gsub local io = io local _G,print,type,tonumber,ipairs,setmetatable = _G,print,type,tonumber,ipairs,setmetatable local data = {} local parse_select local function rstrip(s) return (s:gsub('%s+$','')) end local function strip (s) return (rstrip(s):gsub('^%s*','')) end -- This gives `l` the standard List metatable, -- pulling in the List module. local function makelist(l) return setmetatable(l, require('pl.List')) end local function map(fun,t) local res = {} for i = 1,#t do res[i] = fun(t[i]) end return res end local function split(line,delim,csv,n) local massage -- CSV fields may be double-quoted and may contain commas! if csv and line:match '"' then line = line:gsub('"([^"]+)"',function(str) local s,cnt = str:gsub(',','\001') if cnt > 0 then massage = true end return s end) if massage then massage = function(s) return (s:gsub('\001',',')) end end end local res = (usplit(line,delim,false,n)) if csv then -- restore CSV commas-in-fields if massage then res = map(massage,res) end -- in CSV mode trailiing commas are significant! if line:match ',$' then append(res,'') end end return makelist(res) end local function find(t,v) for i = 1,#t do if v == t[i] then return i end end end local DataMT = { column_by_name = function(self,name) if type(name) == 'number' then name = '$'..name end local arr = {} for res in data.query(self,name) do append(arr,res) end return makelist(arr) end, copy_select = function(self,condn) condn = parse_select(condn,self) local iter = data.query(self,condn) local res = {} local row = makelist{iter()} while #row > 0 do append(res,row) row = makelist{iter()} end res.delim = self.delim return data.new(res,split(condn.fields,',')) end, column_names = function(self) return self.fieldnames end, } local array2d DataMT.__index = function(self,name) local f = DataMT[name] if f then return f end if not array2d then array2d = require 'pl.array2d' end return array2d[name] end --- return a particular column as a list of values (method). -- @param name either name of column, or numerical index. -- @function Data.column_by_name --- return a query iterator on this data (method). -- @string condn the query expression -- @function Data.select -- @see data.query --- return a row iterator on this data (method). -- @string condn the query expression -- @function Data.select_row --- return a new data object based on this query (method). -- @string condn the query expression -- @function Data.copy_select --- return the field names of this data object (method). -- @function Data.column_names --- write out a row (method). -- @param f file-like object -- @function Data.write_row --- write data out to file (method). -- @param f file-like object -- @function Data.write -- [guessing delimiter] We check for comma, tab and spaces in that order. -- [issue] any other delimiters to be checked? local delims = {',', '\t', ' ', ';'} local function guess_delim (line) if line=='' then return ' ' end for _,delim in ipairs(delims) do if line:find(delim) then return delim == ' ' and '%s+' or delim end end return ' ' end -- [file parameter] If it's a string, we try open as a filename. If nil, then -- either stdin or stdout depending on the mode. Otherwise, check if this is -- a file-like object (implements read or write depending) local function open_file (f,mode) local opened, err local reading = mode == 'r' if type(f) == 'string' then if f == 'stdin' then f = io.stdin elseif f == 'stdout' then f = io.stdout else f,err = io.open(f,mode) if not f then return nil,err end opened = true end end if f and ((reading and not f.read) or (not reading and not f.write)) then return nil, "not a file-like object" end return f,nil,opened end --- read a delimited file in a Lua table. -- By default, attempts to treat first line as separated list of fieldnames. -- @param file a filename or a file-like object -- @tab cnfg parsing options -- @string cnfg.delim a string pattern to split fields -- @array cnfg.fieldnames (i.e. don't read from first line) -- @bool cnfg.no_convert (default is to try conversion on first data line) -- @tab cnfg.convert table of custom conversion functions with column keys -- @int cnfg.numfields indices of columns known to be numbers -- @bool cnfg.last_field_collect only split as many fields as fieldnames. -- @int cnfg.thousands_dot thousands separator in Excel CSV is '.' -- @bool cnfg.csv fields may be double-quoted and contain commas; -- Also, empty fields are considered to be equivalent to zero. -- @return `data` object, or `nil` -- @return error message. May be a file error, 'not a file-like object' -- or a conversion error function data.read(file,cnfg) local count,line local D = {} if not cnfg then cnfg = {} end local f,err,opened = open_file(file,'r') if not f then return nil, err end local thousands_dot = cnfg.thousands_dot local csv = cnfg.csv if csv then cnfg.delim = ',' end -- note that using dot as the thousands separator (@thousands_dot) -- requires a special conversion function! For CSV, _empty fields_ are -- considered to default to numerial zeroes. local tonumber = tonumber local function try_number(x) if thousands_dot then x = x:gsub('%.(...)','%1') end if csv and x == '' then x = '0' end local v = tonumber(x) if v == nil then return nil,"not a number" end return v end count = 1 line = f:read() if not line then return nil, "empty file" end -- first question: what is the delimiter? D.delim = cnfg.delim and cnfg.delim or guess_delim(line) local delim = D.delim local conversion local numfields = {} local function append_conversion (idx,conv) conversion = conversion or {} append(numfields,idx) append(conversion,conv) end if cnfg.numfields then for _,n in ipairs(cnfg.numfields) do append_conversion(n,try_number) end end -- some space-delimited data starts with a space. This should not be a column, -- although it certainly would be for comma-separated, etc. local stripper if delim == '%s+' and line:find(delim) == 1 then stripper = function(s) return s:gsub('^%s+','') end line = stripper(line) end -- first line will usually be field names. Unless fieldnames are specified, -- we check if it contains purely numerical values for the case of reading -- plain data files. if not cnfg.fieldnames then local fields,nums fields = split(line,delim,csv) if not cnfg.convert then nums = map(tonumber,fields) if #nums == #fields then -- they're ALL numbers! append(D,nums) -- add the first converted row -- and specify conversions for subsequent rows for i = 1,#nums do append_conversion(i,try_number) end else -- we'll try to check numbers just now.. nums = nil end else -- [explicit column conversions] (any deduced number conversions will be added) for idx,conv in pairs(cnfg.convert) do append_conversion(idx,conv) end end if nums == nil then cnfg.fieldnames = fields end line = f:read() count = count + 1 if stripper then line = stripper(line) end elseif type(cnfg.fieldnames) == 'string' then cnfg.fieldnames = split(cnfg.fieldnames,delim,csv) end local nfields -- at this point, the column headers have been read in. If the first -- row consisted of numbers, it has already been added to the dataset. if cnfg.fieldnames then D.fieldnames = cnfg.fieldnames -- [collecting end field] If @last_field_collect then we'll -- only split as many fields as there are fieldnames if cnfg.last_field_collect then nfields = #D.fieldnames end -- [implicit column conversion] unless @no_convert, we need the numerical field indices -- of the first data row. These can also be specified explicitly by @numfields. if not cnfg.no_convert then local fields = split(line,D.delim,csv,nfields) for i = 1,#fields do if not find(numfields,i) and try_number(fields[i]) then append_conversion(i,try_number) end end end end -- keep going until finished while line do if not line:find ('^%s*$') then -- [blank lines] ignore them! if stripper then line = stripper(line) end local fields = split(line,delim,csv,nfields) if conversion then -- there were field conversions... for k = 1,#numfields do local i,conv = numfields[k],conversion[k] local val,err = conv(fields[i]) if val == nil then return nil, err..": "..fields[i].." at line "..count else fields[i] = val end end end append(D,fields) end line = f:read() count = count + 1 end if opened then f:close() end if delim == '%s+' then D.delim = ' ' end if not D.fieldnames then D.fieldnames = {} end return data.new(D) end local function write_row (data,f,row,delim) data.temp = array_tostring(row,data.temp) f:write(concat(data.temp,delim),'\n') end function DataMT:write_row(f,row) write_row(self,f,row,self.delim) end --- write 2D data to a file. -- Does not assume that the data has actually been -- generated with `new` or `read`. -- @param data 2D array -- @param file filename or file-like object -- @tparam[opt] {string} fieldnames list of fields (optional) -- @string[opt='\t'] delim delimiter (default tab) -- @return true or nil, error function data.write (data,file,fieldnames,delim) local f,err,opened = open_file(file,'w') if not f then return nil, err end if not fieldnames then fieldnames = data.fieldnames end delim = delim or '\t' if fieldnames and #fieldnames > 0 then f:write(concat(fieldnames,delim),'\n') end for i = 1,#data do write_row(data,f,data[i],delim) end if opened then f:close() end return true end function DataMT:write(file) data.write(self,file,self.fieldnames,self.delim) end local function massage_fieldnames (fields,copy) -- fieldnames must be valid Lua identifiers; ignore any surrounding padding -- but keep the original fieldnames... for i = 1,#fields do local f = strip(fields[i]) copy[i] = f fields[i] = f:gsub('%W','_') end end --- create a new dataset from a table of rows. -- Can specify the fieldnames, else the table must have a field called -- 'fieldnames', which is either a string of delimiter-separated names, -- or a table of names.
-- If the table does not have a field called 'delim', then an attempt will be -- made to guess it from the fieldnames string, defaults otherwise to tab. -- @param d the table. -- @tparam[opt] {string} fieldnames optional fieldnames -- @return the table. function data.new (d,fieldnames) d.fieldnames = d.fieldnames or fieldnames or '' if not d.delim and type(d.fieldnames) == 'string' then d.delim = guess_delim(d.fieldnames) d.fieldnames = split(d.fieldnames,d.delim) end d.fieldnames = makelist(d.fieldnames) d.original_fieldnames = {} massage_fieldnames(d.fieldnames,d.original_fieldnames) setmetatable(d,DataMT) -- a query with just the fieldname will return a sequence -- of values, which seq.copy turns into a table. return d end local sorted_query = [[ return function (t) local i = 0 local v local ls = {} for i,v in ipairs(t) do if CONDITION then ls[#ls+1] = v end end table.sort(ls,function(v1,v2) return SORT_EXPR end) local n = #ls return function() i = i + 1 v = ls[i] if i > n then return end return FIELDLIST end end ]] -- question: is this optimized case actually worth the extra code? local simple_query = [[ return function (t) local n = #t local i = 0 local v return function() repeat i = i + 1 v = t[i] until i > n or CONDITION if i > n then return end return FIELDLIST end end ]] local function is_string (s) return type(s) == 'string' end local field_error local function fieldnames_as_string (data) return concat(data.fieldnames,',') end local function massage_fields(data,f) local idx if f:find '^%d+$' then idx = tonumber(f) else idx = find(data.fieldnames,f) end if idx then return 'v['..idx..']' else field_error = f..' not found in '..fieldnames_as_string(data) return f end end local function process_select (data,parms) --- preparing fields ---- field_error = nil local fields = parms.fields local numfields = fields:find '%$' or #data.fieldnames == 0 if fields:find '^%s*%*%s*' then if not numfields then fields = fieldnames_as_string(data) else local ncol = #data[1] fields = {} for i = 1,ncol do append(fields,'$'..i) end fields = concat(fields,',') end end local idpat = patterns.IDEN if numfields then idpat = '%$(%d+)' else -- massage field names to replace non-identifier chars fields = rstrip(fields):gsub('[^,%w]','_') end local massage_fields = utils.bind1(massage_fields,data) local ret = gsub(fields,idpat,massage_fields) if field_error then return nil,field_error end parms.fields = fields parms.proc_fields = ret parms.where = parms.where or 'true' if is_string(parms.where) then parms.where = gsub(parms.where,idpat,massage_fields) field_error = nil end return true end parse_select = function(s,data) local endp local parms = {} local w1,w2 = s:find('where ') local s1,s2 = s:find('sort by ') if w1 then -- where clause! endp = (s1 or 0)-1 parms.where = s:sub(w2+1,endp) end if s1 then -- sort by clause (must be last!) parms.sort_by = s:sub(s2+1) end endp = (w1 or s1 or 0)-1 parms.fields = s:sub(1,endp) local status,err = process_select(data,parms) if not status then return nil,err else return parms end end --- create a query iterator from a select string. -- Select string has this format:
-- FIELDLIST [ where LUA-CONDN [ sort by FIELD] ]
-- FIELDLIST is a comma-separated list of valid fields, or '*'.

-- The condition can also be a table, with fields 'fields' (comma-sep string or -- table), 'sort_by' (string) and 'where' (Lua expression string or function) -- @param data table produced by read -- @param condn select string or table -- @param context a list of tables to be searched when resolving functions -- @param return_row if true, wrap the results in a row table -- @return an iterator over the specified fields, or nil -- @return an error message function data.query(data,condn,context,return_row) local err if is_string(condn) then condn,err = parse_select(condn,data) if not condn then return nil,err end elseif type(condn) == 'table' then if type(condn.fields) == 'table' then condn.fields = concat(condn.fields,',') end if not condn.proc_fields then local status,err = process_select(data,condn) if not status then return nil,err end end else return nil, "condition must be a string or a table" end local query if condn.sort_by then -- use sorted_query query = sorted_query else query = simple_query end local fields = condn.proc_fields or condn.fields if return_row then fields = '{'..fields..'}' end query = query:gsub('FIELDLIST',fields) if is_string(condn.where) then query = query:gsub('CONDITION',condn.where) condn.where = nil else query = query:gsub('CONDITION','_condn(v)') condn.where = function_arg(0,condn.where,'condition.where must be callable') end if condn.sort_by then local expr,sort_var,sort_dir local sort_by = condn.sort_by local i1,i2 = sort_by:find('%s+') if i1 then sort_var,sort_dir = sort_by:sub(1,i1-1),sort_by:sub(i2+1) else sort_var = sort_by sort_dir = 'asc' end if sort_var:match '^%$' then sort_var = sort_var:sub(2) end sort_var = massage_fields(data,sort_var) if field_error then return nil,field_error end if sort_dir == 'asc' then sort_dir = '<' else sort_dir = '>' end expr = ('%s %s %s'):format(sort_var:gsub('v','v1'),sort_dir,sort_var:gsub('v','v2')) query = query:gsub('SORT_EXPR',expr) end if condn.where then query = 'return function(_condn) '..query..' end' end if _DEBUG then print(query) end local fn,err = utils.load(query,'tmp') if not fn then return nil,err end fn = fn() -- get the function if condn.where then fn = fn(condn.where) end local qfun = fn(data) if context then -- [specifying context for condition] @context is a list of tables which are -- 'injected'into the condition's custom context append(context,_G) local lookup = {} utils.setfenv(qfun,lookup) setmetatable(lookup,{ __index = function(tbl,key) -- _G.print(tbl,key) for k,t in ipairs(context) do if t[key] then return t[key] end end end }) end return qfun end DataMT.select = data.query DataMT.select_row = function(d,condn,context) return data.query(d,condn,context,true) end --- Filter input using a query. -- @string Q a query string -- @param infile filename or file-like object -- @param outfile filename or file-like object -- @bool dont_fail true if you want to return an error, not just fail function data.filter (Q,infile,outfile,dont_fail) local d = data.read(infile or 'stdin') local out = open_file(outfile or 'stdout') local iter,err = d:select(Q) local delim = d.delim if not iter then err = 'error: '..err if dont_fail then return nil,err else utils.quit(1,err) end end while true do local res = {iter()} if #res == 0 then break end out:write(concat(res,delim),'\n') end end return data Penlight-1.12.0/lua/pl/dir.lua000066400000000000000000000423421416703176500160410ustar00rootroot00000000000000--- Listing files in directories and creating/removing directory paths. -- -- Dependencies: `pl.utils`, `pl.path` -- -- Soft Dependencies: `alien`, `ffi` (either are used on Windows for copying/moving files) -- @module pl.dir local utils = require 'pl.utils' local path = require 'pl.path' local is_windows = path.is_windows local ldir = path.dir local mkdir = path.mkdir local rmdir = path.rmdir local sub = string.sub local os,pcall,ipairs,pairs,require,setmetatable = os,pcall,ipairs,pairs,require,setmetatable local remove = os.remove local append = table.insert local assert_arg,assert_string,raise = utils.assert_arg,utils.assert_string,utils.raise local exists, isdir = path.exists, path.isdir local sep = path.sep local dir = {} local function makelist(l) return setmetatable(l, require('pl.List')) end local function assert_dir (n,val) assert_arg(n,val,'string',path.isdir,'not a directory',4) end local function filemask(mask) mask = utils.escape(path.normcase(mask)) return '^'..mask:gsub('%%%*','.*'):gsub('%%%?','.')..'$' end --- Test whether a file name matches a shell pattern. -- Both parameters are case-normalized if operating system is -- case-insensitive. -- @string filename A file name. -- @string pattern A shell pattern. The only special characters are -- `'*'` and `'?'`: `'*'` matches any sequence of characters and -- `'?'` matches any single character. -- @treturn bool -- @raise dir and mask must be strings function dir.fnmatch(filename,pattern) assert_string(1,filename) assert_string(2,pattern) return path.normcase(filename):find(filemask(pattern)) ~= nil end --- Return a list of all file names within an array which match a pattern. -- @tab filenames An array containing file names. -- @string pattern A shell pattern (see `fnmatch`). -- @treturn List(string) List of matching file names. -- @raise dir and mask must be strings function dir.filter(filenames,pattern) assert_arg(1,filenames,'table') assert_string(2,pattern) local res = {} local mask = filemask(pattern) for i,f in ipairs(filenames) do if path.normcase(f):find(mask) then append(res,f) end end return makelist(res) end local function _listfiles(dirname,filemode,match) local res = {} local check = utils.choose(filemode,path.isfile,path.isdir) if not dirname then dirname = '.' end for f in ldir(dirname) do if f ~= '.' and f ~= '..' then local p = path.join(dirname,f) if check(p) and (not match or match(f)) then append(res,p) end end end return makelist(res) end --- return a list of all files in a directory which match a shell pattern. -- @string[opt='.'] dirname A directory. -- @string[opt] mask A shell pattern (see `fnmatch`). If not given, all files are returned. -- @treturn {string} list of files -- @raise dirname and mask must be strings function dir.getfiles(dirname,mask) dirname = dirname or '.' assert_dir(1,dirname) if mask then assert_string(2,mask) end local match if mask then mask = filemask(mask) match = function(f) return path.normcase(f):find(mask) end end return _listfiles(dirname,true,match) end --- return a list of all subdirectories of the directory. -- @string[opt='.'] dirname A directory. -- @treturn {string} a list of directories -- @raise dir must be a valid directory function dir.getdirectories(dirname) dirname = dirname or '.' assert_dir(1,dirname) return _listfiles(dirname,false) end local alien,ffi,ffi_checked,CopyFile,MoveFile,GetLastError,win32_errors,cmd_tmpfile local function execute_command(cmd,parms) if not cmd_tmpfile then cmd_tmpfile = path.tmpname () end local err = path.is_windows and ' > ' or ' 2> ' cmd = cmd..' '..parms..err..utils.quote_arg(cmd_tmpfile) local ret = utils.execute(cmd) if not ret then local err = (utils.readfile(cmd_tmpfile):gsub('\n(.*)','')) remove(cmd_tmpfile) return false,err else remove(cmd_tmpfile) return true end end local function find_ffi_copyfile () if not ffi_checked then ffi_checked = true local res res,alien = pcall(require,'alien') if not res then alien = nil res, ffi = pcall(require,'ffi') end if not res then ffi = nil return end else return end if alien then -- register the Win32 CopyFile and MoveFile functions local kernel = alien.load('kernel32.dll') CopyFile = kernel.CopyFileA CopyFile:types{'string','string','int',ret='int',abi='stdcall'} MoveFile = kernel.MoveFileA MoveFile:types{'string','string',ret='int',abi='stdcall'} GetLastError = kernel.GetLastError GetLastError:types{ret ='int', abi='stdcall'} elseif ffi then ffi.cdef [[ int CopyFileA(const char *src, const char *dest, int iovr); int MoveFileA(const char *src, const char *dest); int GetLastError(); ]] CopyFile = ffi.C.CopyFileA MoveFile = ffi.C.MoveFileA GetLastError = ffi.C.GetLastError end win32_errors = { ERROR_FILE_NOT_FOUND = 2, ERROR_PATH_NOT_FOUND = 3, ERROR_ACCESS_DENIED = 5, ERROR_WRITE_PROTECT = 19, ERROR_BAD_UNIT = 20, ERROR_NOT_READY = 21, ERROR_WRITE_FAULT = 29, ERROR_READ_FAULT = 30, ERROR_SHARING_VIOLATION = 32, ERROR_LOCK_VIOLATION = 33, ERROR_HANDLE_DISK_FULL = 39, ERROR_BAD_NETPATH = 53, ERROR_NETWORK_BUSY = 54, ERROR_DEV_NOT_EXIST = 55, ERROR_FILE_EXISTS = 80, ERROR_OPEN_FAILED = 110, ERROR_INVALID_NAME = 123, ERROR_BAD_PATHNAME = 161, ERROR_ALREADY_EXISTS = 183, } end local function two_arguments (f1,f2) return utils.quote_arg(f1)..' '..utils.quote_arg(f2) end local function file_op (is_copy,src,dest,flag) if flag == 1 and path.exists(dest) then return false,"cannot overwrite destination" end if is_windows then -- if we haven't tried to load Alien/LuaJIT FFI before, then do so find_ffi_copyfile() -- fallback if there's no Alien, just use DOS commands *shudder* -- 'rename' involves a copy and then deleting the source. if not CopyFile then if path.is_windows then src = src:gsub("/","\\") dest = dest:gsub("/","\\") end local res, err = execute_command('copy',two_arguments(src,dest)) if not res then return false,err end if not is_copy then return execute_command('del',utils.quote_arg(src)) end return true else if path.isdir(dest) then dest = path.join(dest,path.basename(src)) end local ret if is_copy then ret = CopyFile(src,dest,flag) else ret = MoveFile(src,dest) end if ret == 0 then local err = GetLastError() for name,value in pairs(win32_errors) do if value == err then return false,name end end return false,"Error #"..err else return true end end else -- for Unix, just use cp for now return execute_command(is_copy and 'cp' or 'mv', two_arguments(src,dest)) end end --- copy a file. -- @string src source file -- @string dest destination file or directory -- @bool flag true if you want to force the copy (default) -- @treturn bool operation succeeded -- @raise src and dest must be strings function dir.copyfile (src,dest,flag) assert_string(1,src) assert_string(2,dest) flag = flag==nil or flag return file_op(true,src,dest,flag and 0 or 1) end --- move a file. -- @string src source file -- @string dest destination file or directory -- @treturn bool operation succeeded -- @raise src and dest must be strings function dir.movefile (src,dest) assert_string(1,src) assert_string(2,dest) return file_op(false,src,dest,0) end local function _dirfiles(dirname,attrib) local dirs = {} local files = {} for f in ldir(dirname) do if f ~= '.' and f ~= '..' then local p = path.join(dirname,f) local mode = attrib(p,'mode') if mode=='directory' then append(dirs,f) else append(files,f) end end end return makelist(dirs), makelist(files) end --- return an iterator which walks through a directory tree starting at root. -- The iterator returns (root,dirs,files) -- Note that dirs and files are lists of names (i.e. you must say path.join(root,d) -- to get the actual full path) -- If bottom_up is false (or not present), then the entries at the current level are returned -- before we go deeper. This means that you can modify the returned list of directories before -- continuing. -- This is a clone of os.walk from the Python libraries. -- @string root A starting directory -- @bool bottom_up False if we start listing entries immediately. -- @bool follow_links follow symbolic links -- @return an iterator returning root,dirs,files -- @raise root must be a directory function dir.walk(root,bottom_up,follow_links) assert_dir(1,root) local attrib if path.is_windows or not follow_links then attrib = path.attrib else attrib = path.link_attrib end local to_scan = { root } local to_return = {} local iter = function() while #to_scan > 0 do local current_root = table.remove(to_scan) local dirs,files = _dirfiles(current_root, attrib) for _, d in ipairs(dirs) do table.insert(to_scan, current_root..path.sep..d) end if not bottom_up then return current_root, dirs, files else table.insert(to_return, { current_root, dirs, files }) end end if #to_return > 0 then return utils.unpack(table.remove(to_return)) end end return iter end --- remove a whole directory tree. -- Symlinks in the tree will be deleted without following them. -- @string fullpath A directory path (must be an actual directory, not a symlink) -- @return true or nil -- @return error if failed -- @raise fullpath must be a string function dir.rmtree(fullpath) assert_dir(1,fullpath) if path.islink(fullpath) then return false,'will not follow symlink' end for root,dirs,files in dir.walk(fullpath,true) do if path.islink(root) then -- sub dir is a link, remove link, do not follow if is_windows then -- Windows requires using "rmdir". Deleting the link like a file -- will instead delete all files from the target directory!! local res, err = rmdir(root) if not res then return nil,err .. ": " .. root end else local res, err = remove(root) if not res then return nil,err .. ": " .. root end end else for i,f in ipairs(files) do local res, err = remove(path.join(root,f)) if not res then return nil,err .. ": " .. path.join(root,f) end end local res, err = rmdir(root) if not res then return nil,err .. ": " .. root end end end return true end do local dirpat if path.is_windows then dirpat = '(.+)\\[^\\]+$' else dirpat = '(.+)/[^/]+$' end local _makepath function _makepath(p) -- windows root drive case if p:find '^%a:[\\]*$' then return true end if not path.isdir(p) then local subp = p:match(dirpat) if subp then local ok, err = _makepath(subp) if not ok then return nil, err end end return mkdir(p) else return true end end --- create a directory path. -- This will create subdirectories as necessary! -- @string p A directory path -- @return true on success, nil + errormsg on failure -- @raise failure to create function dir.makepath (p) assert_string(1,p) if path.is_windows then p = p:gsub("/", "\\") end return _makepath(path.abspath(p)) end end --- clone a directory tree. Will always try to create a new directory structure -- if necessary. -- @string path1 the base path of the source tree -- @string path2 the new base path for the destination -- @func file_fun an optional function to apply on all files -- @bool verbose an optional boolean to control the verbosity of the output. -- It can also be a logging function that behaves like print() -- @return true, or nil -- @return error message, or list of failed directory creations -- @return list of failed file operations -- @raise path1 and path2 must be strings -- @usage clonetree('.','../backup',copyfile) function dir.clonetree (path1,path2,file_fun,verbose) assert_string(1,path1) assert_string(2,path2) if verbose == true then verbose = print end local abspath,normcase,isdir,join = path.abspath,path.normcase,path.isdir,path.join local faildirs,failfiles = {},{} if not isdir(path1) then return raise 'source is not a valid directory' end path1 = abspath(normcase(path1)) path2 = abspath(normcase(path2)) if verbose then verbose('normalized:',path1,path2) end -- particularly NB that the new path isn't fully contained in the old path if path1 == path2 then return raise "paths are the same" end local _,i2 = path2:find(path1,1,true) if i2 == #path1 and path2:sub(i2+1,i2+1) == path.sep then return raise 'destination is a subdirectory of the source' end local cp = path.common_prefix (path1,path2) local idx = #cp if idx == 0 then -- no common path, but watch out for Windows paths! if path1:sub(2,2) == ':' then idx = 3 end end for root,dirs,files in dir.walk(path1) do local opath = path2..root:sub(idx) if verbose then verbose('paths:',opath,root) end if not isdir(opath) then local ret = dir.makepath(opath) if not ret then append(faildirs,opath) end if verbose then verbose('creating:',opath,ret) end end if file_fun then for i,f in ipairs(files) do local p1 = join(root,f) local p2 = join(opath,f) local ret = file_fun(p1,p2) if not ret then append(failfiles,p2) end if verbose then verbose('files:',p1,p2,ret) end end end end return true,faildirs,failfiles end -- each entry of the stack is an array with three items: -- 1. the name of the directory -- 2. the lfs iterator function -- 3. the lfs iterator userdata local function treeiter(iterstack) local diriter = iterstack[#iterstack] if not diriter then return -- done end local dirname = diriter[1] local entry = diriter[2](diriter[3]) if not entry then table.remove(iterstack) return treeiter(iterstack) -- tail-call to try next end if entry ~= "." and entry ~= ".." then entry = dirname .. sep .. entry if exists(entry) then -- Just in case a symlink is broken. local is_dir = isdir(entry) if is_dir then table.insert(iterstack, { entry, ldir(entry) }) end return entry, is_dir end end return treeiter(iterstack) -- tail-call to try next end --- return an iterator over all entries in a directory tree -- @string d a directory -- @return an iterator giving pathname and mode (true for dir, false otherwise) -- @raise d must be a non-empty string function dir.dirtree( d ) assert( d and d ~= "", "directory parameter is missing or empty" ) local last = sub ( d, -1 ) if last == sep or last == '/' then d = sub( d, 1, -2 ) end local iterstack = { {d, ldir(d)} } return treeiter, iterstack end --- Recursively returns all the file starting at 'path'. It can optionally take a shell pattern and -- only returns files that match 'shell_pattern'. If a pattern is given it will do a case insensitive search. -- @string[opt='.'] start_path A directory. -- @string[opt='*'] shell_pattern A shell pattern (see `fnmatch`). -- @treturn List(string) containing all the files found recursively starting at 'path' and filtered by 'shell_pattern'. -- @raise start_path must be a directory function dir.getallfiles( start_path, shell_pattern ) start_path = start_path or '.' assert_dir(1,start_path) shell_pattern = shell_pattern or "*" local files = {} local normcase = path.normcase for filename, mode in dir.dirtree( start_path ) do if not mode then local mask = filemask( shell_pattern ) if normcase(filename):find( mask ) then files[#files + 1] = filename end end end return makelist(files) end return dir Penlight-1.12.0/lua/pl/file.lua000066400000000000000000000027371416703176500162060ustar00rootroot00000000000000--- File manipulation functions: reading, writing, moving and copying. -- -- This module wraps a number of functions from other modules into a -- file related module for convenience. -- -- Dependencies: `pl.utils`, `pl.dir`, `pl.path` -- @module pl.file local os = os local utils = require 'pl.utils' local dir = require 'pl.dir' local path = require 'pl.path' local file = {} --- return the contents of a file as a string. -- This function is a copy of `utils.readfile`. -- @function file.read file.read = utils.readfile --- write a string to a file. -- This function is a copy of `utils.writefile`. -- @function file.write file.write = utils.writefile --- copy a file. -- This function is a copy of `dir.copyfile`. -- @function file.copy file.copy = dir.copyfile --- move a file. -- This function is a copy of `dir.movefile`. -- @function file.move file.move = dir.movefile --- Return the time of last access as the number of seconds since the epoch. -- This function is a copy of `path.getatime`. -- @function file.access_time file.access_time = path.getatime ---Return when the file was created. -- This function is a copy of `path.getctime`. -- @function file.creation_time file.creation_time = path.getctime --- Return the time of last modification. -- This function is a copy of `path.getmtime`. -- @function file.modified_time file.modified_time = path.getmtime --- Delete a file. -- This function is a copy of `os.remove`. -- @function file.delete file.delete = os.remove return file Penlight-1.12.0/lua/pl/func.lua000066400000000000000000000255531416703176500162230ustar00rootroot00000000000000--- Functional helpers like composition, binding and placeholder expressions. -- Placeholder expressions are useful for short anonymous functions, and were -- inspired by the Boost Lambda library. -- -- > utils.import 'pl.func' -- > ls = List{10,20,30} -- > = ls:map(_1+1) -- {11,21,31} -- -- They can also be used to _bind_ particular arguments of a function. -- -- > p = bind(print,'start>',_0) -- > p(10,20,30) -- > start> 10 20 30 -- -- See @{07-functional.md.Creating_Functions_from_Functions|the Guide} -- -- Dependencies: `pl.utils`, `pl.tablex` -- @module pl.func local type,setmetatable,getmetatable,rawset = type,setmetatable,getmetatable,rawset local concat,append = table.concat,table.insert local tostring = tostring local utils = require 'pl.utils' local pairs,rawget,unpack,pack = pairs,rawget,utils.unpack,utils.pack local tablex = require 'pl.tablex' local map = tablex.map local _DEBUG = rawget(_G,'_DEBUG') local assert_arg = utils.assert_arg local func = {} -- metatable for Placeholder Expressions (PE) local _PEMT = {} local function P (t) setmetatable(t,_PEMT) return t end func.PE = P local function isPE (obj) return getmetatable(obj) == _PEMT end func.isPE = isPE -- construct a placeholder variable (e.g _1 and _2) local function PH (idx) return P {op='X',repr='_'..idx, index=idx} end -- construct a constant placeholder variable (e.g _C1 and _C2) local function CPH (idx) return P {op='X',repr='_C'..idx, index=idx} end func._1,func._2,func._3,func._4,func._5 = PH(1),PH(2),PH(3),PH(4),PH(5) func._0 = P{op='X',repr='...',index=0} function func.Var (name) local ls = utils.split(name,'[%s,]+') local res = {} for i = 1, #ls do append(res,P{op='X',repr=ls[i],index=0}) end return unpack(res) end function func._ (value) return P{op='X',repr=value,index='wrap'} end local repr func.Nil = func.Var 'nil' function _PEMT.__index(obj,key) return P{op='[]',obj,key} end function _PEMT.__call(fun,...) return P{op='()',fun,...} end function _PEMT.__tostring (e) return repr(e) end function _PEMT.__unm(arg) return P{op='unm',arg} end function func.Not (arg) return P{op='not',arg} end function func.Len (arg) return P{op='#',arg} end local function binreg(context,t) for name,op in pairs(t) do rawset(context,name,function(x,y) return P{op=op,x,y} end) end end local function import_name (name,fun,context) rawset(context,name,function(...) return P{op='()',fun,...} end) end local imported_functions = {} local function is_global_table (n) return type(_G[n]) == 'table' end --- wrap a table of functions. This makes them available for use in -- placeholder expressions. -- @string tname a table name -- @tab context context to put results, defaults to environment of caller function func.import(tname,context) assert_arg(1,tname,'string',is_global_table,'arg# 1: not a name of a global table') local t = _G[tname] context = context or _G for name,fun in pairs(t) do import_name(name,fun,context) imported_functions[fun] = name end end --- register a function for use in placeholder expressions. -- @func fun a function -- @string[opt] name an optional name -- @return a placeholder functiond function func.register (fun,name) assert_arg(1,fun,'function') if name then assert_arg(2,name,'string') imported_functions[fun] = name end return function(...) return P{op='()',fun,...} end end function func.lookup_imported_name (fun) return imported_functions[fun] end local function _arg(...) return ... end function func.Args (...) return P{op='()',_arg,...} end -- binary operators with their precedences (see Lua manual) -- precedences might be incremented by one before use depending on -- left- or right-associativity, space them out local binary_operators = { ['or'] = 0, ['and'] = 2, ['=='] = 4, ['~='] = 4, ['<'] = 4, ['>'] = 4, ['<='] = 4, ['>='] = 4, ['..'] = 6, ['+'] = 8, ['-'] = 8, ['*'] = 10, ['/'] = 10, ['%'] = 10, ['^'] = 14 } -- unary operators with their precedences local unary_operators = { ['not'] = 12, ['#'] = 12, ['unm'] = 12 } -- comparisons (as prefix functions) binreg (func,{And='and',Or='or',Eq='==',Lt='<',Gt='>',Le='<=',Ge='>='}) -- standard binary operators (as metamethods) binreg (_PEMT,{__add='+',__sub='-',__mul='*',__div='/',__mod='%',__pow='^',__concat='..'}) binreg (_PEMT,{__eq='=='}) --- all elements of a table except the first. -- @tab ls a list-like table. function func.tail (ls) assert_arg(1,ls,'table') local res = {} for i = 2,#ls do append(res,ls[i]) end return res end --- create a string representation of a placeholder expression. -- @param e a placeholder expression -- @param lastpred not used function repr (e,lastpred) local tail = func.tail if isPE(e) then local pred = binary_operators[e.op] or unary_operators[e.op] if pred then -- binary or unary operator local s if binary_operators[e.op] then local left_pred = pred local right_pred = pred if e.op == '..' or e.op == '^' then left_pred = left_pred + 1 else right_pred = right_pred + 1 end local left_arg = repr(e[1], left_pred) local right_arg = repr(e[2], right_pred) s = left_arg..' '..e.op..' '..right_arg else local op = e.op == 'unm' and '-' or e.op s = op..' '..repr(e[1], pred) end if lastpred and lastpred > pred then s = '('..s..')' end return s else -- either postfix, or a placeholder local ls = map(repr,e) if e.op == '[]' then return ls[1]..'['..ls[2]..']' elseif e.op == '()' then local fn if ls[1] ~= nil then -- was _args, undeclared! fn = ls[1] else fn = '' end return fn..'('..concat(tail(ls),',')..')' else return e.repr end end elseif type(e) == 'string' then return '"'..e..'"' elseif type(e) == 'function' then local name = func.lookup_imported_name(e) if name then return name else return tostring(e) end else return tostring(e) --should not really get here! end end func.repr = repr -- collect all the non-PE values in this PE into vlist, and replace each occurence -- with a constant PH (_C1, etc). Return the maximum placeholder index found. local collect_values function collect_values (e,vlist) if isPE(e) then if e.op ~= 'X' then local m = 0 for i = 1,#e do local subx = e[i] local pe = isPE(subx) if pe then if subx.op == 'X' and subx.index == 'wrap' then subx = subx.repr pe = false else m = math.max(m,collect_values(subx,vlist)) end end if not pe then append(vlist,subx) e[i] = CPH(#vlist) end end return m else -- was a placeholder, it has an index... return e.index end else -- plain value has no placeholder dependence return 0 end end func.collect_values = collect_values --- instantiate a PE into an actual function. First we find the largest placeholder used, -- e.g. _2; from this a list of the formal parameters can be build. Then we collect and replace -- any non-PE values from the PE, and build up a constant binding list. -- Finally, the expression can be compiled, and e.__PE_function is set. -- @param e a placeholder expression -- @return a function function func.instantiate (e) local consts,values,parms = {},{},{} local rep, err, fun local n = func.collect_values(e,values) for i = 1,#values do append(consts,'_C'..i) if _DEBUG then print(i,values[i]) end end for i =1,n do append(parms,'_'..i) end consts = concat(consts,',') parms = concat(parms,',') rep = repr(e) local fstr = ('return function(%s) return function(%s) return %s end end'):format(consts,parms,rep) if _DEBUG then print(fstr) end fun,err = utils.load(fstr,'fun') if not fun then return nil,err end fun = fun() -- get wrapper fun = fun(unpack(values)) -- call wrapper (values could be empty) e.__PE_function = fun return fun end --- instantiate a PE unless it has already been done. -- @param e a placeholder expression -- @return the function function func.I(e) if rawget(e,'__PE_function') then return e.__PE_function else return func.instantiate(e) end end utils.add_function_factory(_PEMT,func.I) --- bind the first parameter of the function to a value. -- @function func.bind1 -- @func fn a function of one or more arguments -- @param p a value -- @return a function of one less argument -- @usage (bind1(math.max,10))(20) == math.max(10,20) func.bind1 = utils.bind1 func.curry = func.bind1 --- create a function which chains two functions. -- @func f a function of at least one argument -- @func g a function of at least one argument -- @return a function -- @usage printf = compose(io.write,string.format) function func.compose (f,g) return function(...) return f(g(...)) end end --- bind the arguments of a function to given values. -- `bind(fn,v,_2)` is equivalent to `bind1(fn,v)`. -- @func fn a function of at least one argument -- @param ... values or placeholder variables -- @return a function -- @usage (bind(f,_1,a))(b) == f(a,b) -- @usage (bind(f,_2,_1))(a,b) == f(b,a) function func.bind(fn,...) local args = pack(...) local holders,parms,bvalues,values = {},{},{'fn'},{} local nv,maxplace,varargs = 1,0,false for i = 1,args.n do local a = args[i] if isPE(a) and a.op == 'X' then append(holders,a.repr) maxplace = math.max(maxplace,a.index) if a.index == 0 then varargs = true end else local v = '_v'..nv append(bvalues,v) append(holders,v) append(values,a) nv = nv + 1 end end for np = 1,maxplace do append(parms,'_'..np) end if varargs then append(parms,'...') end bvalues = concat(bvalues,',') parms = concat(parms,',') holders = concat(holders,',') local fstr = ([[ return function (%s) return function(%s) return fn(%s) end end ]]):format(bvalues,parms,holders) if _DEBUG then print(fstr) end local res = utils.load(fstr) res = res() return res(fn,unpack(values)) end return func Penlight-1.12.0/lua/pl/import_into.lua000066400000000000000000000056231416703176500176270ustar00rootroot00000000000000-------------- -- PL loader, for loading all PL libraries, only on demand. -- Whenever a module is implicitly accesssed, the table will have the module automatically injected. -- (e.g. `_ENV.tablex`) -- then that module is dynamically loaded. The submodules are all brought into -- the table that is provided as the argument, or returned in a new table. -- If a table is provided, that table's metatable is clobbered, but the values are not. -- This module returns a single function, which is passed the environment. -- If this is `true`, then return a 'shadow table' as the module -- See @{01-introduction.md.To_Inject_or_not_to_Inject_|the Guide} -- @module pl.import_into return function(env) local mod if env == true then mod = {} env = {} end local env = env or {} local modules = { utils = true,path=true,dir=true,tablex=true,stringio=true,sip=true, input=true,seq=true,lexer=true,stringx=true, config=true,pretty=true,data=true,func=true,text=true, operator=true,lapp=true,array2d=true, comprehension=true,xml=true,types=true, test = true, app = true, file = true, class = true, luabalanced = true, permute = true, template = true, url = true, compat = true, -- classes -- List = true, Map = true, Set = true, OrderedMap = true, MultiMap = true, Date = true, } rawset(env,'utils',require 'pl.utils') for name,klass in pairs(env.utils.stdmt) do klass.__index = function(t,key) return require ('pl.'..name)[key] end; end -- ensure that we play nice with libraries that also attach a metatable -- to the global table; always forward to a custom __index if we don't -- match local _hook,_prev_index local gmt = {} local prevenvmt = getmetatable(env) if prevenvmt then _prev_index = prevenvmt.__index if prevenvmt.__newindex then gmt.__newindex = prevenvmt.__newindex end end function gmt.hook(handler) _hook = handler end function gmt.__index(t,name) local found = modules[name] -- either true, or the name of the module containing this class. -- either way, we load the required module and make it globally available. if found then -- e..g pretty.dump causes pl.pretty to become available as 'pretty' rawset(env,name,require('pl.'..name)) return env[name] else local res if _hook then res = _hook(t,name) if res then return res end end if _prev_index then return _prev_index(t,name) end end end if mod then function gmt.__newindex(t,name,value) mod[name] = value rawset(t,name,value) end end setmetatable(env,gmt) return env,mod or env end Penlight-1.12.0/lua/pl/init.lua000066400000000000000000000006601416703176500162230ustar00rootroot00000000000000-------------- -- Entry point for loading all PL libraries only on demand, into the global space. -- Requiring 'pl' means that whenever a module is implicitly accesssed -- (e.g. `utils.split`) -- then that module is dynamically loaded. The submodules are all brought into -- the global space. --Updated to use @{pl.import_into} -- @module pl require'pl.import_into'(_G) if rawget(_G,'PENLIGHT_STRICT') then require 'pl.strict' end Penlight-1.12.0/lua/pl/input.lua000066400000000000000000000120441416703176500164160ustar00rootroot00000000000000--- Iterators for extracting words or numbers from an input source. -- -- require 'pl' -- local total,n = seq.sum(input.numbers()) -- print('average',total/n) -- -- _source_ is defined as a string or a file-like object (i.e. has a read() method which returns the next line) -- -- See @{06-data.md.Reading_Unstructured_Text_Data|here} -- -- Dependencies: `pl.utils` -- @module pl.input local strfind = string.find local strsub = string.sub local strmatch = string.match local utils = require 'pl.utils' local unpack = utils.unpack local pairs,type,tonumber = pairs,type,tonumber local patterns = utils.patterns local io = io local input = {} --- create an iterator over all tokens. -- based on allwords from PiL, 7.1 -- @func getter any function that returns a line of text -- @string pattern -- @string[opt] fn Optionally can pass a function to process each token as it's found. -- @return an iterator function input.alltokens (getter,pattern,fn) local line = getter() -- current line local pos = 1 -- current position in the line return function () -- iterator function while line do -- repeat while there are lines local s, e = strfind(line, pattern, pos) if s then -- found a word? pos = e + 1 -- next position is after this token local res = strsub(line, s, e) -- return the token if fn then res = fn(res) end return res else line = getter() -- token not found; try next line pos = 1 -- restart from first position end end return nil -- no more lines: end of traversal end end local alltokens = input.alltokens -- question: shd this _split_ a string containing line feeds? --- create a function which grabs the next value from a source. If the source is a string, then the getter -- will return the string and thereafter return nil. If not specified then the source is assumed to be stdin. -- @param f a string or a file-like object (i.e. has a read() method which returns the next line) -- @return a getter function function input.create_getter(f) if f then if type(f) == 'string' then local ls = utils.split(f,'\n') local i,n = 0,#ls return function() i = i + 1 if i > n then return nil end return ls[i] end else -- anything that supports the read() method! if not f.read then error('not a file-like object') end return function() return f:read() end end else return io.read -- i.e. just read from stdin end end --- generate a sequence of numbers from a source. -- @param f A source -- @return An iterator function input.numbers(f) return alltokens(input.create_getter(f), '('..patterns.FLOAT..')',tonumber) end --- generate a sequence of words from a source. -- @param f A source -- @return An iterator function input.words(f) return alltokens(input.create_getter(f),"%w+") end local function apply_tonumber (no_fail,...) local args = {...} for i = 1,#args do local n = tonumber(args[i]) if n == nil then if not no_fail then return nil,args[i] end else args[i] = n end end return args end --- parse an input source into fields. -- By default, will fail if it cannot convert a field to a number. -- @param ids a list of field indices, or a maximum field index -- @string delim delimiter to parse fields (default space) -- @param f a source @see create_getter -- @tab opts option table, `{no_fail=true}` -- @return an iterator with the field values -- @usage for x,y in fields {2,3} do print(x,y) end -- 2nd and 3rd fields from stdin function input.fields (ids,delim,f,opts) local sep local s local getter = input.create_getter(f) local no_fail = opts and opts.no_fail local no_convert = opts and opts.no_convert if not delim or delim == ' ' then delim = '%s' sep = '%s+' s = '%s*' else sep = delim s = '' end local max_id = 0 if type(ids) == 'table' then for i,id in pairs(ids) do if id > max_id then max_id = id end end else max_id = ids ids = {} for i = 1,max_id do ids[#ids+1] = i end end local pat = '[^'..delim..']*' local k = 1 for i = 1,max_id do if ids[k] == i then k = k + 1 s = s..'('..pat..')' else s = s..pat end if i < max_id then s = s..sep end end local linecount = 1 return function() local line,results,err repeat line = getter() linecount = linecount + 1 if not line then return nil end if no_convert then results = {strmatch(line,s)} else results,err = apply_tonumber(no_fail,strmatch(line,s)) if not results then utils.quit("line "..(linecount-1)..": cannot convert '"..err.."' to number") end end until #results > 0 return unpack(results) end end return input Penlight-1.12.0/lua/pl/lapp.lua000066400000000000000000000347631416703176500162270ustar00rootroot00000000000000--- Simple command-line parsing using human-readable specification. -- Supports GNU-style parameters. -- -- lapp = require 'pl.lapp' -- local args = lapp [[ -- Does some calculations -- -o,--offset (default 0.0) Offset to add to scaled number -- -s,--scale (number) Scaling factor -- (number) Number to be scaled -- ]] -- -- print(args.offset + args.scale * args.number) -- -- Lines beginning with `'-'` are flags; there may be a short and a long name; -- lines beginning with `''` are arguments. Anything in parens after -- the flag/argument is either a default, a type name or a range constraint. -- -- See @{08-additional.md.Command_line_Programs_with_Lapp|the Guide} -- -- Dependencies: `pl.sip` -- @module pl.lapp local status,sip = pcall(require,'pl.sip') if not status then sip = require 'sip' end local match = sip.match_at_start local append,tinsert = table.insert,table.insert sip.custom_pattern('X','(%a[%w_%-]*)') local function lines(s) return s:gmatch('([^\n]*)\n') end local function lstrip(str) return str:gsub('^%s+','') end local function strip(str) return lstrip(str):gsub('%s+$','') end local function at(s,k) return s:sub(k,k) end local lapp = {} local open_files,parms,aliases,parmlist,usage,script lapp.callback = false -- keep Strict happy local filetypes = { stdin = {io.stdin,'file-in'}, stdout = {io.stdout,'file-out'}, stderr = {io.stderr,'file-out'} } --- controls whether to dump usage on error. -- Defaults to true lapp.show_usage_error = true --- quit this script immediately. -- @string msg optional message -- @bool no_usage suppress 'usage' display function lapp.quit(msg,no_usage) if no_usage == 'throw' then error(msg) end if msg then io.stderr:write(msg..'\n\n') end if not no_usage then io.stderr:write(usage) end os.exit(1) end --- print an error to stderr and quit. -- @string msg a message -- @bool no_usage suppress 'usage' display function lapp.error(msg,no_usage) if not lapp.show_usage_error then no_usage = true elseif lapp.show_usage_error == 'throw' then no_usage = 'throw' end lapp.quit(script..': '..msg,no_usage) end --- open a file. -- This will quit on error, and keep a list of file objects for later cleanup. -- @string file filename -- @string[opt] opt same as second parameter of `io.open` function lapp.open (file,opt) local val,err = io.open(file,opt) if not val then lapp.error(err,true) end append(open_files,val) return val end --- quit if the condition is false. -- @bool condn a condition -- @string msg message text function lapp.assert(condn,msg) if not condn then lapp.error(msg) end end local function range_check(x,min,max,parm) lapp.assert(min <= x and max >= x,parm..' out of range') end local function xtonumber(s) local val = tonumber(s) if not val then lapp.error("unable to convert to number: "..s) end return val end local types = {} local builtin_types = {string=true,number=true,['file-in']='file',['file-out']='file',boolean=true} local function convert_parameter(ps,val) if ps.converter then val = ps.converter(val) end if ps.type == 'number' then val = xtonumber(val) elseif builtin_types[ps.type] == 'file' then val = lapp.open(val,(ps.type == 'file-in' and 'r') or 'w' ) elseif ps.type == 'boolean' then return val end if ps.constraint then ps.constraint(val) end return val end --- add a new type to Lapp. These appear in parens after the value like -- a range constraint, e.g. ' (integer) Process PID' -- @string name name of type -- @param converter either a function to convert values, or a Lua type name. -- @func[opt] constraint optional function to verify values, should use lapp.error -- if failed. function lapp.add_type (name,converter,constraint) types[name] = {converter=converter,constraint=constraint} end local function force_short(short) lapp.assert(#short==1,short..": short parameters should be one character") end -- deducing type of variable from default value; local function process_default (sval,vtype) local val, success if not vtype or vtype == 'number' then val = tonumber(sval) end if val then -- we have a number! return val,'number' elseif filetypes[sval] then local ft = filetypes[sval] return ft[1],ft[2] else if sval == 'true' and not vtype then return true, 'boolean' end if sval:match '^["\']' then sval = sval:sub(2,-2) end local ps = types[vtype] or {} ps.type = vtype local show_usage_error = lapp.show_usage_error lapp.show_usage_error = "throw" success, val = pcall(convert_parameter, ps, sval) lapp.show_usage_error = show_usage_error if success then return val, vtype or 'string' end return sval,vtype or 'string' end end --- process a Lapp options string. -- Usually called as `lapp()`. -- @string str the options text -- @tparam {string} args a table of arguments (default is `_G.arg`) -- @return a table with parameter-value pairs function lapp.process_options_string(str,args) local results = {} local varargs local arg = args or _G.arg open_files = {} parms = {} aliases = {} parmlist = {} local function check_varargs(s) local res,cnt = s:gsub('^%.%.%.%s*','') return res, (cnt > 0) end local function set_result(ps,parm,val) parm = type(parm) == "string" and parm:gsub("%W", "_") or parm -- so foo-bar becomes foo_bar in Lua if not ps.varargs then results[parm] = val else if not results[parm] then results[parm] = { val } else append(results[parm],val) end end end usage = str for _,a in ipairs(arg) do if a == "-h" or a == "--help" then return lapp.quit() end end for line in lines(str) do local res = {} local optparm,defval,vtype,constraint,rest line = lstrip(line) local function check(str) return match(str,line,res) end -- flags: either '-', '-,--' or '--' if check '-$v{short}, --$o{long} $' or check '-$v{short} $' or check '--$o{long} $' then if res.long then optparm = res.long:gsub('[^%w%-]','_') -- I'm not sure the $o pattern will let anything else through? if #res.rest == 1 then optparm = optparm .. res.rest end if res.short then aliases[res.short] = optparm end else optparm = res.short end if res.short and not lapp.slack then force_short(res.short) end res.rest, varargs = check_varargs(res.rest) elseif check '$<{name} $' then -- is it ? -- so becomes input_file ... optparm,rest = res.name:match '([^%.]+)(.*)' optparm = optparm:gsub('%A','_') varargs = rest == '...' append(parmlist,optparm) end -- this is not a pure doc line and specifies the flag/parameter type if res.rest then line = res.rest res = {} local optional -- do we have ([optional] [] [default ])? if match('$({def} $',line,res) or match('$({def}',line,res) then local typespec = strip(res.def) local ftype, rest = typespec:match('^(%S+)(.*)$') rest = strip(rest) if ftype == 'optional' then ftype, rest = rest:match('^(%S+)(.*)$') rest = strip(rest) optional = true end local default if ftype == 'default' then default = true if rest == '' then lapp.error("value must follow default") end else -- a type specification if match('$f{min}..$f{max}',ftype,res) then -- a numerical range like 1..10 local min,max = res.min,res.max vtype = 'number' constraint = function(x) range_check(x,min,max,optparm) end elseif not ftype:match '|' then -- plain type vtype = ftype else -- 'enum' type is a string which must belong to -- one of several distinct values local enums = ftype local enump = '|' .. enums .. '|' vtype = 'string' constraint = function(s) lapp.assert(enump:find('|'..s..'|', 1, true), "value '"..s.."' not in "..enums ) end end end res.rest = rest typespec = res.rest -- optional 'default value' clause. Type is inferred as -- 'string' or 'number' if there's no explicit type if default or match('default $r{rest}',typespec,res) then defval,vtype = process_default(res.rest,vtype) end else -- must be a plain flag, no extra parameter required defval = false vtype = 'boolean' end local ps = { type = vtype, defval = defval, required = defval == nil and not optional, comment = res.rest or optparm, constraint = constraint, varargs = varargs } varargs = nil if types[vtype] then local converter = types[vtype].converter if type(converter) == 'string' then ps.type = converter else ps.converter = converter end ps.constraint = types[vtype].constraint elseif not builtin_types[vtype] and vtype then lapp.error(vtype.." is unknown type") end parms[optparm] = ps end end -- cool, we have our parms, let's parse the command line args local iparm = 1 local iextra = 1 local i = 1 local parm,ps,val local end_of_flags = false local function check_parm (parm) local eqi = parm:find '[=:]' if eqi then tinsert(arg,i+1,parm:sub(eqi+1)) parm = parm:sub(1,eqi-1) end return parm,eqi end local function is_flag (parm) return parms[aliases[parm] or parm] end while i <= #arg do local theArg = arg[i] local res = {} -- after '--' we don't parse args and they end up in -- the array part of the result (args[1] etc) if theArg == '--' then end_of_flags = true iparm = #parmlist + 1 i = i + 1 theArg = arg[i] if not theArg then break end end -- look for a flag, - or -- if not end_of_flags and (match('--$S{long}',theArg,res) or match('-$S{short}',theArg,res)) then if res.long then -- long option parm = check_parm(res.long) elseif #res.short == 1 or is_flag(res.short) then parm = res.short else local parmstr,eq = check_parm(res.short) if not eq then parm = at(parmstr,1) local flag = is_flag(parm) if flag and flag.type ~= 'boolean' then --if isdigit(at(parmstr,2)) then -- a short option followed by a digit is an exception (for AW;)) -- push ahead into the arg array tinsert(arg,i+1,parmstr:sub(2)) else -- push multiple flags into the arg array! for k = 2,#parmstr do tinsert(arg,i+k-1,'-'..at(parmstr,k)) end end else parm = parmstr end end if aliases[parm] then parm = aliases[parm] end if not parms[parm] and (parm == 'h' or parm == 'help') then lapp.quit() end else -- a parameter parm = parmlist[iparm] if not parm then -- extra unnamed parameters are indexed starting at 1 parm = iextra ps = { type = 'string' } parms[parm] = ps iextra = iextra + 1 else ps = parms[parm] end if not ps.varargs then iparm = iparm + 1 end val = theArg end ps = parms[parm] if not ps then lapp.error("unrecognized parameter: "..parm) end if ps.type ~= 'boolean' then -- we need a value! This should follow if not val then i = i + 1 val = arg[i] theArg = val end lapp.assert(val,parm.." was expecting a value") else -- toggle boolean flags (usually false -> true) val = not ps.defval end ps.used = true val = convert_parameter(ps,val) set_result(ps,parm,val) if builtin_types[ps.type] == 'file' then set_result(ps,parm..'_name',theArg) end if lapp.callback then lapp.callback(parm,theArg,res) end i = i + 1 val = nil end -- check unused parms, set defaults and check if any required parameters were missed for parm,ps in pairs(parms) do if not ps.used then if ps.required then lapp.error("missing required parameter: "..parm) end set_result(ps,parm,ps.defval) end end return results end if arg then script = arg[0] script = script or rawget(_G,"LAPP_SCRIPT") or "unknown" -- strip dir and extension to get current script name script = script:gsub('.+[\\/]',''):gsub('%.%a+$','') else script = "inter" end setmetatable(lapp, { __call = function(tbl,str,args) return lapp.process_options_string(str,args) end, }) return lapp Penlight-1.12.0/lua/pl/lexer.lua000066400000000000000000000360751416703176500164100ustar00rootroot00000000000000--- Lexical scanner for creating a sequence of tokens from text. -- `lexer.scan(s)` returns an iterator over all tokens found in the -- string `s`. This iterator returns two values, a token type string -- (such as 'string' for quoted string, 'iden' for identifier) and the value of the -- token. -- -- Versions specialized for Lua and C are available; these also handle block comments -- and classify keywords as 'keyword' tokens. For example: -- -- > s = 'for i=1,n do' -- > for t,v in lexer.lua(s) do print(t,v) end -- keyword for -- iden i -- = = -- number 1 -- , , -- iden n -- keyword do -- -- See the Guide for further @{06-data.md.Lexical_Scanning|discussion} -- @module pl.lexer local strfind = string.find local strsub = string.sub local append = table.insert local function assert_arg(idx,val,tp) if type(val) ~= tp then error("argument "..idx.." must be "..tp, 2) end end local lexer = {} local NUMBER1 = '^[%+%-]?%d+%.?%d*[eE][%+%-]?%d+' local NUMBER1a = '^[%+%-]?%d*%.%d+[eE][%+%-]?%d+' local NUMBER2 = '^[%+%-]?%d+%.?%d*' local NUMBER2a = '^[%+%-]?%d*%.%d+' local NUMBER3 = '^0x[%da-fA-F]+' local NUMBER4 = '^%d+%.?%d*[eE][%+%-]?%d+' local NUMBER4a = '^%d*%.%d+[eE][%+%-]?%d+' local NUMBER5 = '^%d+%.?%d*' local NUMBER5a = '^%d*%.%d+' local IDEN = '^[%a_][%w_]*' local WSPACE = '^%s+' local STRING1 = "^(['\"])%1" -- empty string local STRING2 = [[^(['"])(\*)%2%1]] local STRING3 = [[^(['"]).-[^\](\*)%2%1]] local CHAR1 = "^''" local CHAR2 = [[^'(\*)%1']] local CHAR3 = [[^'.-[^\](\*)%1']] local PREPRO = '^#.-[^\\]\n' local plain_matches,lua_matches,cpp_matches,lua_keyword,cpp_keyword local function tdump(tok) return tok,tok end local function ndump(tok,options) if options and options.number then tok = tonumber(tok) end return "number",tok end -- regular strings, single or double quotes; usually we want them -- without the quotes local function sdump(tok,options) if options and options.string then tok = tok:sub(2,-2) end return "string",tok end -- long Lua strings need extra work to get rid of the quotes local function sdump_l(tok,options,findres) if options and options.string then local quotelen = 3 if findres[3] then quotelen = quotelen + findres[3]:len() end tok = tok:sub(quotelen, -quotelen) if tok:sub(1, 1) == "\n" then tok = tok:sub(2) end end return "string",tok end local function chdump(tok,options) if options and options.string then tok = tok:sub(2,-2) end return "char",tok end local function cdump(tok) return "comment",tok end local function wsdump (tok) return "space",tok end local function pdump (tok) return "prepro",tok end local function plain_vdump(tok) return "iden",tok end local function lua_vdump(tok) if lua_keyword[tok] then return "keyword",tok else return "iden",tok end end local function cpp_vdump(tok) if cpp_keyword[tok] then return "keyword",tok else return "iden",tok end end --- create a plain token iterator from a string or file-like object. -- @tparam string|file s a string or a file-like object with `:read()` method returning lines. -- @tab matches an optional match table - array of token descriptions. -- A token is described by a `{pattern, action}` pair, where `pattern` should match -- token body and `action` is a function called when a token of described type is found. -- @tab[opt] filter a table of token types to exclude, by default `{space=true}` -- @tab[opt] options a table of options; by default, `{number=true,string=true}`, -- which means convert numbers and strip string quotes. function lexer.scan(s,matches,filter,options) local file = type(s) ~= 'string' and s filter = filter or {space=true} options = options or {number=true,string=true} if filter then if filter.space then filter[wsdump] = true end if filter.comments then filter[cdump] = true end end if not matches then if not plain_matches then plain_matches = { {WSPACE,wsdump}, {NUMBER3,ndump}, {IDEN,plain_vdump}, {NUMBER1,ndump}, {NUMBER1a,ndump}, {NUMBER2,ndump}, {NUMBER2a,ndump}, {STRING1,sdump}, {STRING2,sdump}, {STRING3,sdump}, {'^.',tdump} } end matches = plain_matches end local line_nr = 0 local next_line = file and file:read() local sz = file and 0 or #s local idx = 1 local tlist_i local tlist local first_hit = true local function iter(res) local tp = type(res) if tlist then -- returning the inserted token list local cur = tlist[tlist_i] if cur then tlist_i = tlist_i + 1 return cur[1], cur[2] else tlist = nil end end if tp == 'string' then -- search up to some special pattern local i1,i2 = strfind(s,res,idx) if i1 then local tok = strsub(s,i1,i2) idx = i2 + 1 return '', tok else idx = sz + 1 return '', '' end elseif tp == 'table' then -- insert a token list tlist_i = 1 tlist = res return '', '' elseif tp ~= 'nil' then -- return position return line_nr, idx else -- look for next token if first_hit then if not file then line_nr = 1 end first_hit = false end if idx > sz then if file then if not next_line then return -- past the end of file, done end s = next_line line_nr = line_nr + 1 next_line = file:read() if next_line then s = s .. '\n' end idx, sz = 1, #s else return -- past the end of input, done end end for _,m in ipairs(matches) do local pat = m[1] local fun = m[2] local findres = {strfind(s,pat,idx)} local i1, i2 = findres[1], findres[2] if i1 then local tok = strsub(s,i1,i2) idx = i2 + 1 local ret1, ret2 if not (filter and filter[fun]) then lexer.finished = idx > sz ret1, ret2 = fun(tok, options, findres) end if not file and tok:find("\n") then -- Update line number. local _, newlines = tok:gsub("\n", {}) line_nr = line_nr + newlines end if ret1 then return ret1, ret2 -- found a match else return iter() -- tail-call to try again end end end end end return iter end local function isstring (s) return type(s) == 'string' end --- insert tokens into a stream. -- @param tok a token stream -- @param a1 a string is the type, a table is a token list and -- a function is assumed to be a token-like iterator (returns type & value) -- @string a2 a string is the value function lexer.insert (tok,a1,a2) if not a1 then return end local ts if isstring(a1) and isstring(a2) then ts = {{a1,a2}} elseif type(a1) == 'function' then ts = {} for t,v in a1() do append(ts,{t,v}) end else ts = a1 end tok(ts) end --- get everything in a stream upto a newline. -- @param tok a token stream -- @return a string function lexer.getline (tok) local _,v = tok('.-\n') return v end --- get current line number. -- @param tok a token stream -- @return the line number. -- if the input source is a file-like object, -- also return the column. function lexer.lineno (tok) return tok(0) end --- get the rest of the stream. -- @param tok a token stream -- @return a string function lexer.getrest (tok) local _,v = tok('.+') return v end --- get the Lua keywords as a set-like table. -- So `res["and"]` etc would be `true`. -- @return a table function lexer.get_keywords () if not lua_keyword then lua_keyword = { ["and"] = true, ["break"] = true, ["do"] = true, ["else"] = true, ["elseif"] = true, ["end"] = true, ["false"] = true, ["for"] = true, ["function"] = true, ["if"] = true, ["in"] = true, ["local"] = true, ["nil"] = true, ["not"] = true, ["or"] = true, ["repeat"] = true, ["return"] = true, ["then"] = true, ["true"] = true, ["until"] = true, ["while"] = true } end return lua_keyword end --- create a Lua token iterator from a string or file-like object. -- Will return the token type and value. -- @string s the string -- @tab[opt] filter a table of token types to exclude, by default `{space=true,comments=true}` -- @tab[opt] options a table of options; by default, `{number=true,string=true}`, -- which means convert numbers and strip string quotes. function lexer.lua(s,filter,options) filter = filter or {space=true,comments=true} lexer.get_keywords() if not lua_matches then lua_matches = { {WSPACE,wsdump}, {NUMBER3,ndump}, {IDEN,lua_vdump}, {NUMBER4,ndump}, {NUMBER4a,ndump}, {NUMBER5,ndump}, {NUMBER5a,ndump}, {STRING1,sdump}, {STRING2,sdump}, {STRING3,sdump}, {'^%-%-%[(=*)%[.-%]%1%]',cdump}, {'^%-%-.-\n',cdump}, {'^%[(=*)%[.-%]%1%]',sdump_l}, {'^==',tdump}, {'^~=',tdump}, {'^<=',tdump}, {'^>=',tdump}, {'^%.%.%.',tdump}, {'^%.%.',tdump}, {'^.',tdump} } end return lexer.scan(s,lua_matches,filter,options) end --- create a C/C++ token iterator from a string or file-like object. -- Will return the token type type and value. -- @string s the string -- @tab[opt] filter a table of token types to exclude, by default `{space=true,comments=true}` -- @tab[opt] options a table of options; by default, `{number=true,string=true}`, -- which means convert numbers and strip string quotes. function lexer.cpp(s,filter,options) filter = filter or {space=true,comments=true} if not cpp_keyword then cpp_keyword = { ["class"] = true, ["break"] = true, ["do"] = true, ["sizeof"] = true, ["else"] = true, ["continue"] = true, ["struct"] = true, ["false"] = true, ["for"] = true, ["public"] = true, ["void"] = true, ["private"] = true, ["protected"] = true, ["goto"] = true, ["if"] = true, ["static"] = true, ["const"] = true, ["typedef"] = true, ["enum"] = true, ["char"] = true, ["int"] = true, ["bool"] = true, ["long"] = true, ["float"] = true, ["true"] = true, ["delete"] = true, ["double"] = true, ["while"] = true, ["new"] = true, ["namespace"] = true, ["try"] = true, ["catch"] = true, ["switch"] = true, ["case"] = true, ["extern"] = true, ["return"] = true,["default"] = true,['unsigned'] = true,['signed'] = true, ["union"] = true, ["volatile"] = true, ["register"] = true,["short"] = true, } end if not cpp_matches then cpp_matches = { {WSPACE,wsdump}, {PREPRO,pdump}, {NUMBER3,ndump}, {IDEN,cpp_vdump}, {NUMBER4,ndump}, {NUMBER4a,ndump}, {NUMBER5,ndump}, {NUMBER5a,ndump}, {CHAR1,chdump}, {CHAR2,chdump}, {CHAR3,chdump}, {STRING1,sdump}, {STRING2,sdump}, {STRING3,sdump}, {'^//.-\n',cdump}, {'^/%*.-%*/',cdump}, {'^==',tdump}, {'^!=',tdump}, {'^<=',tdump}, {'^>=',tdump}, {'^->',tdump}, {'^&&',tdump}, {'^||',tdump}, {'^%+%+',tdump}, {'^%-%-',tdump}, {'^%+=',tdump}, {'^%-=',tdump}, {'^%*=',tdump}, {'^/=',tdump}, {'^|=',tdump}, {'^%^=',tdump}, {'^::',tdump}, {'^.',tdump} } end return lexer.scan(s,cpp_matches,filter,options) end --- get a list of parameters separated by a delimiter from a stream. -- @param tok the token stream -- @string[opt=')'] endtoken end of list. Can be '\n' -- @string[opt=','] delim separator -- @return a list of token lists. function lexer.get_separated_list(tok,endtoken,delim) endtoken = endtoken or ')' delim = delim or ',' local parm_values = {} local level = 1 -- used to count ( and ) local tl = {} local function tappend (tl,t,val) val = val or t append(tl,{t,val}) end local is_end if endtoken == '\n' then is_end = function(t,val) return t == 'space' and val:find '\n' end else is_end = function (t) return t == endtoken end end local token,value while true do token,value=tok() if not token then return nil,'EOS' end -- end of stream is an error! if is_end(token,value) and level == 1 then append(parm_values,tl) break elseif token == '(' then level = level + 1 tappend(tl,'(') elseif token == ')' then level = level - 1 if level == 0 then -- finished with parm list append(parm_values,tl) break else tappend(tl,')') end elseif token == delim and level == 1 then append(parm_values,tl) -- a new parm tl = {} else tappend(tl,token,value) end end return parm_values,{token,value} end --- get the next non-space token from the stream. -- @param tok the token stream. function lexer.skipws (tok) local t,v = tok() while t == 'space' do t,v = tok() end return t,v end local skipws = lexer.skipws --- get the next token, which must be of the expected type. -- Throws an error if this type does not match! -- @param tok the token stream -- @string expected_type the token type -- @bool no_skip_ws whether we should skip whitespace function lexer.expecting (tok,expected_type,no_skip_ws) assert_arg(1,tok,'function') assert_arg(2,expected_type,'string') local t,v if no_skip_ws then t,v = tok() else t,v = skipws(tok) end if t ~= expected_type then error ("expecting "..expected_type,2) end return v end return lexer Penlight-1.12.0/lua/pl/luabalanced.lua000066400000000000000000000174611416703176500175220ustar00rootroot00000000000000--- Extract delimited Lua sequences from strings. -- Inspired by Damian Conway's Text::Balanced in Perl.
--
    --
  • [1] Lua Wiki Page
  • --
  • [2] http://search.cpan.org/dist/Text-Balanced/lib/Text/Balanced.pm
  • --

--
-- local lb = require "pl.luabalanced"
-- --Extract Lua expression starting at position 4.
--  print(lb.match_expression("if x^2 + x > 5 then print(x) end", 4))
--  --> x^2 + x > 5     16
-- --Extract Lua string starting at (default) position 1.
-- print(lb.match_string([["test\"123" .. "more"]]))
-- --> "test\"123"     12
-- 
-- (c) 2008, David Manura, Licensed under the same terms as Lua (MIT license). -- @class module -- @name pl.luabalanced local M = {} local assert = assert -- map opening brace <-> closing brace. local ends = { ['('] = ')', ['{'] = '}', ['['] = ']' } local begins = {}; for k,v in pairs(ends) do begins[v] = k end -- Match Lua string in string starting at position . -- Returns , , where is the matched -- string (or nil on no match) and is the character -- following the match (or on no match). -- Supports all Lua string syntax: "...", '...', [[...]], [=[...]=], etc. local function match_string(s, pos) pos = pos or 1 local posa = pos local c = s:sub(pos,pos) if c == '"' or c == "'" then pos = pos + 1 while 1 do pos = assert(s:find("[" .. c .. "\\]", pos), 'syntax error') if s:sub(pos,pos) == c then local part = s:sub(posa, pos) return part, pos + 1 else pos = pos + 2 end end else local sc = s:match("^%[(=*)%[", pos) if sc then local _; _, pos = s:find("%]" .. sc .. "%]", pos) assert(pos) local part = s:sub(posa, pos) return part, pos + 1 else return nil, pos end end end M.match_string = match_string -- Match bracketed Lua expression, e.g. "(...)", "{...}", "[...]", "[[...]]", -- [=[...]=], etc. -- Function interface is similar to match_string. local function match_bracketed(s, pos) pos = pos or 1 local posa = pos local ca = s:sub(pos,pos) if not ends[ca] then return nil, pos end local stack = {} while 1 do pos = s:find('[%(%{%[%)%}%]\"\']', pos) assert(pos, 'syntax error: unbalanced') local c = s:sub(pos,pos) if c == '"' or c == "'" then local part; part, pos = match_string(s, pos) assert(part) elseif ends[c] then -- open local mid, posb if c == '[' then mid, posb = s:match('^%[(=*)%[()', pos) end if mid then pos = s:match('%]' .. mid .. '%]()', posb) assert(pos, 'syntax error: long string not terminated') if #stack == 0 then local part = s:sub(posa, pos-1) return part, pos end else stack[#stack+1] = c pos = pos + 1 end else -- close assert(stack[#stack] == assert(begins[c]), 'syntax error: unbalanced') stack[#stack] = nil if #stack == 0 then local part = s:sub(posa, pos) return part, pos+1 end pos = pos + 1 end end end M.match_bracketed = match_bracketed -- Match Lua comment, e.g. "--...\n", "--[[...]]", "--[=[...]=]", etc. -- Function interface is similar to match_string. local function match_comment(s, pos) pos = pos or 1 if s:sub(pos, pos+1) ~= '--' then return nil, pos end pos = pos + 2 local partt, post = match_string(s, pos) if partt then return '--' .. partt, post end local part; part, pos = s:match('^([^\n]*\n?)()', pos) return '--' .. part, pos end -- Match Lua expression, e.g. "a + b * c[e]". -- Function interface is similar to match_string. local wordop = {['and']=true, ['or']=true, ['not']=true} local is_compare = {['>']=true, ['<']=true, ['~']=true} local function match_expression(s, pos) pos = pos or 1 local _ local posa = pos local lastident local poscs, posce while pos do local c = s:sub(pos,pos) if c == '"' or c == "'" or c == '[' and s:find('^[=%[]', pos+1) then local part; part, pos = match_string(s, pos) assert(part, 'syntax error') elseif c == '-' and s:sub(pos+1,pos+1) == '-' then -- note: handle adjacent comments in loop to properly support -- backtracing (poscs/posce). poscs = pos while s:sub(pos,pos+1) == '--' do local part; part, pos = match_comment(s, pos) assert(part) pos = s:match('^%s*()', pos) posce = pos end elseif c == '(' or c == '{' or c == '[' then _, pos = match_bracketed(s, pos) elseif c == '=' and s:sub(pos+1,pos+1) == '=' then pos = pos + 2 -- skip over two-char op containing '=' elseif c == '=' and is_compare[s:sub(pos-1,pos-1)] then pos = pos + 1 -- skip over two-char op containing '=' elseif c:match'^[%)%}%];,=]' then local part = s:sub(posa, pos-1) return part, pos elseif c:match'^[%w_]' then local newident,newpos = s:match('^([%w_]+)()', pos) if pos ~= posa and not wordop[newident] then -- non-first ident local pose = ((posce == pos) and poscs or pos) - 1 while s:match('^%s', pose) do pose = pose - 1 end local ce = s:sub(pose,pose) if ce:match'[%)%}\'\"%]]' or ce:match'[%w_]' and not wordop[lastident] then local part = s:sub(posa, pos-1) return part, pos end end lastident, pos = newident, newpos else pos = pos + 1 end pos = s:find('[%(%{%[%)%}%]\"\';,=%w_%-]', pos) end local part = s:sub(posa, #s) return part, #s+1 end M.match_expression = match_expression -- Match name list (zero or more names). E.g. "a,b,c" -- Function interface is similar to match_string, -- but returns array as match. local function match_namelist(s, pos) pos = pos or 1 local list = {} while 1 do local c = #list == 0 and '^' or '^%s*,%s*' local item, post = s:match(c .. '([%a_][%w_]*)%s*()', pos) if item then pos = post else break end list[#list+1] = item end return list, pos end M.match_namelist = match_namelist -- Match expression list (zero or more expressions). E.g. "a+b,b*c". -- Function interface is similar to match_string, -- but returns array as match. local function match_explist(s, pos) pos = pos or 1 local list = {} while 1 do if #list ~= 0 then local post = s:match('^%s*,%s*()', pos) if post then pos = post else break end end local item; item, pos = match_expression(s, pos) assert(item, 'syntax error') list[#list+1] = item end return list, pos end M.match_explist = match_explist -- Replace snippets of code in Lua code string -- using replacement function f(u,sin) --> sout. -- is the type of snippet ('c' = comment, 's' = string, -- 'e' = any other code). -- Snippet is replaced with (unless is nil or false, in -- which case the original snippet is kept) -- This is somewhat analogous to string.gsub . local function gsub(s, f) local pos = 1 local posa = 1 local sret = '' while 1 do pos = s:find('[%-\'\"%[]', pos) if not pos then break end if s:match('^%-%-', pos) then local exp = s:sub(posa, pos-1) if #exp > 0 then sret = sret .. (f('e', exp) or exp) end local comment; comment, pos = match_comment(s, pos) sret = sret .. (f('c', assert(comment)) or comment) posa = pos else local posb = s:find('^[\'\"%[]', pos) local str if posb then str, pos = match_string(s, posb) end if str then local exp = s:sub(posa, posb-1) if #exp > 0 then sret = sret .. (f('e', exp) or exp) end sret = sret .. (f('s', str) or str) posa = pos else pos = pos + 1 end end end local exp = s:sub(posa) if #exp > 0 then sret = sret .. (f('e', exp) or exp) end return sret end M.gsub = gsub return M Penlight-1.12.0/lua/pl/operator.lua000066400000000000000000000102311416703176500171060ustar00rootroot00000000000000--- Lua operators available as functions. -- -- (similar to the Python module of the same name) -- -- There is a module field `optable` which maps the operator strings -- onto these functions, e.g. `operator.optable['()']==operator.call` -- -- Operator strings like '>' and '{}' can be passed to most Penlight functions -- expecting a function argument. -- -- @module pl.operator local strfind = string.find local operator = {} --- apply function to some arguments **()** -- @param fn a function or callable object -- @param ... arguments function operator.call(fn,...) return fn(...) end --- get the indexed value from a table **[]** -- @param t a table or any indexable object -- @param k the key function operator.index(t,k) return t[k] end --- returns true if arguments are equal **==** -- @param a value -- @param b value function operator.eq(a,b) return a==b end --- returns true if arguments are not equal **~=** -- @param a value -- @param b value function operator.neq(a,b) return a~=b end --- returns true if a is less than b **<** -- @param a value -- @param b value function operator.lt(a,b) return a < b end --- returns true if a is less or equal to b **<=** -- @param a value -- @param b value function operator.le(a,b) return a <= b end --- returns true if a is greater than b **>** -- @param a value -- @param b value function operator.gt(a,b) return a > b end --- returns true if a is greater or equal to b **>=** -- @param a value -- @param b value function operator.ge(a,b) return a >= b end --- returns length of string or table **#** -- @param a a string or a table function operator.len(a) return #a end --- add two values **+** -- @param a value -- @param b value function operator.add(a,b) return a+b end --- subtract b from a **-** -- @param a value -- @param b value function operator.sub(a,b) return a-b end --- multiply two values __*__ -- @param a value -- @param b value function operator.mul(a,b) return a*b end --- divide first value by second **/** -- @param a value -- @param b value function operator.div(a,b) return a/b end --- raise first to the power of second **^** -- @param a value -- @param b value function operator.pow(a,b) return a^b end --- modulo; remainder of a divided by b **%** -- @param a value -- @param b value function operator.mod(a,b) return a%b end --- concatenate two values (either strings or `__concat` defined) **..** -- @param a value -- @param b value function operator.concat(a,b) return a..b end --- return the negative of a value **-** -- @param a value function operator.unm(a) return -a end --- false if value evaluates as true **not** -- @param a value function operator.lnot(a) return not a end --- true if both values evaluate as true **and** -- @param a value -- @param b value function operator.land(a,b) return a and b end --- true if either value evaluate as true **or** -- @param a value -- @param b value function operator.lor(a,b) return a or b end --- make a table from the arguments **{}** -- @param ... non-nil arguments -- @return a table function operator.table (...) return {...} end --- match two strings **~**. -- uses @{string.find} function operator.match (a,b) return strfind(a,b)~=nil end --- the null operation. -- @param ... arguments -- @return the arguments function operator.nop (...) return ... end ---- Map from operator symbol to function. -- Most of these map directly from operators; -- But note these extras -- -- * __'()'__ `call` -- * __'[]'__ `index` -- * __'{}'__ `table` -- * __'~'__ `match` -- -- @table optable -- @field operator operator.optable = { ['+']=operator.add, ['-']=operator.sub, ['*']=operator.mul, ['/']=operator.div, ['%']=operator.mod, ['^']=operator.pow, ['..']=operator.concat, ['()']=operator.call, ['[]']=operator.index, ['<']=operator.lt, ['<=']=operator.le, ['>']=operator.gt, ['>=']=operator.ge, ['==']=operator.eq, ['~=']=operator.neq, ['#']=operator.len, ['and']=operator.land, ['or']=operator.lor, ['{}']=operator.table, ['~']=operator.match, ['']=operator.nop, } return operator Penlight-1.12.0/lua/pl/path.lua000066400000000000000000000405551416703176500162230ustar00rootroot00000000000000--- Path manipulation and file queries. -- -- This is modelled after Python's os.path library (10.1); see @{04-paths.md|the Guide}. -- -- NOTE: the functions assume the paths being dealt with to originate -- from the OS the application is running on. Windows drive letters are not -- to be used when running on a Unix system for example. The one exception -- is Windows paths to allow both forward and backward slashes (since Lua -- also accepts those) -- -- Dependencies: `pl.utils`, `lfs` -- @module pl.path -- imports and locals local _G = _G local sub = string.sub local getenv = os.getenv local tmpnam = os.tmpname local package = package local append, concat, remove = table.insert, table.concat, table.remove local utils = require 'pl.utils' local assert_string,raise = utils.assert_string,utils.raise local res,lfs = _G.pcall(_G.require,'lfs') if not res then error("pl.path requires LuaFileSystem") end local attrib = lfs.attributes local currentdir = lfs.currentdir local link_attrib = lfs.symlinkattributes local path = {} local function err_func(name, param, err, code) local ret = ("%s failed"):format(tostring(name)) if param ~= nil then ret = ret .. (" for '%s'"):format(tostring(param)) end ret = ret .. (": %s"):format(tostring(err)) if code ~= nil then ret = ret .. (" (code %s)"):format(tostring(code)) end return ret end --- Lua iterator over the entries of a given directory. -- Implicit link to [`luafilesystem.dir`](https://keplerproject.github.io/luafilesystem/manual.html#reference) -- @function dir path.dir = lfs.dir --- Creates a directory. -- Implicit link to [`luafilesystem.mkdir`](https://keplerproject.github.io/luafilesystem/manual.html#reference) -- @function mkdir path.mkdir = function(d) local ok, err, code = lfs.mkdir(d) if not ok then return ok, err_func("mkdir", d, err, code), code end return ok, err, code end --- Removes a directory. -- Implicit link to [`luafilesystem.rmdir`](https://keplerproject.github.io/luafilesystem/manual.html#reference) -- @function rmdir path.rmdir = function(d) local ok, err, code = lfs.rmdir(d) if not ok then return ok, err_func("rmdir", d, err, code), code end return ok, err, code end --- Gets attributes. -- Implicit link to [`luafilesystem.attributes`](https://keplerproject.github.io/luafilesystem/manual.html#reference) -- @function attrib path.attrib = function(d, r) local ok, err, code = attrib(d, r) if not ok then return ok, err_func("attrib", d, err, code), code end return ok, err, code end --- Get the working directory. -- Implicit link to [`luafilesystem.currentdir`](https://keplerproject.github.io/luafilesystem/manual.html#reference) -- @function currentdir path.currentdir = function() local ok, err, code = currentdir() if not ok then return ok, err_func("currentdir", nil, err, code), code end return ok, err, code end --- Gets symlink attributes. -- Implicit link to [`luafilesystem.symlinkattributes`](https://keplerproject.github.io/luafilesystem/manual.html#reference) -- @function link_attrib path.link_attrib = function(d, r) local ok, err, code = link_attrib(d, r) if not ok then return ok, err_func("link_attrib", d, err, code), code end return ok, err, code end --- Changes the working directory. -- On Windows, if a drive is specified, it also changes the current drive. If -- only specifying the drive, it will only switch drive, but not modify the path. -- Implicit link to [`luafilesystem.chdir`](https://keplerproject.github.io/luafilesystem/manual.html#reference) -- @function chdir path.chdir = function(d) local ok, err, code = lfs.chdir(d) if not ok then return ok, err_func("chdir", d, err, code), code end return ok, err, code end --- is this a directory? -- @string P A file path function path.isdir(P) assert_string(1,P) if P:match("\\$") then P = P:sub(1,-2) end return attrib(P,'mode') == 'directory' end --- is this a file? -- @string P A file path function path.isfile(P) assert_string(1,P) return attrib(P,'mode') == 'file' end -- is this a symbolic link? -- @string P A file path function path.islink(P) assert_string(1,P) if link_attrib then return link_attrib(P,'mode')=='link' else return false end end --- return size of a file. -- @string P A file path function path.getsize(P) assert_string(1,P) return attrib(P,'size') end --- does a path exist? -- @string P A file path -- @return the file path if it exists (either as file, directory, socket, etc), nil otherwise function path.exists(P) assert_string(1,P) return attrib(P,'mode') ~= nil and P end --- Return the time of last access as the number of seconds since the epoch. -- @string P A file path function path.getatime(P) assert_string(1,P) return attrib(P,'access') end --- Return the time of last modification as the number of seconds since the epoch. -- @string P A file path function path.getmtime(P) assert_string(1,P) return attrib(P,'modification') end ---Return the system's ctime as the number of seconds since the epoch. -- @string P A file path function path.getctime(P) assert_string(1,P) return path.attrib(P,'change') end local function at(s,i) return sub(s,i,i) end path.is_windows = utils.is_windows local sep, other_sep, seps -- constant sep is the directory separator for this platform. -- constant dirsep is the separator in the PATH environment variable if path.is_windows then path.sep = '\\'; other_sep = '/' path.dirsep = ';' seps = { ['/'] = true, ['\\'] = true } else path.sep = '/' path.dirsep = ':' seps = { ['/'] = true } end sep = path.sep --- are we running Windows? -- @class field -- @name path.is_windows --- path separator for this platform. -- @class field -- @name path.sep --- separator for PATH for this platform -- @class field -- @name path.dirsep --- given a path, return the directory part and a file part. -- if there's no directory part, the first value will be empty -- @string P A file path -- @return directory part -- @return file part -- @usage -- local dir, file = path.splitpath("some/dir/myfile.txt") -- assert(dir == "some/dir") -- assert(file == "myfile.txt") -- -- local dir, file = path.splitpath("some/dir/") -- assert(dir == "some/dir") -- assert(file == "") -- -- local dir, file = path.splitpath("some_dir") -- assert(dir == "") -- assert(file == "some_dir") function path.splitpath(P) assert_string(1,P) local i = #P local ch = at(P,i) while i > 0 and ch ~= sep and ch ~= other_sep do i = i - 1 ch = at(P,i) end if i == 0 then return '',P else return sub(P,1,i-1), sub(P,i+1) end end --- return an absolute path. -- @string P A file path -- @string[opt] pwd optional start path to use (default is current dir) function path.abspath(P,pwd) assert_string(1,P) if pwd then assert_string(2,pwd) end local use_pwd = pwd ~= nil if not use_pwd and not currentdir() then return P end P = P:gsub('[\\/]$','') pwd = pwd or currentdir() if not path.isabs(P) then P = path.join(pwd,P) elseif path.is_windows and not use_pwd and at(P,2) ~= ':' and at(P,2) ~= '\\' then P = pwd:sub(1,2)..P -- attach current drive to path like '\\fred.txt' end return path.normpath(P) end --- given a path, return the root part and the extension part. -- if there's no extension part, the second value will be empty -- @string P A file path -- @treturn string root part (everything upto the "."", maybe empty) -- @treturn string extension part (including the ".", maybe empty) -- @usage -- local file_path, ext = path.splitext("/bonzo/dog_stuff/cat.txt") -- assert(file_path == "/bonzo/dog_stuff/cat") -- assert(ext == ".txt") -- -- local file_path, ext = path.splitext("") -- assert(file_path == "") -- assert(ext == "") function path.splitext(P) assert_string(1,P) local i = #P local ch = at(P,i) while i > 0 and ch ~= '.' do if seps[ch] then return P,'' end i = i - 1 ch = at(P,i) end if i == 0 then return P,'' else return sub(P,1,i-1),sub(P,i) end end --- return the directory part of a path -- @string P A file path -- @treturn string everything before the last dir-separator -- @see splitpath -- @usage -- path.dirname("/some/path/file.txt") -- "/some/path" -- path.dirname("file.txt") -- "" (empty string) function path.dirname(P) assert_string(1,P) local p1 = path.splitpath(P) return p1 end --- return the file part of a path -- @string P A file path -- @treturn string -- @see splitpath -- @usage -- path.basename("/some/path/file.txt") -- "file.txt" -- path.basename("/some/path/file/") -- "" (empty string) function path.basename(P) assert_string(1,P) local _,p2 = path.splitpath(P) return p2 end --- get the extension part of a path. -- @string P A file path -- @treturn string -- @see splitext -- @usage -- path.extension("/some/path/file.txt") -- ".txt" -- path.extension("/some/path/file_txt") -- "" (empty string) function path.extension(P) assert_string(1,P) local _,p2 = path.splitext(P) return p2 end --- is this an absolute path? -- @string P A file path -- @usage -- path.isabs("hello/path") -- false -- path.isabs("/hello/path") -- true -- -- Windows; -- path.isabs("hello\path") -- false -- path.isabs("\hello\path") -- true -- path.isabs("C:\hello\path") -- true -- path.isabs("C:hello\path") -- false function path.isabs(P) assert_string(1,P) if path.is_windows and at(P,2) == ":" then return seps[at(P,3)] ~= nil end return seps[at(P,1)] ~= nil end --- return the path resulting from combining the individual paths. -- if the second (or later) path is absolute, we return the last absolute path (joined with any non-absolute paths following). -- empty elements (except the last) will be ignored. -- @string p1 A file path -- @string p2 A file path -- @string ... more file paths -- @treturn string the combined path -- @usage -- path.join("/first","second","third") -- "/first/second/third" -- path.join("first","second/third") -- "first/second/third" -- path.join("/first","/second","third") -- "/second/third" function path.join(p1,p2,...) assert_string(1,p1) assert_string(2,p2) if select('#',...) > 0 then local p = path.join(p1,p2) local args = {...} for i = 1,#args do assert_string(i,args[i]) p = path.join(p,args[i]) end return p end if path.isabs(p2) then return p2 end local endc = at(p1,#p1) if endc ~= path.sep and endc ~= other_sep and endc ~= "" then p1 = p1..path.sep end return p1..p2 end --- normalize the case of a pathname. On Unix, this returns the path unchanged, -- for Windows it converts; -- -- * the path to lowercase -- * forward slashes to backward slashes -- @string P A file path -- @usage path.normcase("/Some/Path/File.txt") -- -- Windows: "\some\path\file.txt" -- -- Others : "/Some/Path/File.txt" function path.normcase(P) assert_string(1,P) if path.is_windows then return P:gsub('/','\\'):lower() else return P end end --- normalize a path name. -- `A//B`, `A/./B`, and `A/foo/../B` all become `A/B`. -- -- An empty path results in '.'. -- @string P a file path function path.normpath(P) assert_string(1,P) -- Split path into anchor and relative path. local anchor = '' if path.is_windows then if P:match '^\\\\' then -- UNC anchor = '\\\\' P = P:sub(3) elseif seps[at(P, 1)] then anchor = '\\' P = P:sub(2) elseif at(P, 2) == ':' then anchor = P:sub(1, 2) P = P:sub(3) if seps[at(P, 1)] then anchor = anchor..'\\' P = P:sub(2) end end P = P:gsub('/','\\') else -- According to POSIX, in path start '//' and '/' are distinct, -- but '///+' is equivalent to '/'. if P:match '^//' and at(P, 3) ~= '/' then anchor = '//' P = P:sub(3) elseif at(P, 1) == '/' then anchor = '/' P = P:match '^/*(.*)$' end end local parts = {} for part in P:gmatch('[^'..sep..']+') do if part == '..' then if #parts ~= 0 and parts[#parts] ~= '..' then remove(parts) else append(parts, part) end elseif part ~= '.' then append(parts, part) end end P = anchor..concat(parts, sep) if P == '' then P = '.' end return P end --- relative path from current directory or optional start point -- @string P a path -- @string[opt] start optional start point (default current directory) function path.relpath (P,start) assert_string(1,P) if start then assert_string(2,start) end local split,min,append = utils.split, math.min, table.insert P = path.abspath(P,start) start = start or currentdir() local compare if path.is_windows then P = P:gsub("/","\\") start = start:gsub("/","\\") compare = function(v) return v:lower() end else compare = function(v) return v end end local startl, Pl = split(start,sep), split(P,sep) local n = min(#startl,#Pl) if path.is_windows and n > 0 and at(Pl[1],2) == ':' and Pl[1] ~= startl[1] then return P end local k = n+1 -- default value if this loop doesn't bail out! for i = 1,n do if compare(startl[i]) ~= compare(Pl[i]) then k = i break end end local rell = {} for i = 1, #startl-k+1 do rell[i] = '..' end if k <= #Pl then for i = k,#Pl do append(rell,Pl[i]) end end return table.concat(rell,sep) end --- Replace a starting '~' with the user's home directory. -- In windows, if HOME isn't set, then USERPROFILE is used in preference to -- HOMEDRIVE HOMEPATH. This is guaranteed to be writeable on all versions of Windows. -- @string P A file path function path.expanduser(P) assert_string(1,P) if at(P,1) == '~' then local home = getenv('HOME') if not home then -- has to be Windows home = getenv 'USERPROFILE' or (getenv 'HOMEDRIVE' .. getenv 'HOMEPATH') end return home..sub(P,2) else return P end end ---Return a suitable full path to a new temporary file name. -- unlike os.tmpname(), it always gives you a writeable path (uses TEMP environment variable on Windows) function path.tmpname () local res = tmpnam() -- On Windows if Lua is compiled using MSVC14 os.tmpname -- already returns an absolute path within TEMP env variable directory, -- no need to prepend it. if path.is_windows and not res:find(':') then res = getenv('TEMP')..res end return res end --- return the largest common prefix path of two paths. -- @string path1 a file path -- @string path2 a file path -- @return the common prefix (Windows: separators will be normalized, casing will be original) function path.common_prefix (path1,path2) assert_string(1,path1) assert_string(2,path2) -- get them in order! if #path1 > #path2 then path2,path1 = path1,path2 end local compare if path.is_windows then path1 = path1:gsub("/", "\\") path2 = path2:gsub("/", "\\") compare = function(v) return v:lower() end else compare = function(v) return v end end for i = 1,#path1 do if compare(at(path1,i)) ~= compare(at(path2,i)) then local cp = path1:sub(1,i-1) if at(path1,i-1) ~= sep then cp = path.dirname(cp) end return cp end end if at(path2,#path1+1) ~= sep then path1 = path.dirname(path1) end return path1 --return '' end --- return the full path where a particular Lua module would be found. -- Both package.path and package.cpath is searched, so the result may -- either be a Lua file or a shared library. -- @string mod name of the module -- @return on success: path of module, lua or binary -- @return on error: nil, error string listing paths tried function path.package_path(mod) assert_string(1,mod) local res, err1, err2 res, err1 = package.searchpath(mod,package.path) if res then return res,true end res, err2 = package.searchpath(mod,package.cpath) if res then return res,false end return raise ('cannot find module on path\n' .. err1 .. "\n" .. err2) end ---- finis ----- return path Penlight-1.12.0/lua/pl/permute.lua000066400000000000000000000120321416703176500167350ustar00rootroot00000000000000--- Permutation operations. -- -- Dependencies: `pl.utils`, `pl.tablex` -- @module pl.permute local tablex = require 'pl.tablex' local utils = require 'pl.utils' local copy = tablex.deepcopy local append = table.insert local assert_arg = utils.assert_arg local permute = {} --- an iterator over all order-permutations of the elements of a list. -- Please note that the same list is returned each time, so do not keep references! -- @param a list-like table -- @return an iterator which provides the next permutation as a list function permute.order_iter(a) assert_arg(1,a,'table') local t = #a local stack = { 1 } local function iter() local h = #stack local n = t - h + 1 local i = stack[h] if i > t then return end if n == 0 then table.remove(stack) h = h - 1 stack[h] = stack[h] + 1 return a elseif i <= n then -- put i-th element as the last one a[n], a[i] = a[i], a[n] -- generate all permutations of the other elements table.insert(stack, 1) else table.remove(stack) h = h - 1 n = n + 1 i = stack[h] -- restore i-th element a[n], a[i] = a[i], a[n] stack[h] = stack[h] + 1 end return iter() -- tail-call end return iter end --- construct a table containing all the order-permutations of a list. -- @param a list-like table -- @return a table of tables -- @usage permute.order_table {1,2,3} --> {{2,3,1},{3,2,1},{3,1,2},{1,3,2},{2,1,3},{1,2,3}} function permute.order_table (a) assert_arg(1,a,'table') local res = {} for t in permute.iter(a) do append(res,copy(t)) end return res end --- an iterator over all permutations of the elements of the given lists. -- @param ... list-like tables, they are nil-safe if a length-field `n` is provided (see `utils.pack`) -- @return an iterator which provides the next permutation as return values in the same order as the provided lists, preceeded by an index -- @usage -- local strs = utils.pack("one", nil, "three") -- adds an 'n' field for nil-safety -- local bools = utils.pack(true, false) -- local iter = permute.list_iter(strs, bools) -- -- print(iter()) --> 1, one, true -- print(iter()) --> 2, nil, true -- print(iter()) --> 3, three, true -- print(iter()) --> 4, one, false -- print(iter()) --> 5, nil, false -- print(iter()) --> 6, three, false function permute.list_iter(...) local elements = {...} local pointers = {} local sizes = {} local size = #elements for i, list in ipairs(elements) do assert_arg(i,list,'table') pointers[i] = 1 sizes[i] = list.n or #list end local count = 0 return function() if pointers[size] > sizes[size] then return end -- we're done count = count + 1 local r = { n = #elements } local cascade_up = true for i = 1, size do r[i] = elements[i][pointers[i]] if cascade_up then pointers[i] = pointers[i] + 1 if pointers[i] <= sizes[i] then -- this list is not done yet, stop cascade cascade_up = false else -- this list is done if i ~= size then -- reset pointer pointers[i] = 1 end end end end return count, utils.unpack(r) end end --- construct a table containing all the permutations of a set of lists. -- @param ... list-like tables, they are nil-safe if a length-field `n` is provided -- @return a list of lists, the sub-lists have an 'n' field for nil-safety -- @usage -- local strs = utils.pack("one", nil, "three") -- adds an 'n' field for nil-safety -- local bools = utils.pack(true, false) -- local results = permute.list_table(strs, bools) -- -- results = { -- -- { "one, true, n = 2 } -- -- { nil, true, n = 2 }, -- -- { "three, true, n = 2 }, -- -- { "one, false, n = 2 }, -- -- { nil, false, n = 2 }, -- -- { "three", false, n = 2 }, -- -- } function permute.list_table(...) local iter = permute.list_iter(...) local results = {} local i = 1 while true do local values = utils.pack(iter()) if values[1] == nil then return results end for i = 1, values.n do values[i] = values[i+1] end values.n = values.n - 1 results[i] = values i = i + 1 end end -- backward compat, to be deprecated --- deprecated. -- @param ... -- @see permute.order_iter function permute.iter(...) utils.raise_deprecation { source = "Penlight " .. utils._VERSION, message = "function 'iter' was renamed to 'order_iter'", version_removed = "2.0.0", deprecated_after = "1.9.2", } return permute.order_iter(...) end --- deprecated. -- @param ... -- @see permute.order_iter function permute.table(...) utils.raise_deprecation { source = "Penlight " .. utils._VERSION, message = "function 'table' was renamed to 'order_table'", version_removed = "2.0.0", deprecated_after = "1.9.2", } return permute.order_table(...) end return permute Penlight-1.12.0/lua/pl/pretty.lua000066400000000000000000000315011416703176500166050ustar00rootroot00000000000000--- Pretty-printing Lua tables. -- Also provides a sandboxed Lua table reader and -- a function to present large numbers in human-friendly format. -- -- Dependencies: `pl.utils`, `pl.lexer`, `pl.stringx`, `debug` -- @module pl.pretty local append = table.insert local concat = table.concat local mfloor, mhuge = math.floor, math.huge local mtype = math.type local utils = require 'pl.utils' local lexer = require 'pl.lexer' local debug = require 'debug' local quote_string = require'pl.stringx'.quote_string local assert_arg = utils.assert_arg local original_tostring = tostring -- Patch tostring to format numbers with better precision -- and to produce cross-platform results for -- infinite values and NaN. local function tostring(value) if type(value) ~= "number" then return original_tostring(value) elseif value ~= value then return "NaN" elseif value == mhuge then return "Inf" elseif value == -mhuge then return "-Inf" elseif (_VERSION ~= "Lua 5.3" or mtype(value) == "integer") and mfloor(value) == value then return ("%d"):format(value) else local res = ("%.14g"):format(value) if _VERSION == "Lua 5.3" and mtype(value) == "float" and not res:find("%.") then -- Number is internally a float but looks like an integer. -- Insert ".0" after first run of digits. res = res:gsub("%d+", "%0.0", 1) end return res end end local pretty = {} local function save_global_env() local env = {} env.hook, env.mask, env.count = debug.gethook() -- env.hook is "external hook" if is a C hook function if env.hook~="external hook" then debug.sethook() end env.string_mt = getmetatable("") debug.setmetatable("", nil) return env end local function restore_global_env(env) if env then debug.setmetatable("", env.string_mt) if env.hook~="external hook" then debug.sethook(env.hook, env.mask, env.count) end end end --- Read a string representation of a Lua table. -- This function loads and runs the string as Lua code, but bails out -- if it contains a function definition. -- Loaded string is executed in an empty environment. -- @string s string to read in `{...}` format, possibly with some whitespace -- before or after the curly braces. A single line comment may be present -- at the beginning. -- @return a table in case of success. -- If loading the string failed, return `nil` and error message. -- If executing loaded string failed, return `nil` and the error it raised. function pretty.read(s) assert_arg(1,s,'string') if s:find '^%s*%-%-' then -- may start with a comment.. s = s:gsub('%-%-.-\n','') end if not s:find '^%s*{' then return nil,"not a Lua table" end if s:find '[^\'"%w_]function[^\'"%w_]' then local tok = lexer.lua(s) for t,v in tok do if t == 'keyword' and v == 'function' then return nil,"cannot have functions in table definition" end end end s = 'return '..s local chunk,err = utils.load(s,'tbl','t',{}) if not chunk then return nil,err end local global_env = save_global_env() local ok,ret = pcall(chunk) restore_global_env(global_env) if ok then return ret else return nil,ret end end --- Read a Lua chunk. -- @string s Lua code. -- @tab[opt] env environment used to run the code, empty by default. -- @bool[opt] paranoid abort loading if any looping constructs a found in the code -- and disable string methods. -- @return the environment in case of success or `nil` and syntax or runtime error -- if something went wrong. function pretty.load (s, env, paranoid) env = env or {} if paranoid then local tok = lexer.lua(s) for t,v in tok do if t == 'keyword' and (v == 'for' or v == 'repeat' or v == 'function' or v == 'goto') then return nil,"looping not allowed" end end end local chunk,err = utils.load(s,'tbl','t',env) if not chunk then return nil,err end local global_env = paranoid and save_global_env() local ok,err = pcall(chunk) restore_global_env(global_env) if not ok then return nil,err end return env end local function quote_if_necessary (v) if not v then return '' else --AAS if v:find ' ' then v = quote_string(v) end end return v end local keywords local function is_identifier (s) return type(s) == 'string' and s:find('^[%a_][%w_]*$') and not keywords[s] end local function quote (s) if type(s) == 'table' then return pretty.write(s,'') else --AAS return quote_string(s)-- ('%q'):format(tostring(s)) end end local function index (numkey,key) --AAS if not numkey then key = quote(key) key = key:find("^%[") and (" " .. key .. " ") or key end return '['..key..']' end --- Create a string representation of a Lua table. -- This function never fails, but may complain by returning an -- extra value. Normally puts out one item per line, using -- the provided indent; set the second parameter to an empty string -- if you want output on one line. -- -- *NOTE:* this is NOT a serialization function, not a full blown -- debug function. Checkout out respectively the -- [serpent](https://github.com/pkulchenko/serpent) -- or [inspect](https://github.com/kikito/inspect.lua) -- Lua modules for that if you need them. -- @tab tbl Table to serialize to a string. -- @string[opt] space The indent to use. -- Defaults to two spaces; pass an empty string for no indentation. -- @bool[opt] not_clever Pass `true` for plain output, e.g `{['key']=1}`. -- Defaults to `false`. -- @return a string -- @return an optional error message function pretty.write (tbl,space,not_clever) if type(tbl) ~= 'table' then local res = tostring(tbl) if type(tbl) == 'string' then return quote(tbl) end return res, 'not a table' end if not keywords then keywords = lexer.get_keywords() end local set = ' = ' if space == '' then set = '=' end space = space or ' ' local lines = {} local line = '' local tables = {} local function put(s) if #s > 0 then line = line..s end end local function putln (s) if #line > 0 then line = line..s append(lines,line) line = '' else append(lines,s) end end local function eat_last_comma () local n = #lines local lastch = lines[n]:sub(-1,-1) if lastch == ',' then lines[n] = lines[n]:sub(1,-2) end end -- safe versions for iterators since 5.3+ honors metamethods that can throw -- errors local ipairs = function(t) local i = 0 local ok, v local getter = function() return t[i] end return function() i = i + 1 ok, v = pcall(getter) if v == nil or not ok then return end return i, t[i] end end local pairs = function(t) local k, v, ok local getter = function() return next(t, k) end return function() ok, k, v = pcall(getter) if not ok then return end return k, v end end local writeit writeit = function (t,oldindent,indent) local tp = type(t) if tp ~= 'string' and tp ~= 'table' then putln(quote_if_necessary(tostring(t))..',') elseif tp == 'string' then -- if t:find('\n') then -- putln('[[\n'..t..']],') -- else -- putln(quote(t)..',') -- end --AAS putln(quote_string(t) ..",") elseif tp == 'table' then if tables[t] then putln(',') return end tables[t] = true local newindent = indent..space putln('{') local used = {} if not not_clever then for i,val in ipairs(t) do put(indent) writeit(val,indent,newindent) used[i] = true end end local ordered_keys = {} for k,v in pairs(t) do if type(k) ~= 'number' then ordered_keys[#ordered_keys + 1] = k end end table.sort(ordered_keys, function (a, b) if type(a) == type(b) and type(a) == 'string' then return a < b end return type(a) == 'boolean' or (type(b) ~= 'boolean' and type(a) == 'table') end) local function write_entry (key, val) local tkey = type(key) local numkey = tkey == 'number' if not_clever then key = tostring(key) put(indent..index(numkey,key)..set) writeit(val,indent,newindent) else if not numkey or not used[key] then -- non-array indices if tkey ~= 'string' then key = tostring(key) end if numkey or not is_identifier(key) then key = index(numkey,key) end put(indent..key..set) writeit(val,indent,newindent) end end end for i = 1, #ordered_keys do local key = ordered_keys[i] local val = t[key] write_entry(key, val) end for key,val in pairs(t) do if type(key) == 'number' then write_entry(key, val) end end tables[t] = nil eat_last_comma() putln(oldindent..'},') else putln(tostring(t)..',') end end writeit(tbl,'',space) eat_last_comma() return concat(lines,#space > 0 and '\n' or '') end --- Dump a Lua table out to a file or stdout. -- @tab t The table to write to a file or stdout. -- @string[opt] filename File name to write too. Defaults to writing -- to stdout. function pretty.dump (t, filename) if not filename then print(pretty.write(t)) return true else return utils.writefile(filename, pretty.write(t)) end end --- Dump a series of arguments to stdout for debug purposes. -- This function is attached to the module table `__call` method, to make it -- extra easy to access. So the full: -- -- print(require("pl.pretty").write({...})) -- -- Can be shortened to: -- -- require"pl.pretty" (...) -- -- Any `nil` entries will be printed as `""` to make them explicit. -- @param ... the parameters to dump to stdout. -- @usage -- -- example debug output -- require"pl.pretty" ("hello", nil, "world", { bye = "world", true} ) -- -- -- output: -- { -- ["arg 1"] = "hello", -- ["arg 2"] = "", -- ["arg 3"] = "world", -- ["arg 4"] = { -- true, -- bye = "world" -- } -- } function pretty.debug(...) local n = select("#", ...) local t = { ... } for i = 1, n do local value = t[i] if value == nil then value = "" end t[i] = nil t["arg " .. i] = value end print(pretty.write(t)) return true end local memp,nump = {'B','KiB','MiB','GiB'},{'','K','M','B'} local function comma (val) local thou = math.floor(val/1000) if thou > 0 then return comma(thou)..','.. tostring(val % 1000) else return tostring(val) end end --- Format large numbers nicely for human consumption. -- @number num a number. -- @string[opt] kind one of `'M'` (memory in `KiB`, `MiB`, etc.), -- `'N'` (postfixes are `'K'`, `'M'` and `'B'`), -- or `'T'` (use commas as thousands separator), `'N'` by default. -- @int[opt] prec number of digits to use for `'M'` and `'N'`, `1` by default. function pretty.number (num,kind,prec) local fmt = '%.'..(prec or 1)..'f%s' if kind == 'T' then return comma(num) else local postfixes, fact if kind == 'M' then fact = 1024 postfixes = memp else fact = 1000 postfixes = nump end local div = fact local k = 1 while num >= div and k <= #postfixes do div = div * fact k = k + 1 end div = div / fact if k > #postfixes then k = k - 1; div = div/fact end if k > 1 then return fmt:format(num/div,postfixes[k] or 'duh') else return num..postfixes[1] end end end return setmetatable(pretty, { __call = function(self, ...) return self.debug(...) end }) Penlight-1.12.0/lua/pl/seq.lua000066400000000000000000000342461416703176500160570ustar00rootroot00000000000000--- Manipulating iterators as sequences. -- See @{07-functional.md.Sequences|The Guide} -- -- Dependencies: `pl.utils`, `pl.types`, `debug` -- @module pl.seq local next,assert,pairs,tonumber,type,setmetatable = next,assert,pairs,tonumber,type,setmetatable local strfind,format = string.find,string.format local mrandom = math.random local tsort,tappend = table.sort,table.insert local io = io local utils = require 'pl.utils' local callable = require 'pl.types'.is_callable local function_arg = utils.function_arg local assert_arg = utils.assert_arg local debug = require 'debug' local seq = {} -- given a number, return a function(y) which returns true if y > x -- @param x a number function seq.greater_than(x) return function(v) return tonumber(v) > x end end -- given a number, returns a function(y) which returns true if y < x -- @param x a number function seq.less_than(x) return function(v) return tonumber(v) < x end end -- given any value, return a function(y) which returns true if y == x -- @param x a value function seq.equal_to(x) if type(x) == "number" then return function(v) return tonumber(v) == x end else return function(v) return v == x end end end --- given a string, return a function(y) which matches y against the string. -- @param s a string function seq.matching(s) return function(v) return strfind(v,s) end end local nexti --- sequence adaptor for a table. Note that if any generic function is -- passed a table, it will automatically use seq.list() -- @param t a list-like table -- @usage sum(list(t)) is the sum of all elements of t -- @usage for x in list(t) do...end function seq.list(t) assert_arg(1,t,'table') if not nexti then nexti = ipairs{} end local key,value = 0 return function() key,value = nexti(t,key) return value end end --- return the keys of the table. -- @param t an arbitrary table -- @return iterator over keys function seq.keys(t) assert_arg(1,t,'table') local key return function() key = next(t,key) return key end end local list = seq.list local function default_iter(iter) if type(iter) == 'table' then return list(iter) else return iter end end seq.iter = default_iter --- create an iterator over a numerical range. Like the standard Python function xrange. -- @param start a number -- @param finish a number greater than start function seq.range(start,finish) local i = start - 1 return function() i = i + 1 if i > finish then return nil else return i end end end -- count the number of elements in the sequence which satisfy the predicate -- @param iter a sequence -- @param condn a predicate function (must return either true or false) -- @param optional argument to be passed to predicate as second argument. -- @return count function seq.count(iter,condn,arg) local i = 0 seq.foreach(iter,function(val) if condn(val,arg) then i = i + 1 end end) return i end --- return the minimum and the maximum value of the sequence. -- @param iter a sequence -- @return minimum value -- @return maximum value function seq.minmax(iter) local vmin,vmax = 1e70,-1e70 for v in default_iter(iter) do v = tonumber(v) if v < vmin then vmin = v end if v > vmax then vmax = v end end return vmin,vmax end --- return the sum and element count of the sequence. -- @param iter a sequence -- @param fn an optional function to apply to the values function seq.sum(iter,fn) local s = 0 local i = 0 for v in default_iter(iter) do if fn then v = fn(v) end s = s + v i = i + 1 end return s,i end --- create a table from the sequence. (This will make the result a List.) -- @param iter a sequence -- @return a List -- @usage copy(list(ls)) is equal to ls -- @usage copy(list {1,2,3}) == List{1,2,3} function seq.copy(iter) local res,k = {},1 for v in default_iter(iter) do res[k] = v k = k + 1 end setmetatable(res, require('pl.List')) return res end --- create a table of pairs from the double-valued sequence. -- @param iter a double-valued sequence -- @param i1 used to capture extra iterator values -- @param i2 as with pairs & ipairs -- @usage copy2(ipairs{10,20,30}) == {{1,10},{2,20},{3,30}} -- @return a list-like table function seq.copy2 (iter,i1,i2) local res,k = {},1 for v1,v2 in iter,i1,i2 do res[k] = {v1,v2} k = k + 1 end return res end --- create a table of 'tuples' from a multi-valued sequence. -- A generalization of copy2 above -- @param iter a multiple-valued sequence -- @return a list-like table function seq.copy_tuples (iter) iter = default_iter(iter) local res = {} local row = {iter()} while #row > 0 do tappend(res,row) row = {iter()} end return res end --- return an iterator of random numbers. -- @param n the length of the sequence -- @param l same as the first optional argument to math.random -- @param u same as the second optional argument to math.random -- @return a sequence function seq.random(n,l,u) local rand assert(type(n) == 'number') if u then rand = function() return mrandom(l,u) end elseif l then rand = function() return mrandom(l) end else rand = mrandom end return function() if n == 0 then return nil else n = n - 1 return rand() end end end --- return an iterator to the sorted elements of a sequence. -- @param iter a sequence -- @param comp an optional comparison function (comp(x,y) is true if x < y) function seq.sort(iter,comp) local t = seq.copy(iter) tsort(t,comp) return list(t) end --- return an iterator which returns elements of two sequences. -- @param iter1 a sequence -- @param iter2 a sequence -- @usage for x,y in seq.zip(ls1,ls2) do....end function seq.zip(iter1,iter2) iter1 = default_iter(iter1) iter2 = default_iter(iter2) return function() return iter1(),iter2() end end --- Makes a table where the key/values are the values and value counts of the sequence. -- This version works with 'hashable' values like strings and numbers. -- `pl.tablex.count_map` is more general. -- @param iter a sequence -- @return a map-like table -- @return a table -- @see pl.tablex.count_map function seq.count_map(iter) local t = {} local v for s in default_iter(iter) do v = t[s] if v then t[s] = v + 1 else t[s] = 1 end end return setmetatable(t, require('pl.Map')) end -- given a sequence, return all the unique values in that sequence. -- @param iter a sequence -- @param returns_table true if we return a table, not a sequence -- @return a sequence or a table; defaults to a sequence. function seq.unique(iter,returns_table) local t = seq.count_map(iter) local res,k = {},1 for key in pairs(t) do res[k] = key; k = k + 1 end table.sort(res) if returns_table then return res else return list(res) end end --- print out a sequence iter with a separator. -- @param iter a sequence -- @param sep the separator (default space) -- @param nfields maximum number of values per line (default 7) -- @param fmt optional format function for each value function seq.printall(iter,sep,nfields,fmt) local write = io.write if not sep then sep = ' ' end if not nfields then if sep == '\n' then nfields = 1e30 else nfields = 7 end end if fmt then local fstr = fmt fmt = function(v) return format(fstr,v) end end local k = 1 for v in default_iter(iter) do if fmt then v = fmt(v) end if k < nfields then write(v,sep) k = k + 1 else write(v,'\n') k = 1 end end write '\n' end -- return an iterator running over every element of two sequences (concatenation). -- @param iter1 a sequence -- @param iter2 a sequence function seq.splice(iter1,iter2) iter1 = default_iter(iter1) iter2 = default_iter(iter2) local iter = iter1 return function() local ret = iter() if ret == nil then if iter == iter1 then iter = iter2 return iter() else return nil end else return ret end end end --- return a sequence where every element of a sequence has been transformed -- by a function. If you don't supply an argument, then the function will -- receive both values of a double-valued sequence, otherwise behaves rather like -- tablex.map. -- @param fn a function to apply to elements; may take two arguments -- @param iter a sequence of one or two values -- @param arg optional argument to pass to function. function seq.map(fn,iter,arg) fn = function_arg(1,fn) iter = default_iter(iter) return function() local v1,v2 = iter() if v1 == nil then return nil end return fn(v1,arg or v2) or false end end --- filter a sequence using a predicate function. -- @param iter a sequence of one or two values -- @param pred a boolean function; may take two arguments -- @param arg optional argument to pass to function. function seq.filter (iter,pred,arg) pred = function_arg(2,pred) return function () local v1,v2 while true do v1,v2 = iter() if v1 == nil then return nil end if pred(v1,arg or v2) then return v1,v2 end end end end --- 'reduce' a sequence using a binary function. -- @func fn a function of two arguments -- @param iter a sequence -- @param initval optional initial value -- @usage seq.reduce(operator.add,seq.list{1,2,3,4}) == 10 -- @usage seq.reduce('-',{1,2,3,4,5}) == -13 function seq.reduce (fn,iter,initval) fn = function_arg(1,fn) iter = default_iter(iter) local val = initval or iter() if val == nil then return nil end for v in iter do val = fn(val,v) end return val end --- take the first n values from the sequence. -- @param iter a sequence of one or two values -- @param n number of items to take -- @return a sequence of at most n items function seq.take (iter,n) iter = default_iter(iter) return function() if n < 1 then return end local val1,val2 = iter() if not val1 then return end n = n - 1 return val1,val2 end end --- skip the first n values of a sequence -- @param iter a sequence of one or more values -- @param n number of items to skip function seq.skip (iter,n) n = n or 1 for i = 1,n do if iter() == nil then return list{} end end return iter end --- a sequence with a sequence count and the original value. -- enum(copy(ls)) is a roundabout way of saying ipairs(ls). -- @param iter a single or double valued sequence -- @return sequence of (i,v), i = 1..n and v is from iter. function seq.enum (iter) local i = 0 iter = default_iter(iter) return function () local val1,val2 = iter() if not val1 then return end i = i + 1 return i,val1,val2 end end --- map using a named method over a sequence. -- @param iter a sequence -- @param name the method name -- @param arg1 optional first extra argument -- @param arg2 optional second extra argument function seq.mapmethod (iter,name,arg1,arg2) iter = default_iter(iter) return function() local val = iter() if not val then return end local fn = val[name] if not fn then error(type(val).." does not have method "..name) end return fn(val,arg1,arg2) end end --- a sequence of (last,current) values from another sequence. -- This will return S(i-1),S(i) if given S(i) -- @param iter a sequence function seq.last (iter) iter = default_iter(iter) local val, l = iter(), nil if val == nil then return list{} end return function () val,l = iter(),val if val == nil then return nil end return val,l end end --- call the function on each element of the sequence. -- @param iter a sequence with up to 3 values -- @param fn a function function seq.foreach(iter,fn) fn = function_arg(2,fn) for i1,i2,i3 in default_iter(iter) do fn(i1,i2,i3) end end ---------------------- Sequence Adapters --------------------- local SMT local function SW (iter,...) if callable(iter) then return setmetatable({iter=iter},SMT) else return iter,... end end -- can't directly look these up in seq because of the wrong argument order... local map,reduce,mapmethod = seq.map, seq.reduce, seq.mapmethod local overrides = { map = function(self,fun,arg) return map(fun,self,arg) end, reduce = function(self,fun,initval) return reduce(fun,self,initval) end } SMT = { __index = function (tbl,key) local fn = overrides[key] or seq[key] if fn then return function(sw,...) return SW(fn(sw.iter,...)) end else return function(sw,...) return SW(mapmethod(sw.iter,key,...)) end end end, __call = function (sw) return sw.iter() end, } setmetatable(seq,{ __call = function(tbl,iter,extra) if not callable(iter) then if type(iter) == 'table' then iter = seq.list(iter) else return iter end end if extra then return setmetatable({iter=function() return iter(extra) end},SMT) else return setmetatable({iter=iter},SMT) end end }) --- create a wrapped iterator over all lines in the file. -- @param f either a filename, file-like object, or 'STDIN' (for standard input) -- @param ... for Lua 5.2 only, optional format specifiers, as in `io.read`. -- @return a sequence wrapper function seq.lines (f,...) local iter,obj if f == 'STDIN' then f = io.stdin elseif type(f) == 'string' then iter,obj = io.lines(f,...) elseif not f.read then error("Pass either a string or a file-like object",2) end if not iter then iter,obj = f:lines(...) end if obj then -- LuaJIT version returns a function operating on a file local lines,file = iter,obj iter = function() return lines(file) end end return SW(iter) end function seq.import () debug.setmetatable(function() end,{ __index = function(tbl,key) local s = overrides[key] or seq[key] if s then return s else return function(s,...) return seq.mapmethod(s,key,...) end end end }) end return seq Penlight-1.12.0/lua/pl/sip.lua000066400000000000000000000240701416703176500160540ustar00rootroot00000000000000--- Simple Input Patterns (SIP). -- SIP patterns start with '$', then a -- one-letter type, and then an optional variable in curly braces. -- -- sip.match('$v=$q','name="dolly"',res) -- ==> res=={'name','dolly'} -- sip.match('($q{first},$q{second})','("john","smith")',res) -- ==> res=={second='smith',first='john'} -- -- Type names: -- -- v identifier -- i integer -- f floating-point -- q quoted string -- ([{< match up to closing bracket -- -- See @{08-additional.md.Simple_Input_Patterns|the Guide} -- -- @module pl.sip local loadstring = rawget(_G,'loadstring') or load local unpack = rawget(_G,'unpack') or rawget(table,'unpack') local append,concat = table.insert,table.concat local ipairs,type = ipairs,type local io,_G = io,_G local print,rawget = print,rawget local patterns = { FLOAT = '[%+%-%d]%d*%.?%d*[eE]?[%+%-]?%d*', INTEGER = '[+%-%d]%d*', IDEN = '[%a_][%w_]*', OPTION = '[%a_][%w_%-]*', } local function assert_arg(idx,val,tp) if type(val) ~= tp then error("argument "..idx.." must be "..tp, 2) end end local sip = {} local brackets = {['<'] = '>', ['('] = ')', ['{'] = '}', ['['] = ']' } local stdclasses = {a=1,c=0,d=1,l=1,p=0,u=1,w=1,x=1,s=0} local function group(s) return '('..s..')' end -- escape all magic characters except $, which has special meaning -- Also, un-escape any characters after $, so $( and $[ passes through as is. local function escape (spec) return (spec:gsub('[%-%.%+%[%]%(%)%^%%%?%*]','%%%0'):gsub('%$%%(%S)','$%1')) end -- Most spaces within patterns can match zero or more spaces. -- Spaces between alphanumeric characters or underscores or between -- patterns that can match these characters, however, must match at least -- one space. Otherwise '$v $v' would match 'abcd' as {'abc', 'd'}. -- This function replaces continuous spaces within a pattern with either -- '%s*' or '%s+' according to this rule. The pattern has already -- been stripped of pattern names by now. local function compress_spaces(patt) return (patt:gsub("()%s+()", function(i1, i2) local before = patt:sub(i1 - 2, i1 - 1) if before:match('%$[vifadxlu]') or before:match('^[^%$]?[%w_]$') then local after = patt:sub(i2, i2 + 1) if after:match('%$[vifadxlu]') or after:match('^[%w_]') then return '%s+' end end return '%s*' end)) end local pattern_map = { v = group(patterns.IDEN), i = group(patterns.INTEGER), f = group(patterns.FLOAT), o = group(patterns.OPTION), r = '(%S.*)', p = '([%a]?[:]?[\\/%.%w_]+)' } function sip.custom_pattern(flag,patt) pattern_map[flag] = patt end --- convert a SIP pattern into the equivalent Lua string pattern. -- @param spec a SIP pattern -- @param options a table; only the at_start field is -- currently meaningful and ensures that the pattern is anchored -- at the start of the string. -- @return a Lua string pattern. function sip.create_pattern (spec,options) assert_arg(1,spec,'string') local fieldnames,fieldtypes = {},{} if type(spec) == 'string' then spec = escape(spec) else local res = {} for i,s in ipairs(spec) do res[i] = escape(s) end spec = concat(res,'.-') end local kount = 1 local function addfield (name,type) name = name or kount append(fieldnames,name) fieldtypes[name] = type kount = kount + 1 end local named_vars = spec:find('{%a+}') if options and options.at_start then spec = '^'..spec end if spec:sub(-1,-1) == '$' then spec = spec:sub(1,-2)..'$r' if named_vars then spec = spec..'{rest}' end end local names if named_vars then names = {} spec = spec:gsub('{(%a+)}',function(name) append(names,name) return '' end) end spec = compress_spaces(spec) local k = 1 local err local r = (spec:gsub('%$%S',function(s) local type,name type = s:sub(2,2) if names then name = names[k]; k=k+1 end -- this kludge is necessary because %q generates two matches, and -- we want to ignore the first. Not a problem for named captures. if not names and type == 'q' then addfield(nil,'Q') else addfield(name,type) end local res if pattern_map[type] then res = pattern_map[type] elseif type == 'q' then -- some Lua pattern matching voodoo; we want to match '...' as -- well as "...", and can use the fact that %n will match a -- previous capture. Adding the extra field above comes from needing -- to accommodate the extra spurious match (which is either ' or ") addfield(name,type) res = '(["\'])(.-)%'..(kount-2) else local endbracket = brackets[type] if endbracket then res = '(%b'..type..endbracket..')' elseif stdclasses[type] or stdclasses[type:lower()] then res = '(%'..type..'+)' else err = "unknown format type or character class" end end return res end)) if err then return nil,err else return r,fieldnames,fieldtypes end end local function tnumber (s) return s == 'd' or s == 'i' or s == 'f' end function sip.create_spec_fun(spec,options) local fieldtypes,fieldnames local ls = {} spec,fieldnames,fieldtypes = sip.create_pattern(spec,options) if not spec then return spec,fieldnames end local named_vars = type(fieldnames[1]) == 'string' for i = 1,#fieldnames do append(ls,'mm'..i) end ls[1] = ls[1] or "mm1" -- behave correctly if there are no patterns local fun = ('return (function(s,res)\n\tlocal %s = s:match(%q)\n'):format(concat(ls,','),spec) fun = fun..'\tif not mm1 then return false end\n' local k=1 for i,f in ipairs(fieldnames) do if f ~= '_' then local var = 'mm'..i if tnumber(fieldtypes[f]) then var = 'tonumber('..var..')' elseif brackets[fieldtypes[f]] then var = var..':sub(2,-2)' end if named_vars then fun = ('%s\tres.%s = %s\n'):format(fun,f,var) else if fieldtypes[f] ~= 'Q' then -- we skip the string-delim capture fun = ('%s\tres[%d] = %s\n'):format(fun,k,var) k = k + 1 end end end end return fun..'\treturn true\nend)\n', named_vars end --- convert a SIP pattern into a matching function. -- The returned function takes two arguments, the line and an empty table. -- If the line matched the pattern, then this function returns true -- and the table is filled with field-value pairs. -- @param spec a SIP pattern -- @param options optional table; {at_start=true} ensures that the pattern -- is anchored at the start of the string. -- @return a function if successful, or nil,error function sip.compile(spec,options) assert_arg(1,spec,'string') local fun,names = sip.create_spec_fun(spec,options) if not fun then return nil,names end if rawget(_G,'_DEBUG') then print(fun) end local chunk,err = loadstring(fun,'tmp') if err then return nil,err end return chunk(),names end local cache = {} --- match a SIP pattern against a string. -- @param spec a SIP pattern -- @param line a string -- @param res a table to receive values -- @param options (optional) option table -- @return true or false function sip.match (spec,line,res,options) assert_arg(1,spec,'string') assert_arg(2,line,'string') assert_arg(3,res,'table') if not cache[spec] then cache[spec] = sip.compile(spec,options) end return cache[spec](line,res) end --- match a SIP pattern against the start of a string. -- @param spec a SIP pattern -- @param line a string -- @param res a table to receive values -- @return true or false function sip.match_at_start (spec,line,res) return sip.match(spec,line,res,{at_start=true}) end --- given a pattern and a file object, return an iterator over the results -- @param spec a SIP pattern -- @param f a file-like object. function sip.fields (spec,f) assert_arg(1,spec,'string') if not f then return nil,"no file object" end local fun,err = sip.compile(spec) if not fun then return nil,err end local res = {} return function() while true do local line = f:read() if not line then return end if fun(line,res) then local values = res res = {} return unpack(values) end end end end local read_patterns = {} --- register a match which will be used in the read function. -- @string spec a SIP pattern -- @func fun a function to be called with the results of the match -- @see read function sip.pattern (spec,fun) assert_arg(1,spec,'string') local pat,named = sip.compile(spec) append(read_patterns,{pat=pat,named=named,callback=fun}) end --- enter a loop which applies all registered matches to the input file. -- @param f a file-like object -- @array matches optional list of `{spec,fun}` pairs, as for `pattern` above. function sip.read (f,matches) local owned,err if not f then return nil,"no file object" end if type(f) == 'string' then f,err = io.open(f) if not f then return nil,err end owned = true end if matches then for _,p in ipairs(matches) do sip.pattern(p[1],p[2]) end end local res = {} for line in f:lines() do for _,item in ipairs(read_patterns) do if item.pat(line,res) then if item.callback then if item.named then item.callback(res) else item.callback(unpack(res)) end end res = {} break end end end if owned then f:close() end end return sip Penlight-1.12.0/lua/pl/strict.lua000066400000000000000000000105151416703176500165700ustar00rootroot00000000000000--- Checks uses of undeclared global variables. -- All global variables must be 'declared' through a regular assignment -- (even assigning `nil` will do) in a main chunk before being used -- anywhere or assigned to inside a function. Existing metatables `__newindex` and `__index` -- metamethods are respected. -- -- You can set any table to have strict behaviour using `strict.module`. Creating a new -- module with `strict.closed_module` makes the module immune to monkey-patching, if -- you don't wish to encourage monkey business. -- -- If the global `PENLIGHT_NO_GLOBAL_STRICT` is defined, then this module won't make the -- global environment strict - if you just want to explicitly set table strictness. -- -- @module pl.strict require 'debug' -- for Lua 5.2 local getinfo, error, rawset, rawget = debug.getinfo, error, rawset, rawget local strict = {} local function what () local d = getinfo(3, "S") return d and d.what or "C" end --- make an existing table strict. -- @string[opt] name name of table -- @tab[opt] mod the table to protect - if `nil` then we'll return a new table -- @tab[opt] predeclared - table of variables that are to be considered predeclared. -- @return the given table, or a new table -- @usage -- local M = { hello = "world" } -- strict.module ("Awesome_Module", M, { -- Lua = true, -- defines allowed keys -- }) -- -- assert(M.hello == "world") -- assert(M.Lua == nil) -- access allowed, but has no value yet -- M.Lua = "Rocks" -- assert(M.Lua == "Rocks") -- M.not_allowed = "bad boy" -- throws an error function strict.module (name,mod,predeclared) local mt, old_newindex, old_index, old_index_type, global if predeclared then global = predeclared.__global end if type(mod) == 'table' then mt = getmetatable(mod) if mt and rawget(mt,'__declared') then return end -- already patched... else mod = {} end if mt == nil then mt = {} setmetatable(mod, mt) else old_newindex = mt.__newindex old_index = mt.__index old_index_type = type(old_index) end mt.__declared = predeclared or {} mt.__newindex = function(t, n, v) if old_newindex then old_newindex(t, n, v) if rawget(t,n)~=nil then return end end if not mt.__declared[n] then if global then local w = what() if w ~= "main" and w ~= "C" then error("assign to undeclared global '"..n.."'", 2) end end mt.__declared[n] = true end rawset(t, n, v) end mt.__index = function(t,n) if not mt.__declared[n] and what() ~= "C" then if old_index then if old_index_type == "table" then local fallback = old_index[n] if fallback ~= nil then return fallback end else local res = old_index(t, n) if res ~= nil then return res end end end local msg = "variable '"..n.."' is not declared" if name then msg = msg .. " in '"..tostring(name).."'" end error(msg, 2) end return rawget(t, n) end return mod end --- make all tables in a table strict. -- So `strict.make_all_strict(_G)` prevents monkey-patching -- of any global table -- @tab T the table containing the tables to protect. Table `T` itself will NOT be protected. function strict.make_all_strict (T) for k,v in pairs(T) do if type(v) == 'table' and v ~= T then strict.module(k,v) end end end --- make a new module table which is closed to further changes. -- @tab mod module table -- @string name module name function strict.closed_module (mod,name) -- No clue to what this is useful for? see tests -- Deprecate this and remove??? local M = {} mod = mod or {} local mt = getmetatable(mod) if not mt then mt = {} setmetatable(mod,mt) end mt.__newindex = function(t,k,v) M[k] = v end return strict.module(name,M) end if not rawget(_G,'PENLIGHT_NO_GLOBAL_STRICT') then strict.module(nil,_G,{_PROMPT=true,_PROMPT2=true,__global=true}) end return strict Penlight-1.12.0/lua/pl/stringio.lua000066400000000000000000000072761416703176500171300ustar00rootroot00000000000000--- Reading and writing strings using file-like objects.
-- -- f = stringio.open(text) -- l1 = f:read() -- read first line -- n,m = f:read ('*n','*n') -- read two numbers -- for line in f:lines() do print(line) end -- iterate over all lines -- f = stringio.create() -- f:write('hello') -- f:write('dolly') -- assert(f:value(),'hellodolly') -- -- See @{03-strings.md.File_style_I_O_on_Strings|the Guide}. -- @module pl.stringio local unpack = rawget(_G,'unpack') or rawget(table,'unpack') local tonumber = tonumber local concat,append = table.concat,table.insert local stringio = {} -- Writer class local SW = {} SW.__index = SW local function xwrite(self,...) local args = {...} --arguments may not be nil! for i = 1, #args do append(self.tbl,args[i]) end end function SW:write(arg1,arg2,...) if arg2 then xwrite(self,arg1,arg2,...) else append(self.tbl,arg1) end end function SW:writef(fmt,...) self:write(fmt:format(...)) end function SW:value() return concat(self.tbl) end function SW:__tostring() return self:value() end function SW:close() -- for compatibility only end function SW:seek() end -- Reader class local SR = {} SR.__index = SR function SR:_read(fmt) local i,str = self.i,self.str local sz = #str if i > sz then return nil end local res if fmt == '*l' or fmt == '*L' then local idx = str:find('\n',i) or (sz+1) res = str:sub(i,fmt == '*l' and idx-1 or idx) self.i = idx+1 elseif fmt == '*a' then res = str:sub(i) self.i = sz elseif fmt == '*n' then local _,i2,idx _,idx = str:find ('%s*%d+',i) _,i2 = str:find ('^%.%d+',idx+1) if i2 then idx = i2 end _,i2 = str:find ('^[eE][%+%-]*%d+',idx+1) if i2 then idx = i2 end local val = str:sub(i,idx) res = tonumber(val) self.i = idx+1 elseif type(fmt) == 'number' then res = str:sub(i,i+fmt-1) self.i = i + fmt else error("bad read format",2) end return res end function SR:read(...) if select('#',...) == 0 then return self:_read('*l') else local res, fmts = {},{...} for i = 1, #fmts do res[i] = self:_read(fmts[i]) end return unpack(res) end end function SR:seek(whence,offset) local base whence = whence or 'cur' offset = offset or 0 if whence == 'set' then base = 1 elseif whence == 'cur' then base = self.i elseif whence == 'end' then base = #self.str end self.i = base + offset return self.i end function SR:lines(...) local n, args = select('#',...) if n > 0 then args = {...} end return function() if n == 0 then return self:_read '*l' else return self:read(unpack(args)) end end end function SR:close() -- for compatibility only end --- create a file-like object which can be used to construct a string. -- The resulting object has an extra `value()` method for -- retrieving the string value. Implements `file:write`, `file:seek`, `file:lines`, -- plus an extra `writef` method which works like `utils.printf`. -- @usage f = create(); f:write('hello, dolly\n'); print(f:value()) function stringio.create() return setmetatable({tbl={}},SW) end --- create a file-like object for reading from a given string. -- Implements `file:read`. -- @string s The input string. -- @usage fs = open '20 10'; x,y = f:read ('*n','*n'); assert(x == 20 and y == 10) function stringio.open(s) return setmetatable({str=s,i=1},SR) end function stringio.lines(s,...) return stringio.open(s):lines(...) end return stringio Penlight-1.12.0/lua/pl/stringx.lua000066400000000000000000000630461416703176500167650ustar00rootroot00000000000000--- Python-style extended string library. -- -- see 3.6.1 of the Python reference. -- If you want to make these available as string methods, then say -- `stringx.import()` to bring them into the standard `string` table. -- -- See @{03-strings.md|the Guide} -- -- Dependencies: `pl.utils`, `pl.types` -- @module pl.stringx local utils = require 'pl.utils' local is_callable = require 'pl.types'.is_callable local string = string local find = string.find local type,setmetatable,ipairs = type,setmetatable,ipairs local error = error local gsub = string.gsub local rep = string.rep local sub = string.sub local reverse = string.reverse local concat = table.concat local append = table.insert local remove = table.remove local escape = utils.escape local ceil, max = math.ceil, math.max local assert_arg,usplit = utils.assert_arg,utils.split local lstrip local unpack = utils.unpack local pack = utils.pack local function assert_string (n,s) assert_arg(n,s,'string') end local function non_empty(s) return #s > 0 end local function assert_nonempty_string(n,s) assert_arg(n,s,'string',non_empty,'must be a non-empty string') end local function makelist(l) return setmetatable(l, require('pl.List')) end local stringx = {} ------------------ -- String Predicates -- @section predicates --- does s only contain alphabetic characters? -- @string s a string function stringx.isalpha(s) assert_string(1,s) return find(s,'^%a+$') == 1 end --- does s only contain digits? -- @string s a string function stringx.isdigit(s) assert_string(1,s) return find(s,'^%d+$') == 1 end --- does s only contain alphanumeric characters? -- @string s a string function stringx.isalnum(s) assert_string(1,s) return find(s,'^%w+$') == 1 end --- does s only contain whitespace? -- Matches on pattern '%s' so matches space, newline, tabs, etc. -- @string s a string function stringx.isspace(s) assert_string(1,s) return find(s,'^%s+$') == 1 end --- does s only contain lower case characters? -- @string s a string function stringx.islower(s) assert_string(1,s) return find(s,'^[%l%s]+$') == 1 end --- does s only contain upper case characters? -- @string s a string function stringx.isupper(s) assert_string(1,s) return find(s,'^[%u%s]+$') == 1 end local function raw_startswith(s, prefix) return find(s,prefix,1,true) == 1 end local function raw_endswith(s, suffix) return #s >= #suffix and find(s, suffix, #s-#suffix+1, true) and true or false end local function test_affixes(s, affixes, fn) if type(affixes) == 'string' then return fn(s,affixes) elseif type(affixes) == 'table' then for _,affix in ipairs(affixes) do if fn(s,affix) then return true end end return false else error(("argument #2 expected a 'string' or a 'table', got a '%s'"):format(type(affixes))) end end --- does s start with prefix or one of prefixes? -- @string s a string -- @param prefix a string or an array of strings function stringx.startswith(s,prefix) assert_string(1,s) return test_affixes(s,prefix,raw_startswith) end --- does s end with suffix or one of suffixes? -- @string s a string -- @param suffix a string or an array of strings function stringx.endswith(s,suffix) assert_string(1,s) return test_affixes(s,suffix,raw_endswith) end --- Strings and Lists -- @section lists --- concatenate the strings using this string as a delimiter. -- Note that the arguments are reversed from `string.concat`. -- @string s the string -- @param seq a table of strings or numbers -- @usage stringx.join(' ', {1,2,3}) == '1 2 3' function stringx.join(s,seq) assert_string(1,s) return concat(seq,s) end --- Split a string into a list of lines. -- `"\r"`, `"\n"`, and `"\r\n"` are considered line ends. -- They are not included in the lines unless `keepends` is passed. -- Terminal line end does not produce an extra line. -- Splitting an empty string results in an empty list. -- @string s the string. -- @bool[opt] keep_ends include line ends. -- @return List of lines function stringx.splitlines(s, keep_ends) assert_string(1, s) local res = {} local pos = 1 while true do local line_end_pos = find(s, '[\r\n]', pos) if not line_end_pos then break end local line_end = sub(s, line_end_pos, line_end_pos) if line_end == '\r' and sub(s, line_end_pos + 1, line_end_pos + 1) == '\n' then line_end = '\r\n' end local line = sub(s, pos, line_end_pos - 1) if keep_ends then line = line .. line_end end append(res, line) pos = line_end_pos + #line_end end if pos <= #s then append(res, sub(s, pos)) end return makelist(res) end --- split a string into a list of strings using a delimiter. -- @function split -- @string s the string -- @string[opt] re a delimiter (defaults to whitespace) -- @int[opt] n maximum number of results -- @return List -- @usage #(stringx.split('one two')) == 2 -- @usage stringx.split('one,two,three', ',') == List{'one','two','three'} -- @usage stringx.split('one,two,three', ',', 2) == List{'one','two,three'} function stringx.split(s,re,n) assert_string(1,s) local plain = true if not re then -- default spaces s = lstrip(s) plain = false end local res = usplit(s,re,plain,n) if re and re ~= '' and find(s,re,-#re,true) and (n or math.huge) > #res then res[#res+1] = "" end return makelist(res) end --- replace all tabs in s with tabsize spaces. If not specified, tabsize defaults to 8. -- Tab stops will be honored. -- @string s the string -- @int tabsize[opt=8] number of spaces to expand each tab -- @return expanded string -- @usage stringx.expandtabs('\tone,two,three', 4) == ' one,two,three' -- @usage stringx.expandtabs(' \tone,two,three', 4) == ' one,two,three' function stringx.expandtabs(s,tabsize) assert_string(1,s) tabsize = tabsize or 8 return (s:gsub("([^\t\r\n]*)\t", function(before_tab) if tabsize == 0 then return before_tab else return before_tab .. (" "):rep(tabsize - #before_tab % tabsize) end end)) end --- Finding and Replacing -- @section find local function _find_all(s,sub,first,last,allow_overlap) first = first or 1 last = last or #s if sub == '' then return last+1,last-first+1 end local i1,i2 = find(s,sub,first,true) local res local k = 0 while i1 do if last and i2 > last then break end res = i1 k = k + 1 if allow_overlap then i1,i2 = find(s,sub,i1+1,true) else i1,i2 = find(s,sub,i2+1,true) end end return res,k end --- find index of first instance of sub in s from the left. -- @string s the string -- @string sub substring -- @int[opt] first first index -- @int[opt] last last index -- @return start index, or nil if not found function stringx.lfind(s,sub,first,last) assert_string(1,s) assert_string(2,sub) local i1, i2 = find(s,sub,first,true) if i1 and (not last or i2 <= last) then return i1 else return nil end end --- find index of first instance of sub in s from the right. -- @string s the string -- @string sub substring -- @int[opt] first first index -- @int[opt] last last index -- @return start index, or nil if not found function stringx.rfind(s,sub,first,last) assert_string(1,s) assert_string(2,sub) return (_find_all(s,sub,first,last,true)) end --- replace up to n instances of old by new in the string s. -- If n is not present, replace all instances. -- @string s the string -- @string old the target substring -- @string new the substitution -- @int[opt] n optional maximum number of substitutions -- @return result string function stringx.replace(s,old,new,n) assert_string(1,s) assert_string(2,old) assert_string(3,new) return (gsub(s,escape(old),new:gsub('%%','%%%%'),n)) end --- count all instances of substring in string. -- @string s the string -- @string sub substring -- @bool[opt] allow_overlap allow matches to overlap -- @usage -- assert(stringx.count('banana', 'ana') == 1) -- assert(stringx.count('banana', 'ana', true) == 2) function stringx.count(s,sub,allow_overlap) assert_string(1,s) local _,k = _find_all(s,sub,1,false,allow_overlap) return k end --- Stripping and Justifying -- @section strip local function _just(s,w,ch,left,right) local n = #s if w > n then if not ch then ch = ' ' end local f1,f2 if left and right then local rn = ceil((w-n)/2) local ln = w - n - rn f1 = rep(ch,ln) f2 = rep(ch,rn) elseif right then f1 = rep(ch,w-n) f2 = '' else f2 = rep(ch,w-n) f1 = '' end return f1..s..f2 else return s end end --- left-justify s with width w. -- @string s the string -- @int w width of justification -- @string[opt=' '] ch padding character -- @usage stringx.ljust('hello', 10, '*') == '*****hello' function stringx.ljust(s,w,ch) assert_string(1,s) assert_arg(2,w,'number') return _just(s,w,ch,true,false) end --- right-justify s with width w. -- @string s the string -- @int w width of justification -- @string[opt=' '] ch padding character -- @usage stringx.rjust('hello', 10, '*') == 'hello*****' function stringx.rjust(s,w,ch) assert_string(1,s) assert_arg(2,w,'number') return _just(s,w,ch,false,true) end --- center-justify s with width w. -- @string s the string -- @int w width of justification -- @string[opt=' '] ch padding character -- @usage stringx.center('hello', 10, '*') == '**hello***' function stringx.center(s,w,ch) assert_string(1,s) assert_arg(2,w,'number') return _just(s,w,ch,true,true) end local function _strip(s,left,right,chrs) if not chrs then chrs = '%s' else chrs = '['..escape(chrs)..']' end local f = 1 local t if left then local i1,i2 = find(s,'^'..chrs..'*') if i2 >= i1 then f = i2+1 end end if right then if #s < 200 then local i1,i2 = find(s,chrs..'*$',f) if i2 >= i1 then t = i1-1 end else local rs = reverse(s) local i1,i2 = find(rs, '^'..chrs..'*') if i2 >= i1 then t = -i2-1 end end end return sub(s,f,t) end --- trim any characters on the left of s. -- @string s the string -- @string[opt='%s'] chrs default any whitespace character, -- but can be a string of characters to be trimmed function stringx.lstrip(s,chrs) assert_string(1,s) return _strip(s,true,false,chrs) end lstrip = stringx.lstrip --- trim any characters on the right of s. -- @string s the string -- @string[opt='%s'] chrs default any whitespace character, -- but can be a string of characters to be trimmed function stringx.rstrip(s,chrs) assert_string(1,s) return _strip(s,false,true,chrs) end --- trim any characters on both left and right of s. -- @string s the string -- @string[opt='%s'] chrs default any whitespace character, -- but can be a string of characters to be trimmed -- @usage stringx.strip(' --== Hello ==-- ', "- =") --> 'Hello' function stringx.strip(s,chrs) assert_string(1,s) return _strip(s,true,true,chrs) end --- Partitioning Strings -- @section partitioning --- split a string using a pattern. Note that at least one value will be returned! -- @string s the string -- @string[opt='%s'] re a Lua string pattern (defaults to whitespace) -- @return the parts of the string -- @usage a,b = line:splitv('=') -- @see utils.splitv function stringx.splitv(s,re) assert_string(1,s) return utils.splitv(s,re) end -- The partition functions split a string using a delimiter into three parts: -- the part before, the delimiter itself, and the part afterwards local function _partition(p,delim,fn) local i1,i2 = fn(p,delim) if not i1 or i1 == -1 then return p,'','' else if not i2 then i2 = i1 end return sub(p,1,i1-1),sub(p,i1,i2),sub(p,i2+1) end end --- partition the string using first occurance of a delimiter -- @string s the string -- @string ch delimiter (match as plain string, no patterns) -- @return part before ch -- @return ch -- @return part after ch -- @usage {stringx.partition('a,b,c', ','))} == {'a', ',', 'b,c'} -- @usage {stringx.partition('abc', 'x'))} == {'abc', '', ''} function stringx.partition(s,ch) assert_string(1,s) assert_nonempty_string(2,ch) return _partition(s,ch,stringx.lfind) end --- partition the string p using last occurance of a delimiter -- @string s the string -- @string ch delimiter (match as plain string, no patterns) -- @return part before ch -- @return ch -- @return part after ch -- @usage {stringx.rpartition('a,b,c', ','))} == {'a,b', ',', 'c'} -- @usage {stringx.rpartition('abc', 'x'))} == {'', '', 'abc'} function stringx.rpartition(s,ch) assert_string(1,s) assert_nonempty_string(2,ch) local a,b,c = _partition(s,ch,stringx.rfind) if a == s then -- no match found return c,b,a end return a,b,c end --- return the 'character' at the index. -- @string s the string -- @int idx an index (can be negative) -- @return a substring of length 1 if successful, empty string otherwise. function stringx.at(s,idx) assert_string(1,s) assert_arg(2,idx,'number') return sub(s,idx,idx) end --- Text handling -- @section text --- indent a multiline string. -- @tparam string s the (multiline) string -- @tparam integer n the size of the indent -- @tparam[opt=' '] string ch the character to use when indenting -- @return indented string function stringx.indent (s,n,ch) assert_arg(1,s,'string') assert_arg(2,n,'number') local lines = usplit(s ,'\n') local prefix = string.rep(ch or ' ',n) for i, line in ipairs(lines) do lines[i] = prefix..line end return concat(lines,'\n')..'\n' end --- dedent a multiline string by removing any initial indent. -- useful when working with [[..]] strings. -- Empty lines are ignored. -- @tparam string s the (multiline) string -- @return a string with initial indent zero. -- @usage -- local s = dedent [[ -- One -- -- Two -- -- Three -- ]] -- assert(s == [[ -- One -- -- Two -- -- Three -- ]]) function stringx.dedent (s) assert_arg(1,s,'string') local lst = usplit(s,'\n') if #lst>0 then local ind_size = math.huge for i, line in ipairs(lst) do local i1, i2 = lst[i]:find('^%s*[^%s]') if i1 and i2 < ind_size then ind_size = i2 end end for i, line in ipairs(lst) do lst[i] = lst[i]:sub(ind_size, -1) end end return concat(lst,'\n')..'\n' end do local buildline = function(words, size, breaklong) -- if overflow is set, a word longer than size, will overflow the size -- otherwise it will be chopped in line-length pieces local line = {} if #words[1] > size then -- word longer than line if not breaklong then line[1] = words[1] remove(words, 1) else line[1] = words[1]:sub(1, size) words[1] = words[1]:sub(size + 1, -1) end else local len = 0 while words[1] and (len + #words[1] <= size) or (len == 0 and #words[1] == size) do if words[1] ~= "" then line[#line+1] = words[1] len = len + #words[1] + 1 end remove(words, 1) end end return stringx.strip(concat(line, " ")), words end --- format a paragraph into lines so that they fit into a line width. -- It will not break long words by default, so lines can be over the length -- to that extent. -- @tparam string s the string to format -- @tparam[opt=70] integer width the margin width -- @tparam[opt=false] boolean breaklong if truthy, words longer than the width given will be forced split. -- @return a list of lines (List object), use `fill` to return a string instead of a `List`. -- @see pl.List -- @see fill stringx.wrap = function(s, width, breaklong) s = s:gsub('\n',' ') -- remove line breaks s = stringx.strip(s) -- remove leading/trailing whitespace if s == "" then return { "" } end width = width or 70 local out = {} local words = usplit(s, "%s") while words[1] do out[#out+1], words = buildline(words, width, breaklong) end return makelist(out) end end --- format a paragraph so that it fits into a line width. -- @tparam string s the string to format -- @tparam[opt=70] integer width the margin width -- @tparam[opt=false] boolean breaklong if truthy, words longer than the width given will be forced split. -- @return a string, use `wrap` to return a list of lines instead of a string. -- @see wrap function stringx.fill (s,width,breaklong) return concat(stringx.wrap(s,width,breaklong),'\n') .. '\n' end --- Template -- @section Template local function _substitute(s,tbl,safe) local subst if is_callable(tbl) then subst = tbl else function subst(f) local s = tbl[f] if not s then if safe then return f else error("not present in table "..f) end else return s end end end local res = gsub(s,'%${([%w_]+)}',subst) return (gsub(res,'%$([%w_]+)',subst)) end local Template = {} stringx.Template = Template Template.__index = Template setmetatable(Template, { __call = function(obj,tmpl) return Template.new(tmpl) end }) --- Creates a new Template class. -- This is a shortcut to `Template.new(tmpl)`. -- @tparam string tmpl the template string -- @function Template -- @treturn Template function Template.new(tmpl) assert_arg(1,tmpl,'string') local res = {} res.tmpl = tmpl setmetatable(res,Template) return res end --- substitute values into a template, throwing an error. -- This will throw an error if no name is found. -- @tparam table tbl a table of name-value pairs. -- @return string with place holders substituted function Template:substitute(tbl) assert_arg(1,tbl,'table') return _substitute(self.tmpl,tbl,false) end --- substitute values into a template. -- This version just passes unknown names through. -- @tparam table tbl a table of name-value pairs. -- @return string with place holders substituted function Template:safe_substitute(tbl) assert_arg(1,tbl,'table') return _substitute(self.tmpl,tbl,true) end --- substitute values into a template, preserving indentation.
-- If the value is a multiline string _or_ a template, it will insert -- the lines at the correct indentation.
-- Furthermore, if a template, then that template will be substituted -- using the same table. -- @tparam table tbl a table of name-value pairs. -- @return string with place holders substituted function Template:indent_substitute(tbl) assert_arg(1,tbl,'table') if not self.strings then self.strings = usplit(self.tmpl,'\n') end -- the idea is to substitute line by line, grabbing any spaces as -- well as the $var. If the value to be substituted contains newlines, -- then we split that into lines and adjust the indent before inserting. local function subst(line) return line:gsub('(%s*)%$([%w_]+)',function(sp,f) local subtmpl local s = tbl[f] if not s then error("not present in table "..f) end if getmetatable(s) == Template then subtmpl = s s = s.tmpl else s = tostring(s) end if s:find '\n' then local lines = usplit(s, '\n') for i, line in ipairs(lines) do lines[i] = sp..line end s = concat(lines, '\n') .. '\n' end if subtmpl then return _substitute(s, tbl) else return s end end) end local lines = {} for i, line in ipairs(self.strings) do lines[i] = subst(line) end return concat(lines,'\n')..'\n' end --- Miscelaneous -- @section misc --- return an iterator over all lines in a string -- @string s the string -- @return an iterator -- @usage -- local line_no = 1 -- for line in stringx.lines(some_text) do -- print(line_no, line) -- line_no = line_no + 1 -- end function stringx.lines(s) assert_string(1,s) if not s:find '\n$' then s = s..'\n' end return s:gmatch('([^\n]*)\n') end --- inital word letters uppercase ('title case'). -- Here 'words' mean chunks of non-space characters. -- @string s the string -- @return a string with each word's first letter uppercase -- @usage stringx.title("hello world") == "Hello World") function stringx.title(s) assert_string(1,s) return (s:gsub('(%S)(%S*)',function(f,r) return f:upper()..r:lower() end)) end stringx.capitalize = stringx.title do local ellipsis = '...' local n_ellipsis = #ellipsis --- Return a shortened version of a string. -- Fits string within w characters. Removed characters are marked with ellipsis. -- @string s the string -- @int w the maxinum size allowed -- @bool tail true if we want to show the end of the string (head otherwise) -- @usage ('1234567890'):shorten(8) == '12345...' -- @usage ('1234567890'):shorten(8, true) == '...67890' -- @usage ('1234567890'):shorten(20) == '1234567890' function stringx.shorten(s,w,tail) assert_string(1,s) if #s > w then if w < n_ellipsis then return ellipsis:sub(1,w) end if tail then local i = #s - w + 1 + n_ellipsis return ellipsis .. s:sub(i) else return s:sub(1,w-n_ellipsis) .. ellipsis end end return s end end do -- Utility function that finds any patterns that match a long string's an open or close. -- Note that having this function use the least number of equal signs that is possible is a harder algorithm to come up with. -- Right now, it simply returns the greatest number of them found. -- @param s The string -- @return 'nil' if not found. If found, the maximum number of equal signs found within all matches. local function has_lquote(s) local lstring_pat = '([%[%]])(=*)%1' local equals, new_equals, _ local finish = 1 repeat _, finish, _, new_equals = s:find(lstring_pat, finish) if new_equals then equals = max(equals or 0, #new_equals) end until not new_equals return equals end --- Quote the given string and preserve any control or escape characters, such that reloading the string in Lua returns the same result. -- @param s The string to be quoted. -- @return The quoted string. function stringx.quote_string(s) assert_string(1,s) -- Find out if there are any embedded long-quote sequences that may cause issues. -- This is important when strings are embedded within strings, like when serializing. -- Append a closing bracket to catch unfinished long-quote sequences at the end of the string. local equal_signs = has_lquote(s .. "]") -- Note that strings containing "\r" can't be quoted using long brackets -- as Lua lexer converts all newlines to "\n" within long strings. if (s:find("\n") or equal_signs) and not s:find("\r") then -- If there is an embedded sequence that matches a long quote, then -- find the one with the maximum number of = signs and add one to that number. equal_signs = ("="):rep((equal_signs or -1) + 1) -- Long strings strip out leading newline. We want to retain that, when quoting. if s:find("^\n") then s = "\n" .. s end local lbracket, rbracket = "[" .. equal_signs .. "[", "]" .. equal_signs .. "]" s = lbracket .. s .. rbracket else -- Escape funny stuff. Lua 5.1 does not handle "\r" correctly. s = ("%q"):format(s):gsub("\r", "\\r") end return s end end --- Python-style formatting operator. -- Calling `text.format_operator()` overloads the % operator for strings to give -- Python/Ruby style formated output. -- This is extended to also do template-like substitution for map-like data. -- -- Note this goes further than the original, and will allow these cases: -- -- 1. a single value -- 2. a list of values -- 3. a map of var=value pairs -- 4. a function, as in gsub -- -- For the second two cases, it uses $-variable substituion. -- -- When called, this function will monkey-patch the global `string` metatable by -- adding a `__mod` method. -- -- See the lua-users wiki -- -- @usage -- require 'pl.text'.format_operator() -- local out1 = '%s = %5.3f' % {'PI',math.pi} --> 'PI = 3.142' -- local out2 = '$name = $value' % {name='dog',value='Pluto'} --> 'dog = Pluto' function stringx.format_operator() local format = string.format -- a more forgiving version of string.format, which applies -- tostring() to any value with a %s format. local function formatx (fmt,...) local args = pack(...) local i = 1 for p in fmt:gmatch('%%.') do if p == '%s' and type(args[i]) ~= 'string' then args[i] = tostring(args[i]) end i = i + 1 end return format(fmt,unpack(args)) end local function basic_subst(s,t) return (s:gsub('%$([%w_]+)',t)) end getmetatable("").__mod = function(a, b) if b == nil then return a elseif type(b) == "table" and getmetatable(b) == nil then if #b == 0 then -- assume a map-like table return _substitute(a,b,true) else return formatx(a,unpack(b)) end elseif type(b) == 'function' then return basic_subst(a,b) else return formatx(a,b) end end end --- import the stringx functions into the global string (meta)table function stringx.import() utils.import(stringx,string) end return stringx Penlight-1.12.0/lua/pl/tablex.lua000066400000000000000000000730671416703176500165520ustar00rootroot00000000000000--- Extended operations on Lua tables. -- -- See @{02-arrays.md.Useful_Operations_on_Tables|the Guide} -- -- Dependencies: `pl.utils`, `pl.types` -- @module pl.tablex local utils = require ('pl.utils') local types = require ('pl.types') local getmetatable,setmetatable,require = getmetatable,setmetatable,require local tsort,append,remove = table.sort,table.insert,table.remove local min = math.min local pairs,type,unpack,select,tostring = pairs,type,utils.unpack,select,tostring local function_arg = utils.function_arg local assert_arg = utils.assert_arg local tablex = {} -- generally, functions that make copies of tables try to preserve the metatable. -- However, when the source has no obvious type, then we attach appropriate metatables -- like List, Map, etc to the result. local function setmeta (res,tbl,pl_class) local mt = getmetatable(tbl) or pl_class and require('pl.' .. pl_class) return mt and setmetatable(res, mt) or res end local function makelist(l) return setmetatable(l, require('pl.List')) end local function makemap(m) return setmetatable(m, require('pl.Map')) end local function complain (idx,msg) error(('argument %d is not %s'):format(idx,msg),3) end local function assert_arg_indexable (idx,val) if not types.is_indexable(val) then complain(idx,"indexable") end end local function assert_arg_iterable (idx,val) if not types.is_iterable(val) then complain(idx,"iterable") end end local function assert_arg_writeable (idx,val) if not types.is_writeable(val) then complain(idx,"writeable") end end --- copy a table into another, in-place. -- @within Copying -- @tab t1 destination table -- @tab t2 source (actually any iterable object) -- @return first table function tablex.update (t1,t2) assert_arg_writeable(1,t1) assert_arg_iterable(2,t2) for k,v in pairs(t2) do t1[k] = v end return t1 end --- total number of elements in this table. -- Note that this is distinct from `#t`, which is the number -- of values in the array part; this value will always -- be greater or equal. The difference gives the size of -- the hash part, for practical purposes. Works for any -- object with a __pairs metamethod. -- @tab t a table -- @return the size function tablex.size (t) assert_arg_iterable(1,t) local i = 0 for k in pairs(t) do i = i + 1 end return i end --- make a shallow copy of a table -- @within Copying -- @tab t an iterable source -- @return new table function tablex.copy (t) assert_arg_iterable(1,t) local res = {} for k,v in pairs(t) do res[k] = v end return res end local function cycle_aware_copy(t, cache) if type(t) ~= 'table' then return t end if cache[t] then return cache[t] end assert_arg_iterable(1,t) local res = {} cache[t] = res local mt = getmetatable(t) for k,v in pairs(t) do k = cycle_aware_copy(k, cache) v = cycle_aware_copy(v, cache) res[k] = v end setmetatable(res,mt) return res end --- make a deep copy of a table, recursively copying all the keys and fields. -- This supports cycles in tables; cycles will be reproduced in the copy. -- This will also set the copied table's metatable to that of the original. -- @within Copying -- @tab t A table -- @return new table function tablex.deepcopy(t) return cycle_aware_copy(t,{}) end local abs = math.abs local function cycle_aware_compare(t1,t2,ignore_mt,eps,cache) if cache[t1] and cache[t1][t2] then return true end local ty1 = type(t1) local ty2 = type(t2) if ty1 ~= ty2 then return false end -- non-table types can be directly compared if ty1 ~= 'table' then if ty1 == 'number' and eps then return abs(t1-t2) < eps end return t1 == t2 end -- as well as tables which have the metamethod __eq local mt = getmetatable(t1) if not ignore_mt and mt and mt.__eq then return t1 == t2 end for k1 in pairs(t1) do if t2[k1]==nil then return false end end for k2 in pairs(t2) do if t1[k2]==nil then return false end end cache[t1] = cache[t1] or {} cache[t1][t2] = true for k1,v1 in pairs(t1) do local v2 = t2[k1] if not cycle_aware_compare(v1,v2,ignore_mt,eps,cache) then return false end end return true end --- compare two values. -- if they are tables, then compare their keys and fields recursively. -- @within Comparing -- @param t1 A value -- @param t2 A value -- @bool[opt] ignore_mt if true, ignore __eq metamethod (default false) -- @number[opt] eps if defined, then used for any number comparisons -- @return true or false function tablex.deepcompare(t1,t2,ignore_mt,eps) return cycle_aware_compare(t1,t2,ignore_mt,eps,{}) end --- compare two arrays using a predicate. -- @within Comparing -- @array t1 an array -- @array t2 an array -- @func cmp A comparison function; `bool = cmp(t1_value, t2_value)` -- @return true or false -- @usage -- assert(tablex.compare({ 1, 2, 3 }, { 1, 2, 3 }, "==")) -- -- assert(tablex.compare( -- {1,2,3, hello = "world"}, -- fields are not compared! -- {1,2,3}, function(v1, v2) return v1 == v2 end) function tablex.compare (t1,t2,cmp) assert_arg_indexable(1,t1) assert_arg_indexable(2,t2) if #t1 ~= #t2 then return false end cmp = function_arg(3,cmp) for k = 1,#t1 do if not cmp(t1[k],t2[k]) then return false end end return true end --- compare two list-like tables using an optional predicate, without regard for element order. -- @within Comparing -- @array t1 a list-like table -- @array t2 a list-like table -- @param cmp A comparison function (may be nil) function tablex.compare_no_order (t1,t2,cmp) assert_arg_indexable(1,t1) assert_arg_indexable(2,t2) if cmp then cmp = function_arg(3,cmp) end if #t1 ~= #t2 then return false end local visited = {} for i = 1,#t1 do local val = t1[i] local gotcha for j = 1,#t2 do if not visited[j] then local match if cmp then match = cmp(val,t2[j]) else match = val == t2[j] end if match then gotcha = j break end end end if not gotcha then return false end visited[gotcha] = true end return true end --- return the index of a value in a list. -- Like string.find, there is an optional index to start searching, -- which can be negative. -- @within Finding -- @array t A list-like table -- @param val A value -- @int idx index to start; -1 means last element,etc (default 1) -- @return index of value or nil if not found -- @usage find({10,20,30},20) == 2 -- @usage find({'a','b','a','c'},'a',2) == 3 function tablex.find(t,val,idx) assert_arg_indexable(1,t) idx = idx or 1 if idx < 0 then idx = #t + idx + 1 end for i = idx,#t do if t[i] == val then return i end end return nil end --- return the index of a value in a list, searching from the end. -- Like string.find, there is an optional index to start searching, -- which can be negative. -- @within Finding -- @array t A list-like table -- @param val A value -- @param idx index to start; -1 means last element,etc (default `#t`) -- @return index of value or nil if not found -- @usage rfind({10,10,10},10) == 3 function tablex.rfind(t,val,idx) assert_arg_indexable(1,t) idx = idx or #t if idx < 0 then idx = #t + idx + 1 end for i = idx,1,-1 do if t[i] == val then return i end end return nil end --- return the index (or key) of a value in a table using a comparison function. -- -- *NOTE*: the 2nd return value of this function, the value returned -- by the comparison function, has a limitation that it cannot be `false`. -- Because if it is, then it indicates the comparison failed, and the -- function will continue the search. See examples. -- @within Finding -- @tab t A table -- @func cmp A comparison function -- @param arg an optional second argument to the function -- @return index of value, or nil if not found -- @return value returned by comparison function (cannot be `false`!) -- @usage -- -- using an operator -- local lst = { "Rudolph", true, false, 15 } -- local idx, cmp_result = tablex.rfind(lst, "==", "Rudolph") -- assert(idx == 1) -- assert(cmp_result == true) -- -- local idx, cmp_result = tablex.rfind(lst, "==", false) -- assert(idx == 3) -- assert(cmp_result == true) -- looking up 'false' works! -- -- -- using a function returning the value looked up -- local cmp = function(v1, v2) return v1 == v2 and v2 end -- local idx, cmp_result = tablex.rfind(lst, cmp, "Rudolph") -- assert(idx == 1) -- assert(cmp_result == "Rudolph") -- the value is returned -- -- -- NOTE: this fails, since 'false' cannot be returned! -- local idx, cmp_result = tablex.rfind(lst, cmp, false) -- assert(idx == nil) -- looking up 'false' failed! -- assert(cmp_result == nil) function tablex.find_if(t,cmp,arg) assert_arg_iterable(1,t) cmp = function_arg(2,cmp) for k,v in pairs(t) do local c = cmp(v,arg) if c then return k,c end end return nil end --- return a list of all values in a table indexed by another list. -- @tab tbl a table -- @array idx an index table (a list of keys) -- @return a list-like table -- @usage index_by({10,20,30,40},{2,4}) == {20,40} -- @usage index_by({one=1,two=2,three=3},{'one','three'}) == {1,3} function tablex.index_by(tbl,idx) assert_arg_indexable(1,tbl) assert_arg_indexable(2,idx) local res = {} for i = 1,#idx do res[i] = tbl[idx[i]] end return setmeta(res,tbl,'List') end --- apply a function to all values of a table. -- This returns a table of the results. -- Any extra arguments are passed to the function. -- @within MappingAndFiltering -- @func fun A function that takes at least one argument -- @tab t A table -- @param ... optional arguments -- @usage map(function(v) return v*v end, {10,20,30,fred=2}) is {100,400,900,fred=4} function tablex.map(fun,t,...) assert_arg_iterable(1,t) fun = function_arg(1,fun) local res = {} for k,v in pairs(t) do res[k] = fun(v,...) end return setmeta(res,t) end --- apply a function to all values of a list. -- This returns a table of the results. -- Any extra arguments are passed to the function. -- @within MappingAndFiltering -- @func fun A function that takes at least one argument -- @array t a table (applies to array part) -- @param ... optional arguments -- @return a list-like table -- @usage imap(function(v) return v*v end, {10,20,30,fred=2}) is {100,400,900} function tablex.imap(fun,t,...) assert_arg_indexable(1,t) fun = function_arg(1,fun) local res = {} for i = 1,#t do res[i] = fun(t[i],...) or false end return setmeta(res,t,'List') end --- apply a named method to values from a table. -- @within MappingAndFiltering -- @string name the method name -- @array t a list-like table -- @param ... any extra arguments to the method -- @return a `List` with the results of the method (1st result only) -- @usage -- local Car = {} -- Car.__index = Car -- function Car.new(car) -- return setmetatable(car or {}, Car) -- end -- Car.speed = 0 -- function Car:faster(increase) -- self.speed = self.speed + increase -- return self.speed -- end -- -- local ferrari = Car.new{ name = "Ferrari" } -- local lamborghini = Car.new{ name = "Lamborghini", speed = 50 } -- local cars = { ferrari, lamborghini } -- -- assert(ferrari.speed == 0) -- assert(lamborghini.speed == 50) -- tablex.map_named_method("faster", cars, 10) -- assert(ferrari.speed == 10) -- assert(lamborghini.speed == 60) function tablex.map_named_method (name,t,...) utils.assert_string(1,name) assert_arg_indexable(2,t) local res = {} for i = 1,#t do local val = t[i] local fun = val[name] res[i] = fun(val,...) end return setmeta(res,t,'List') end --- apply a function to all values of a table, in-place. -- Any extra arguments are passed to the function. -- @func fun A function that takes at least one argument -- @tab t a table -- @param ... extra arguments passed to `fun` -- @see tablex.foreach function tablex.transform (fun,t,...) assert_arg_iterable(1,t) fun = function_arg(1,fun) for k,v in pairs(t) do t[k] = fun(v,...) end end --- generate a table of all numbers in a range. -- This is consistent with a numerical for loop. -- @int start number -- @int finish number -- @int[opt=1] step make this negative for start < finish function tablex.range (start,finish,step) local res step = step or 1 if start == finish then res = {start} elseif (start > finish and step > 0) or (finish > start and step < 0) then res = {} else local k = 1 res = {} for i=start,finish,step do res[k]=i; k=k+1 end end return makelist(res) end --- apply a function to values from two tables. -- @within MappingAndFiltering -- @func fun a function of at least two arguments -- @tab t1 a table -- @tab t2 a table -- @param ... extra arguments -- @return a table -- @usage map2('+',{1,2,3,m=4},{10,20,30,m=40}) is {11,22,23,m=44} function tablex.map2 (fun,t1,t2,...) assert_arg_iterable(1,t1) assert_arg_iterable(2,t2) fun = function_arg(1,fun) local res = {} for k,v in pairs(t1) do res[k] = fun(v,t2[k],...) end return setmeta(res,t1,'List') end --- apply a function to values from two arrays. -- The result will be the length of the shortest array. -- @within MappingAndFiltering -- @func fun a function of at least two arguments -- @array t1 a list-like table -- @array t2 a list-like table -- @param ... extra arguments -- @usage imap2('+',{1,2,3,m=4},{10,20,30,m=40}) is {11,22,23} function tablex.imap2 (fun,t1,t2,...) assert_arg_indexable(2,t1) assert_arg_indexable(3,t2) fun = function_arg(1,fun) local res,n = {},math.min(#t1,#t2) for i = 1,n do res[i] = fun(t1[i],t2[i],...) end return res end --- 'reduce' a list using a binary function. -- @func fun a function of two arguments -- @array t a list-like table -- @array memo optional initial memo value. Defaults to first value in table. -- @return the result of the function -- @usage reduce('+',{1,2,3,4}) == 10 function tablex.reduce (fun,t,memo) assert_arg_indexable(2,t) fun = function_arg(1,fun) local n = #t if n == 0 then return memo end local res = memo and fun(memo, t[1]) or t[1] for i = 2,n do res = fun(res,t[i]) end return res end --- apply a function to all elements of a table. -- The arguments to the function will be the value, -- the key and _finally_ any extra arguments passed to this function. -- Note that the Lua 5.0 function table.foreach passed the _key_ first. -- @within Iterating -- @tab t a table -- @func fun a function on the elements; `function(value, key, ...)` -- @param ... extra arguments passed to `fun` -- @see tablex.transform function tablex.foreach(t,fun,...) assert_arg_iterable(1,t) fun = function_arg(2,fun) for k,v in pairs(t) do fun(v,k,...) end end --- apply a function to all elements of a list-like table in order. -- The arguments to the function will be the value, -- the index and _finally_ any extra arguments passed to this function -- @within Iterating -- @array t a table -- @func fun a function with at least one argument -- @param ... optional arguments function tablex.foreachi(t,fun,...) assert_arg_indexable(1,t) fun = function_arg(2,fun) for i = 1,#t do fun(t[i],i,...) end end --- Apply a function to a number of tables. -- A more general version of map -- The result is a table containing the result of applying that function to the -- ith value of each table. Length of output list is the minimum length of all the lists -- @within MappingAndFiltering -- @func fun a function of n arguments -- @tab ... n tables -- @usage mapn(function(x,y,z) return x+y+z end, {1,2,3},{10,20,30},{100,200,300}) is {111,222,333} -- @usage mapn(math.max, {1,20,300},{10,2,3},{100,200,100}) is {100,200,300} -- @param fun A function that takes as many arguments as there are tables function tablex.mapn(fun,...) fun = function_arg(1,fun) local res = {} local lists = {...} local minn = 1e40 for i = 1,#lists do minn = min(minn,#(lists[i])) end for i = 1,minn do local args,k = {},1 for j = 1,#lists do args[k] = lists[j][i] k = k + 1 end res[#res+1] = fun(unpack(args)) end return res end --- call the function with the key and value pairs from a table. -- The function can return a value and a key (note the order!). If both -- are not nil, then this pair is inserted into the result: if the key already exists, we convert the value for that -- key into a table and append into it. If only value is not nil, then it is appended to the result. -- @within MappingAndFiltering -- @func fun A function which will be passed each key and value as arguments, plus any extra arguments to pairmap. -- @tab t A table -- @param ... optional arguments -- @usage pairmap(function(k,v) return v end,{fred=10,bonzo=20}) is {10,20} _or_ {20,10} -- @usage pairmap(function(k,v) return {k,v},k end,{one=1,two=2}) is {one={'one',1},two={'two',2}} function tablex.pairmap(fun,t,...) assert_arg_iterable(1,t) fun = function_arg(1,fun) local res = {} for k,v in pairs(t) do local rv,rk = fun(k,v,...) if rk then if res[rk] then if type(res[rk]) == 'table' then table.insert(res[rk],rv) else res[rk] = {res[rk], rv} end else res[rk] = rv end else res[#res+1] = rv end end return res end local function keys_op(i,v) return i end --- return all the keys of a table in arbitrary order. -- @within Extraction -- @tab t A table function tablex.keys(t) assert_arg_iterable(1,t) return makelist(tablex.pairmap(keys_op,t)) end local function values_op(i,v) return v end --- return all the values of the table in arbitrary order -- @within Extraction -- @tab t A table function tablex.values(t) assert_arg_iterable(1,t) return makelist(tablex.pairmap(values_op,t)) end local function index_map_op (i,v) return i,v end --- create an index map from a list-like table. The original values become keys, -- and the associated values are the indices into the original list. -- @array t a list-like table -- @return a map-like table function tablex.index_map (t) assert_arg_indexable(1,t) return makemap(tablex.pairmap(index_map_op,t)) end local function set_op(i,v) return true,v end --- create a set from a list-like table. A set is a table where the original values -- become keys, and the associated values are all true. -- @array t a list-like table -- @return a set (a map-like table) function tablex.makeset (t) assert_arg_indexable(1,t) return setmetatable(tablex.pairmap(set_op,t),require('pl.Set')) end --- combine two tables, either as union or intersection. Corresponds to -- set operations for sets () but more general. Not particularly -- useful for list-like tables. -- @within Merging -- @tab t1 a table -- @tab t2 a table -- @bool dup true for a union, false for an intersection. -- @usage merge({alice=23,fred=34},{bob=25,fred=34}) is {fred=34} -- @usage merge({alice=23,fred=34},{bob=25,fred=34},true) is {bob=25,fred=34,alice=23} -- @see tablex.index_map function tablex.merge (t1,t2,dup) assert_arg_iterable(1,t1) assert_arg_iterable(2,t2) local res = {} for k,v in pairs(t1) do if dup or t2[k] then res[k] = v end end if dup then for k,v in pairs(t2) do res[k] = v end end return setmeta(res,t1,'Map') end --- the union of two map-like tables. -- If there are duplicate keys, the second table wins. -- @tab t1 a table -- @tab t2 a table -- @treturn tab -- @see tablex.merge function tablex.union(t1, t2) return tablex.merge(t1, t2, true) end --- the intersection of two map-like tables. -- @tab t1 a table -- @tab t2 a table -- @treturn tab -- @see tablex.merge function tablex.intersection(t1, t2) return tablex.merge(t1, t2, false) end --- a new table which is the difference of two tables. -- With sets (where the values are all true) this is set difference and -- symmetric difference depending on the third parameter. -- @within Merging -- @tab s1 a map-like table or set -- @tab s2 a map-like table or set -- @bool symm symmetric difference (default false) -- @return a map-like table or set function tablex.difference (s1,s2,symm) assert_arg_iterable(1,s1) assert_arg_iterable(2,s2) local res = {} for k,v in pairs(s1) do if s2[k] == nil then res[k] = v end end if symm then for k,v in pairs(s2) do if s1[k] == nil then res[k] = v end end end return setmeta(res,s1,'Map') end --- A table where the key/values are the values and value counts of the table. -- @array t a list-like table -- @func cmp a function that defines equality (otherwise uses ==) -- @return a map-like table -- @see seq.count_map function tablex.count_map (t,cmp) assert_arg_indexable(1,t) local res,mask = {},{} cmp = function_arg(2,cmp or '==') local n = #t for i = 1,#t do local v = t[i] if not mask[v] then mask[v] = true -- check this value against all other values res[v] = 1 -- there's at least one instance for j = i+1,n do local w = t[j] local ok = cmp(v,w) if ok then res[v] = res[v] + 1 mask[w] = true end end end end return makemap(res) end --- filter an array's values using a predicate function -- @within MappingAndFiltering -- @array t a list-like table -- @func pred a boolean function -- @param arg optional argument to be passed as second argument of the predicate function tablex.filter (t,pred,arg) assert_arg_indexable(1,t) pred = function_arg(2,pred) local res,k = {},1 for i = 1,#t do local v = t[i] if pred(v,arg) then res[k] = v k = k + 1 end end return setmeta(res,t,'List') end --- return a table where each element is a table of the ith values of an arbitrary -- number of tables. It is equivalent to a matrix transpose. -- @within Merging -- @usage zip({10,20,30},{100,200,300}) is {{10,100},{20,200},{30,300}} -- @array ... arrays to be zipped function tablex.zip(...) return tablex.mapn(function(...) return {...} end,...) end local _copy function _copy (dest,src,idest,isrc,nsrc,clean_tail) idest = idest or 1 isrc = isrc or 1 local iend if not nsrc then nsrc = #src iend = #src else iend = isrc + min(nsrc-1,#src-isrc) end if dest == src then -- special case if idest > isrc and iend >= idest then -- overlapping ranges src = tablex.sub(src,isrc,nsrc) isrc = 1; iend = #src end end for i = isrc,iend do dest[idest] = src[i] idest = idest + 1 end if clean_tail then tablex.clear(dest,idest) end return dest end --- copy an array into another one, clearing `dest` after `idest+nsrc`, if necessary. -- @within Copying -- @array dest a list-like table -- @array src a list-like table -- @int[opt=1] idest where to start copying values into destination -- @int[opt=1] isrc where to start copying values from source -- @int[opt=#src] nsrc number of elements to copy from source function tablex.icopy (dest,src,idest,isrc,nsrc) assert_arg_indexable(1,dest) assert_arg_indexable(2,src) return _copy(dest,src,idest,isrc,nsrc,true) end --- copy an array into another one. -- @within Copying -- @array dest a list-like table -- @array src a list-like table -- @int[opt=1] idest where to start copying values into destination -- @int[opt=1] isrc where to start copying values from source -- @int[opt=#src] nsrc number of elements to copy from source function tablex.move (dest,src,idest,isrc,nsrc) assert_arg_indexable(1,dest) assert_arg_indexable(2,src) return _copy(dest,src,idest,isrc,nsrc,false) end function tablex._normalize_slice(self,first,last) local sz = #self if not first then first=1 end if first<0 then first=sz+first+1 end -- make the range _inclusive_! if not last then last=sz end if last < 0 then last=sz+1+last end return first,last end --- Extract a range from a table, like 'string.sub'. -- If first or last are negative then they are relative to the end of the list -- eg. sub(t,-2) gives last 2 entries in a list, and -- sub(t,-4,-2) gives from -4th to -2nd -- @within Extraction -- @array t a list-like table -- @int first An index -- @int last An index -- @return a new List function tablex.sub(t,first,last) assert_arg_indexable(1,t) first,last = tablex._normalize_slice(t,first,last) local res={} for i=first,last do append(res,t[i]) end return setmeta(res,t,'List') end --- set an array range to a value. If it's a function we use the result -- of applying it to the indices. -- @array t a list-like table -- @param val a value -- @int[opt=1] i1 start range -- @int[opt=#t] i2 end range function tablex.set (t,val,i1,i2) assert_arg_indexable(1,t) i1,i2 = i1 or 1,i2 or #t if types.is_callable(val) then for i = i1,i2 do t[i] = val(i) end else for i = i1,i2 do t[i] = val end end end --- create a new array of specified size with initial value. -- @int n size -- @param val initial value (can be `nil`, but don't expect `#` to work!) -- @return the table function tablex.new (n,val) local res = {} tablex.set(res,val,1,n) return res end --- clear out the contents of a table. -- @array t a list -- @param istart optional start position function tablex.clear(t,istart) istart = istart or 1 for i = istart,#t do remove(t) end end --- insert values into a table. -- similar to `table.insert` but inserts values from given table `values`, -- not the object itself, into table `t` at position `pos`. -- @within Copying -- @array t the list -- @int[opt] position (default is at end) -- @array values function tablex.insertvalues(t, ...) assert_arg(1,t,'table') local pos, values if select('#', ...) == 1 then pos,values = #t+1, ... else pos,values = ... end if #values > 0 then for i=#t,pos,-1 do t[i+#values] = t[i] end local offset = 1 - pos for i=pos,pos+#values-1 do t[i] = values[i + offset] end end return t end --- remove a range of values from a table. -- End of range may be negative. -- @array t a list-like table -- @int i1 start index -- @int i2 end index -- @return the table function tablex.removevalues (t,i1,i2) assert_arg(1,t,'table') i1,i2 = tablex._normalize_slice(t,i1,i2) for i = i1,i2 do remove(t,i1) end return t end local _find _find = function (t,value,tables) for k,v in pairs(t) do if v == value then return k end end for k,v in pairs(t) do if not tables[v] and type(v) == 'table' then tables[v] = true local res = _find(v,value,tables) if res then res = tostring(res) if type(k) ~= 'string' then return '['..k..']'..res else return k..'.'..res end end end end end --- find a value in a table by recursive search. -- @within Finding -- @tab t the table -- @param value the value -- @array[opt] exclude any tables to avoid searching -- @return a fieldspec, e.g. 'a.b' or 'math.sin' -- @usage search(_G,math.sin,{package.path}) == 'math.sin' function tablex.search (t,value,exclude) assert_arg_iterable(1,t) local tables = {[t]=true} if exclude then for _,v in pairs(exclude) do tables[v] = true end end return _find(t,value,tables) end --- return an iterator to a table sorted by its keys -- @within Iterating -- @tab t the table -- @func f an optional comparison function (f(x,y) is true if x < y) -- @usage for k,v in tablex.sort(t) do print(k,v) end -- @return an iterator to traverse elements sorted by the keys function tablex.sort(t,f) local keys = {} for k in pairs(t) do keys[#keys + 1] = k end tsort(keys,f) local i = 0 return function() i = i + 1 return keys[i], t[keys[i]] end end --- return an iterator to a table sorted by its values -- @within Iterating -- @tab t the table -- @func f an optional comparison function (f(x,y) is true if x < y) -- @usage for k,v in tablex.sortv(t) do print(k,v) end -- @return an iterator to traverse elements sorted by the values function tablex.sortv(t,f) f = function_arg(2, f or '<') local keys = {} for k in pairs(t) do keys[#keys + 1] = k end tsort(keys,function(x, y) return f(t[x], t[y]) end) local i = 0 return function() i = i + 1 return keys[i], t[keys[i]] end end --- modifies a table to be read only. -- This only offers weak protection. Tables can still be modified with -- `table.insert` and `rawset`. -- -- *NOTE*: for Lua 5.1 length, pairs and ipairs will not work, since the -- equivalent metamethods are only available in Lua 5.2 and newer. -- @tab t the table -- @return the table read only (a proxy). function tablex.readonly(t) local mt = { __index=t, __newindex=function(t, k, v) error("Attempt to modify read-only table", 2) end, __pairs=function() return pairs(t) end, __ipairs=function() return ipairs(t) end, __len=function() return #t end, __metatable=false } return setmetatable({}, mt) end return tablex Penlight-1.12.0/lua/pl/template.lua000066400000000000000000000167441416703176500171050ustar00rootroot00000000000000--- A template preprocessor. -- Originally by [Ricki Lake](http://lua-users.org/wiki/SlightlyLessSimpleLuaPreprocessor) -- -- There are two rules: -- -- * lines starting with # are Lua -- * otherwise, `$(expr)` is the result of evaluating `expr` -- -- Example: -- -- # for i = 1,3 do -- $(i) Hello, Word! -- # end -- ===> -- 1 Hello, Word! -- 2 Hello, Word! -- 3 Hello, Word! -- -- Other escape characters can be used, when the defaults conflict -- with the output language. -- -- > for _,n in pairs{'one','two','three'} do -- static int l_${n} (luaState *state); -- > end -- -- See @{03-strings.md.Another_Style_of_Template|the Guide}. -- -- Dependencies: `pl.utils` -- @module pl.template local utils = require 'pl.utils' local append,format,strsub,strfind,strgsub = table.insert,string.format,string.sub,string.find,string.gsub local APPENDER = "\n__R_size = __R_size + 1; __R_table[__R_size] = " local function parseDollarParen(pieces, chunk, exec_pat, newline) local s = 1 for term, executed, e in chunk:gmatch(exec_pat) do executed = '('..strsub(executed,2,-2)..')' append(pieces, APPENDER..format("%q", strsub(chunk,s, term - 1))) append(pieces, APPENDER..format("__tostring(%s or '')", executed)) s = e end local r if newline then r = format("%q", strgsub(strsub(chunk,s),"\n","")) else r = format("%q", strsub(chunk,s)) end if r ~= '""' then append(pieces, APPENDER..r) end end local function parseHashLines(chunk,inline_escape,brackets,esc,newline) local exec_pat = "()"..inline_escape.."(%b"..brackets..")()" local esc_pat = esc.."+([^\n]*\n?)" local esc_pat1, esc_pat2 = "^"..esc_pat, "\n"..esc_pat local pieces, s = {"return function()\nlocal __R_size, __R_table, __tostring = 0, {}, __tostring", n = 1}, 1 while true do local _, e, lua = strfind(chunk,esc_pat1, s) if not e then local ss ss, e, lua = strfind(chunk,esc_pat2, s) parseDollarParen(pieces, strsub(chunk,s, ss), exec_pat, newline) if not e then break end end if strsub(lua, -1, -1) == "\n" then lua = strsub(lua, 1, -2) end append(pieces, "\n"..lua) s = e + 1 end append(pieces, "\nreturn __R_table\nend") -- let's check for a special case where there is nothing to template, but it's -- just a single static string local short = false if (#pieces == 3) and (pieces[2]:find(APPENDER, 1, true) == 1) then pieces = { "return " .. pieces[2]:sub(#APPENDER+1,-1) } short = true end -- if short == true, the generated function will not return a table of strings, -- but a single string return table.concat(pieces), short end local template = {} --- expand the template using the specified environment. -- This function will compile and render the template. For more performant -- recurring usage use the two step approach by using `compile` and `ct:render`. -- There are six special fields in the environment table `env` -- -- * `_parent`: continue looking up in this table (e.g. `_parent=_G`). -- * `_brackets`: bracket pair that wraps inline Lua expressions, default is '()'. -- * `_escape`: character marking Lua lines, default is '#' -- * `_inline_escape`: character marking inline Lua expression, default is '$'. -- * `_chunk_name`: chunk name for loaded templates, used if there -- is an error in Lua code. Default is 'TMP'. -- * `_debug`: if truthy, the generated code will be printed upon a render error -- -- @string str the template string -- @tab[opt] env the environment -- @return `rendered template + nil + source_code`, or `nil + error + source_code`. The last -- return value (`source_code`) is only returned if the debug option is used. function template.substitute(str,env) env = env or {} local t, err = template.compile(str, { chunk_name = rawget(env,"_chunk_name"), escape = rawget(env,"_escape"), inline_escape = rawget(env,"_inline_escape"), inline_brackets = rawget(env,"_brackets"), newline = nil, debug = rawget(env,"_debug") }) if not t then return t, err end return t:render(env, rawget(env,"_parent"), rawget(env,"_debug")) end --- executes the previously compiled template and renders it. -- @function ct:render -- @tab[opt] env the environment. -- @tab[opt] parent continue looking up in this table (e.g. `parent=_G`). -- @bool[opt] db if thruthy, it will print the code upon a render error -- (provided the template was compiled with the debug option). -- @return `rendered template + nil + source_code`, or `nil + error + source_code`. The last return value -- (`source_code`) is only returned if the template was compiled with the debug option. -- @usage -- local ct, err = template.compile(my_template) -- local rendered , err = ct:render(my_env, parent) local render = function(self, env, parent, db) env = env or {} if parent then -- parent is a bit silly, but for backward compatibility retained setmetatable(env, {__index = parent}) end setmetatable(self.env, {__index = env}) local res, out = xpcall(self.fn, debug.traceback) if not res then if self.code and db then print(self.code) end return nil, out, self.code end return table.concat(out), nil, self.code end --- compiles the template. -- Returns an object that can repeatedly be rendered without parsing/compiling -- the template again. -- The options passed in the `opts` table support the following options: -- -- * `chunk_name`: chunk name for loaded templates, used if there -- is an error in Lua code. Default is 'TMP'. -- * `escape`: character marking Lua lines, default is '#' -- * `inline_escape`: character marking inline Lua expression, default is '$'. -- * `inline_brackets`: bracket pair that wraps inline Lua expressions, default is '()'. -- * `newline`: string to replace newline characters, default is `nil` (not replacing newlines). -- * `debug`: if truthy, the generated source code will be retained within the compiled template object, default is `nil`. -- -- @string str the template string -- @tab[opt] opts the compilation options to use -- @return template object, or `nil + error + source_code` -- @usage -- local ct, err = template.compile(my_template) -- local rendered , err = ct:render(my_env, parent) function template.compile(str, opts) opts = opts or {} local chunk_name = opts.chunk_name or 'TMP' local escape = opts.escape or '#' local inline_escape = opts.inline_escape or '$' local inline_brackets = opts.inline_brackets or '()' local code, short = parseHashLines(str,inline_escape,inline_brackets,escape,opts.newline) local env = { __tostring = tostring } local fn, err = utils.load(code, chunk_name,'t',env) if not fn then return nil, err, code end if short then -- the template returns a single constant string, let's optimize for that local constant_string = fn() return { fn = fn(), env = env, render = function(self) -- additional params can be ignored -- skip the metatable magic and error handling in the render -- function above for this special case return constant_string, nil, self.code end, code = opts.debug and code or nil, } end return { fn = fn(), env = env, render = render, code = opts.debug and code or nil, } end return template Penlight-1.12.0/lua/pl/test.lua000066400000000000000000000120321416703176500162330ustar00rootroot00000000000000--- Useful test utilities. -- -- test.asserteq({1,2},{1,2}) -- can compare tables -- test.asserteq(1.2,1.19,0.02) -- compare FP numbers within precision -- T = test.tuple -- used for comparing multiple results -- test.asserteq(T(string.find(" me","me")),T(2,3)) -- -- Dependencies: `pl.utils`, `pl.tablex`, `pl.pretty`, `pl.path`, `debug` -- @module pl.test local tablex = require 'pl.tablex' local utils = require 'pl.utils' local pretty = require 'pl.pretty' local path = require 'pl.path' local type,unpack,pack = type,utils.unpack,utils.pack local clock = os.clock local debug = require 'debug' local io = io local function dump(x) if type(x) == 'table' and not (getmetatable(x) and getmetatable(x).__tostring) then return pretty.write(x,' ',true) elseif type(x) == 'string' then return '"'..x..'"' else return tostring(x) end end local test = {} ---- error handling for test results. -- By default, this writes to stderr and exits the program. -- Re-define this function to raise an error and/or redirect output function test.error_handler(file,line,got_text, needed_text,msg) local err = io.stderr err:write(path.basename(file)..':'..line..': assertion failed\n') err:write("got:\t",got_text,'\n') err:write("needed:\t",needed_text,'\n') utils.quit(1,msg or "these values were not equal") end local function complain (x,y,msg,where) local i = debug.getinfo(3 + (where or 0)) test.error_handler(i.short_src,i.currentline,dump(x),dump(y),msg) end --- general test complain message. -- Useful for composing new test functions (see tests/tablex.lua for an example) -- @param x a value -- @param y value to compare first value against -- @param msg message -- @param where extra level offset for errors -- @function complain test.complain = complain --- like assert, except takes two arguments that must be equal and can be tables. -- If they are plain tables, it will use tablex.deepcompare. -- @param x any value -- @param y a value equal to x -- @param eps an optional tolerance for numerical comparisons -- @param where extra level offset function test.asserteq (x,y,eps,where) local res = x == y if not res then res = tablex.deepcompare(x,y,true,eps) end if not res then complain(x,y,nil,where) end end --- assert that the first string matches the second. -- @param s1 a string -- @param s2 a string -- @param where extra level offset function test.assertmatch (s1,s2,where) if not s1:match(s2) then complain (s1,s2,"these strings did not match",where) end end --- assert that the function raises a particular error. -- @param fn a function or a table of the form {function,arg1,...} -- @param e a string to match the error against -- @param where extra level offset function test.assertraise(fn,e,where) local ok, err if type(fn) == 'table' then ok, err = pcall(unpack(fn)) else ok, err = pcall(fn) end if ok or err:match(e)==nil then complain (err,e,"these errors did not match",where) end end --- a version of asserteq that takes two pairs of values. -- x1==y1 and x2==y2 must be true. Useful for functions that naturally -- return two values. -- @param x1 any value -- @param x2 any value -- @param y1 any value -- @param y2 any value -- @param where extra level offset function test.asserteq2 (x1,x2,y1,y2,where) if x1 ~= y1 then complain(x1,y1,nil,where) end if x2 ~= y2 then complain(x2,y2,nil,where) end end -- tuple type -- local tuple_mt = { unpack = unpack } tuple_mt.__index = tuple_mt function tuple_mt.__tostring(self) local ts = {} for i=1, self.n do local s = self[i] ts[i] = type(s) == 'string' and ('%q'):format(s) or tostring(s) end return 'tuple(' .. table.concat(ts, ', ') .. ')' end function tuple_mt.__eq(a, b) if a.n ~= b.n then return false end for i=1, a.n do if a[i] ~= b[i] then return false end end return true end function tuple_mt.__len(self) return self.n end --- encode an arbitrary argument list as a tuple. -- This can be used to compare to other argument lists, which is -- very useful for testing functions which return a number of values. -- Unlike regular array-like tables ('sequences') they may contain nils. -- Tuples understand equality and know how to print themselves out. -- The # operator is defined to be the size, irrespecive of any nils, -- and there is an `unpack` method. -- @usage asserteq(tuple( ('ab'):find 'a'), tuple(1,1)) function test.tuple(...) return setmetatable(pack(...), tuple_mt) end --- Time a function. Call the function a given number of times, and report the number of seconds taken, -- together with a message. Any extra arguments will be passed to the function. -- @string msg a descriptive message -- @int n number of times to call the function -- @func fun the function -- @param ... optional arguments to fun function test.timer(msg,n,fun,...) local start = clock() for i = 1,n do fun(...) end utils.printf("%s: took %7.2f sec\n",msg,clock()-start) end return test Penlight-1.12.0/lua/pl/text.lua000066400000000000000000000014401416703176500162410ustar00rootroot00000000000000--- Text processing utilities. -- -- This provides a Template class (modeled after the same from the Python -- libraries, see string.Template). It also provides similar functions to those -- found in the textwrap module. -- -- IMPORTANT: this module has been deprecated and will be removed in a future -- version (2.0). The contents of this module have moved to the `pl.stringx` -- module. -- -- See @{03-strings.md.String_Templates|the Guide}. -- -- Dependencies: `pl.stringx`, `pl.utils` -- @module pl.text local utils = require("pl.utils") utils.raise_deprecation { source = "Penlight " .. utils._VERSION, message = "the contents of module 'pl.text' has moved into 'pl.stringx'", version_removed = "2.0.0", deprecated_after = "1.11.0", no_trace = true, } return require "pl.stringx" Penlight-1.12.0/lua/pl/types.lua000066400000000000000000000133041416703176500164230ustar00rootroot00000000000000---- Dealing with Detailed Type Information -- Dependencies `pl.utils` -- @module pl.types local utils = require 'pl.utils' local math_ceil = math.ceil local assert_arg = utils.assert_arg local types = {} --- is the object either a function or a callable object?. -- @param obj Object to check. function types.is_callable (obj) return type(obj) == 'function' or getmetatable(obj) and getmetatable(obj).__call and true end --- is the object of the specified type?. -- If the type is a string, then use type, otherwise compare with metatable. -- -- NOTE: this function is imported from `utils.is_type`. -- @param obj An object to check -- @param tp The expected type -- @function is_type -- @see utils.is_type types.is_type = utils.is_type local fileMT = getmetatable(io.stdout) --- a string representation of a type. -- For tables and userdata with metatables, we assume that the metatable has a `_name` -- field. If the field is not present it will return 'unknown table' or -- 'unknown userdata'. -- Lua file objects return the type 'file'. -- @param obj an object -- @return a string like 'number', 'table', 'file' or 'List' function types.type (obj) local t = type(obj) if t == 'table' or t == 'userdata' then local mt = getmetatable(obj) if mt == fileMT then return 'file' elseif mt == nil then return t else -- TODO: the "unknown" is weird, it should just return the type return mt._name or "unknown "..t end else return t end end --- is this number an integer? -- @param x a number -- @raise error if x is not a number -- @return boolean function types.is_integer (x) return math_ceil(x)==x end --- Check if the object is "empty". -- An object is considered empty if it is: -- -- - `nil` -- - a table without any items (key-value pairs or indexes) -- - a string with no content ("") -- - not a nil/table/string -- @param o The object to check if it is empty. -- @param ignore_spaces If the object is a string and this is true the string is -- considered empty if it only contains spaces. -- @return `true` if the object is empty, otherwise a falsy value. function types.is_empty(o, ignore_spaces) if o == nil then return true elseif type(o) == "table" then return next(o) == nil elseif type(o) == "string" then return o == "" or (not not ignore_spaces and (not not o:find("^%s+$"))) else return true end end local function check_meta (val) if type(val) == 'table' then return true end return getmetatable(val) end --- is an object 'array-like'? -- An object is array like if: -- -- - it is a table, or -- - it has a metatable with `__len` and `__index` methods -- -- NOTE: since `__len` is 5.2+, on 5.1 is usually returns `false` for userdata -- @param val any value. -- @return `true` if the object is array-like, otherwise a falsy value. function types.is_indexable (val) local mt = check_meta(val) if mt == true then return true end return mt and mt.__len and mt.__index and true end --- can an object be iterated over with `pairs`? -- An object is iterable if: -- -- - it is a table, or -- - it has a metatable with a `__pairs` meta method -- -- NOTE: since `__pairs` is 5.2+, on 5.1 is usually returns `false` for userdata -- @param val any value. -- @return `true` if the object is iterable, otherwise a falsy value. function types.is_iterable (val) local mt = check_meta(val) if mt == true then return true end return mt and mt.__pairs and true end --- can an object accept new key/pair values? -- An object is iterable if: -- -- - it is a table, or -- - it has a metatable with a `__newindex` meta method -- -- @param val any value. -- @return `true` if the object is writeable, otherwise a falsy value. function types.is_writeable (val) local mt = check_meta(val) if mt == true then return true end return mt and mt.__newindex and true end -- Strings that should evaluate to true. -- TODO: add on/off ??? local trues = { yes=true, y=true, ["true"]=true, t=true, ["1"]=true } -- Conditions types should evaluate to true. local true_types = { boolean=function(o, true_strs, check_objs) return o end, string=function(o, true_strs, check_objs) o = o:lower() if trues[o] then return true end -- Check alternative user provided strings. for _,v in ipairs(true_strs or {}) do if type(v) == "string" and o == v:lower() then return true end end return false end, number=function(o, true_strs, check_objs) return o ~= 0 end, table=function(o, true_strs, check_objs) if check_objs and next(o) ~= nil then return true end return false end } --- Convert to a boolean value. -- True values are: -- -- * boolean: true. -- * string: 'yes', 'y', 'true', 't', '1' or additional strings specified by `true_strs`. -- * number: Any non-zero value. -- * table: Is not empty and `check_objs` is true. -- * everything else: Is not `nil` and `check_objs` is true. -- -- @param o The object to evaluate. -- @param[opt] true_strs optional Additional strings that when matched should evaluate to true. Comparison is case insensitive. -- This should be a List of strings. E.g. "ja" to support German. -- @param[opt] check_objs True if objects should be evaluated. -- @return true if the input evaluates to true, otherwise false. function types.to_bool(o, true_strs, check_objs) local true_func if true_strs then assert_arg(2, true_strs, "table") end true_func = true_types[type(o)] if true_func then return true_func(o, true_strs, check_objs) elseif check_objs and o ~= nil then return true end return false end return types Penlight-1.12.0/lua/pl/url.lua000066400000000000000000000022541416703176500160630ustar00rootroot00000000000000--- Python-style URL quoting library. -- -- @module pl.url local url = {} local function quote_char(c) return string.format("%%%02X", string.byte(c)) end --- Quote the url, replacing special characters using the '%xx' escape. -- @string s the string -- @bool quote_plus Also escape slashes and replace spaces by plus signs. -- @return The quoted string, or if `s` wasn't a string, just plain unaltered `s`. function url.quote(s, quote_plus) if type(s) ~= "string" then return s end s = s:gsub("\n", "\r\n") s = s:gsub("([^A-Za-z0-9 %-_%./])", quote_char) if quote_plus then s = s:gsub(" ", "+") s = s:gsub("/", quote_char) else s = s:gsub(" ", "%%20") end return s end local function unquote_char(h) return string.char(tonumber(h, 16)) end --- Unquote the url, replacing '%xx' escapes and plus signs. -- @string s the string -- @return The unquoted string, or if `s` wasn't a string, just plain unaltered `s`. function url.unquote(s) if type(s) ~= "string" then return s end s = s:gsub("+", " ") s = s:gsub("%%(%x%x)", unquote_char) s = s:gsub("\r\n", "\n") return s end return url Penlight-1.12.0/lua/pl/utils.lua000066400000000000000000000647101416703176500164260ustar00rootroot00000000000000--- Generally useful routines. -- See @{01-introduction.md.Generally_useful_functions|the Guide}. -- -- Dependencies: `pl.compat`, all exported fields and functions from -- `pl.compat` are also available in this module. -- -- @module pl.utils local format = string.format local compat = require 'pl.compat' local stdout = io.stdout local append = table.insert local concat = table.concat local _unpack = table.unpack -- always injected by 'compat' local find = string.find local sub = string.sub local is_windows = compat.is_windows local err_mode = 'default' local raise local operators local _function_factories = {} local utils = { _VERSION = "1.12.0" } for k, v in pairs(compat) do utils[k] = v end --- Some standard patterns -- @table patterns utils.patterns = { FLOAT = '[%+%-%d]%d*%.?%d*[eE]?[%+%-]?%d*', -- floating point number INTEGER = '[+%-%d]%d*', -- integer number IDEN = '[%a_][%w_]*', -- identifier FILE = '[%a%.\\][:%][%w%._%-\\]*', -- file } --- Standard meta-tables as used by other Penlight modules -- @table stdmt -- @field List the List metatable -- @field Map the Map metatable -- @field Set the Set metatable -- @field MultiMap the MultiMap metatable utils.stdmt = { List = {_name='List'}, Map = {_name='Map'}, Set = {_name='Set'}, MultiMap = {_name='MultiMap'}, } --- pack an argument list into a table. -- @param ... any arguments -- @return a table with field `n` set to the length -- @function utils.pack -- @see compat.pack -- @see utils.npairs -- @see utils.unpack utils.pack = table.pack -- added here to be symmetrical with unpack --- unpack a table and return its contents. -- -- NOTE: this implementation differs from the Lua implementation in the way -- that this one DOES honor the `n` field in the table `t`, such that it is 'nil-safe'. -- @param t table to unpack -- @param[opt] i index from which to start unpacking, defaults to 1 -- @param[opt] j index of the last element to unpack, defaults to `t.n` or else `#t` -- @return multiple return values from the table -- @function utils.unpack -- @see compat.unpack -- @see utils.pack -- @see utils.npairs -- @usage -- local t = table.pack(nil, nil, nil, 4) -- local a, b, c, d = table.unpack(t) -- this `unpack` is NOT nil-safe, so d == nil -- -- local a, b, c, d = utils.unpack(t) -- this is nil-safe, so d == 4 function utils.unpack(t, i, j) return _unpack(t, i or 1, j or t.n or #t) end --- print an arbitrary number of arguments using a format. -- Output will be sent to `stdout`. -- @param fmt The format (see `string.format`) -- @param ... Extra arguments for format function utils.printf(fmt, ...) utils.assert_string(1, fmt) utils.fprintf(stdout, fmt, ...) end --- write an arbitrary number of arguments to a file using a format. -- @param f File handle to write to. -- @param fmt The format (see `string.format`). -- @param ... Extra arguments for format function utils.fprintf(f,fmt,...) utils.assert_string(2,fmt) f:write(format(fmt,...)) end do local function import_symbol(T,k,v,libname) local key = rawget(T,k) -- warn about collisions! if key and k ~= '_M' and k ~= '_NAME' and k ~= '_PACKAGE' and k ~= '_VERSION' then utils.fprintf(io.stderr,"warning: '%s.%s' will not override existing symbol\n",libname,k) return end rawset(T,k,v) end local function lookup_lib(T,t) for k,v in pairs(T) do if v == t then return k end end return '?' end local already_imported = {} --- take a table and 'inject' it into the local namespace. -- @param t The table (table), or module name (string), defaults to this `utils` module table -- @param T An optional destination table (defaults to callers environment) function utils.import(t,T) T = T or _G t = t or utils if type(t) == 'string' then t = require (t) end local libname = lookup_lib(T,t) if already_imported[t] then return end already_imported[t] = libname for k,v in pairs(t) do import_symbol(T,k,v,libname) end end end --- return either of two values, depending on a condition. -- @param cond A condition -- @param value1 Value returned if cond is truthy -- @param value2 Value returned if cond is falsy function utils.choose(cond, value1, value2) return cond and value1 or value2 end --- convert an array of values to strings. -- @param t a list-like table -- @param[opt] temp (table) buffer to use, otherwise allocate -- @param[opt] tostr custom tostring function, called with (value,index). Defaults to `tostring`. -- @return the converted buffer function utils.array_tostring (t,temp,tostr) temp, tostr = temp or {}, tostr or tostring for i = 1,#t do temp[i] = tostr(t[i],i) end return temp end --- is the object of the specified type? -- If the type is a string, then use type, otherwise compare with metatable -- @param obj An object to check -- @param tp String of what type it should be -- @return boolean -- @usage utils.is_type("hello world", "string") --> true -- -- or check metatable -- local my_mt = {} -- local my_obj = setmetatable(my_obj, my_mt) -- utils.is_type(my_obj, my_mt) --> true function utils.is_type (obj,tp) if type(tp) == 'string' then return type(obj) == tp end local mt = getmetatable(obj) return tp == mt end --- an iterator with indices, similar to `ipairs`, but with a range. -- This is a nil-safe index based iterator that will return `nil` when there -- is a hole in a list. To be safe ensure that table `t.n` contains the length. -- @tparam table t the table to iterate over -- @tparam[opt=1] integer i_start start index -- @tparam[opt=t.n or #t] integer i_end end index -- @tparam[opt=1] integer step step size -- @treturn integer index -- @treturn any value at index (which can be `nil`!) -- @see utils.pack -- @see utils.unpack -- @usage -- local t = utils.pack(nil, 123, nil) -- adds an `n` field when packing -- -- for i, v in utils.npairs(t, 2) do -- start at index 2 -- t[i] = tostring(t[i]) -- end -- -- -- t = { n = 3, [2] = "123", [3] = "nil" } function utils.npairs(t, i_start, i_end, step) step = step or 1 if step == 0 then error("iterator step-size cannot be 0", 2) end local i = (i_start or 1) - step i_end = i_end or t.n or #t if step < 0 then return function() i = i + step if i < i_end then return nil end return i, t[i] end else return function() i = i + step if i > i_end then return nil end return i, t[i] end end end --- Error handling -- @section Error-handling --- assert that the given argument is in fact of the correct type. -- @param n argument index -- @param val the value -- @param tp the type -- @param verify an optional verification function -- @param msg an optional custom message -- @param lev optional stack position for trace, default 2 -- @return the validated value -- @raise if `val` is not the correct type -- @usage -- local param1 = assert_arg(1,"hello",'table') --> error: argument 1 expected a 'table', got a 'string' -- local param4 = assert_arg(4,'!@#$%^&*','string',path.isdir,'not a directory') -- --> error: argument 4: '!@#$%^&*' not a directory function utils.assert_arg (n,val,tp,verify,msg,lev) if type(val) ~= tp then error(("argument %d expected a '%s', got a '%s'"):format(n,tp,type(val)),lev or 2) end if verify and not verify(val) then error(("argument %d: '%s' %s"):format(n,val,msg),lev or 2) end return val end --- creates an Enum table. -- This helps prevent magic strings in code by throwing errors for accessing -- non-existing values. -- -- Calling on the object does the same, but returns a soft error; `nil + err`. -- -- The values are equal to the keys. The enum object is -- read-only. -- @param ... strings that make up the enumeration. -- @return Enum object -- @usage -- accessing at runtime -- local obj = {} -- obj.MOVEMENT = utils.enum("FORWARD", "REVERSE", "LEFT", "RIGHT") -- -- if current_movement == obj.MOVEMENT.FORWARD then -- -- do something -- -- elseif current_movement == obj.MOVEMENT.REVERES then -- -- throws error due to typo 'REVERES', so a silent mistake becomes a hard error -- -- "'REVERES' is not a valid value (expected one of: 'FORWARD', 'REVERSE', 'LEFT', 'RIGHT')" -- -- end -- @usage -- validating user-input -- local parameter = "...some user provided option..." -- local ok, err = obj.MOVEMENT(parameter) -- calling on the object -- if not ok then -- print("bad 'parameter', " .. err) -- os.exit(1) -- end function utils.enum(...) local lst = utils.pack(...) utils.assert_arg(1, lst[1], "string") -- at least 1 string local enum = {} for i, value in ipairs(lst) do utils.assert_arg(i, value, "string") enum[value] = value end local valid = "(expected one of: '" .. concat(lst, "', '") .. "')" setmetatable(enum, { __index = function(self, key) error(("'%s' is not a valid value %s"):format(tostring(key), valid), 2) end, __newindex = function(self, key, value) error("the Enum object is read-only", 2) end, __call = function(self, key) if type(key) == "string" then local v = rawget(self, key) if v then return v end end return nil, ("'%s' is not a valid value %s"):format(tostring(key), valid) end }) return enum end --- process a function argument. -- This is used throughout Penlight and defines what is meant by a function: -- Something that is callable, or an operator string as defined by pl.operator, -- such as '>' or '#'. If a function factory has been registered for the type, it will -- be called to get the function. -- @param idx argument index -- @param f a function, operator string, or callable object -- @param msg optional error message -- @return a callable -- @raise if idx is not a number or if f is not callable function utils.function_arg (idx,f,msg) utils.assert_arg(1,idx,'number') local tp = type(f) if tp == 'function' then return f end -- no worries! -- ok, a string can correspond to an operator (like '==') if tp == 'string' then if not operators then operators = require 'pl.operator'.optable end local fn = operators[f] if fn then return fn end local fn, err = utils.string_lambda(f) if not fn then error(err..': '..f) end return fn elseif tp == 'table' or tp == 'userdata' then local mt = getmetatable(f) if not mt then error('not a callable object',2) end local ff = _function_factories[mt] if not ff then if not mt.__call then error('not a callable object',2) end return f else return ff(f) -- we have a function factory for this type! end end if not msg then msg = " must be callable" end if idx > 0 then error("argument "..idx..": "..msg,2) else error(msg,2) end end --- assert the common case that the argument is a string. -- @param n argument index -- @param val a value that must be a string -- @return the validated value -- @raise val must be a string -- @usage -- local val = 42 -- local param2 = utils.assert_string(2, val) --> error: argument 2 expected a 'string', got a 'number' function utils.assert_string (n, val) return utils.assert_arg(n,val,'string',nil,nil,3) end --- control the error strategy used by Penlight. -- This is a global setting that controls how `utils.raise` behaves: -- -- - 'default': return `nil + error` (this is the default) -- - 'error': throw a Lua error -- - 'quit': exit the program -- -- @param mode either 'default', 'quit' or 'error' -- @see utils.raise function utils.on_error (mode) mode = tostring(mode) if ({['default'] = 1, ['quit'] = 2, ['error'] = 3})[mode] then err_mode = mode else -- fail loudly local err = "Bad argument expected string; 'default', 'quit', or 'error'. Got '"..tostring(mode).."'" if err_mode == 'default' then error(err, 2) -- even in 'default' mode fail loud in this case end raise(err) end end --- used by Penlight functions to return errors. Its global behaviour is controlled -- by `utils.on_error`. -- To use this function you MUST use it in conjunction with `return`, since it might -- return `nil + error`. -- @param err the error string. -- @see utils.on_error -- @usage -- if some_condition then -- return utils.raise("some condition was not met") -- MUST use 'return'! -- end function utils.raise (err) if err_mode == 'default' then return nil, err elseif err_mode == 'quit' then return utils.quit(err) else error(err, 2) end end raise = utils.raise --- File handling -- @section files --- return the contents of a file as a string -- @param filename The file path -- @param is_bin open in binary mode -- @return file contents function utils.readfile(filename,is_bin) local mode = is_bin and 'b' or '' utils.assert_string(1,filename) local f,open_err = io.open(filename,'r'..mode) if not f then return raise (open_err) end local res,read_err = f:read('*a') f:close() if not res then -- Errors in io.open have "filename: " prefix, -- error in file:read don't, add it. return raise (filename..": "..read_err) end return res end --- write a string to a file -- @param filename The file path -- @param str The string -- @param is_bin open in binary mode -- @return true or nil -- @return error message -- @raise error if filename or str aren't strings function utils.writefile(filename,str,is_bin) local mode = is_bin and 'b' or '' utils.assert_string(1,filename) utils.assert_string(2,str) local f,err = io.open(filename,'w'..mode) if not f then return raise(err) end local ok, write_err = f:write(str) f:close() if not ok then -- Errors in io.open have "filename: " prefix, -- error in file:write don't, add it. return raise (filename..": "..write_err) end return true end --- return the contents of a file as a list of lines -- @param filename The file path -- @return file contents as a table -- @raise error if filename is not a string function utils.readlines(filename) utils.assert_string(1,filename) local f,err = io.open(filename,'r') if not f then return raise(err) end local res = {} for line in f:lines() do append(res,line) end f:close() return res end --- OS functions -- @section OS-functions --- execute a shell command and return the output. -- This function redirects the output to tempfiles and returns the content of those files. -- @param cmd a shell command -- @param bin boolean, if true, read output as binary file -- @return true if successful -- @return actual return code -- @return stdout output (string) -- @return errout output (string) function utils.executeex(cmd, bin) local outfile = os.tmpname() local errfile = os.tmpname() if is_windows and not outfile:find(':') then outfile = os.getenv('TEMP')..outfile errfile = os.getenv('TEMP')..errfile end cmd = cmd .. " > " .. utils.quote_arg(outfile) .. " 2> " .. utils.quote_arg(errfile) local success, retcode = utils.execute(cmd) local outcontent = utils.readfile(outfile, bin) local errcontent = utils.readfile(errfile, bin) os.remove(outfile) os.remove(errfile) return success, retcode, (outcontent or ""), (errcontent or "") end --- Quote and escape an argument of a command. -- Quotes a single (or list of) argument(s) of a command to be passed -- to `os.execute`, `pl.utils.execute` or `pl.utils.executeex`. -- @param argument (string or table/list) the argument to quote. If a list then -- all arguments in the list will be returned as a single string quoted. -- @return quoted and escaped argument. -- @usage -- local options = utils.quote_arg { -- "-lluacov", -- "-e", -- "utils = print(require('pl.utils')._VERSION", -- } -- -- returns: -lluacov -e 'utils = print(require('\''pl.utils'\'')._VERSION' function utils.quote_arg(argument) if type(argument) == "table" then -- encode an entire table local r = {} for i, arg in ipairs(argument) do r[i] = utils.quote_arg(arg) end return concat(r, " ") end -- only a single argument if is_windows then if argument == "" or argument:find('[ \f\t\v]') then -- Need to quote the argument. -- Quotes need to be escaped with backslashes; -- additionally, backslashes before a quote, escaped or not, -- need to be doubled. -- See documentation for CommandLineToArgvW Windows function. argument = '"' .. argument:gsub([[(\*)"]], [[%1%1\"]]):gsub([[\+$]], "%0%0") .. '"' end -- os.execute() uses system() C function, which on Windows passes command -- to cmd.exe. Escape its special characters. return (argument:gsub('["^<>!|&%%]', "^%0")) else if argument == "" or argument:find('[^a-zA-Z0-9_@%+=:,./-]') then -- To quote arguments on posix-like systems use single quotes. -- To represent an embedded single quote close quoted string ('), -- add escaped quote (\'), open quoted string again ('). argument = "'" .. argument:gsub("'", [['\'']]) .. "'" end return argument end end --- error out of this program gracefully. -- @param[opt] code The exit code, defaults to -`1` if omitted -- @param msg The exit message will be sent to `stderr` (will be formatted with the extra parameters) -- @param ... extra arguments for message's format' -- @see utils.fprintf -- @usage utils.quit(-1, "Error '%s' happened", "42") -- -- is equivalent to -- utils.quit("Error '%s' happened", "42") --> Error '42' happened function utils.quit(code, msg, ...) if type(code) == 'string' then utils.fprintf(io.stderr, code, msg, ...) io.stderr:write('\n') code = -1 -- TODO: this is odd, see the test. Which returns 255 as exit code elseif msg then utils.fprintf(io.stderr, msg, ...) io.stderr:write('\n') end os.exit(code, true) end --- String functions -- @section string-functions --- escape any Lua 'magic' characters in a string -- @param s The input string function utils.escape(s) utils.assert_string(1,s) return (s:gsub('[%-%.%+%[%]%(%)%$%^%%%?%*]','%%%1')) end --- split a string into a list of strings separated by a delimiter. -- @param s The input string -- @param re optional A Lua string pattern; defaults to '%s+' -- @param plain optional If truthy don't use Lua patterns -- @param n optional maximum number of elements (if there are more, the last will remian un-split) -- @return a list-like table -- @raise error if s is not a string -- @see splitv function utils.split(s,re,plain,n) utils.assert_string(1,s) local i1,ls = 1,{} if not re then re = '%s+' end if re == '' then return {s} end while true do local i2,i3 = find(s,re,i1,plain) if not i2 then local last = sub(s,i1) if last ~= '' then append(ls,last) end if #ls == 1 and ls[1] == '' then return {} else return ls end end append(ls,sub(s,i1,i2-1)) if n and #ls == n then ls[#ls] = sub(s,i1) return ls end i1 = i3+1 end end --- split a string into a number of return values. -- Identical to `split` but returns multiple sub-strings instead of -- a single list of sub-strings. -- @param s the string -- @param re A Lua string pattern; defaults to '%s+' -- @param plain don't use Lua patterns -- @param n optional maximum number of splits -- @return n values -- @usage first,next = splitv('user=jane=doe','=', false, 2) -- assert(first == "user") -- assert(next == "jane=doe") -- @see split function utils.splitv (s,re, plain, n) return _unpack(utils.split(s,re, plain, n)) end --- Functional -- @section functional --- 'memoize' a function (cache returned value for next call). -- This is useful if you have a function which is relatively expensive, -- but you don't know in advance what values will be required, so -- building a table upfront is wasteful/impossible. -- @param func a function of at least one argument -- @return a function with at least one argument, which is used as the key. function utils.memoize(func) local cache = {} return function(k) local res = cache[k] if res == nil then res = func(k) cache[k] = res end return res end end --- associate a function factory with a type. -- A function factory takes an object of the given type and -- returns a function for evaluating it -- @tab mt metatable -- @func fun a callable that returns a function function utils.add_function_factory (mt,fun) _function_factories[mt] = fun end local function _string_lambda(f) if f:find '^|' or f:find '_' then local args,body = f:match '|([^|]*)|(.+)' if f:find '_' then args = '_' body = f else if not args then return raise 'bad string lambda' end end local fstr = 'return function('..args..') return '..body..' end' local fn,err = utils.load(fstr) if not fn then return raise(err) end fn = fn() return fn else return raise 'not a string lambda' end end --- an anonymous function as a string. This string is either of the form -- '|args| expression' or is a function of one argument, '_' -- @param lf function as a string -- @return a function -- @function utils.string_lambda -- @usage -- string_lambda '|x|x+1' (2) == 3 -- string_lambda '_+1' (2) == 3 utils.string_lambda = utils.memoize(_string_lambda) --- bind the first argument of the function to a value. -- @param fn a function of at least two values (may be an operator string) -- @param p a value -- @return a function such that f(x) is fn(p,x) -- @raise same as @{function_arg} -- @see func.bind1 -- @usage local function f(msg, name) -- print(msg .. " " .. name) -- end -- -- local hello = utils.bind1(f, "Hello") -- -- print(hello("world")) --> "Hello world" -- print(hello("sunshine")) --> "Hello sunshine" function utils.bind1 (fn,p) fn = utils.function_arg(1,fn) return function(...) return fn(p,...) end end --- bind the second argument of the function to a value. -- @param fn a function of at least two values (may be an operator string) -- @param p a value -- @return a function such that f(x) is fn(x,p) -- @raise same as @{function_arg} -- @usage local function f(a, b, c) -- print(a .. " " .. b .. " " .. c) -- end -- -- local hello = utils.bind1(f, "world") -- -- print(hello("Hello", "!")) --> "Hello world !" -- print(hello("Bye", "?")) --> "Bye world ?" function utils.bind2 (fn,p) fn = utils.function_arg(1,fn) return function(x,...) return fn(x,p,...) end end --- Deprecation -- @section deprecation do -- the default implementation local deprecation_func = function(msg, trace) if trace then warn(msg, "\n", trace) -- luacheck: ignore else warn(msg) -- luacheck: ignore end end --- Sets a deprecation warning function. -- An application can override this function to support proper output of -- deprecation warnings. The warnings can be generated from libraries or -- functions by calling `utils.raise_deprecation`. The default function -- will write to the 'warn' system (introduced in Lua 5.4, or the compatibility -- function from the `compat` module for earlier versions). -- -- Note: only applications should set/change this function, libraries should not. -- @param func a callback with signature: `function(msg, trace)` both arguments are strings, the latter being optional. -- @see utils.raise_deprecation -- @usage -- -- write to the Nginx logs with OpenResty -- utils.set_deprecation_func(function(msg, trace) -- ngx.log(ngx.WARN, msg, (trace and (" " .. trace) or nil)) -- end) -- -- -- disable deprecation warnings -- utils.set_deprecation_func() function utils.set_deprecation_func(func) if func == nil then deprecation_func = function() end else utils.assert_arg(1, func, "function") deprecation_func = func end end --- raises a deprecation warning. -- For options see the usage example below. -- -- Note: the `opts.deprecated_after` field is the last version in which -- a feature or option was NOT YET deprecated! Because when writing the code it -- is quite often not known in what version the code will land. But the last -- released version is usually known. -- @param opts options table -- @see utils.set_deprecation_func -- @usage -- warn("@on") -- enable Lua warnings, they are usually off by default -- -- function stringx.islower(str) -- raise_deprecation { -- source = "Penlight " .. utils._VERSION, -- optional -- message = "function 'islower' was renamed to 'is_lower'", -- required -- version_removed = "2.0.0", -- optional -- deprecated_after = "1.2.3", -- optional -- no_trace = true, -- optional -- } -- return stringx.is_lower(str) -- end -- -- output: "[Penlight 1.9.2] function 'islower' was renamed to 'is_lower' (deprecated after 1.2.3, scheduled for removal in 2.0.0)" function utils.raise_deprecation(opts) utils.assert_arg(1, opts, "table") if type(opts.message) ~= "string" then error("field 'message' of the options table must be a string", 2) end local trace if not opts.no_trace then trace = debug.traceback("", 2):match("[\n%s]*(.-)$") end local msg if opts.deprecated_after and opts.version_removed then msg = (" (deprecated after %s, scheduled for removal in %s)"):format( tostring(opts.deprecated_after), tostring(opts.version_removed)) elseif opts.deprecated_after then msg = (" (deprecated after %s)"):format(tostring(opts.deprecated_after)) elseif opts.version_removed then msg = (" (scheduled for removal in %s)"):format(tostring(opts.version_removed)) else msg = "" end msg = opts.message .. msg if opts.source then msg = "[" .. opts.source .."] " .. msg else if msg:sub(1,1) == "@" then -- in Lua 5.4 "@" prefixed messages are control messages to the warn system error("message cannot start with '@'", 2) end end deprecation_func(msg, trace) end end return utils Penlight-1.12.0/lua/pl/xml.lua000066400000000000000000001044321416703176500160620ustar00rootroot00000000000000--- XML LOM Utilities. -- -- This implements some useful things on [LOM](http://matthewwild.co.uk/projects/luaexpat/lom.html) documents, such as returned by `lxp.lom.parse`. -- In particular, it can convert LOM back into XML text, with optional pretty-printing control. -- It is based on stanza.lua from [Prosody](http://hg.prosody.im/trunk/file/4621c92d2368/util/stanza.lua) -- -- > d = xml.parse "alice" -- > = d -- alice -- > = xml.tostring(d,'',' ') -- -- alice -- -- -- Can be used as a lightweight one-stop-shop for simple XML processing; a simple XML parser is included -- but the default is to use `lxp.lom` if it can be found. --
-- Prosody IM
-- Copyright (C) 2008-2010 Matthew Wild
-- Copyright (C) 2008-2010 Waqas Hussain--
-- classic Lua XML parser by Roberto Ierusalimschy.
-- modified to output LOM format.
-- http://lua-users.org/wiki/LuaXml
-- 
-- See @{06-data.md.XML|the Guide} -- -- Dependencies: `pl.utils` -- -- Soft Dependencies: `lxp.lom` (fallback is to use basic Lua parser) -- @module pl.xml local utils = require 'pl.utils' local split = utils.split local t_insert = table.insert local t_concat = table.concat local t_remove = table.remove local s_match = string.match local tostring = tostring local setmetatable = setmetatable local getmetatable = getmetatable local pairs = pairs local ipairs = ipairs local type = type local next = next local print = print local unpack = utils.unpack local s_gsub = string.gsub local s_sub = string.sub local s_find = string.find local pcall = pcall local require = require utils.raise_deprecation { source = "Penlight " .. utils._VERSION, message = "the contents of module 'pl.xml' has been deprecated, please use a more specialized library instead", version_removed = "2.0.0", deprecated_after = "1.11.0", no_trace = true, } local _M = {} local Doc = { __type = "doc" }; Doc.__index = Doc; local function is_text(s) return type(s) == 'string' end local function is_tag(d) return type(d) == 'table' and is_text(d.tag) end --- create a new document node. -- @tparam string tag the tag name -- @tparam[opt={}] table attr attributes (table of name-value pairs) -- @return the Node object -- @see xml.elem -- @usage -- local doc = xml.new("main", { hello = "world", answer = "42" }) -- print(doc) -->
function _M.new(tag, attr) if type(tag) ~= "string" then error("expected 'tag' to be a string value, got: " .. type(tag), 2) end attr = attr or {} if type(attr) ~= "table" then error("expected 'attr' to be a table value, got: " .. type(attr), 2) end local doc = { tag = tag, attr = attr, last_add = {}}; return setmetatable(doc, Doc); end --- parse an XML document. By default, this uses lxp.lom.parse, but -- falls back to basic_parse, or if `use_basic` is truthy -- @param text_or_filename file or string representation -- @param is_file whether text_or_file is a file name or not -- @param use_basic do a basic parse -- @return a parsed LOM document with the document metatatables set -- @return nil, error the error can either be a file error or a parse error function _M.parse(text_or_filename, is_file, use_basic) local parser,status,lom if use_basic then parser = _M.basic_parse else status,lom = pcall(require,'lxp.lom') if not status then parser = _M.basic_parse else parser = lom.parse end end if is_file then local text_or_filename, err = utils.readfile(text_or_filename) if not text_or_filename then return nil, err end end local doc, err = parser(text_or_filename) if not doc then return nil, err end if lom then _M.walk(doc, false, function(_, d) setmetatable(d, Doc) end) end return doc end --- Create a Node with a set of children (text or Nodes) and attributes. -- @tparam string tag a tag name -- @tparam table|string items either a single child (text or Node), or a table where the hash -- part is the attributes and the list part is the children (text or Nodes). -- @return the new Node -- @see xml.new -- @see xml.tags -- @usage -- local doc = xml.elem("top", "hello world") -- hello world -- local doc = xml.elem("main", xml.new("child")) --
-- local doc = xml.elem("main", { "this ", "is ", "nice" }) --
this is nice
-- local doc = xml.elem("main", { xml.new "this", -- xml.new "is", -- xml.new "nice" }) --
-- local doc = xml.elem("main", { hello = "world" }) --
-- local doc = xml.elem("main", { -- "prefix", -- xml.elem("child", { "this ", "is ", "nice"}), -- "postfix", -- attrib = "value" -- }) --
prefixthis is nicepostfix
" function _M.elem(tag, items) local s = _M.new(tag) if is_text(items) then items = {items} end if is_tag(items) then t_insert(s,items) elseif type(items) == 'table' then for k,v in pairs(items) do if is_text(k) then s.attr[k] = v t_insert(s.attr,k) else s[k] = v end end end return s end --- given a list of names, return a number of element constructors. -- If passing a comma-separated string, then whitespace surrounding the values -- will be stripped. -- -- The returned constructor functions are a shortcut to `xml.elem` where you -- no longer provide the tag-name, but only the `items` table. -- @tparam string|table list a list of names, or a comma-separated string. -- @return (multiple) constructor functions; `function(items)`. For the `items` -- parameter see `xml.elem`. -- @see xml.elem -- @usage -- local new_parent, new_child = xml.tags 'mom, kid' -- doc = new_parent {new_child 'Bob', new_child 'Annie'} -- -- BobAnnie function _M.tags(list) local ctors = {} if is_text(list) then list = split(list:match("^%s*(.-)%s*$"),'%s*,%s*') end for i,tag in ipairs(list) do local function ctor(items) return _M.elem(tag,items) end ctors[i] = ctor end return unpack(ctors) end --- Adds a document Node, at current position. -- This updates the last inserted position to the new Node. -- @tparam string tag the tag name -- @tparam[opt={}] table attrs attributes (table of name-value pairs) -- @return the current node (`self`) -- @usage -- local doc = xml.new("main") -- doc:addtag("penlight", { hello = "world"}) -- doc:addtag("expat") -- added to 'penlight' since position moved -- print(doc) -->
function Doc:addtag(tag, attrs) local s = _M.new(tag, attrs) self:add_child(s) t_insert(self.last_add, s) return self end --- Adds a text node, at current position. -- @tparam string text a string -- @return the current node (`self`) -- @usage -- local doc = xml.new("main") -- doc:text("penlight") -- doc:text("expat") -- print(doc) -->
function Doc:text(text) self:add_child(text) return self end --- Moves current position up one level. -- @return the current node (`self`) function Doc:up() t_remove(self.last_add) return self end --- Resets current position to top level. -- Resets to the `self` node. -- @return the current node (`self`) function Doc:reset() local last_add = self.last_add for i = 1,#last_add do last_add[i] = nil end return self end --- Append a child to the currrent Node (ignoring current position). -- @param child a child node (either text or a document) -- @return the current node (`self`) -- @usage -- local doc = xml.new("main") -- doc:add_direct_child("dog") -- doc:add_direct_child(xml.new("child")) -- doc:add_direct_child("cat") -- print(doc) -->
dogcat
function Doc:add_direct_child(child) t_insert(self, child) return self end --- Append a child at the current position (without changing position). -- @param child a child node (either text or a document) -- @return the current node (`self`) -- @usage -- local doc = xml.new("main") -- doc:addtag("one") -- doc:add_child(xml.new("item1")) -- doc:add_child(xml.new("item2")) -- doc:add_child(xml.new("item3")) -- print(doc) -->
function Doc:add_child(child) (self.last_add[#self.last_add] or self):add_direct_child(child) return self end --accessing attributes: useful not to have to expose implementation (attr) --but also can allow attr to be nil in any future optimizations --- Set attributes of a document node. -- Will add/overwite values, but will not remove existing ones. -- Operates on the Node itself, will not take position into account. -- @tparam table t a table containing attribute/value pairs -- @return the current node (`self`) function Doc:set_attribs(t) -- TODO: keep array part in sync for k,v in pairs(t) do self.attr[k] = v end return self end --- Set a single attribute of a document node. -- Operates on the Node itself, will not take position into account. -- @param a attribute -- @param v its value, pass in `nil` to delete the attribute -- @return the current node (`self`) function Doc:set_attrib(a,v) -- TODO: keep array part in sync self.attr[a] = v return self end --- Gets the attributes of a document node. -- Operates on the Node itself, will not take position into account. -- @return table with attributes (attribute/value pairs) function Doc:get_attribs() return self.attr end local template_cache do local templ_cache = {} -- @param templ a template, a string being valid xml to be parsed, or a Node object function template_cache(templ) if is_text(templ) then if templ_cache[templ] then -- cache hit return templ_cache[templ] else -- parse and cache local ptempl, err = _M.parse(templ,false,true) if not ptempl then return nil, err end templ_cache[templ] = ptempl return ptempl end end if is_tag(templ) then return templ end return nil, "template is not a document" end end do local function is_data(data) return #data == 0 or type(data[1]) ~= 'table' end local function prepare_data(data) -- a hack for ensuring that $1 maps to first element of data, etc. -- Either this or could change the gsub call just below. for i,v in ipairs(data) do data[tostring(i)] = v end end --- create a substituted copy of a document, -- @param template may be a document or a string representation which will be parsed and cached -- @param data a table of name-value pairs or a list of such tables -- @return an XML document function Doc.subst(template, data) if type(data) ~= 'table' or not next(data) then return nil, "data must be a non-empty table" end if is_data(data) then prepare_data(data) end local templ, err = template_cache(template) if err then return nil, err end local function _subst(item) return _M.clone(templ, function(s) return s:gsub('%$(%w+)', item) end) end if is_data(data) then return _subst(data) end local list = {} for _, item in ipairs(data) do prepare_data(item) t_insert(list, _subst(item)) end if data.tag then list = _M.elem(data.tag,list) end return list end end --- Return the first child with a given tag name (non-recursive). -- @param tag the tag name -- @return the child Node found or `nil` if not found function Doc:child_with_name(tag) for _, child in ipairs(self) do if child.tag == tag then return child end end end do -- @param self document node to traverse -- @param tag tag-name to look for -- @param list array table to add the matching ones to -- @param recurse if truthy, recursivly search the node local function _children_with_name(self, tag, list, recurse) -- TODO: protect against recursion for _, child in ipairs(self) do if type(child) == 'table' then if child.tag == tag then t_insert(list, child) end if recurse then _children_with_name(child, tag, list, recurse) end end end end --- Returns all elements in a document that have a given tag. -- @tparam string tag a tag name -- @tparam[opt=false] boolean dont_recurse optionally only return the immediate children with this tag name -- @return a list of elements found, list will be empty if none was found. function Doc:get_elements_with_name(tag, dont_recurse) local res = {} _children_with_name(self, tag, res, not dont_recurse) return res end end --- Iterator over all children of a document node, including text nodes. -- This function is not recursive, so returns only direct child nodes. -- @return iterator that returns a single Node per iteration. function Doc:children() local i = 0; return function (a) i = i + 1 return a[i]; end, self, i; end --- Return the first child element of a node, if it exists. -- This will skip text nodes. -- @return first child Node or `nil` if there is none. function Doc:first_childtag() if #self == 0 then return end for _, t in ipairs(self) do if is_tag(t) then return t end end end --- Iterator that matches tag names, and a namespace (non-recursive). -- @tparam[opt=nil] string tag tag names to return. Returns all tags if not provided. -- @tparam[opt=nil] string xmlns the namespace value ('xmlns' attribute) to return. If not -- provided will match all namespaces. -- @return iterator that returns a single Node per iteration. function Doc:matching_tags(tag, xmlns) -- TODO: this doesn't make sense??? namespaces are not "xmnls", as matched below -- but "xmlns:name"... so should be a string-prefix match if anything... xmlns = xmlns or self.attr.xmlns; local tags = self local next_i = 1 local max_i = #tags local node return function () for i = next_i, max_i do node = tags[i]; if (not tag or node.tag == tag) and (not xmlns or xmlns == node.attr.xmlns) then next_i = i + 1 return node end end end, tags, next_i end --- Iterator over all child tags of a document node. This will skip over -- text nodes. -- @return iterator that returns a single Node per iteration. function Doc:childtags() local i = 0; return function (a) local v repeat i = i + 1 v = self[i] if v and type(v) == 'table' then return v end until not v end, self[1], i; end --- Visit child Nodes of a node and call a function, possibly modifying the document. -- Text elements will be skipped. -- This is not recursive, so only direct children will be passed. -- @tparam function callback a function with signature `function(node)`, passed the node. -- The element will be updated with the returned value, or deleted if it returns `nil`. function Doc:maptags(callback) local i = 1; while i <= #self do if is_tag(self[i]) then local ret = callback(self[i]); if ret == nil then -- remove it t_remove(self, i); else -- update it self[i] = ret; i = i + 1; end else i = i + 1 end end return self; end do local escape_table = { ["'"] = "'", ['"'] = """, ["<"] = "<", [">"] = ">", ["&"] = "&", } --- Escapes a string for safe use in xml. -- Handles quotes(single+double), less-than, greater-than, and ampersand. -- @tparam string str string value to escape -- @return escaped string -- @usage -- local esc = xml.xml_escape([["'<>&]]) --> ""'<>&" function _M.xml_escape(str) return (s_gsub(str, "['&<>\"]", escape_table)) end end local xml_escape = _M.xml_escape do local escape_table = { quot = '"', apos = "'", lt = "<", gt = ">", amp = "&", } --- Unescapes a string from xml. -- Handles quotes(single+double), less-than, greater-than, and ampersand. -- @tparam string str string value to unescape -- @return unescaped string -- @usage -- local unesc = xml.xml_escape(""'<>&") --> [["'<>&]] function _M.xml_unescape(str) return (str:gsub( "&(%a+);", escape_table)) end end local xml_unescape = _M.xml_unescape -- pretty printing -- if indent, then put each new tag on its own line -- if attr_indent, put each new attribute on its own line local function _dostring(t, buf, parentns, block_indent, tag_indent, attr_indent) local nsid = 0 local tag = t.tag local lf = "" if tag_indent then lf = '\n'..block_indent end local alf = " " if attr_indent then alf = '\n'..block_indent..attr_indent end t_insert(buf, lf.."<"..tag) local function write_attr(k,v) if s_find(k, "\1", 1, true) then nsid = nsid + 1 local ns, attrk = s_match(k, "^([^\1]*)\1?(.*)$") t_insert(buf, " xmlns:ns"..nsid.."='"..xml_escape(ns).."' ".."ns"..nsid..":"..attrk.."='"..xml_escape(v).."'") elseif not (k == "xmlns" and v == parentns) then t_insert(buf, alf..k.."='"..xml_escape(v).."'"); end end -- it's useful for testing to have predictable attribute ordering, if available if #t.attr > 0 then -- TODO: the key-value list is leading, what if they are not in-sync for _,k in ipairs(t.attr) do write_attr(k,t.attr[k]) end else for k, v in pairs(t.attr) do write_attr(k,v) end end local len = #t local has_children if len == 0 then t_insert(buf, attr_indent and '\n'..block_indent.."/>" or "/>") else t_insert(buf, ">"); for n = 1, len do local child = t[n] if child.tag then has_children = true _dostring(child, buf, t.attr.xmlns, block_indent and block_indent..tag_indent, tag_indent, attr_indent) else -- text element t_insert(buf, xml_escape(child)) end end t_insert(buf, (has_children and lf or '')..""); end end --- Function to pretty-print an XML document. -- @param doc an XML document -- @tparam[opt] string|int b_ind an initial block-indent (required when `t_ind` is set) -- @tparam[opt] string|int t_ind an tag-indent for each level (required when `a_ind` is set) -- @tparam[opt] string|int a_ind if given, indent each attribute pair and put on a separate line -- @tparam[opt] string|bool xml_preface force prefacing with default or custom , if truthy then `<?xml version='1.0'?>` will be used as default. -- @return a string representation -- @see Doc:tostring function _M.tostring(doc, b_ind, t_ind, a_ind, xml_preface) local buf = {} if type(b_ind) == "number" then b_ind = (" "):rep(b_ind) end if type(t_ind) == "number" then t_ind = (" "):rep(t_ind) end if type(a_ind) == "number" then a_ind = (" "):rep(a_ind) end if xml_preface then if type(xml_preface) == "string" then buf[1] = xml_preface else buf[1] = "" end end _dostring(doc, buf, nil, b_ind, t_ind, a_ind, xml_preface) return t_concat(buf) end Doc.__tostring = _M.tostring --- Method to pretty-print an XML document. -- Invokes `xml.tostring`. -- @tparam[opt] string|int b_ind an initial indent (required when `t_ind` is set) -- @tparam[opt] string|int t_ind an indent for each level (required when `a_ind` is set) -- @tparam[opt] string|int a_ind if given, indent each attribute pair and put on a separate line -- @tparam[opt="<?xml version='1.0'?>"] string xml_preface force prefacing with default or custom -- @return a string representation -- @see xml.tostring function Doc:tostring(b_ind, t_ind, a_ind, xml_preface) return _M.tostring(self, b_ind, t_ind, a_ind, xml_preface) end --- get the full text value of an element. -- @return a single string with all text elements concatenated -- @usage -- local doc = xml.new("main") -- doc:text("one") -- doc:add_child(xml.elem "two") -- doc:text("three") -- -- local t = doc:get_text() --> "onethree" function Doc:get_text() local res = {} for i,el in ipairs(self) do if is_text(el) then t_insert(res,el) end end return t_concat(res); end do local function _copy(object, kind, parent, strsubst, lookup_table) if type(object) ~= "table" then if strsubst and is_text(object) then return strsubst(object, kind, parent) else return object end end if lookup_table[object] then error("recursion detected") end lookup_table[object] = true local new_table = {} lookup_table[object] = new_table local tag = object.tag new_table.tag = _copy(tag, '*TAG', parent, strsubst, lookup_table) if object.attr then local res = {} for attr, value in pairs(object.attr) do if type(attr) == "string" then res[attr] = _copy(value, attr, object, strsubst, lookup_table) end end new_table.attr = res end for index = 1, #object do local v = _copy(object[index], '*TEXT', object, strsubst, lookup_table) t_insert(new_table,v) end return setmetatable(new_table, getmetatable(object)) end --- Returns a copy of a document. -- The `strsubst` parameter is a callback with signature `function(object, kind, parent)`. -- -- Param `kind` has the following values, and parameters: -- -- - `"*TAG"`: `object` is the tag-name, `parent` is the Node object. Returns the new tag name. -- -- - `"*TEXT"`: `object` is the text-element, `parent` is the Node object. Returns the new text value. -- -- - other strings not prefixed with `*`: `kind` is the attribute name, `object` is the -- attribute value, `parent` is the Node object. Returns the new attribute value. -- -- @tparam Node|string doc a Node object or string (text node) -- @tparam[opt] function strsubst an optional function for handling string copying -- which could do substitution, etc. -- @return copy of the document -- @see Doc:filter function _M.clone(doc, strsubst) return _copy(doc, nil, nil, strsubst, {}) end end --- Returns a copy of a document. -- This is the method version of `xml.clone`. -- @see xml.clone -- @name Doc:filter -- @tparam[opt] function strsubst an optional function for handling string copying Doc.filter = _M.clone -- also available as method do local function _compare(t1, t2, recurse_check) local ty1 = type(t1) local ty2 = type(t2) if ty1 ~= ty2 then return false, 'type mismatch' end if ty1 == 'string' then if t1 == t2 then return true else return false, 'text '..t1..' ~= text '..t2 end end if ty1 ~= 'table' or ty2 ~= 'table' then return false, 'not a document' end if recurse_check[t1] then return false, "recursive document" end recurse_check[t1] = true if t1.tag ~= t2.tag then return false, 'tag '..t1.tag..' ~= tag '..t2.tag end if #t1 ~= #t2 then return false, 'size '..#t1..' ~= size '..#t2..' for tag '..t1.tag end -- compare attributes for k,v in pairs(t1.attr) do local t2_value = t2.attr[k] if type(k) == "string" then if t2_value ~= v then return false, 'mismatch attrib' end else if t2_value ~= nil and t2_value ~= v then return false, "mismatch attrib order" end end end for k,v in pairs(t2.attr) do local t1_value = t1.attr[k] if type(k) == "string" then if t1_value ~= v then return false, 'mismatch attrib' end else if t1_value ~= nil and t1_value ~= v then return false, "mismatch attrib order" end end end -- compare children for i = 1, #t1 do local ok, err = _compare(t1[i], t2[i], recurse_check) if not ok then return ok, err end end return true end --- Compare two documents or elements. -- Equality is based on tag, child nodes (text and tags), attributes and order -- of those (order only fails if both are given, and not equal). -- @tparam Node|string t1 a Node object or string (text node) -- @tparam Node|string t2 a Node object or string (text node) -- @treturn boolean `true` when the Nodes are equal. function _M.compare(t1,t2) return _compare(t1, t2, {}) end end --- is this value a document element? -- @param d any value -- @treturn boolean `true` if it is a `table` with property `tag` being a string value. -- @name is_tag _M.is_tag = is_tag do local function _walk(doc, depth_first, operation, recurse_check) if not depth_first then operation(doc.tag, doc) end for _,d in ipairs(doc) do if is_tag(d) then assert(not recurse_check[d], "recursion detected") recurse_check[d] = true _walk(d, depth_first, operation, recurse_check) end end if depth_first then operation(doc.tag, doc) end end --- Calls a function recursively over Nodes in the document. -- Will only call on tags, it will skip text nodes. -- The function signature for `operation` is `function(tag_name, Node)`. -- @tparam Node|string doc a Node object or string (text node) -- @tparam boolean depth_first visit child nodes first, then the current node -- @tparam function operation a function which will receive the current tag name and current node. function _M.walk(doc, depth_first, operation) return _walk(doc, depth_first, operation, {}) end end local html_empty_elements = { --lists all HTML empty (void) elements br = true, img = true, meta = true, frame = true, area = true, hr = true, base = true, col = true, link = true, input = true, option = true, param = true, isindex = true, embed = true, } --- Parse a well-formed HTML file as a string. -- Tags are case-insenstive, DOCTYPE is ignored, and empty elements can be .. empty. -- @param s the HTML function _M.parsehtml(s) return _M.basic_parse(s,false,true) end --- Parse a simple XML document using a pure Lua parser based on Robero Ierusalimschy's original version. -- @param s the XML document to be parsed. -- @param all_text if true, preserves all whitespace. Otherwise only text containing non-whitespace is included. -- @param html if true, uses relaxed HTML rules for parsing function _M.basic_parse(s, all_text, html) local stack = {} local top = {} local function parseargs(s) local arg = {} s:gsub("([%w:%-_]+)%s*=%s*([\"'])(.-)%2", function (w, _, a) if html then w = w:lower() end arg[w] = xml_unescape(a) end) if html then s:gsub("([%w:%-_]+)%s*=%s*([^\"']+)%s*", function (w, a) w = w:lower() arg[w] = xml_unescape(a) end) end return arg end t_insert(stack, top) local ni,c,label,xarg, empty, _, istart local i = 1 local j -- we're not interested in _,istart = s_find(s,'^%s*<%?[^%?]+%?>%s*') if not istart then -- or _,istart = s_find(s,'^%s*%s*') end if istart then i = istart+1 end while true do ni,j,c,label,xarg, empty = s_find(s, "<([%/!]?)([%w:%-_]+)(.-)(%/?)>", i) if not ni then break end if c == "!" then -- comment -- case where there's no space inside comment if not (label:match '%-%-$' and xarg == '') then if xarg:match '%-%-$' then -- we've grabbed it all j = j - 2 end -- match end of comment _,j = s_find(s, "-->", j, true) end else local text = s_sub(s, i, ni-1) if html then label = label:lower() if html_empty_elements[label] then empty = "/" end end if all_text or not s_find(text, "^%s*$") then t_insert(top, xml_unescape(text)) end if empty == "/" then -- empty element tag t_insert(top, setmetatable({tag=label, attr=parseargs(xarg), empty=1},Doc)) elseif c == "" then -- start tag top = setmetatable({tag=label, attr=parseargs(xarg)},Doc) t_insert(stack, top) -- new level else -- end tag local toclose = t_remove(stack) -- remove top top = stack[#stack] if #stack < 1 then error("nothing to close with "..label..':'..text) end if toclose.tag ~= label then error("trying to close "..toclose.tag.." with "..label.." "..text) end t_insert(top, toclose) end end i = j+1 end local text = s_sub(s, i) if all_text or not s_find(text, "^%s*$") then t_insert(stack[#stack], xml_unescape(text)) end if #stack > 1 then error("unclosed "..stack[#stack].tag) end local res = stack[1] return is_text(res[1]) and res[2] or res[1] end do local match do local function empty(attr) return not attr or not next(attr) end local append_capture do -- returns the key,value pair from a table if it has exactly one entry local function has_one_element(t) local key,value = next(t) if next(t,key) ~= nil then return false end return key,value end function append_capture(res,tbl) if not empty(tbl) then -- no point in capturing empty tables... local key if tbl._ then -- if $_ was set then it is meant as the top-level key for the captured table key = tbl._ tbl._ = nil if empty(tbl) then return end end -- a table with only one pair {[0]=value} shall be reduced to that value local numkey,val = has_one_element(tbl) if numkey == 0 then tbl = val end if key then res[key] = tbl else -- otherwise, we append the captured table t_insert(res,tbl) end end end end local function make_number(pat) if pat:find '^%d+$' then -- $1 etc means use this as an array location pat = tonumber(pat) end return pat end local function capture_attrib(res,pat,value) pat = make_number(pat:sub(2)) res[pat] = value return true end function match(d,pat,res,keep_going) local ret = true if d == nil then d = '' end --return false end -- attribute string matching is straight equality, except if the pattern is a $ capture, -- which always succeeds. if is_text(d) then if not is_text(pat) then return false end if _M.debug then print(d,pat) end if pat:find '^%$' then return capture_attrib(res,pat,d) else return d == pat end else if _M.debug then print(d.tag,pat.tag) end -- this is an element node. For a match to succeed, the attributes must -- match as well. -- a tagname in the pattern ending with '-' is a wildcard and matches like an attribute local tagpat = pat.tag:match '^(.-)%-$' if tagpat then tagpat = make_number(tagpat) res[tagpat] = d.tag end if d.tag == pat.tag or tagpat then if not empty(pat.attr) then if empty(d.attr) then ret = false else for prop,pval in pairs(pat.attr) do local dval = d.attr[prop] if not match(dval,pval,res) then ret = false; break end end end end -- the pattern may have child nodes. We match partially, so that {P1,P2} shall match {X,P1,X,X,P2,..} if ret and #pat > 0 then local i,j = 1,1 local function next_elem() j = j + 1 -- next child element of data if is_text(d[j]) then j = j + 1 end return j <= #d end repeat local p = pat[i] -- repeated {{<...>}} patterns shall match one or more elements -- so e.g. {P+} will match {X,X,P,P,X,P,X,X,X} if is_tag(p) and p.repeated then local found repeat local tbl = {} ret = match(d[j],p,tbl,false) if ret then found = false --true append_capture(res,tbl) end until not next_elem() or (found and not ret) i = i + 1 else ret = match(d[j],p,res,false) if ret then i = i + 1 end end until not next_elem() or i > #pat -- run out of elements or patterns to match -- if every element in our pattern matched ok, then it's been a successful match if i > #pat then return true end end if ret then return true end else ret = false end -- keep going anyway - look at the children! if keep_going then for child in d:childtags() do ret = match(child,pat,res,keep_going) if ret then break end end end end return ret end end --- does something... function Doc:match(pat) local err pat,err = template_cache(pat) if not pat then return nil, err end _M.walk(pat,false,function(_,d) if is_text(d[1]) and is_tag(d[2]) and is_text(d[3]) and d[1]:find '%s*{{' and d[3]:find '}}%s*' then t_remove(d,1) t_remove(d,2) d[1].repeated = true end end) local res = {} local ret = match(self,pat,res,true) return res,ret end end return _M Penlight-1.12.0/penlight-dev-1.rockspec000066400000000000000000000055431416703176500176450ustar00rootroot00000000000000local package_name = "penlight" local package_version = "dev" local rockspec_revision = "1" local github_account_name = "lunarmodules" local github_repo_name = package_name local git_checkout = package_version == "dev" and "master" or package_version rockspec_format = "3.0" package = package_name version = package_version .. "-" .. rockspec_revision source = { url = "git://github.com/"..github_account_name.."/"..github_repo_name..".git", branch = git_checkout } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", detailed = [[ Penlight is a set of pure Lua libraries focusing on input data handling (such as reading configuration files), functional programming (such as map, reduce, placeholder expressions,etc), and OS path management. Much of the functionality is inspired by the Python standard libraries. ]], license = "MIT/X11", homepage = "https://"..github_account_name..".github.io/"..github_repo_name, issues_url = "https://github.com/"..github_account_name.."/"..github_repo_name.."/issues", maintainer = "thijs@thijsschreijer.nl", } dependencies = { "lua >= 5.1", "luafilesystem" } test_dependencies = { "busted", } test = { type = "busted", } build = { type = "builtin", modules = { ["pl"] = "lua/pl/init.lua", ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.compat"] = "lua/pl/compat.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.types"] = "lua/pl/types.lua", ["pl.import_into"] = "lua/pl/import_into.lua" }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/000077500000000000000000000000001416703176500153535ustar00rootroot00000000000000Penlight-1.12.0/rockspecs/penlight-1.10.0-1.rockspec000066400000000000000000000051101416703176500215660ustar00rootroot00000000000000local package_name = "penlight" local package_version = "1.10.0" local rockspec_revision = "1" local github_account_name = "lunarmodules" local github_repo_name = package_name local git_checkout = package_version == "dev" and "master" or package_version package = package_name version = package_version .. "-" .. rockspec_revision source = { url = "git://github.com/"..github_account_name.."/"..github_repo_name..".git", branch = git_checkout } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "https://"..github_account_name..".github.io/"..github_repo_name, license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.11.0-1.rockspec000066400000000000000000000051101416703176500215670ustar00rootroot00000000000000local package_name = "penlight" local package_version = "1.11.0" local rockspec_revision = "1" local github_account_name = "lunarmodules" local github_repo_name = package_name local git_checkout = package_version == "dev" and "master" or package_version package = package_name version = package_version .. "-" .. rockspec_revision source = { url = "git://github.com/"..github_account_name.."/"..github_repo_name..".git", branch = git_checkout } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "https://"..github_account_name..".github.io/"..github_repo_name, license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.12.0-1.rockspec000066400000000000000000000051101416703176500215700ustar00rootroot00000000000000local package_name = "penlight" local package_version = "1.12.0" local rockspec_revision = "1" local github_account_name = "lunarmodules" local github_repo_name = package_name local git_checkout = package_version == "dev" and "master" or package_version package = package_name version = package_version .. "-" .. rockspec_revision source = { url = "git://github.com/"..github_account_name.."/"..github_repo_name..".git", branch = git_checkout } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "https://"..github_account_name..".github.io/"..github_repo_name, license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.6.0-1.rockspec000066400000000000000000000043401416703176500215170ustar00rootroot00000000000000package = "penlight" version = "1.6.0-1" source = { url = "git://github.com/Tieske/Penlight.git", branch = "1.6.0" } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "http://tieske.github.io/Penlight", license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.7.0-1.rockspec000066400000000000000000000043401416703176500215200ustar00rootroot00000000000000package = "penlight" version = "1.7.0-1" source = { url = "git://github.com/Tieske/Penlight.git", branch = "1.7.0" } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "http://tieske.github.io/Penlight", license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.8.0-1.rockspec000066400000000000000000000043401416703176500215210ustar00rootroot00000000000000package = "penlight" version = "1.8.0-1" source = { url = "git://github.com/Tieske/Penlight.git", branch = "1.8.0" } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "http://tieske.github.io/Penlight", license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.8.1-1.rockspec000066400000000000000000000043521416703176500215250ustar00rootroot00000000000000package = "penlight" version = "1.8.1-1" source = { url = "git://github.com/lunarmodules/Penlight.git", tag = "1.8.1" } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "https://lunarmodules.github.io/Penlight", license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.9.1-1.rockspec000066400000000000000000000043521416703176500215260ustar00rootroot00000000000000package = "penlight" version = "1.9.1-1" source = { url = "git://github.com/lunarmodules/Penlight.git", tag = "1.9.1" } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "https://lunarmodules.github.io/Penlight", license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/rockspecs/penlight-1.9.2-1.rockspec000066400000000000000000000043521416703176500215270ustar00rootroot00000000000000package = "penlight" version = "1.9.2-1" source = { url = "git://github.com/lunarmodules/Penlight.git", tag = "1.9.2" } description = { summary = "Lua utility libraries loosely based on the Python standard libraries", homepage = "https://lunarmodules.github.io/Penlight", license = "MIT/X11", maintainer = "thijs@thijsschreijer.nl", detailed = [[ Penlight is a set of pure Lua libraries for making it easier to work with common tasks like iterating over directories, reading configuration files and the like. Provides functional operations on tables and sequences. ]] } dependencies = { "luafilesystem", } build = { type = "builtin", modules = { ["pl.strict"] = "lua/pl/strict.lua", ["pl.dir"] = "lua/pl/dir.lua", ["pl.operator"] = "lua/pl/operator.lua", ["pl.input"] = "lua/pl/input.lua", ["pl.config"] = "lua/pl/config.lua", ["pl.compat"] = "lua/pl/config.lua", ["pl.seq"] = "lua/pl/seq.lua", ["pl.stringio"] = "lua/pl/stringio.lua", ["pl.text"] = "lua/pl/text.lua", ["pl.test"] = "lua/pl/test.lua", ["pl.tablex"] = "lua/pl/tablex.lua", ["pl.app"] = "lua/pl/app.lua", ["pl.stringx"] = "lua/pl/stringx.lua", ["pl.lexer"] = "lua/pl/lexer.lua", ["pl.utils"] = "lua/pl/utils.lua", ["pl.sip"] = "lua/pl/sip.lua", ["pl.permute"] = "lua/pl/permute.lua", ["pl.pretty"] = "lua/pl/pretty.lua", ["pl.class"] = "lua/pl/class.lua", ["pl.List"] = "lua/pl/List.lua", ["pl.data"] = "lua/pl/data.lua", ["pl.Date"] = "lua/pl/Date.lua", ["pl.init"] = "lua/pl/init.lua", ["pl.luabalanced"] = "lua/pl/luabalanced.lua", ["pl.comprehension"] = "lua/pl/comprehension.lua", ["pl.path"] = "lua/pl/path.lua", ["pl.array2d"] = "lua/pl/array2d.lua", ["pl.func"] = "lua/pl/func.lua", ["pl.lapp"] = "lua/pl/lapp.lua", ["pl.file"] = "lua/pl/file.lua", ['pl.template'] = "lua/pl/template.lua", ["pl.Map"] = "lua/pl/Map.lua", ["pl.MultiMap"] = "lua/pl/MultiMap.lua", ["pl.OrderedMap"] = "lua/pl/OrderedMap.lua", ["pl.Set"] = "lua/pl/Set.lua", ["pl.xml"] = "lua/pl/xml.lua", ["pl.url"] = "lua/pl/url.lua", ["pl.import_into"] = "lua/pl/import_into.lua", ["pl.types"] = "lua/pl/types.lua", }, copy_directories = {"docs", "tests"} } Penlight-1.12.0/run.lua000066400000000000000000000031161416703176500146670ustar00rootroot00000000000000-- Running tests and/or examples. local lfs = require "lfs" local directories = {} local luacov = false for _, argument in ipairs(arg) do if argument == "--help" then print("Usage: lua run.lua [--luacov] []...") os.exit(0) elseif argument == "--luacov" then luacov = true else table.insert(directories, argument) end end if #directories == 0 then directories = {"tests", "examples"} end local lua = "lua" local i = -1 while arg[i] do lua = arg[i] i = i - 1 end if luacov then lua = lua .. " -lluacov" end local dir_sep = package.config:sub(1, 1) local quote = dir_sep == "/" and "'" or '"' local pl_src = "lua/?.lua;lua/?/init.lua" lua = lua .. " -e " .. quote .. "package.path=[[" .. pl_src .. ";]]..package.path" .. quote local function run_directory(dir) local files = {} for path in lfs.dir(dir) do local full_path = dir .. dir_sep .. path if path:find("%.lua$") and lfs.attributes(full_path, "mode") == "file" then table.insert(files, full_path) end end table.sort(files) for _, file in ipairs(files) do local cmd = lua .. " " .. file print("Running " .. file) local code1, _, code2 = os.execute(cmd) local code = type(code1) == "number" and code1 or code2 if code ~= 0 then print(("Running %s failed with code %d"):format(file, code)) os.exit(1) end end end for _, dir in ipairs(directories) do print("Running files in " .. dir) run_directory(dir) end print("Run completed successfully") Penlight-1.12.0/spec/000077500000000000000000000000001416703176500143115ustar00rootroot00000000000000Penlight-1.12.0/spec/app_spec.lua000066400000000000000000000014141416703176500166060ustar00rootroot00000000000000local app = require("pl.app") describe("pl.app.lua", function () local invocation = app.lua() it("should pick up the arguments used to run this test", function () assert.is.truthy(invocation:match("lua.+package.+busted")) end) it("should be reusable to invoke Lua", function () assert.is.truthy(os.execute(app.lua()..' -e "n=1;os.exit(n-1)"')) end) end) describe("pl.app.platform", function () -- TODO: Find a reliable alternate way to determine platform to check that -- this is returning the right answer, not just any old answer. it("should at least return a valid platform", function () local platforms = { Linux = true, OSX = true, Windows = true } local detected = app.platform() assert.is.truthy(platforms[detected]) end) end) Penlight-1.12.0/spec/array2d_spec.lua000066400000000000000000000361361416703176500174030ustar00rootroot00000000000000local array2d = require("pl.array2d") describe("pl.array2d", function() describe("new()", function() it("creates an empty 2d array", function() assert.same({{},{},{}}, array2d.new(3,3,nil)) end) it("creates a value-filled 2d array", function() assert.same({{99,99,99}, {99,99,99}, {99,99,99}}, array2d.new(3,3,99)) end) it("creates a function-filled 2d array", function() assert.same({{2,3,4}, {3,4,5}, {4,5,6}}, array2d.new(3,3,function(i,j) return i+j end)) end) end) describe("size()", function() it("returns array size", function() local a = array2d.new(3,5,99) assert.same({3,5}, {array2d.size(a)}) end) it("returns 0 columns for nil arrays", function() local a = array2d.new(3,5,nil) assert.same({3,0}, {array2d.size(a)}) end) end) describe("column()", function() it("returns a column copy", function() local a = {{1,2}, {3,4}, {5,6}} assert.same({1,3,5}, array2d.column(a,1)) assert.same({2,4,6}, array2d.column(a,2)) end) end) describe("row()", function() it("returns a row copy", function() local a = {{1,2}, {3,4}, {5,6}} assert.same({1,2}, array2d.row(a,1)) -- next test: need to remove the metatable to prevent comparison by -- metamethods in Lua 5.3 and 5.4 assert.not_equal(a[1], setmetatable(array2d.row(a,1),nil)) assert.same({3,4}, array2d.row(a,2)) assert.same({5,6}, array2d.row(a,3)) end) end) describe("map()", function() it("maps a function on an array", function() local a1 = array2d.new(2,3,function(i,j) return i+j end) local a2 = array2d.map(function(a,b) return a .. b end, a1, "x") assert.same({{"2x","3x","4x"}, {"3x","4x","5x"}}, a2) end) end) describe("reduce_rows()", function() it("reduces rows", function() local a = {{ 1, 2, 3, 4}, { 10, 20, 30, 40}, { 100, 200, 300, 400}, {1000,2000,3000,4000}} assert.same({10,100,1000,10000},array2d.reduce_rows('+',a)) end) end) describe("reduce_cols()", function() it("reduces columns", function() local a = {{ 1, 2, 3, 4}, { 10, 20, 30, 40}, { 100, 200, 300, 400}, {1000,2000,3000,4000}} assert.same({1111,2222,3333,4444},array2d.reduce_cols('+',a)) end) end) describe("reduce2()", function() it("recuces array to scalar", function() local a = {{1,10}, {2,10}, {3,10}} assert.same(60, array2d.reduce2('+','*',a)) end) end) describe("map2()", function() it("maps over 2 arrays", function() local b = {{10,20}, {30,40}} local a = {{1,2}, {3,4}} -- 2 2d arrays assert.same({{11,22},{33,44}}, array2d.map2('+',2,2,a,b)) -- 1d, 2d assert.same({{11,102},{13,104}}, array2d.map2('+',1,2,{10,100},a)) -- 2d, 1d assert.same({{1,-2},{3,-4}},array2d.map2('*',2,1,a,{1,-1})) end) end) describe("product()", function() it("creates a product array", function() local a = array2d.product('..',{1,2,3},{'a','b','c'}) assert.same({{'1a','2a','3a'},{'1b','2b','3b'},{'1c','2c','3c'}}, a) local a = array2d.product('{}',{1,2},{'a','b','c'}) assert.same({{{1,'a'},{2,'a'}},{{1,'b'},{2,'b'}},{{1,'c'},{2,'c'}}}, a) end) end) describe("flatten()", function() it("flattens a 2darray", function() local a = {{1,2}, {3,4}, {5,6}} assert.same( {1,2,3,4,5,6}, array2d.flatten(a)) end) it("keeps a nil-array 'square'", function() local a = {{ 1,2}, {nil,4}, {nil,6}} assert.same( {1,2,nil,4,nil,6}, array2d.flatten(a)) end) end) describe("reshape()", function() it("reshapes array in new nr of rows", function() local a = {{ 1, 2, 3}, { 4, 5, 6}, { 7, 8, 9}, {10,11,12}} local b = array2d.reshape(a, 2, false) assert.same({{ 1, 2, 3, 4, 5, 6}, { 7, 8, 9,10,11,12}}, b) local c = array2d.reshape(b, 4, false) assert.same(a, c) end) it("reshapes array in new nr of rows, column order", function() local a = {{ 1, 2, 3}, { 4, 5, 6}, { 7, 8, 9}} local b = array2d.reshape(a, 3, true) assert.same({{ 1, 4, 7}, { 2, 5, 8}, { 3, 6, 9}}, b) end) end) describe("transpose()", function() it("transposes a 2d array", function() local a = {{ 1, 2, 3}, { 4, 5, 6}, { 7, 8, 9}} local b = array2d.transpose(a) assert.same({{ 1, 4, 7}, { 2, 5, 8}, { 3, 6, 9}}, b) local a = {{ 1, 2, 3, 4, 5}, { 6, 7, 8, 9, 10}} local b = array2d.transpose(a) assert.same({{ 1, 6}, { 2, 7}, { 3, 8}, { 4, 9}, { 5,10}}, b) end) end) describe("swap_rows()", function() it("swaps 2 rows, in-place", function() local a = {{1,2}, {3,4}, {5,6}} local b = array2d.swap_rows(a, 1, 3) assert.same({{5,6}, {3,4}, {1,2}}, b) assert.equal(a, b) end) end) describe("swap_cols()", function() it("swaps 2 columns, in-place", function() local a = {{1,2,3}, {4,5,6}, {7,8,9}} local b = array2d.swap_cols(a, 1, 3) assert.same({{3,2,1}, {6,5,4}, {9,8,7}}, b) assert.equal(a, b) end) end) describe("extract_rows()", function() it("extracts rows", function() local a = {{ 1, 2, 3}, { 4, 5, 6}, { 7, 8, 9}, {10,11,12}} local b = array2d.extract_rows(a, {1, 3}) assert.same({{1,2,3}, {7,8,9}}, b) end) end) describe("extract_cols()", function() it("extracts colums", function() local a = {{ 1, 2, 3}, { 4, 5, 6}, { 7, 8, 9}, {10,11,12}} local b = array2d.extract_cols(a, {1, 2}) assert.same({{ 1, 2}, { 4, 5}, { 7, 8}, {10,11}}, b) end) end) describe("remove_row()", function() it("removes a row", function() local a = {{ 1, 2, 3}, { 4, 5, 6}, { 7, 8, 9}, {10,11,12}} array2d.remove_row(a, 2) assert.same({{ 1, 2, 3}, { 7, 8, 9}, {10,11,12}}, a) end) end) describe("remove_col()", function() it("removes a colum", function() local a = {{ 1, 2, 3}, { 4, 5, 6}, { 7, 8, 9}, {10,11,12}} array2d.remove_col(a, 2) assert.same({{ 1, 3}, { 4, 6}, { 7, 9}, {10,12}}, a) end) end) describe("parse_range()", function() it("parses A1:B2 format", function() assert.same({4,11,7,12},{array2d.parse_range("K4:L7")}) assert.same({4,28,7,54},{array2d.parse_range("AB4:BB7")}) -- test Col R since it might be mixed up with RxCx format assert.same({4,18,7,18},{array2d.parse_range("R4:R7")}) end) it("parses A1 format", function() assert.same({4,11},{array2d.parse_range("K4")}) -- test Col R since it might be mixed up with RxCx format assert.same({4,18},{array2d.parse_range("R4")}) end) it("parses R1C1:R2C2 format", function() assert.same({4,11,7,12},{array2d.parse_range("R4C11:R7C12")}) end) it("parses R1C1 format", function() assert.same({4,11},{array2d.parse_range("R4C11")}) end) end) describe("range()", function() it("returns a range", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local b = array2d.range(a, "B3:C4") assert.same({{ 8, 9}, {11,12}}, b) end) end) describe("default_range()", function() it("returns the default range", function() local a = array2d.new(4,6,1) assert.same({1,1,4,6}, {array2d.default_range(a, nil, nil, nil, nil)}) end) it("accepts negative indices", function() local a = array2d.new(4,6,1) assert.same({2,2,3,5}, {array2d.default_range(a, -3, -5, -2, -2)}) end) it("corrects out of bounds indices", function() local a = array2d.new(4,6,1) assert.same({1,1,4,6}, {array2d.default_range(a, -100, -100, 100, 100)}) end) end) describe("slice()", function() it("returns a slice", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local b = array2d.slice(a,3,2,4,3) assert.same({{ 8, 9}, {11,12}}, b) end) it("returns a single row if rows are equal", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local b = array2d.slice(a,4,1,4,3) assert.same({10,11,12}, b) end) it("returns a single column if columns are equal", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local b = array2d.slice(a,1,3,4,3) assert.same({3,6,9,12}, b) end) it("returns a single value if rows and columns are equal", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local b = array2d.slice(a,2,2,2,2) assert.same(5, b) end) end) describe("set()", function() it("sets a range to a value", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} array2d.set(a,0,2,2,3,3) assert.same({{1 ,2 ,3}, {4 ,0 ,0}, {7 ,0 ,0}, {10,11,12}}, a) end) it("sets a range to a function value", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local x = 10 local args = {} local f = function(r,c) args[#args+1] = {r,c} x = x + 1 return x end array2d.set(a,f,3,1,4,3) assert.same({{1 ,2 ,3}, {4 ,5 ,6}, {11,12,13}, {14,15,16}}, a) -- validate args used to call the function assert.same({{3,1},{3,2},{3,3},{4,1},{4,2},{4,3}}, args) end) end) describe("write()", function() it("writes array to a file", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local f = setmetatable({}, { __index = { write = function(self,str) self[#self+1] = str end } }) array2d.write(a,f,"(%s)") f = table.concat(f) assert.equal([[(1)(2)(3) (4)(5)(6) (7)(8)(9) (10)(11)(12) ]],f) end) it("writes partial array to a file", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local f = setmetatable({}, { __index = { write = function(self,str) self[#self+1] = str end } }) array2d.write(a,f,"(%s)", 1,1,2,2) f = table.concat(f) assert.equal([[(1)(2) (4)(5) ]],f) end) end) describe("forall()", function() it("runs all value and row functions", function() local r = {} local t = 0 local fval = function(row, j) t = t + row[j] end local frow = function(i) r[#r+1] = t; t = 0 end local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} array2d.forall(a, fval, frow) assert.same({6, 15, 24, 33}, r) r = {} array2d.forall(a, fval, frow, 2,2,4,3) assert.same({11, 17, 23}, r) end) end) describe("move()", function() it("moves block to destination aray", function() local a = array2d.new(4,4,0) local b = array2d.new(3,3,1) array2d.move(a,2,2,b) assert.same({{0,0,0,0}, {0,1,1,1}, {0,1,1,1}, {0,1,1,1}}, a) end) end) describe("iter()", function() it("iterates all values", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local r = {} for v, i, j in array2d.iter(a) do r[#r+1] = v assert.is_nil(i) assert.is_nil(j) end assert.same({1,2,3,4,5,6,7,8,9,10,11,12}, r) end) it("iterates all values and indices", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local r = {} local ri = {} local rj = {} for i, j, v in array2d.iter(a,true) do r[#r+1] = v ri[#ri+1] = i rj[#rj+1] = j end assert.same({1,2,3,4,5,6,7,8,9,10,11,12}, r) assert.same({1,1,1,2,2,2,3,3,3,4,4,4}, ri) assert.same({1,2,3,1,2,3,1,2,3,1,2,3}, rj) end) it("iterates all values of a 2d array part", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local r = {} for v, i, j in array2d.iter(a,false,2,2,4,3) do r[#r+1] = v assert.is_nil(i) assert.is_nil(j) end assert.same({5,6,8,9,11,12}, r) end) it("iterates all values and indices of a 2d array part", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local r = {} local ri = {} local rj = {} for i, j, v in array2d.iter(a,true,2,2,4,3) do r[#r+1] = v ri[#ri+1] = i rj[#rj+1] = j end assert.same({5,6,8,9,11,12}, r) assert.same({2,2,3,3,4,4}, ri) assert.same({2,3,2,3,2,3}, rj) end) end) describe("columns()", function() it("iterates all columns", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local r = {} for col, idx in array2d.columns(a) do r[#r+1] = col col.idx = idx end assert.same({{1,4,7,10, idx=1},{2,5,8,11, idx=2},{3,6,9,12, idx=3}}, r) end) end) describe("rows()", function() it("iterates all columns", function() local a = {{1 ,2 ,3}, {4 ,5 ,6}, {7 ,8 ,9}, {10,11,12}} local r = {} for row, idx in array2d.rows(a) do r[#r+1] = row row.idx = idx end assert.same({{1,2,3, idx=1},{4,5,6, idx=2}, {7,8,9, idx=3},{10,11,12, idx=4}}, r) end) end) end) Penlight-1.12.0/spec/date_spec.lua000066400000000000000000000024751416703176500167530ustar00rootroot00000000000000local Date = require("pl.Date") describe("pl.Date", function () describe("function", function () describe("Format()", function () it("should output parsable inputs", function () local function assert_date_format(expected, format) local df = Date.Format(format) local d = df:parse(expected) assert.is.equal(expected, df:tostring(d)) end assert_date_format('02/04/10', 'dd/mm/yy') assert_date_format('04/02/2010', 'mm/dd/yyyy') assert_date_format('2011-02-20', 'yyyy-mm-dd') assert_date_format('20070320', 'yyyymmdd') assert_date_format('23:10', 'HH:MM') end) it("should parse 'slack' fields", function () local df = Date.Format("m/d/yy") -- TODO: Re-enable when issue #359 fixed -- assert.is.equal('01/05/99', df:tostring(df:parse('1/5/99'))) assert.is.equal('01/05/01', df:tostring(df:parse('1/5/01'))) assert.is.equal('01/05/32', df:tostring(df:parse('1/5/32'))) end) end) end) describe("meta method", function () describe("__tostring()", function () it("should be suitable for serialization", function () local df = Date.Format() local du = df:parse("2008-07-05") assert.is.equal(du, du:toUTC()) end) end) end) end) Penlight-1.12.0/spec/multimap_spec.lua000066400000000000000000000005071416703176500176600ustar00rootroot00000000000000local MultiMap = require("pl.MultiMap") describe("pl.MultiMap", function () it("should hold multiple values per key", function () local map = MultiMap() map:set('foo', 1) map:set('bar', 3) map:set('foo', 2) local expected = { foo = { 1, 2 }, bar = { 3 } } assert.is.same(expected, map) end) end) Penlight-1.12.0/spec/permute_spec.lua000066400000000000000000000131431416703176500175110ustar00rootroot00000000000000local permute = require("pl.permute") local tcopy = require("pl.tablex").copy local utils = require("pl.utils") describe("pl.permute", function() describe("order_iter", function() it("returns all order combinations", function() local result = {} for list in permute.order_iter({"one", "two", "three"}) do result[#result+1] = tcopy(list) end assert.same({ [1] = { [1] = 'two', [2] = 'three', [3] = 'one' }, [2] = { [1] = 'three', [2] = 'two', [3] = 'one' }, [3] = { [1] = 'three', [2] = 'one', [3] = 'two' }, [4] = { [1] = 'one', [2] = 'three', [3] = 'two' }, [5] = { [1] = 'two', [2] = 'one', [3] = 'three' }, [6] = { [1] = 'one', [2] = 'two', [3] = 'three' } }, result) end) it("returns nil on empty list", function() local result = {} for list in permute.order_iter({}) do result[#result+1] = tcopy(list) end assert.equal(0, #result) end) end) describe("order_table", function() it("returns all order combinations", function() local result = permute.order_table({"one", "two", "three"}) assert.same({ [1] = { [1] = 'two', [2] = 'three', [3] = 'one' }, [2] = { [1] = 'three', [2] = 'two', [3] = 'one' }, [3] = { [1] = 'three', [2] = 'one', [3] = 'two' }, [4] = { [1] = 'one', [2] = 'three', [3] = 'two' }, [5] = { [1] = 'two', [2] = 'one', [3] = 'three' }, [6] = { [1] = 'one', [2] = 'two', [3] = 'three' } }, result) end) it("returns empty table on empty input list", function() local result = permute.order_table({}) assert.same({}, result) end) end) describe("list_iter", function() it("returns all combinations from sub-lists", function() local result = {} local strs = {"one", "two", "three"} local ints = { 1,2,3 } local bools = { true, false } for count, str, int, bool in permute.list_iter(strs, ints, bools) do result[#result+1] = {count, str, int, bool} end assert.same({ [1] = {1, 'one', 1, true }, [2] = {2, 'two', 1, true }, [3] = {3, 'three', 1, true }, [4] = {4, 'one', 2, true }, [5] = {5, 'two', 2, true }, [6] = {6, 'three', 2, true }, [7] = {7, 'one', 3, true }, [8] = {8, 'two', 3, true }, [9] = {9, 'three', 3, true }, [10] = {10, 'one', 1, false }, [11] = {11, 'two', 1, false }, [12] = {12, 'three', 1, false }, [13] = {13, 'one', 2, false }, [14] = {14, 'two', 2, false }, [15] = {15, 'three', 2, false }, [16] = {16, 'one', 3, false }, [17] = {17, 'two', 3, false }, [18] = {18, 'three', 3, false }, }, result) end) it("is nil-safe, given 'n' is set", function() local result = {} local bools = utils.pack(nil, true, false) local strs = utils.pack("one", "two", nil) for count, bool, str in permute.list_iter(bools, strs) do result[#result+1] = {count, bool, str} end assert.same({ [1] = {1, nil, 'one' }, [2] = {2, true, 'one' }, [3] = {3, false, 'one' }, [4] = {4, nil, 'two' }, [5] = {5, true, 'two' }, [6] = {6, false, 'two' }, [7] = {7, nil, nil }, [8] = {8, true, nil }, [9] = {9, false, nil }, }, result) end) it("returns nil on empty list", function() local count = 0 for list in permute.list_iter({}) do count = count + 1 end assert.equal(0, count) end) end) describe("list_table", function() it("returns all combinations from sub-lists", function() local strs = {"one", "two", "three"} local ints = { 1,2,3 } local bools = { true, false } assert.same({ [1] = {'one', 1, true, n = 3 }, [2] = {'two', 1, true, n = 3 }, [3] = {'three', 1, true, n = 3 }, [4] = {'one', 2, true, n = 3 }, [5] = {'two', 2, true, n = 3 }, [6] = {'three', 2, true, n = 3 }, [7] = {'one', 3, true, n = 3 }, [8] = {'two', 3, true, n = 3 }, [9] = {'three', 3, true, n = 3 }, [10] = {'one', 1, false, n = 3 }, [11] = {'two', 1, false, n = 3 }, [12] = {'three', 1, false, n = 3 }, [13] = {'one', 2, false, n = 3 }, [14] = {'two', 2, false, n = 3 }, [15] = {'three', 2, false, n = 3 }, [16] = {'one', 3, false, n = 3 }, [17] = {'two', 3, false, n = 3 }, [18] = {'three', 3, false, n = 3 }, }, permute.list_table(strs, ints, bools)) end) it("is nil-safe, given 'n' is set", function() local bools = utils.pack(nil, true, false) local strs = utils.pack("one", "two", nil) assert.same({ [1] = {nil, 'one', n = 2 }, [2] = {true, 'one', n = 2 }, [3] = {false, 'one', n = 2 }, [4] = {nil, 'two', n = 2 }, [5] = {true, 'two', n = 2 }, [6] = {false, 'two', n = 2 }, [7] = {nil, nil, n = 2 }, [8] = {true, nil, n = 2 }, [9] = {false, nil, n = 2 }, }, permute.list_table(bools, strs)) end) it("returns nil on empty list", function() assert.same({}, permute.list_table({})) end) end) end) Penlight-1.12.0/spec/pretty_spec.lua000066400000000000000000000022641416703176500173610ustar00rootroot00000000000000local pretty = require("pl.pretty") describe("pl.pretty.number", function () it("should format memory", function () local function assert_memory (expected, input) assert.is.equal(expected, pretty.number(input, "M")) end assert_memory("123B", 123) assert_memory("1.2KiB", 1234) assert_memory("10.0KiB", 10*1024) assert_memory("1.0MiB", 1024*1024) assert_memory("1.0GiB", 1024*1024*1024) end) it("should format postfixes", function () local function assert_postfix(expected, input) assert.is.equal(expected, pretty.number(input, "N", 2)) end assert_postfix("123", 123) assert_postfix("1.23K", 1234) assert_postfix("10.24K", 10*1024) assert_postfix("1.05M", 1024*1024) assert_postfix("1.07B", 1024*1024*1024) end) it("should format postfixes", function () local function assert_separator(expected, input) assert.is.equal(expected, pretty.number(input, "T")) end assert_separator('123', 123) assert_separator('1,234', 1234) assert_separator('12,345', 12345) assert_separator('123,456', 123456) assert_separator('1,234,567', 1234567) assert_separator('12,345,678', 12345678) end) end) Penlight-1.12.0/spec/set_spec.lua000066400000000000000000000036141416703176500166250ustar00rootroot00000000000000local Set = require("pl.Set") describe("pl.Set", function () local s = Set() local s1_2 = Set({ 1, 2 }) local s1_2_3 = Set({ 1, 2, 3 }) local s1_3 = Set({ 1, 3 }) local s2 = Set({ 2 }) local s2_1 = Set({ 2, 1 }) local s2_3 = Set({ 2, 3 }) local s3 = Set({ 3 }) local sm = Set({ "foo", "bar" }) it("should produce a set object", function () assert.is.same({ true, true }, s1_2) end) it("should produce identical sets for any ordered input", function () assert.is.same(s1_2, s2_1) end) describe("should have an operator for", function () it("union", function () assert.is.same(s1_2_3, s1_2 + s3) assert.is.same(s1_2_3, s1_2 + 3) end) it("intersection", function () assert.is.same(s2, s1_2 * s2_3) end) it("difference", function () assert.is.same(s2_1, s1_2_3 - s3) assert.is.same(s2_3, s1_2_3 - 1) end) it("symmetric difference", function () assert.is.same(s1_3, s1_2 ^ s2_3) end) it("tostring", function () -- Cannot test multi-entry sets because of non-deterministic key order assert.is.same('[2]', tostring(s2)) end) end) describe("should provide functions", function () it("isempty", function () assert.is.truthy(Set.isempty(s)) assert.is.falsy(Set.isempty(s3)) end) it("set", function () local m = Set() Set.set(m, 'foo', true) m.bar = true assert.is.same(m, sm) assert.is_not.same(m, s1_2) end) end) describe("should have a comparison operator for", function () it("supersets/subsets than", function () assert.is.truthy(s1_2 > s2) assert.is.falsy(s1_3 > s2) assert.is.falsy(s1_2 > s2_3) assert.is.truthy(s1_2 < s1_2_3) assert.is.falsy(s1_2_3 < s1_2) end) it("equality", function () assert.is.truthy(s1_2 == s2_1) assert.is.falsy(s1_2 == s2_3) end) end) end) Penlight-1.12.0/spec/stringx_spec.lua000066400000000000000000000546471416703176500175440ustar00rootroot00000000000000describe("stringx", function() local stringx = require "pl.stringx" it("isalpha()", function() assert.equal(false, stringx.isalpha '') assert.equal(false, stringx.isalpha' ') assert.equal(false, stringx.isalpha'0') assert.equal(false, stringx.isalpha'\0') assert.equal(true, stringx.isalpha'azAZ') assert.equal(false, stringx.isalpha'az9AZ') end) it("isdigit()", function() assert.equal(false, stringx.isdigit'') assert.equal(false, stringx.isdigit' ') assert.equal(false, stringx.isdigit'a') assert.equal(true, stringx.isdigit'0123456789') end) it("isalnum()", function() assert.equal(false, stringx.isalnum'') assert.equal(false, stringx.isalnum' ') assert.equal(true, stringx.isalnum'azAZ01234567890') end) it("isspace()", function() assert.equal(false, stringx.isspace'') assert.equal(true, stringx.isspace' ') assert.equal(true, stringx.isspace' \r\n\f\t') assert.equal(false, stringx.isspace' \r\n-\f\t') end) it("islower()", function() assert.equal(false, stringx.islower'') assert.equal(true, stringx.islower'az') assert.equal(false, stringx.islower'aMz') assert.equal(true, stringx.islower'a z') end) it("isupper()", function() assert.equal(false, stringx.isupper'') assert.equal(true, stringx.isupper'AZ') assert.equal(false, stringx.isupper'AmZ') assert.equal(true, stringx.isupper'A Z') end) it("startswith()", function() local startswith = stringx.startswith assert.equal(true, startswith('', '')) assert.equal(false, startswith('', 'a')) assert.equal(true, startswith('a', '')) assert.equal(true, startswith('a', 'a')) assert.equal(false, startswith('a', 'b')) assert.equal(false, startswith('a', 'ab')) assert.equal(true, startswith('abc', 'ab')) assert.equal(false, startswith('abc', 'bc')) -- off by one assert.equal(false, startswith('abc', '.')) -- Lua pattern char assert.equal(true, startswith('a\0bc', 'a\0b')) -- '\0' assert.equal(true, startswith('abcfoo',{'abc','def'})) assert.equal(true, startswith('deffoo',{'abc','def'})) assert.equal(false, startswith('cdefoo',{'abc','def'})) end) it("endswith()", function() local endswith = stringx.endswith assert.equal(true, endswith("", "")) assert.equal(false, endswith("", "a")) assert.equal(true, endswith("a", "")) assert.equal(true, endswith("a", "a")) assert.equal(false, endswith("a", "A")) -- case sensitive assert.equal(false, endswith("a", "aa")) assert.equal(true, endswith("abc", "")) assert.equal(false, endswith("abc", "ab")) -- off by one assert.equal(true, endswith("abc", "c")) assert.equal(true, endswith("abc", "bc")) assert.equal(true, endswith("abc", "abc")) assert.equal(false, endswith("abc", " abc")) assert.equal(false, endswith("abc", "a")) assert.equal(false, endswith("abc", ".")) -- Lua pattern char assert.equal(true, endswith("ab\0c", "b\0c")) -- \0 assert.equal(false, endswith("ab\0c", "b\0d")) -- \0 assert.equal(true, endswith('dollar.dot',{'.dot','.txt'})) assert.equal(true, endswith('dollar.txt',{'.dot','.txt'})) assert.equal(false, endswith('dollar.rtxt',{'.dot','.txt'})) end) it("join()", function() assert.equal('1 2 3', stringx.join(' ', {1,2,3})) end) it("splitlines", function() assert.same({}, stringx.splitlines('')) assert.same({'a'}, stringx.splitlines('a')) assert.same({''}, stringx.splitlines('\n')) assert.same({'', ''}, stringx.splitlines('\n\n')) assert.same({'', ''}, stringx.splitlines('\r\r')) assert.same({''}, stringx.splitlines('\r\n')) assert.same({'ab', 'cd'}, stringx.splitlines('ab\ncd\n')) assert.same({'ab\n', 'cd\n'}, stringx.splitlines('ab\ncd\n', true)) assert.same({'\n', 'ab\r', '\r\n', 'cd\n'}, stringx.splitlines('\nab\r\r\ncd\n', true)) end) it("split()", function() local split = stringx.split assert.same({''}, split('', '')) assert.same({}, split('', 'z')) --FIX:intended and specified behavior? assert.same({'a'}, split('a', '')) --FIX:intended and specified behavior? assert.same({''}, split('a', 'a')) -- stringx.split now follows the Python pattern, so it uses a substring, not a pattern. -- If you need to split on a pattern, use utils.split() -- asserteq(split('ab1cd23ef%d', '%d+'), {'ab', 'cd', 'ef%d'}) -- pattern chars -- note that leading space is ignored by the default assert.same({'1','2','3'}, split(' 1 2 3 ')) assert.same({'a','bb','c','ddd'}, split('a*bb*c*ddd','*')) assert.same({'dog','fred','bonzo:alice'}, split('dog:fred:bonzo:alice',':',3)) assert.same({'dog','fred','bonzo:alice:'}, split('dog:fred:bonzo:alice:',':',3)) assert.same({'','','',''}, split('///','/')) end) it("expandtabs()", function() assert.equal('', stringx.expandtabs('',0)) assert.equal('', stringx.expandtabs('',1)) assert.equal(' ', stringx.expandtabs(' ',1)) assert.equal((' '):rep(1+8), stringx.expandtabs(' \t ')) assert.equal((' '):rep(3), stringx.expandtabs(' \t ',2)) assert.equal((' '):rep(2), stringx.expandtabs(' \t ',0)) assert.equal(' hi there folks!', stringx.expandtabs('\thi\tthere\tfolks!')) end) it("lfind()", function() assert.equal(1, stringx.lfind('', '')) assert.equal(1, stringx.lfind('a', '')) assert.equal(2, stringx.lfind('ab', 'b')) assert.is_nil(stringx.lfind('abc', 'cd')) assert.equal(2, stringx.lfind('abcbc', 'bc')) assert.equal(3, stringx.lfind('ab..cd', '.')) -- pattern char assert.equal(4, stringx.lfind('abcbcbbc', 'bc', 3)) assert.is_nil(stringx.lfind('abcbcbbc', 'bc', 3, 4)) assert.equal(4, stringx.lfind('abcbcbbc', 'bc', 3, 5)) assert.equal(2, stringx.lfind('abcbcbbc', 'bc', nil, 5)) end) it("rfind()", function() assert.equal(1, stringx.rfind('', '')) assert.equal(3, stringx.rfind('ab', '')) assert.is_nil(stringx.rfind('abc', 'cd')) assert.equal(4, stringx.rfind('abcbc', 'bc')) assert.equal(4, stringx.rfind('abcbcb', 'bc')) assert.equal(4, stringx.rfind('ab..cd', '.')) -- pattern char assert.equal(7, stringx.rfind('abcbcbbc', 'bc', 3)) assert.is_nil(stringx.rfind('abcbcbbc', 'bc', 3, 4)) assert.equal(4, stringx.rfind('abcbcbbc', 'bc', 3, 5)) assert.equal(4, stringx.rfind('abcbcbbc', 'bc', nil, 5)) assert.equal(4, stringx.rfind('banana', 'ana')) end) it("replace()", function() assert.equal('', stringx.replace('', '', '')) assert.equal(' ', stringx.replace(' ', '', '')) assert.equal(' ', stringx.replace(' ', '', ' ')) assert.equal('', stringx.replace(' ', ' ', '')) assert.equal('aBCaBCaBC', stringx.replace('abcabcabc', 'bc', 'BC')) assert.equal('aBCabcabc', stringx.replace('abcabcabc', 'bc', 'BC', 1)) assert.equal('abcabcabc', stringx.replace('abcabcabc', 'bc', 'BC', 0)) assert.equal('abc', stringx.replace('abc', 'd', 'e')) assert.equal('a%db', stringx.replace('a.b', '.', '%d')) end) it("count()", function() assert.equal(0, stringx.count('', '')) --infinite loop]] assert.equal(2, stringx.count(' ', '')) --infinite loop]] assert.equal(2, stringx.count('a..c', '.')) -- pattern chars assert.equal(0, stringx.count('a1c', '%d')) -- pattern chars assert.equal(3, stringx.count('Anna Anna Anna', 'Anna')) -- no overlap assert.equal(1, stringx.count('banana', 'ana', false)) -- no overlap assert.equal(2, stringx.count('banana', 'ana', true)) -- overlap end) it("ljust()", function() assert.equal('', stringx.ljust('', 0)) assert.equal(' ', stringx.ljust('', 2)) assert.equal('ab ', stringx.ljust('ab', 3)) assert.equal('ab%', stringx.ljust('ab', 3, '%')) assert.equal('abcd', stringx.ljust('abcd', 3)) -- agrees with Python end) it("rjust()", function() assert.equal('', stringx.rjust('', 0)) assert.equal(' ', stringx.rjust('', 2)) assert.equal(' ab', stringx.rjust('ab', 3)) assert.equal('%ab', stringx.rjust('ab', 3, '%')) assert.equal('abcd', stringx.rjust('abcd', 3)) -- agrees with Python end) it("center()", function() assert.equal('', stringx.center('', 0)) assert.equal(' ', stringx.center('', 1)) assert.equal(' ', stringx.center('', 2)) assert.equal('a', stringx.center('a', 1)) assert.equal('a ', stringx.center('a', 2)) assert.equal(' a ', stringx.center('a', 3)) end) it("lstrip()", function() local trim = stringx.lstrip assert.equal('', trim'') assert.equal('', trim' ') assert.equal('', trim' ') assert.equal('a', trim'a') assert.equal('a', trim' a') assert.equal('a ', trim'a ') assert.equal('a ', trim' a ') assert.equal('a ', trim' a ') assert.equal('ab cd ', trim' ab cd ') assert.equal('a\000b \r\t\n\f\v', trim' \t\r\n\f\va\000b \r\t\n\f\v') assert.equal('hello] -- - ', trim(' - -- [hello] -- - ','-[] ')) end) it("rstrip()", function() local trim = stringx.rstrip assert.equal('', trim'') assert.equal('', trim' ') assert.equal('', trim' ') assert.equal('a', trim'a') assert.equal(' a', trim' a') assert.equal('a', trim'a ') assert.equal(' a', trim' a ') assert.equal(' a', trim' a ') assert.equal(' ab cd', trim' ab cd ') assert.equal(' \t\r\n\f\va\000b', trim' \t\r\n\f\va\000b \r\t\n\f\v') assert.equal(' - -- [hello', trim(' - -- [hello] -- - ','-[] ')) end) it("strip()", function() local trim = stringx.strip assert.equal('', trim'') assert.equal('', trim' ') assert.equal('', trim' ') assert.equal('a', trim'a') assert.equal('a', trim' a') assert.equal('a', trim'a ') assert.equal('a', trim' a ') assert.equal('a', trim' a ') assert.equal('ab cd', trim' ab cd ') assert.equal('a\000b', trim' \t\r\n\f\va\000b \r\t\n\f\v') assert.equal('hello', trim(' - -- [hello] -- - ','-[] ')) local long = 'a' .. string.rep(' ', 200000) .. 'a' assert.equal(long, trim(long)) end) it("splitv()", function() -- is actually 'utils.splitv' assert.same({"hello", "dolly"}, {stringx.splitv("hello dolly")}) end) it("partition()", function() assert.has.error(function() stringx.partition('a', '') end) assert.same({'', 'a', ''}, {stringx.partition('a', 'a')}) assert.same({'a', 'b', 'c'}, {stringx.partition('abc', 'b')}) assert.same({'abc','',''}, {stringx.partition('abc', '.+')}) assert.same({'ab','.','c'}, {stringx.partition('ab.c', '.')}) assert.same({'a',',','b,c'}, {stringx.partition('a,b,c', ',')}) assert.same({'abc', '', ''}, {stringx.partition('abc', '/')}) end) it("rpartition()", function() assert.has.error(function() stringx.rpartition('a', '') end) assert.same({'a/b', '/', 'c'}, {stringx.rpartition('a/b/c', '/')}) assert.same({'a', 'b', 'c'}, {stringx.rpartition('abc', 'b')}) assert.same({'', 'a', ''}, {stringx.rpartition('a', 'a')}) assert.same({'', '', 'abc'}, {stringx.rpartition('abc', '/')}) end) it("at()", function() -- at (works like s:sub(idx,idx), so negative indices allowed assert.equal('a', stringx.at('a', 1)) assert.equal('b', stringx.at('ab', 2)) assert.equal('d', stringx.at('abcd', -1)) assert.equal('', stringx.at('abcd', 10)) -- not found end) describe("indent()", function() it("adds an indent", function() local t = "a whole lot\nof love" assert.equal([[ a whole lot of love ]], stringx.indent(t, 4)) assert.equal([[ **easy ** **enough! ]], stringx.indent("easy\n\nenough!", 2 ,'*')) end) it("appends a newline if not present", function() assert.equal(" hello\n world\n", stringx.indent("hello\nworld", 2)) assert.equal(" hello\n world\n", stringx.indent("hello\nworld\n", 2)) end) end) describe("dedent()", function() it("removes prefixed whitespace", function() assert.equal([[ one two three ]], stringx.dedent [[ one two three ]]) end) it("removes prefixed whitespace, retains structure", function() assert.equal([[ one two three ]], stringx.dedent [[ one two three ]]) end) it("appends a newline if not present", function() assert.equal("hello\nworld\n", stringx.dedent(" hello\n world")) assert.equal("hello\nworld\n", stringx.dedent(" hello\n world\n")) end) end) describe("fill()/wrap()", function() it("wraps width over limit", function() assert.same({ "abc", "def" }, stringx.wrap("abc def", 2)) end) it("wraps width at limit", function() assert.same({ "abc", "def" }, stringx.wrap("abc def", 3)) assert.same({ "a c", "d f" }, stringx.wrap("a c d f", 3)) end) it("wraps single letters", function() assert.same({"a"}, stringx.wrap("a")) end) it("wraps empty strings", function() assert.same({""}, stringx.wrap("")) assert.same({""}, stringx.wrap(" ")) end) it("handles leading/trailing whitespace", function() assert.same({"hello"}, stringx.wrap(" hello ", 10)) assert.same({"hello"}, stringx.wrap(" hello ", 2)) assert.same({"he", "ll", "o"}, stringx.wrap(" hello ", 2, true)) end) it("handles line-breaks", function() assert.same({"Hello", "Dolly"}, stringx.wrap("Hello\nDolly", 10)) assert.same({"Hello Dolly"}, stringx.wrap("Hello\nDolly", 20)) end) it("doesn't split on accented characters", function() assert.same({"àbcdéfghîj"}, stringx.wrap("àbcdéfghîj")) end) it("word-wraps a text", function() -- local binstring = require("luassert.formatters.binarystring") -- assert:add_formatter(binstring) assert.equal([[ It is often said of Lua that it does not include batteries. That is because the goal of Lua is to produce a lean expressive language that will be used on all sorts of machines, (some of which don't even have hierarchical filesystems). The Lua language is the equivalent of an operating system kernel; the creators of Lua do not see it as their responsibility to create a full software ecosystem around the language. That is the role of the community. ]], stringx.fill("It is often said of Lua that it does not include batteries. That is because the goal of Lua is to produce a lean expressive language that will be used on all sorts of machines, (some of which don't even have hierarchical filesystems). The Lua language is the equivalent of an operating system kernel; the creators of Lua do not see it as their responsibility to create a full software ecosystem around the language. That is the role of the community.", 20)) end) it("generic wrap test", function() local t = [[ hello "world" 'this' -is- a bb ccc dddd test... but wouldn't it pass??? final. word-that-can-be-broken ]] assert.same({ "hello", '"world"', "'this'", "-is-", "a", "bb", "ccc", "dddd", "test...", "but", "wouldn't", "it", "pass???", "final.", "word-that-can-be-broken", }, stringx.wrap(t, 3)) end) it("generic wrap test, with overflow breaking", function() local t = [[ hello "world" 'this' -is- a bb ccc dddd test... but wouldn't it pass??? final. word-that-can-be-broken ]] assert.same({ "hel", "lo", '"wo', 'rld', '"', "'th", "is'", "-is", "- a", "bb", "ccc", "ddd", "d", "tes", "t..", ".", "but", "wou", "ldn", "'t", "it", "pas", "s??", "?", "fin", "al.", "wor", "d-t", "hat", "-ca", "n-b", "e-b", "rok", "en", }, stringx.wrap(t, 3, true)) end) end) describe("Template", function() local Template = stringx.Template it("substitute() replaces placeholders", function() local t1 = Template [[ while true do $contents end ]] assert.equal([[ while true do print "hello" end ]], t1:substitute {contents = 'print "hello"'}) end) it("substitute() replaces multiple placeholders", function () local template = Template("${here} is the $answer") local out = template:substitute({ here = 'one', answer = 'two' }) assert.is.equal('one is the two', out) end) it("indent_substitute() indents replaced multi-lines", function() local t1 = Template [[ while true do $contents end ]] assert.equal( "while true do\n".. " for i = 1,10 do\n".. " gotcha(i)\n".. " end\n".. "\n".. "end\n" , t1:indent_substitute {contents = [[ for i = 1,10 do gotcha(i) end ]]}) end) end) it("lines()", function() local function merge(it, ...) assert(select('#', ...) == 0) local ts = {} for val in it do ts[#ts+1] = val end return ts end assert.same({''}, merge(stringx.lines(''))) assert.same({'ab'}, merge(stringx.lines('ab'))) assert.same({'ab', 'cd'}, merge(stringx.lines('ab\ncd'))) end) it("title()", function() assert.equal('', stringx.title('')) assert.equal('Abc Def1', stringx.title('abC deF1')) -- Python behaviour assert.equal('Hello World', stringx.capitalize('hello world')) end) it("capitalize()", function() -- old name for 'title' assert.equal(stringx.title, stringx.capitalize) end) it("shorten()", function() assert.equal('', stringx.shorten('', 0)) assert.equal('a', stringx.shorten('a', 1)) assert.equal('.', stringx.shorten('ab', 1)) --FIX:ok? assert.equal('abc', stringx.shorten('abc', 3)) assert.equal('...', stringx.shorten('abcd', 3)) assert.equal('abcde', stringx.shorten('abcde', 5)) assert.equal('a...', stringx.shorten('abcde', 4)) assert.equal('...', stringx.shorten('abcde', 3)) assert.equal('..', stringx.shorten('abcde', 2)) assert.equal('', stringx.shorten('abcde', 0)) assert.equal('', stringx.shorten('', 0, true)) assert.equal('a', stringx.shorten('a', 1, true)) assert.equal('.', stringx.shorten('ab', 1, true)) assert.equal('abcde', stringx.shorten('abcde', 5, true)) assert.equal('...e', stringx.shorten('abcde', 4, true)) assert.equal('...', stringx.shorten('abcde', 3, true)) assert.equal('..', stringx.shorten('abcde', 2, true)) assert.equal('', stringx.shorten('abcde', 0, true)) end) it("quote_string()", function() local assert_str_round_trip = function(s) local qs = stringx.quote_string(s) local compiled, err = require("pl.utils").load("return "..qs) if not compiled then print( ("stringx.quote_string assert failed: invalid string created: Received:\n%s\n\nCompiled to\n%s\n\nError:\t%s\n"): format(s, qs, err) ) error() else compiled = compiled() end if compiled ~= s then print("stringx.quote_string assert Failed: String compiled but did not round trip.") print("input string:\t\t",s, #s) print("compiled string:\t", compiled, #compiled) print("output string:\t\t",qs, #qs) error() -- else -- print("input string:\t\t",s) -- print("compiled string:\t", compiled) -- print("output string:\t\t",qs) end end assert_str_round_trip( "normal string with nothing weird.") assert_str_round_trip( "Long string quoted with escaped quote \\\" and a long string pattern match [==[ found near the end.") assert_str_round_trip( "Unescapped quote \" in the middle") assert_str_round_trip( "[[Embedded long quotes \\\". Escaped must stay! ]]") assert_str_round_trip( [[Long quoted string with a slash prior to quote \\\". ]]) assert_str_round_trip( "[[Completely normal\n long quote. ]]") assert_str_round_trip( "String with a newline\nending with a closing bracket]") assert_str_round_trip( "[[String with opening brackets ending with part of a long closing bracket]=") assert_str_round_trip( "\n[[Completely normal\n long quote. Except that we lead with a return! Tricky! ]]") assert_str_round_trip( '"balance [======[ doesn\'t ]====] mater when searching for embedded long-string quotes.') assert_str_round_trip( "Any\0 \t control character other than a return will be handled by the %q mechanism.") assert_str_round_trip( "This\tincludes\ttabs.") assert_str_round_trip( "But not returns.\n Returns are easier to see using long quotes.") assert_str_round_trip( "The \z escape does not trigger a control pattern, however.") assert_str_round_trip( "[==[If a string is long-quoted, escaped \\\" quotes have to stay! ]==]") assert_str_round_trip('"A quoted string looks like what?"') assert_str_round_trip( "'I think that it should be quoted, anyway.'") assert_str_round_trip( "[[Even if they're long quoted.]]") assert_str_round_trip( "]=]==]") assert_str_round_trip( "\"\\\"\\' pathalogical:starts with a quote ]\"\\']=]]==][[]]]=========]") assert_str_round_trip( "\\\"\\\"\\' pathalogical: quote is after this text with a quote ]\"\\']=]]==][[]]]=========]") assert_str_round_trip( "\\\"\\\"\\' pathalogical: quotes are all escaped. ]\\\"\\']=]]==][[]]]=========]") assert_str_round_trip( "") assert_str_round_trip( " ") assert_str_round_trip( "\n") --tricky. assert_str_round_trip( "\r") assert_str_round_trip( "\r\n") assert_str_round_trip( "\r1\n") assert_str_round_trip( "[[") assert_str_round_trip( "''") assert_str_round_trip( '""') end) describe("format_operator()", function() setup(function() stringx.format_operator() end) it("handles plain substitutions", function() assert.equal('[home]', '[%s]' % 'home') assert.equal('fred = 42', '%s = %d' % {'fred',42}) end) it("invokes tostring on %s formats", function() -- mostly works like string.format, except that %s forces use of tostring() -- rather than throwing an error local List = require 'pl.List' assert.equal('TBL:{1,2,3}', 'TBL:%s' % List{1,2,3}) end) it("replaces '$field' references", function() -- table with keys and format with $ assert.equal('<1>', '<$one>' % {one=1}) end) it("accepts replacement functions", function() local function subst(k) if k == 'A' then return 'ay' elseif k == 'B' then return 'bee' else return '?' end end assert.equal('ay & bee', '$A & $B' % subst) end) end) end) Penlight-1.12.0/spec/text_spec.lua000066400000000000000000000002411416703176500170070ustar00rootroot00000000000000describe("pl.text", function() it("forwarded to stringx", function() assert.equal( require "pl.stringx", require "pl.text" ) end) end) Penlight-1.12.0/spec/utils-deprecate_spec.lua000066400000000000000000000062131416703176500211220ustar00rootroot00000000000000local utils = require("pl.utils") describe("pl.utils", function () local old_fn, last_msg, last_trace before_each(function() old_fn = function() end last_msg = nil last_trace = nil utils.set_deprecation_func(function(msg, trace) last_msg = msg last_trace = trace end) end) after_each(function() utils.deprecation_warning = old_fn end) describe("set_deprecation_func", function () it("accepts nil as callback", function() assert.has.no.error(function() utils.set_deprecation_func() end) end) it("accepts function as callback", function() assert.has.no.error(function() utils.set_deprecation_func(function() end) end) end) it("fails on non-functions", function() assert.has.error(function() utils.set_deprecation_func("not a function") end, "argument 1 expected a 'function', got a 'string'") end) end) describe("raise_deprecation", function () it("requires the opts table", function() assert.has.error(function() utils.raise_deprecation(nil) end, "argument 1 expected a 'table', got a 'nil'") end) it("requires the opts.message field", function() assert.has.error(function() utils.raise_deprecation({}) end, "field 'message' of the options table must be a string") end) it("should output the message", function () utils.raise_deprecation { message = "hello world" } assert.equal("hello world", last_msg) end) it("should output the deprecated version", function () utils.raise_deprecation { message = "hello world", deprecated_after = "2.0.0", } assert.equal("hello world (deprecated after 2.0.0)", last_msg) end) it("should output the removal version", function () utils.raise_deprecation { message = "hello world", version_removed = "3.0.0", } assert.equal("hello world (scheduled for removal in 3.0.0)", last_msg) end) it("should output the deprecated and removal versions", function () utils.raise_deprecation { message = "hello world", deprecated_after = "2.0.0", version_removed = "3.0.0", } assert.equal("hello world (deprecated after 2.0.0, scheduled for removal in 3.0.0)", last_msg) end) it("should output the application/module name", function () utils.raise_deprecation { source = "MyApp 1.2.3", message = "hello world", deprecated_after = "2.0.0", version_removed = "3.0.0", } assert.equal("[MyApp 1.2.3] hello world (deprecated after 2.0.0, scheduled for removal in 3.0.0)", last_msg) end) it("should add a stracktrace", function () local function my_function_name() utils.raise_deprecation { source = "MyApp 1.2.3", message = "hello world", deprecated_after = "2.0.0", version_removed = "3.0.0", } end my_function_name() assert.Not.match("raise_deprecation", last_trace) assert.match("my_function_name", last_trace) end) end) end) Penlight-1.12.0/spec/utils-enum_spec.lua000066400000000000000000000035731416703176500201400ustar00rootroot00000000000000describe("pl.utils", function () describe("enum()", function () local enum, t before_each(function() enum = require("pl.utils").enum t = enum("ONE", "two", "THREE") end) it("holds enumerated values", function() assert.equal("ONE", t.ONE) assert.equal("two", t.two) assert.equal("THREE", t.THREE) end) describe("accessing", function() it("errors on unknown values", function() assert.has.error(function() print(t.four) end, "'four' is not a valid value (expected one of: 'ONE', 'two', 'THREE')") end) it("errors on setting new keys", function() assert.has.error(function() t.four = "four" end, "the Enum object is read-only") end) it("entries must be strings", function() assert.has.error(function() t = enum("hello", true, "world") end, "argument 2 expected a 'string', got a 'boolean'") end) it("requires at least 1 entry", function() assert.has.error(function() t = enum() end, "argument 1 expected a 'string', got a 'nil'") end) it("keys can have 'format' placeholders", function() t = enum("hello", "contains: %s") assert.has.error(function() print(t["%s"]) -- should still format error properly end, "'%s' is not a valid value (expected one of: 'hello', 'contains: %s')") end) end) describe("calling", function() it("returns error on unknown values", function() local ok, err = t("four") assert.equal(err, "'four' is not a valid value (expected one of: 'ONE', 'two', 'THREE')") assert.equal(nil, ok) end) it("returns value on success", function() local ok, err = t("THREE") assert.equal(nil, err) assert.equal("THREE", ok) end) end) end) end) Penlight-1.12.0/spec/utils-npairs_spec.lua000066400000000000000000000055271416703176500204710ustar00rootroot00000000000000local utils = require("pl.utils") describe("pl.utils", function () describe("npairs", function () local npairs = utils.npairs it("start index defaults to 1", function() local t1 = { 1, 2, 3 } local t2 = {} for i, v in npairs(t1, nil, 2) do t2[i] = v end assert.are.same({ 1, 2 }, t2) end) it("end index defaults to `t.n`", function() local t1 = { n = 2, 1, 2, 3 } local t2 = {} for i, v in npairs(t1) do t2[i] = v end assert.are.same({1, 2}, t2) end) it("step size defaults to 1", function() local t1 = { 1, 2, 3 } local t2 = {} for i, v in npairs(t1) do t2[i] = v end assert.are.same({1, 2, 3}, t2) end) it("step size cannot be 0", function() local t1 = { 1, 2, 3 } assert.has.error(function() npairs(t1, nil, nil, 0) end, "iterator step-size cannot be 0") end) it("end index defaults to `#t` if there is no `t.n`", function() local t1 = { 1, 2, 3 } local t2 = {} for i, v in npairs(t1) do t2[i] = v end assert.are.same({1, 2, 3}, t2) end) it("returns nothing if start index is beyond end index", function() local t1 = { 1, 2, 3 } local t2 = {} for i, v in npairs(t1, 5, 3) do t2[i] = v end assert.are.same({}, t2) end) it("returns nothing if start index is beyond end index, with negative step size", function() local t1 = { 1, 2, 3 } local t2 = {} for i, v in npairs(t1, 3, 1, -1) do t2[#t2+1] = v end assert.are.same({ 3, 2, 1}, t2) end) it("returns 1 key/value if end == start index", function() local t1 = { 1, 2, 3 } local t2 = {} for i, v in npairs(t1, 2, 2) do t2[i] = v end assert.are.same({ [2] = 2 }, t2) end) it("returns negative to positive ranges", function() local t1 = { [-5] = -5, [-4] = -4, [-3] = -3, [-2] = -2, [-1] = -1, [0] = 0, 1, 2, 3 } local t2 = {} for i, v in npairs(t1, -4, 1) do t2[i] = v end assert.are.same({ [-4] = -4, [-3] = -3, [-2] = -2, [-1] = -1, [0] = 0, 1 }, t2) end) it("returns nil values with the range", function() local t1 = { n = 3 } local t2 = {} for i, v in npairs(t1) do t2[i] = tostring(v) end assert.are.same({ "nil", "nil", "nil" }, t2) end) it("honours positive step size", function() local t1 = { [-5] = -5, [-4] = -4, [-3] = -3, [-2] = -2, [-1] = -1, [0] = 0, 1, 2, 3 } local t2 = {} for i, v in npairs(t1, -4, 1, 2) do t2[#t2+1] = v end assert.are.same({ -4, -2, 0}, t2) end) it("honours negative step size", function() local t1 = { [-5] = -5, [-4] = -4, [-3] = -3, [-2] = -2, [-1] = -1, [0] = 0, 1, 2, 3 } local t2 = {} for i, v in npairs(t1, 0, -5, -2) do t2[#t2+1] = v end assert.are.same({ 0, -2, -4 }, t2) end) end) end) Penlight-1.12.0/spec/xml_spec.lua000066400000000000000000000530131416703176500166300ustar00rootroot00000000000000local xml = require "pl.xml" describe("xml", function() describe("new()", function() it("creates a new xml-document", function() local doc = xml.new("main") assert.equal("
", doc:tostring()) end) it("fails without a tag", function() assert.has.error(function() xml.new() end, "expected 'tag' to be a string value, got: nil") end) it("fails with bad attributes", function() assert.has.error(function() xml.new("tag", "not a table...") end, "expected 'attr' to be a table value, got: string") end) it("adds attributes if given", function() local doc = xml.new("main", { hello = "world" }) assert.equal("
", doc:tostring()) end) end) describe("parse()", function() pending("todo", function() -- TODO: implement end) end) describe("elem()", function() it("creates a node", function() local doc = xml.elem("main") assert.equal("
", doc:tostring()) end) it("creates a node, with single text element", function() local doc = xml.elem("main", "oh my") assert.equal("
oh my
", doc:tostring()) end) it("creates a node, with single child tag/Node", function() local doc = xml.elem("main", xml.new("child")) assert.equal("
", doc:tostring()) end) it("creates a node, with multiple text elements", function() local doc = xml.elem("main", { "this ", "is ", "nice" }) assert.equal("
this is nice
", doc:tostring()) end) it("creates a node, with multiple child tags/Nodes", function() local doc = xml.elem("main", { xml.new "this", xml.new "is", xml.new "nice" }) assert.equal("
", doc:tostring()) end) it("creates a node, with attributes", function() local doc = xml.elem("main", { hello = "world" }) assert.equal("
", doc:tostring()) end) it("creates a node, with text/Node children and attributes", function() local doc = xml.elem("main", { "prefix", xml.elem("child", { "this ", "is ", "nice"}), "postfix", attrib = "value" }) assert.equal("
prefixthis is nicepostfix
", doc:tostring()) end) it("creates a node, with text/Node nested children and attributes", function() local doc = xml.elem("main", { "prefix", xml.elem("child", { "this", xml.elem "is", "nice", }), "postfix", attrib = "value" }) assert.equal("
prefixthisnicepostfix
", doc:tostring()) end) end) describe("tags()", function() it("creates constructors", function() local parent, child = xml.tags({ "mom" , "kid" }) local doc = parent {child 'Bob', child 'Annie'} assert.equal("BobAnnie", doc:tostring()) end) it("creates constructors from CSV values", function() local parent, child = xml.tags("mom,kid" ) local doc = parent {child 'Bob', child 'Annie'} assert.equal("BobAnnie", doc:tostring()) end) it("creates constructors from CSV values, ignores surrounding whitespace", function() local parent, child = xml.tags(" mom , kid " ) local doc = parent {child 'Bob', child 'Annie'} assert.equal("BobAnnie", doc:tostring()) end) end) describe("addtag()", function() it("adds a Node", function() local doc = xml.new("main") doc:addtag("penlight", { hello = "world" }) assert.equal("
", doc:tostring()) -- moves position doc:addtag("expat") assert.equal("
", doc:tostring()) end) end) describe("text()", function() it("adds text", function() local doc = xml.new("main") doc:text("penlight") assert.equal("
penlight
", doc:tostring()) -- moves position doc:text("expat") assert.equal("
penlightexpat
", doc:tostring()) end) end) describe("up()", function() it("moves position up 1 level", function() local doc = xml.new("main") doc:addtag("one") doc:addtag("two-a") doc:up() doc:addtag("two-b") assert.equal("
", doc:tostring()) -- doesn't move beyond top level for i = 1, 10 do doc:up() end doc:addtag("solong") assert.equal("
", doc:tostring()) end) end) describe("reset()", function() it("resets position to top Node", function() local doc = xml.new("main") doc:addtag("one") doc:addtag("two") doc:addtag("three") doc:reset() doc:addtag("solong") assert.equal("
", doc:tostring()) end) end) describe("add_direct_child", function() it("adds a child node", function() local doc = xml.new("main") doc:add_direct_child(xml.new("child")) assert.equal("
", doc:tostring()) doc:add_direct_child(xml.new("child")) assert.equal("
", doc:tostring()) end) it("adds a text node", function() local doc = xml.new("main") doc:add_direct_child("child") assert.equal("
child
", doc:tostring()) doc:add_direct_child("child") assert.equal("
childchild
", doc:tostring()) end) end) describe("add_child()", function() it("adds a child at the current position", function() local doc = xml.new("main") doc:addtag("one") doc:add_child(xml.new("item1")) doc:add_child(xml.new("item2")) doc:add_child(xml.new("item3")) assert.equal("
", doc:tostring()) end) end) describe("set_attribs()", function() it("sets attributes on the Node", function() local doc = xml.new("main") doc:addtag("one") -- moves position doc:set_attribs( { one = "a" }) assert.equal("
", doc:tostring()) -- overwrites and adds doc:set_attribs( { one = "1", two = "2" }) assert.matches("one='1'", doc:tostring()) assert.matches("two='2'", doc:tostring()) -- 'two' doesn't get removed doc:set_attribs( { one = "a" }) assert.matches("one='a'", doc:tostring()) assert.matches("two='2'", doc:tostring()) end) end) describe("set_attrib()", function() it("sets/deletes a single attribute on the Node", function() local doc = xml.new("main") doc:addtag("one") -- moves position doc:set_attrib("one", "a") assert.equal("
", doc:tostring()) -- deletes doc:set_attrib("one", nil) assert.equal("
", doc:tostring()) end) end) describe("get_attribs()", function() it("gets attributes on the Node", function() local doc = xml.new("main") doc:addtag("one") -- moves position doc:set_attribs( { one = "1", two = "2" }) assert.same({ one = "1", two = "2" }, doc:get_attribs()) end) end) describe("subst()", function() pending("todo", function() -- TODO: implement end) end) describe("child_with_name()", function() it("returns the first child", function() local doc = xml.new("main") doc:add_child(xml.elem "one") doc:text("hello") doc:add_child(xml.elem "two") doc:text("goodbye") doc:add_child(xml.elem "three") local child = doc:child_with_name("two") assert.not_nil(child) assert.equal(doc[3], child) end) end) describe("tostring()", function() pending("todo still...", function() -- TODO: implement end) end) describe("get_elements_with_name()", function() it("returns matching nodes", function() local doc = assert(xml.parse[[ John Bob Bob junior Annie Melissa Noel ]]) local list = doc:get_elements_with_name("name") for i, entry in ipairs(list) do list[i] = entry:get_text() end assert.same({"John", "Bob", "Bob junior", "Annie", "Melissa", "Noel"}, list) -- if tag not found, returns empty table local list = doc:get_elements_with_name("unknown") assert.same({}, list) end) end) describe("children()", function() it("iterates over all children", function() local doc = xml.elem("main", { "prefix", xml.elem("child"), "postfix", attrib = "value" }) local lst = {} for node in doc:children() do lst[#lst+1] = tostring(node) end assert.same({ "prefix", "", "postfix"}, lst) end) it("doesn't fail on empty node", function() local doc = xml.elem("main") local lst = {} for node in doc:children() do lst[#lst+1] = tostring(node) end assert.same({}, lst) end) end) describe("first_childtag()", function() it("returns first non-text tag", function() local doc = xml.elem("main", { "prefix", xml.elem("child"), "postfix", attrib = "value" }) local node = doc:first_childtag() assert.same("", tostring(node)) end) it("returns nil if there is none", function() local doc = xml.elem("main", { "prefix", "postfix", attrib = "value" }) local node = doc:first_childtag() assert.is_nil(node) end) end) describe("matching_tags()", function() local _ = [[ Apples Bananas African Coffee Table 80 120 ]] pending("xmlns is weird...", function() -- the xmlns stuff doesn't make sense end) end) describe("childtags()", function() it("returns the first child", function() local doc = xml.new("main") doc:add_child(xml.elem "one") doc:text("hello") doc:add_child(xml.elem "two") doc:text("goodbye") doc:add_child(xml.elem "three") local lst = {} for node in doc:childtags() do lst[#lst+1] = tostring(node) end assert.same({"", "", ""},lst) end) end) describe("maptags()", function() it("updates nodes", function() local doc = xml.new("main") doc:add_child(xml.elem "one") doc:text("hello") doc:add_child(xml.elem "two") doc:text("goodbye") doc:add_child(xml.elem "three") doc:maptags(function(node) if node.tag then -- return a new object so we know it got replaced return xml.new(node.tag:upper()) end return node end) assert.same("
hellogoodbye
", doc:tostring()) end) it("removes nodes", function() local doc = xml.new("main") doc:add_child(xml.elem "one") doc:text("hello") doc:add_child(xml.elem "two") doc:text("goodbye") doc:add_child(xml.elem "three") doc:maptags(function(node) if node.tag then return nil -- remove it end return node end) assert.same("
hellogoodbye
", doc:tostring()) end) end) describe("xml_escape()", function() it("escapes reserved characters", function() local esc = xml.xml_escape([["'<>&]]) assert.same(""'<>&", esc) end) end) describe("xml_unescape()", function() it("escapes reserved characters", function() local unesc = xml.xml_unescape(""'<>&") assert.same([["'<>&]], unesc) end) end) describe("get_text()", function() it("returns all text concatenated", function() local doc = xml.new("main") doc:text("one") doc:add_child(xml.elem "two") doc:text("three") assert.same("onethree", doc:get_text()) end) it("returns empty string if no text", function() local doc = xml.new("main") doc:add_child(xml.elem "two") assert.same("", doc:get_text()) end) end) describe("clone()", function() it("clones a document", function() local doc1 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) local doc2 = xml.clone(doc1) assert.are.same(doc1:tostring(), doc2:tostring()) assert.not_equal(doc1, doc2) for i, elem1 in ipairs(doc1) do assert.are.same(tostring(elem1), tostring(doc2[i])) if type(elem1) == "table" then assert.not_equal(elem1, doc2[i]) end end end) it("calls substitution callback and updates", function() local doc1 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) local repl = { ["*TAG"] = { main = "top", this = "that", is = "was", a = "a", b = "b", nice = "bad", }, ["*TEXT"] = { ["this content"] = "that content", }, hello = { world = "universe", }, } local subst = function(object, kind, parent) if repl[kind] then if repl[kind][object] then return repl[kind][object] else error(("object '%s' of kind '%s' not found"):format(object,kind)) end else error(("kind '%s' not found"):format(kind)) end end local doc2 = xml.clone(doc1, subst) assert.equal("that content", doc2:tostring()) end) it("clones text nodes", function() assert.equal("hello", xml.clone("hello")) end) it("errors on recursion", function() local doc = xml.elem("main", { hello = "world", xml.elem("this", "this content"), "some", xml.elem "is", "text", xml.elem "nice", }) doc[#doc+1] = doc -- add recursion assert.has.error(function() xml.clone(doc) end, "recursion detected") end) end) describe("compare()", function() it("returns true on equal docs", function() local doc1 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) local doc2 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) local ok, err = xml.compare(doc1, doc2) assert.is_nil(err) assert.is_true(ok) end) it("compares types", function() local ok, err = xml.compare(nil, true) assert.equal("type mismatch", err) assert.is_false(ok) local ok, err = xml.compare("true", true) assert.equal("type mismatch", err) assert.is_false(ok) local ok, err = xml.compare(true, true) assert.equal("not a document", err) assert.is_false(ok) local ok, err = xml.compare("text", "text") assert.is_nil(err) assert.is_true(ok) local ok, err = xml.compare("text1", "text2") assert.equal("text text1 ~= text text2", err) assert.is_false(ok) end) it("compares element size (array part)", function() local doc1 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) local doc2 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", "plain text", }) local ok, err = xml.compare(doc1, doc2) assert.equal("size 3 ~= size 4 for tag main", err) assert.is_false(ok) end) it("compares children", function() local doc1 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) local doc2 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "c" }), xml.elem "nice", }) local ok, err = xml.compare(doc1, doc2) assert.equal("tag b ~= tag c", err) assert.is_false(ok) end) it("compares attributes", function() local doc1 = xml.new("main", { hello = "world", goodbye = "universe" }) local ok, err = xml.compare(doc1, xml.new("main", { hello = "world", goodbye = "universe" })) assert.equal(nil, err) assert.is_true(ok) local ok, err = xml.compare(doc1, xml.new("main", { -- hello = "world", -- one less attribute goodbye = "universe" })) assert.equal("mismatch attrib", err) assert.is_false(ok) local ok, err = xml.compare(doc1, xml.new("main", { hello = "world", goodbye = "universe", one = "more", -- one more attribute })) assert.equal("mismatch attrib", err) assert.is_false(ok) end) it("compares attributes order", function() local doc1 = xml.new("main", { [1] = "hello", [2] = "goodbye", hello = "world", goodbye = "universe" }) local ok, err = xml.compare(doc1, xml.new("main", { -- no order, this compares ok hello = "world", goodbye = "universe" })) assert.equal(nil, err) assert.is_true(ok) local ok, err = xml.compare(doc1, xml.new("main", { [2] = "hello", -- order reversed, this should fail [1] = "goodbye", hello = "world", goodbye = "universe" })) assert.equal("mismatch attrib order", err) assert.is_false(ok) end) it("handles recursion", function() local doc1 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) local doc2 = xml.elem("main", { hello = "world", xml.elem("this", "this content"), xml.elem("is", { xml.elem "a", xml.elem "b" }), xml.elem "nice", }) doc1[#doc1 + 1] = doc1 -- add recursion doc2[#doc2 + 1] = xml.elem "main" -- add tag by same name local ok, err = xml.compare(doc1, doc2) assert.equal("recursive document", err) assert.is_false(ok) end) end) describe("walk()", function() it("calls on all tags", function() local doc = xml.elem("main", { hello = "world", xml.elem("this", "this content"), "some", xml.elem "is", "text", xml.elem "nice", }) assert.equal("
this contentsometext
", doc:tostring()) local list = {} xml.walk(doc, nil, function(tag_name, node) list[#list+1] = assert(tag_name) end) assert.same({"main", "this", "is", "nice"}, list) -- now depth_first local list = {} xml.walk(doc, true, function(tag_name, node) list[#list+1] = assert(tag_name) end) assert.same({"this", "is", "nice", "main"}, list) end) it("errors on recursion", function() local doc = xml.elem("main", { hello = "world", xml.elem("this", "this content"), "some", xml.elem "is", "text", xml.elem "nice", }) doc[#doc+1] = doc -- add recursion assert.has.error(function() xml.walk(doc, nil, function() end) end, "recursion detected") end) end) describe("parsehtml()", function() pending("to be deprecated...", function() -- TODO: implement end) end) describe("basic_parse()", function() pending("to be deprecated...", function() -- TODO: implement end) end) describe("match()", function() pending("figure out what it does...", function() -- TODO: implement end) end) end) Penlight-1.12.0/tests/000077500000000000000000000000001416703176500145215ustar00rootroot00000000000000Penlight-1.12.0/tests/lua/000077500000000000000000000000001416703176500153025ustar00rootroot00000000000000Penlight-1.12.0/tests/lua/animal.lua000066400000000000000000000016141416703176500172500ustar00rootroot00000000000000-- Module containing classes local class = require 'pl.class' local utils = require 'pl.utils' local error = error if utils.lua51 then module 'animal' else _ENV = {} end class.Animal() function Animal:_init(name) self.name = name end function Animal:__tostring() return self.name..': '..self:speak() end class.Dog(Animal) function Dog:speak() return 'bark' end class.Cat(Animal) function Cat:_init(name,breed) self:super(name) -- must init base! self.breed = breed end function Cat:speak() return 'meow' end -- you may declare the methods in-line like so; -- note the meaning of `_base`! class.Lion { _base = Cat; speak = function(self) return 'roar' end } -- a class may handle unknown methods with `catch`: Lion:catch(function(self,name) return function() error("no such method "..name,2) end end) if not utils.lua51 then return _ENV end Penlight-1.12.0/tests/lua/bar.lua000066400000000000000000000002071416703176500165500ustar00rootroot00000000000000--- test module for demonstrating app.require_here() local bar = {} function bar.name () return 'bar' end return bar Penlight-1.12.0/tests/lua/foo/000077500000000000000000000000001416703176500160655ustar00rootroot00000000000000Penlight-1.12.0/tests/lua/foo/args.lua000066400000000000000000000002071416703176500175230ustar00rootroot00000000000000--- test module for demonstrating app.require_here() local args = {} function args.answer () return 42 end return args Penlight-1.12.0/tests/lua/mod52.lua000066400000000000000000000016131416703176500167340ustar00rootroot00000000000000local test = require 'pl.test' local LUA_VERSION = _VERSION print(LUA_VERSION) -- if STRICT is true, then M is distinct from _ENV, and ONLY contains -- the exported functions! local _ENV,M = require 'pl.import_into' (rawget(_G,'STRICT')) function answer () -- of course, you don't have the usual global environment available -- so define it as a local up above, or use utils.import(_G). local versioned_errors = { ["1"] = "attempt to call global 'print'", ["2"] = "attempt to call global 'print'", ["3"] = "attempt to call a nil value", ["4"] = "a nil value", } local expected = versioned_errors[LUA_VERSION:match("Lua 5.(%d)")] test.assertraise(function() print 'hello' end, expected) -- but all the Penlight modules are available return pretty.write(utils.split '10 20 30', '') end return M Penlight-1.12.0/tests/lua/mymod.lua000066400000000000000000000007441416703176500171370ustar00rootroot00000000000000local strict = require 'pl.strict' local test = require 'pl.test' local M = strict.module (...) function M.answer () Boo = false -- fine, it's a declared global -- in strict mode, you cannot assign to globals if you aren't in main test.assertraise(function() Foo = true end," assign to undeclared global 'Foo'") return 42 end function M.question () return 'what is the answer to Life, the Universe and Everything?' end return M Penlight-1.12.0/tests/test-__vector.lua000066400000000000000000000054301416703176500200030ustar00rootroot00000000000000---- deriving specialized classes from List -- illustrating covariance of List methods local test = require 'pl.test' local class = require 'pl.class' local types = require 'pl.types' local operator = require 'pl.operator' local List = require 'pl.List' local asserteq = test.asserteq class.Vector(List) function Vector.range (x1,x2,delta) return Vector(List.range(x1,x2,delta)) end local function vbinop (op,v1,v2,scalar) if not Vector:class_of(v1) then v2, v1 = v1, v2 end if type(v2) ~= 'table' then return v1:map(op,v2) else if scalar then error("operation not permitted on two vectors",3) end if #v1 ~= #v2 then error("vectors have different lengths",3) end return v1:map2(op,v2) end end function Vector.__add (v1,v2) return vbinop(operator.add,v1,v2) end function Vector.__sub (v1,v2) return vbinop(operator.sub,v1,v2) end function Vector.__mul (v1,v2) return vbinop(operator.mul,v1,v2,true) end function Vector.__div (v1,v2) return vbinop(operator.div,v1,v2,true) end function Vector.__unm (v) return v:map(operator.unm) end Vector:catch(List.default_map_with(math)) v = Vector() assert(v:is_a(Vector)) assert(Vector:class_of(v)) v:append(10) v:append(20) asserteq(1+v,v+1) -- covariance: the inherited Vector.map returns a Vector asserteq(List{1,2} + v:map '2*_',{21,42}) u = Vector{1,2} asserteq(v + u,{11,22}) asserteq(v - u,{9,18}) asserteq (v - 1, {9,19}) asserteq(2 * v, {20,40}) -- print(v * v) -- throws error: not permitted -- print(v + Vector{1,2,3}) -- throws error: different lengths asserteq(2*v + u, {21,42}) asserteq(-v, {-10,-20}) -- Vector.slice returns the Right Thing due to covariance asserteq( Vector.range(0,1,0.1):slice(1,3)+1, {1,1.1,1.2}, 1e-8) u:transform '_+1' asserteq(u,{2,3}) u = Vector.range(0,1,0.1) asserteq( u:map(math.sin), {0,0.0998,0.1986,0.2955,0.3894,0.4794,0.5646,0.6442,0.7173,0.7833,0.8414}, 0.001) -- unknown Vector methods are assumed to be math.* functions asserteq(Vector{-1,2,3,-4}:abs(),Vector.range(1,4)) local R = Vector.range -- concatenating two Vectors returns another vector (covariance again) -- note the operator precedence here... asserteq(( R(0,1,0.1)..R(1.2,2,0.2)) + 1, {1,1.1,1.2,1.3,1.4,1.5,1.6,1.7,1.8,1.9,2,2.2,2.4,2.6,2.8,3}, 1e-8) class.Strings(List) Strings:catch(List.default_map_with(string)) ls = Strings{'one','two','three'} asserteq(ls:upper(),{'ONE','TWO','THREE'}) asserteq(ls:sub(1,2),{'on','tw','th'}) -- all map operations on specialized lists -- results in another list of that type! This isn't necessarily -- what you want. local sizes = ls:map '#' asserteq(sizes, {3,3,5}) asserteq(types.type(sizes),'Strings') asserteq(sizes:is_a(Strings),true) sizes = Vector:cast(sizes) asserteq(types.type(sizes),'Vector') asserteq(sizes+1,{4,4,6}) Penlight-1.12.0/tests/test-app.lua000066400000000000000000000177461416703176500170000ustar00rootroot00000000000000local app = require "pl.app" local utils = require "pl.utils" local path = require "pl.path" local asserteq = require 'pl.test'.asserteq local quote = utils.quote_arg local _, cmd = app.lua() cmd = cmd .. " " .. quote({"-e", "package.path=[[./lua/?.lua;./lua/?/init.lua;]]..package.path"}) local function run_script(s, fname) local tmpname = path.tmpname() if fname then tmpname = path.join(path.dirname(tmpname), fname) end assert(utils.writefile(tmpname, s)) local success, code, stdout, stderr = utils.executeex(cmd.." "..tmpname) os.remove(tmpname) return success, code, stdout, stderr end do -- app.script_name local success, code, stdout, stderr = run_script([[ print(require("pl.app").script_name()) ]], "justsomescriptname.lua") asserteq(stderr, "") asserteq(stdout:match("(justsome.+)$"), "justsomescriptname.lua\n") -- commandline, no scriptname local success, code, stdout, stderr = run_script([[ arg[0] = nil -- simulate no scriptname local name, err = require("pl.app").script_name() io.stdout:write(tostring(name)) io.stderr:write(err) ]]) assert(stderr:find("No script name found")) asserteq(stdout, "nil") -- commandline, no args table local success, code, stdout, stderr = run_script([[ arg = nil -- simulate no arg table local name, err = require("pl.app").script_name() io.stdout:write(tostring(name)) io.stderr:write(err) ]]) assert(stderr:find("No script name found")) asserteq(stdout, "nil") end do -- app.require_here local cd = path.currentdir() --path.dirname(path.tmpname()) -- plain script name local success, code, stdout, stderr = run_script([[ arg[0] = "justsomescriptname.lua" local p = package.path require("pl.app").require_here() print(package.path:sub(1, -#p-1)) ]]) asserteq(stderr, "") stdout = path.normcase(stdout) assert(stdout:find(path.normcase(cd.."/?.lua;"), 1, true)) assert(stdout:find(path.normcase(cd.."/?/init.lua;"), 1, true)) -- plain script name, with a relative base name local success, code, stdout, stderr = run_script([[ arg[0] = "justsomescriptname.lua" local p = package.path require("pl.app").require_here("basepath/to/somewhere") print(package.path:sub(1, -#p-1)) ]]) asserteq(stderr, "") stdout = path.normcase(stdout) assert(stdout:find(path.normcase(cd.."/basepath/to/somewhere/?.lua;"), 1, true)) assert(stdout:find(path.normcase(cd.."/basepath/to/somewhere/?/init.lua;"), 1, true)) -- plain script name, with an absolute base name local success, code, stdout, stderr = run_script([[ arg[0] = "justsomescriptname.lua" local p = package.path require("pl.app").require_here("/basepath/to/somewhere") print(package.path:sub(1, -#p-1)) ]]) asserteq(stderr, "") stdout = path.normcase(stdout) asserteq(stdout, path.normcase("/basepath/to/somewhere/?.lua;/basepath/to/somewhere/?/init.lua;\n")) -- scriptname with a relative path local success, code, stdout, stderr = run_script([[ arg[0] = "relative/prefix/justsomescriptname.lua" local p = package.path require("pl.app").require_here() print(package.path:sub(1, -#p-1)) os.exit() ]]) asserteq(stderr, "") stdout = path.normcase(stdout) assert(stdout:find(path.normcase(cd.."/relative/prefix/?.lua;"), 1, true)) assert(stdout:find(path.normcase(cd.."/relative/prefix/?/init.lua;"), 1, true)) -- script with an absolute path local success, code, stdout, stderr = run_script([[ arg[0] = "/fixed/justsomescriptname.lua" local p = package.path require("pl.app").require_here() print(package.path:sub(1, -#p-1)) ]]) asserteq(stderr, "") stdout = path.normcase(stdout) asserteq(stdout, path.normcase("/fixed/?.lua;/fixed/?/init.lua;\n")) end do -- app.appfile local success, code, stdout, stderr = run_script([[ arg[0] = "some/path/justsomescriptname_for_penlight_testing.lua" print(require("pl.app").appfile("filename.data")) ]]) asserteq(stderr, "") stdout = path.normcase(stdout) local fname = path.normcase(path.expanduser("~/.justsomescriptname_for_penlight_testing/filename.data")) asserteq(stdout, fname .."\n") assert(path.isdir(path.dirname(fname))) path.rmdir(path.dirname(fname)) end do -- app.lua local success, code, stdout, stderr = run_script([[ arg[0] = "justsomescriptname.lua" local a,b = require("pl.app").lua() print(a) ]]) asserteq(stderr, "") asserteq(stdout, cmd .."\n") end do -- app.parse_args -- no value specified local args = utils.split("-a -b") local t,s = app.parse_args(args, { a = true}) asserteq(t, nil) asserteq(s, "no value for 'a'") -- flag that take a value, space separated local args = utils.split("-a -b value -c") local t,s = app.parse_args(args, { b = true}) asserteq(t, { a = true, b = "value", c = true, }) asserteq(s, {}) -- flag_with_values specified as a list local args = utils.split("-a -b value -c") local t,s = app.parse_args(args, { "b" }) asserteq(t, { a = true, b = "value", c = true, }) asserteq(s, {}) -- flag_with_values missing value at end local args = utils.split("-a -b") local t,s = app.parse_args(args, { "b" }) asserteq(t, nil) asserteq(s, "no value for 'b'") -- error on an unknown flag local args = utils.split("-a -b value -c") local t,s = app.parse_args(args, { b = true }, { "b", "c" }) asserteq(t, nil) asserteq(s, "unknown flag 'a'") -- flag that doesn't take a value local args = utils.split("-a -b:value") local t,s = app.parse_args(args, {}) asserteq(t, { ["a"] = true, ["b"] = "value" }) asserteq(s, {}) -- correctly parsed values, spaces, :, =, and multiple : or = local args = utils.split("-a value -b value:one=two -c=value2:2") local t,s = app.parse_args(args, { "a", "b", "c" }) asserteq(t, { ["a"] = "value", ["b"] = "value:one=two", ["c"] = "value2:2", }) asserteq(s, {}) -- many values, duplicates, and parameters mixed local args = utils.split( "-a -b -cde --long1 --ff:ffvalue --gg=ggvalue -h:hvalue -i=ivalue " .. "-i=2ndvalue param -i:3rdvalue -j1 -k2 -1:hello remaining values") local t,s = app.parse_args(args) asserteq({ i = "3rdvalue", ["1"] = "hello", ff = "ffvalue", long1 = true, c = true, b = true, gg = "ggvalue", j = "1", k = "2", d = true, h = "hvalue", a = true, e = true }, t) asserteq({ "param", "remaining", "values" }, s) -- specify valid flags and aliasses local args = utils.split("-a -b value -e -f3") local t,s = app.parse_args(args, { "b", f = true, }, { bully = "b", -- b with value will be reported as 'bully', alias as string a = true, -- hash-type value c = { "d", "e" }, -- e will be reported as c, aliasses as list/table }) asserteq(t, { a = true, bully = "value", c = true, f = "3", }) asserteq(s, {}) -- error on an unknown flag, in a chain of short ones local args = utils.split("-b value -cd") local t,s = app.parse_args(args, { b = true }, { "b", "c" }) asserteq(t, nil) asserteq(s, "unknown flag 'd'") -- flag, in a chain of short ones, gets converted to alias local args = utils.split("-dbc") local t,s = app.parse_args(args, nil, { "d", full_name = "b", "c" }) asserteq(t, { full_name = true, -- specified as b in a chain of short ones c = true, d = true, }) asserteq(s, {}) end Penlight-1.12.0/tests/test-class.lua000066400000000000000000000061061416703176500173110ustar00rootroot00000000000000local class = require 'pl.class' local test = require 'pl.test' asserteq = test.asserteq T = test.tuple A = class() function A:_init () self.a = 1 end -- calling base class' ctor automatically A1 = class(A) asserteq(A1(),{a=1}) -- explicitly calling base ctor with super B = class(A) function B:_init () self:super() self.b = 2 end function B:foo () self.eee = 1 end function B:foo2 () self.g = 8 end asserteq(B(),{a=1,b=2}) -- can continue this chain C = class(B) function C:_init () self:super() self.c = 3 end function C:foo () -- recommended way to call inherited version of method... B.foo(self) end c = C() c:foo() asserteq(c,{a=1,b=2,c=3,eee=1}) -- test indirect inherit D = class(C) E = class(D) function E:_init () self:super() self.e = 4 end function E:foo () -- recommended way to call inherited version of method... self.eeee = 5 C.foo(self) end F = class(E) function F:_init () self:super() self.f = 6 end f = F() f:foo() f:foo2() -- Test : invocation inherits this function from all the way up in B asserteq(f,{a=1,b=2,c=3,eee=1,e=4,eeee=5,f=6,g=8}) -- Test that inappropriate calls to super() fail gracefully G = class() -- Class with no init H = class(G) -- Class with an init that wrongly calls super() function H:_init() self:super() -- Notice: G has no _init end I = class(H) -- Inherits the init with a bad super J = class(I) -- Grandparent-inits the init with a bad super K = class(J) -- Has an init, which calls the init with a bad super function K:_init() self:super() end local function createG() return G() end local function createH() -- Wrapper function for pcall return H() end local function createJ() return J() end local function createK() return K() end assert(pcall(createG)) -- Should succeed assert(not pcall(createH)) -- These three should fail assert(not pcall(createJ)) assert(not pcall(createK)) --- class methods! assert(c:is_a(C)) assert(c:is_a(B)) assert(c:is_a(A)) assert(c:is_a() == C) assert(C:class_of(c)) assert(B:class_of(c)) assert(A:class_of(c)) --- metamethods! function C:__tostring () return ("%d:%d:%d"):format(self.a,self.b,self.c) end function C.__eq (c1,c2) return c1.a == c2.a and c1.b == c2.b and c1.c == c2.c end asserteq(C(),{a=1,b=2,c=3}) asserteq(tostring(C()),"1:2:3") asserteq(C()==C(),true) ----- properties ----- local MyProps = class(class.properties) local setted_a, got_b function MyProps:_init () self._a = 1 self._b = 2 end function MyProps:set_a (v) setted_a = true self._a = v end function MyProps:get_b () got_b = true return self._b end function MyProps:set (a,b) self._a = a self._b = b end local mp = MyProps() mp.a = 10 asserteq(mp.a,10) asserteq(mp.b,2) asserteq(setted_a and got_b, true) class.MoreProps(MyProps) local setted_c function MoreProps:_init() self:super() self._c = 3 end function MoreProps:set_c (c) setted_c = true self._c = c end mm = MoreProps() mm:set(10,20) mm.c = 30 asserteq(setted_c, true) asserteq(T(mm.a, mm.b, mm.c),T(10,20,30)) Penlight-1.12.0/tests/test-class2.lua000066400000000000000000000012251416703176500173700ustar00rootroot00000000000000-- animal.lua require 'pl.app'.require_here 'lua' local test = require 'pl.test' local asserteq = test.asserteq local A = require 'animal' local fido, felix, leo fido = A.Dog('Fido') felix = A.Cat('Felix','Tabby') leo = A.Lion('Leo','African') asserteq(fido:speak(),'bark') asserteq(felix:speak(),'meow') asserteq(leo:speak(),'roar') asserteq(tostring(leo),'Leo: roar') test.assertraise(function() leo:circus_act() end, "no such method circus_act") asserteq(leo:is_a(A.Animal),true) asserteq(leo:is_a(A.Dog),false) asserteq(leo:is_a(A.Cat),true) asserteq(A.Dog:class_of(leo),false) asserteq(A.Cat:class_of(leo),true) asserteq(A.Lion:class_of(leo),true) Penlight-1.12.0/tests/test-class3.lua000066400000000000000000000010741416703176500173730ustar00rootroot00000000000000-- another way to define classes. Works particularly well -- with Moonscript local class = require('pl.class') local A = class{ _init = function(self, name) self.name = name end, greet = function(self) return "hello " .. self.name end, __tostring = function(self) return self.name end } local B = class{ _base = A, greet = function(self) return "hola " .. self.name end } local a = A('john') assert(a:greet()=="hello john") assert(tostring(a) == "john") local b = B('juan') assert(b:greet()=="hola juan") assert(tostring(b)=="juan") Penlight-1.12.0/tests/test-class4.lua000066400000000000000000000007571416703176500174030ustar00rootroot00000000000000local class = require 'pl.class' local A = class() function A:_init() self.init_chain = "A" end local B = class(A) local C = class(B) function C:_init() self:super() self.init_chain = self.init_chain.."C" end local D = class(C) local E = class(D) function E:_init() self:super() self.init_chain = self.init_chain.."E" end local F = class(E) local G = class(F) function G:_init() self:super() self.init_chain = self.init_chain.."G" end local i = G() assert(i.init_chain == "ACEG") Penlight-1.12.0/tests/test-compat.lua000066400000000000000000000017101416703176500174630ustar00rootroot00000000000000local test = require 'pl.test' local asserteq = test.asserteq local compat = require "pl.compat" local coroutine = require "coroutine" local code_generator = coroutine.wrap(function() local result = {"ret", "urn \"Hello World!\""} for _,v in ipairs(result) do coroutine.yield(v) end coroutine.yield(nil) end) local f, err = compat.load(code_generator) asserteq(err, nil) asserteq(f(), "Hello World!") -- package.searchpath if compat.lua51 and not compat.jit then assert(package.searchpath("pl.compat", package.path):match("lua[/\\]pl[/\\]compat")) local path = "some/?/nice.path;another/?.path" local ok, err = package.searchpath("my.file.name", path, ".", "/") asserteq(err, "\tno file 'some/my/file/name/nice.path'\n\tno file 'another/my/file/name.path'") local ok, err = package.searchpath("my/file/name", path, "/", ".") asserteq(err, "\tno file 'some/my.file.name/nice.path'\n\tno file 'another/my.file.name.path'") endPenlight-1.12.0/tests/test-comprehension.lua000066400000000000000000000046121416703176500210550ustar00rootroot00000000000000-- test-comprehension.lua -- test of comprehension.lua local utils = require 'pl.utils' local comp = require 'pl.comprehension' . new() local asserteq = require 'pl.test' . asserteq -- test of list building asserteq(comp 'x for x' {}, {}) asserteq(comp 'x for x' {2,3}, {2,3}) asserteq(comp 'x^2 for x' {2,3}, {2^2,3^2}) asserteq(comp 'x for x if x % 2 == 0' {4,5,6,7}, {4,6}) asserteq(comp '{x,y} for x for y if x>2 if y>4' ({2,3},{4,5}), {{3,5}}) -- test of table building local t = comp 'table(x,x+1 for x)' {3,4} assert(t[3] == 3+1 and t[4] == 4+1) local t = comp 'table(x,x+y for x for y)' ({3,4}, {2}) assert(t[3] == 3+2 and t[4] == 4+2) local t = comp 'table(v,k for k,v in pairs(_1))' {[3]=5, [5]=7} assert(t[5] == 3 and t[7] == 5) -- test of sum assert(comp 'sum(x for x)' {} == 0) assert(comp 'sum(x for x)' {2,3} == 2+3) assert(comp 'sum(x^2 for x)' {2,3} == 2^2+3^2) assert(comp 'sum(x*y for x for y)' ({2,3}, {4,5}) == 2*4+2*5+3*4+3*5) assert(comp 'sum(x^2 for x if x % 2 == 0)' {4,5,6,7} == 4^2+6^2) assert(comp 'sum(x*y for x for y if x>2 if y>4)' ({2,3}, {4,5}) == 3*5) -- test of min/max assert(comp 'min(x for x)' {3,5,2,4} == 2) assert(comp 'max(x for x)' {3,5,2,4} == 5) -- test of placeholder parameters -- assert(comp 'sum(x^_1 + _3 for x if x >= _4)' (2, nil, 3, 4, {3,4,5}) == 4^2+3 + 5^2+3) -- test of for = assert(comp 'sum(x^2 for x=2,3)' () == 2^2+3^2) assert(comp 'sum(x^2 for x=2,6,1+1)' () == 2^2+4^2+6^2) assert(comp 'sum(x*y*z for x=1,2 for y=3,3 for z)' {5,6} == 1*3*5 + 2*3*5 + 1*3*6 + 2*3*6) assert(comp 'sum(x*y*z for z for x=1,2 for y=3,3)' {5,6} == 1*3*5 + 2*3*5 + 1*3*6 + 2*3*6) -- test of for in assert(comp 'sum(i*v for i,v in ipairs(_1))' {2,3} == 1*2+2*3) assert(comp 'sum(i*v for i,v in _1,_2,_3)' (ipairs{2,3}) == 1*2+2*3) -- test of difficult syntax asserteq(comp '" x for x " for x' {2}, {' x for x '}) asserteq(comp 'x --[=[for x\n\n]=] for x' {2}, {2}) asserteq(comp '(function() for i = 1,1 do return x*2 end end)() for x' {2}, {4}) assert(comp 'sum(("_5" and x)^_1 --[[_6]] for x)' (2, {4,5}) == 4^2 + 5^2) -- error checking assert(({pcall(function() comp 'x for __result' end)})[2] :find'not contain __ prefix') -- environment. -- Note: generated functions are set to the environment of the 'new' call. asserteq(5,(function() local env = {d = 5} local comp = comp.new(env) return comp 'sum(d for x)' {1} end)()); print 'DONE' Penlight-1.12.0/tests/test-config.lua000066400000000000000000000144031416703176500174500ustar00rootroot00000000000000local config = require 'pl.config' local stringio = require 'pl.stringio' asserteq = require 'pl.test'.asserteq function testconfig(test,tbl,cfg) local f = stringio.open(test) local c = config.read(f,cfg) f:close() if not tbl then print(pretty.write(c)) else asserteq(c,tbl) end end testconfig ([[ ; comment 2 (an ini file) [section!] bonzo.dog=20,30 config_parm=here we go again depth = 2 [another] felix="cat" ]],{ section_ = { bonzo_dog = { -- comma-sep values get split by default 20, 30 }, depth = 2, config_parm = "here we go again" }, another = { felix = "\"cat\"" } }) testconfig ([[ # this is a more Unix-y config file fred = 1 alice = 2 home.dog = /bonzo/dog/etc ]],{ home_dog = "/bonzo/dog/etc", -- note the default is {variablilize = true} fred = 1, alice = 2 }) -- backspace line continuation works, thanks to config.lines function testconfig ([[ foo=frodo,a,c,d, \ frank, alice, boyo ]], { foo = { "frodo", "a", "c", "d", "frank", "alice", "boyo" } } ) ------ options to control default behaviour ----- -- want to keep key names as is! testconfig ([[ alpha.dog=10 # comment here ]],{ ["alpha.dog"]=10 },{variabilize=false}) -- don't convert strings to numbers testconfig ([[ alpha.dog=10 ; comment here ]],{ alpha_dog="10" },{convert_numbers=false}) -- convert strings to booleans testconfig ([[ alpha.dog=false alpha.cat=true ; comment here ]],{ alpha_dog=false, alpha_cat=true },{convert_boolean=true}) -- don't split comma-lists by setting the list delimiter to something else testconfig ([[ extra=10,'hello',42 ]],{ extra="10,'hello',42" },{list_delim='@'}) -- Unix-style password file testconfig([[ lp:x:7:7:lp:/var/spool/lpd:/bin/sh mail:x:8:8:mail:/var/mail:/bin/sh news:x:9:9:news:/var/spool/news:/bin/sh ]], { { "lp", "x", 7, 7, "lp", "/var/spool/lpd", "/bin/sh" }, { "mail", "x", 8, 8, "mail", "/var/mail", "/bin/sh" }, { "news", "x", 9, 9, "news", "/var/spool/news", "/bin/sh" } }, {list_delim=':'}) -- Unix updatedb.conf is in shell script form, but config.read -- copes by extracting the variables as keys and the export -- commands as the array part; there is an option to remove quotes -- from values testconfig([[ # Global options for invocations of find(1) FINDOPTIONS='-ignore_readdir_race' export FINDOPTIONS ]],{ "export FINDOPTIONS", FINDOPTIONS = "-ignore_readdir_race" },{trim_quotes=true}) -- Unix fstab format. No key/value assignments so use `ignore_assign`; -- list values are separated by a number of spaces testconfig([[ # proc /proc proc defaults 0 0 /dev/sda1 / ext3 defaults,errors=remount-ro 0 1 ]], { { "proc", "/proc", "proc", "defaults", 0, 0 }, { "/dev/sda1", "/", "ext3", "defaults,errors=remount-ro", 0, 1 } }, {list_delim='%s+',ignore_assign=true} ) -- Linux procfs 'files' often use ':' as the key/pair separator; -- a custom convert_numbers handles the units properly! -- Here is the first two lines from /proc/meminfo testconfig([[ MemTotal: 1024748 kB MemFree: 220292 kB ]], { MemTotal = 1024748, MemFree = 220292 }, { keysep = ':', convert_numbers = function(s) s = s:gsub(' kB$','') return tonumber(s) end } ) -- altho this works, rather use pl.data.read for this kind of purpose. testconfig ([[ # this is just a set of comma-separated values 1000,444,222 44,555,224 ]],{ { 1000, 444, 222 }, { 44, 555, 224 } }) --- new with 1.0.3: smart configuration file reading -- handles a number of common Unix file formats automatically function smart(f) f = stringio.open(f) return config.read(f,{smart=true}) end -- /etc/fstab asserteq (smart[[ # /etc/fstab: static file system information. # # Use 'blkid -o value -s UUID' to print the universally unique identifier # for a device; this may be used with UUID= as a more robust way to name # devices that works even if disks are added and removed. See fstab(5). # # proc /proc proc nodev,noexec,nosuid 0 0 /dev/sdb2 / ext2 errors=remount-ro 0 1 /dev/fd0 /media/floppy0 auto rw,user,noauto,exec,utf8 0 0 ]],{ proc = { "/proc", "proc", "nodev,noexec,nosuid", 0, 0 }, ["/dev/sdb2"] = { "/", "ext2", "errors=remount-ro", 0, 1 }, ["/dev/fd0"] = { "/media/floppy0", "auto", "rw,user,noauto,exec,utf8", 0, 0 } }) -- /proc/XXXX/status asserteq (smart[[ Name: bash State: S (sleeping) Tgid: 30071 Pid: 30071 PPid: 1587 TracerPid: 0 Uid: 1000 1000 1000 1000 Gid: 1000 1000 1000 1000 FDSize: 256 Groups: 4 20 24 46 105 119 122 1000 VmPeak: 6780 kB VmSize: 6716 kB ]],{ Pid = 30071, VmSize = 6716, PPid = 1587, Tgid = 30071, State = "S (sleeping)", Uid = "1000 1000 1000 1000", Name = "bash", Gid = "1000 1000 1000 1000", Groups = "4 20 24 46 105 119 122 1000", FDSize = 256, VmPeak = 6780, TracerPid = 0 }) -- ssh_config asserteq (smart[[ Host * # ForwardAgent no # ForwardX11 no # Tunnel no # TunnelDevice any:any # PermitLocalCommand no # VisualHostKey no SendEnv LANG LC_* HashKnownHosts yes GSSAPIAuthentication yes GSSAPIDelegateCredentials no ]],{ Host = "*", GSSAPIAuthentication = "yes", SendEnv = "LANG LC_*", HashKnownHosts = "yes", GSSAPIDelegateCredentials = "no" }) -- updatedb.conf asserteq (smart[[ PRUNE_BIND_MOUNTS="yes" # PRUNENAMES=".git .bzr .hg .svn" PRUNEPATHS="/tmp /var/spool /media" PRUNEFS="NFS nfs nfs4 rpc_pipefs afs binfmt_misc proc smbfs autofs iso9660 ncpfs coda devpts ftpfs devfs mfs shfs sysfs cifs lustre_lite tmpfs usbfs udf fuse.glusterfs fuse.sshfs ecryptfs fusesmb devtmpfs" ]],{ PRUNEPATHS = "/tmp /var/spool /media", PRUNE_BIND_MOUNTS = "yes", PRUNEFS = "NFS nfs nfs4 rpc_pipefs afs binfmt_misc proc smbfs autofs iso9660 ncpfs coda devpts ftpfs devfs mfs shfs sysfs cifs lustre_lite tmpfs usbfs udf fuse.glusterfs fuse.sshfs ecryptfs fusesmb devtmpfs" }) Penlight-1.12.0/tests/test-data.lua000066400000000000000000000161571416703176500171240ustar00rootroot00000000000000local data = require 'pl.data' local List = require 'pl.List' local array = require 'pl.array2d' local func = require 'pl.func' local seq = require 'pl.seq' local stringio = require 'pl.stringio' local open = stringio. open local asserteq = require 'pl.test' . asserteq local T = require 'pl.test'. tuple --[=[ dat,err = data.read(open [[ 1.0 0.1 0.2 1.3 ]]) if err then print(err) end require 'pl.pretty'.dump(dat) os.exit(0) --]=] -- tab-separated data, explicit column names local t1f = open [[ EventID Magnitude LocationX LocationY LocationZ LocationError EventDate DataFile 981124001 2.0 18988.4 10047.1 4149.7 33.8 24/11/1998 11:18:05 981124DF.AAB 981125001 0.8 19104.0 9970.4 5088.7 3.0 25/11/1998 05:44:54 981125DF.AAB 981127003 0.5 19012.5 9946.9 3831.2 46.0 27/11/1998 17:15:17 981127DF.AAD 981127005 0.6 18676.4 10606.2 3761.9 4.4 27/11/1998 17:46:36 981127DF.AAF 981127006 0.2 19109.9 9716.5 3612.0 11.8 27/11/1998 19:29:51 981127DF.AAG ]] local t1 = data.read (t1f) -- column_by_name returns a List asserteq(t1:column_by_name 'Magnitude',List{2,0.8,0.5,0.6,0.2}) -- can use array.column as well asserteq(array.column(t1,2),{2,0.8,0.5,0.6,0.2}) -- only numerical columns (deduced from first data row) are converted by default -- can look up indices in the list fieldnames. local EDI = t1.fieldnames:index 'EventDate' assert(type(t1[1][EDI]) == 'string') -- select method returns a sequence, in this case single-valued. -- (Note that seq.copy returns a List) asserteq(seq(t1:select 'LocationX where Magnitude > 0.5'):copy(),List{18988.4,19104,18676.4}) --[[ --a common select usage pattern: for event,mag in t1:select 'EventID,Magnitude sort by Magnitude desc' do print(event,mag) end --]] -- space-separated, but with last field containing spaces. local t2f = open [[ USER PID %MEM %CPU COMMAND sdonovan 2333 0.3 0.1 background --n=2 root 2332 0.4 0.2 fred --start=yes root 2338 0.2 0.1 backyard-process ]] local t2,err = data.read(t2f,{last_field_collect=true}) if not t2 then return print (err) end -- the last_field_collect option is useful with space-delimited data where the last -- field may contain spaces. Otherwise, a record count mismatch should be an error! local lt2 = List(t2[2]) asserteq(lt2:join ',','root,2332,0.4,0.2,fred --start=yes') -- fieldnames are converted into valid identifiers by substituting _ -- (we do this to make select queries parseable by Lua) asserteq(t2.fieldnames,List{'USER','PID','_MEM','_CPU','COMMAND'}) -- select queries are NOT SQL so remember to use == ! (and no 'between' operator, sorry) --s,err = t2:select('_MEM where USER="root"') --assert(err == [[[string "tmp"]:9: unexpected symbol near '=']]) local s = t2:select('_MEM where USER=="root"') assert(s() == 0.4) assert(s() == 0.2) assert(s() == nil) -- CSV, Excel style. Double-quoted fields are allowed, and they may contain commas! local t3f = open [[ "Department Name","Employee ID",Project,"Hours Booked" sales,1231,overhead,4 sales,1255,overhead,3 engineering,1501,development,5 engineering,1501,maintenance,3 engineering,1433,maintenance,10 ]] local t3 = data.read(t3f,{csv=true}) -- although fieldnames are turned in valid Lua identifiers, there is always `original_fieldnames` asserteq(t3.fieldnames,List{'Department_Name','Employee_ID','Project','Hours_Booked'}) asserteq(t3.original_fieldnames,List{'Department Name','Employee ID','Project','Hours Booked'}) -- a common operation is to select using a given list of columns, and each row -- on some explicit condition. The select() method can take a table with these -- parameters local keepcols = {'Employee_ID','Hours_Booked'} local q = t3:select { fields = keepcols, where = function(row) return row[1]=='engineering' end } asserteq(seq.copy2(q),{{1501,5},{1501,3},{1433,10}}) -- another pattern is doing a select to restrict rows & columns, process some -- fields and write out the modified rows. local outf = stringio.create() local names = {[1501]='don',[1433]='dilbert'} t3:write_row (outf,{'Employee','Hours_Booked'}) q = t3:select_row {fields=keepcols,where=func.Eq(func._1[1],'engineering')} for row in q do row[1] = names[row[1]] t3:write_row(outf,row) end asserteq(outf:value(), [[ Employee,Hours_Booked don,5 don,3 dilbert,10 ]]) -- data may not always have column headers. When creating a data object -- from a two-dimensional array, may specify the fieldnames, as a list or a string. -- The delimiter is deduced from the fieldname string, so a string just containing -- the delimiter will set it, and the fieldnames will be empty. local dat = List() local row = List.range(1,10) for i = 1,10 do dat:append(row:map('*',i)) end dat = data.new(dat,',') local out = stringio.create() dat:write(out,',') asserteq(out:value(), [[ 1,2,3,4,5,6,7,8,9,10 2,4,6,8,10,12,14,16,18,20 3,6,9,12,15,18,21,24,27,30 4,8,12,16,20,24,28,32,36,40 5,10,15,20,25,30,35,40,45,50 6,12,18,24,30,36,42,48,54,60 7,14,21,28,35,42,49,56,63,70 8,16,24,32,40,48,56,64,72,80 9,18,27,36,45,54,63,72,81,90 10,20,30,40,50,60,70,80,90,100 ]]) -- you can always use numerical field indices, AWK-style; -- note how the copy_select method gives you a data object instead of an -- iterator over the fields local res = dat:copy_select '$1,$3 where $1 > 5' local L = List asserteq(L(res),L{ L{6, 18}, L{7,21}, L{8,24}, L{9,27}, L{10,30}, }) -- the column_by_name method may take a fieldname or an index asserteq(dat:column_by_name(2), L{2,4,6,8,10,12,14,16,18,20}) -- the field list may contain expressions or even constants local q = dat:select '$3,2*$4 where $1 == 8' asserteq(T(q()),T(24,64)) dat,err = data.read(open [[ 1.0 0.1 0.2 1.3 ]]) if err then print(err) end -- if a method cannot be found, then we look up in array2d -- array2d.flatten(t) makes a 1D list out of a 2D array, -- and then List.minmax() gets the extrema. asserteq(T(dat:flatten():minmax()),T(0.1,1.3)) local f = open [[ Time Message 1266840760 +# EE7C0600006F0D00C00F06010302054000000308010A00002B00407B00 1266840760 closure data 0.000000 1972 1972 0 1266840760 ++ 1266840760 EE 1 1266840760 +# EE7C0600006F0D00C00F06010302054000000408020A00002B00407B00 1266840764 closure data 0.000000 1972 1972 0 1266840764 ++ 1266840764 EE 1 1266840764 +# EE7C0600006F0D00C00F06010302054000000508030A00002B00407B00 1266840768 duplicate? 1266840768 +# EE7C0600006F0D00C00F06010302054000000508030A00002B00407B00 1266840768 closure data 0.000000 1972 1972 0 ]] -- the `convert` option provides custom converters for each specified column. -- Here we convert the timestamps into Date objects and collect everything -- else into one field local Date = require 'pl.Date' local function date_convert (ds) return Date(tonumber(ds)) end local d = data.read(f,{convert={[1]=date_convert},last_field_collect=true}) asserteq(#d[1],2) asserteq(d[2][1]:year(),2010) d = {{1,2,3},{10,20,30}} out = stringio.create() data.write(d,out,{'A','B','C'},',') asserteq(out:value(), [[ A,B,C 1,2,3 10,20,30 ]]) out = stringio.create() d.fieldnames = {'A','B','C'} data.write(d,out) asserteq(out:value(), [[ A B C 1 2 3 10 20 30 ]]) d = data.read(stringio.open 'One,Two\n1,\n,20\n',{csv=true}) asserteq(d,{ {1,0},{0,20}, original_fieldnames={"One","Two"},fieldnames={"One","Two"},delim="," }) Penlight-1.12.0/tests/test-data2.lua000066400000000000000000000010601416703176500171710ustar00rootroot00000000000000local utils = require 'pl.utils' local stringio = require 'pl.stringio' local data = require 'pl.data' local test = require 'pl.test' utils.on_error 'quit' stuff = [[ Department Name,Employee ID,Project,Hours Booked sales, 1231,overhead,4 sales,1255,overhead,3 engineering,1501,development,5 engineering,1501,maintenance,3 engineering,1433,maintenance,10 ]] t = data.read(stringio.open(stuff)) q = t:select 'Employee_ID,Hours_Booked where Department_Name == "engineering"' test.asserteq2(1501,5,q()) test.asserteq2(1501,3,q()) test.asserteq2(1433,10,q()) Penlight-1.12.0/tests/test-date.lua000066400000000000000000000052721416703176500171240ustar00rootroot00000000000000local test = require 'pl.test' local asserteq, assertmatch = test.asserteq, test.assertmatch local dump = require 'pl.pretty'.dump local T = require 'pl.test'.tuple local Date = require 'pl.Date' iso = Date.Format 'yyyy-mm-dd' -- ISO date d = iso:parse '2010-04-10' asserteq(T(d:day(),d:month(),d:year()),T(10,4,2010)) amer = Date.Format 'mm/dd/yyyy' -- American style s = amer:tostring(d) dc = amer:parse(s) asserteq(d,dc) d = Date() -- today d:add { day = 1 } -- tomorrow assert(d > Date()) --------- Time intervals ----- -- new explicit Date.Interval class; also returned by Date:diff d1 = Date.Interval(1202) d2 = Date.Interval(1500) asserteq(tostring(d2:diff(d1)),"4 min 58 sec ") -------- testing 'flexible' date parsing --------- local df = Date.Format() function parse_date (s) return df:parse(s) end -- ISO 8601 -- specified as UTC plus/minus offset function parse_utc (s) local d = parse_date(s) return d:toUTC() end asserteq(parse_utc '2010-05-10 12:35:23Z', Date(2010,05,10,12,35,23)) asserteq(parse_utc '2008-10-03T14:30+02', Date(2008,10,03,12,30)) asserteq(parse_utc '2008-10-03T14:00-02:00',Date(2008,10,03,16,0)) ---- can't do anything before 1970, which is somewhat unfortunate.... --parse_date '20/03/59' asserteq(parse_date '15:30', Date {hour=15,min=30}) asserteq(parse_date '8.05pm', Date {hour=20,min=5}) asserteq(parse_date '28/10/02', Date {year=2002,month=10,day=28}) asserteq(parse_date ' 5 Feb 2012 ', Date {year=2012,month=2,day=5}) asserteq(parse_date '20 Jul ', Date {month=7,day=20}) asserteq(parse_date '05/04/02 15:30:43', Date{year=2002,month=4,day=5,hour=15,min=30,sec=43}) asserteq(parse_date 'march', Date {month=3}) asserteq(parse_date '2010-05-23T0130', Date{year=2010,month=5,day=23,hour=1,min=30}) asserteq(parse_date '2008-10-03T14:30:45', Date{year=2008,month=10,day=3,hour=14,min=30,sec=45}) -- allow for a comma after the month... asserteq(parse_date '18 July, 2013 12:00:00', Date{year=2013,month=07,day=18,hour=12,min=0,sec=0}) -- This ISO format must result in a UTC date local d = parse_date '2016-05-01T14:30:00Z' asserteq(d:year(),2016) asserteq(d:month(),5) asserteq(d:day(),1) asserteq(d:hour(),14) asserteq(d:min(),30) asserteq(d:sec(),0) function err (status,e) return e end assertmatch(err(parse_date('2005-10-40 01:30')),'40 is not between 1 and 31') assertmatch(err(parse_date('14.20pm')),'14 is not between 0 and 12') local d = parse_date '2007-08-10' -- '+' works like add, but can also work with intervals local nxt = d + {month=1} -- '-' is an alias for diff method asserteq(tostring(nxt - d), '1 month ') --- Can explicitly get UTC date; these of course refer to same time local now,utc = Date(), Date 'utc' asserteq(tostring(now - utc),'zero') Penlight-1.12.0/tests/test-dir.lua000066400000000000000000000135571416703176500167720ustar00rootroot00000000000000-- This test file expects to be ran from 'run.lua' in the root Penlight directory. local dir = require( "pl.dir" ) local file = require( "pl.file" ) local path = require( "pl.path" ) local asserteq = require( "pl.test" ).asserteq local lfs = require("lfs") asserteq(dir.fnmatch("foobar", "foo*bar"), true) asserteq(dir.fnmatch("afoobar", "foo*bar"), false) asserteq(dir.fnmatch("foobars", "foo*bar"), false) asserteq(dir.fnmatch("foonbar", "foo*bar"), true) asserteq(dir.fnmatch("foo'n'bar", "foo*bar"), true) asserteq(dir.fnmatch("foonbar", "foo?bar"), true) asserteq(dir.fnmatch("foo'n'bar", "foo?bar"), false) asserteq(dir.fnmatch("foo", "FOO"), path.is_windows) asserteq(dir.fnmatch("FOO", "foo"), path.is_windows) local filtered = dir.filter({"foobar", "afoobar", "foobars", "foonbar"}, "foo*bar") asserteq(filtered, {"foobar", "foonbar"}) local normpath = path.normpath local doc_files = dir.getfiles(normpath "docs/", "*.css") asserteq(doc_files, {normpath "docs/ldoc_fixed.css"}) local all_doc_files = dir.getallfiles(normpath "docs/", "*.css") asserteq(all_doc_files, {normpath "docs/ldoc_fixed.css"}) local test_samples = dir.getallfiles(normpath "tests/lua") table.sort(test_samples) asserteq(test_samples, { normpath "tests/lua/animal.lua", normpath "tests/lua/bar.lua", normpath "tests/lua/foo/args.lua", normpath "tests/lua/mod52.lua", normpath "tests/lua/mymod.lua" }) -- Test move files ----------------------------------------- -- Create a dummy file local fileName = path.tmpname() .. "Xx" file.write( fileName, string.rep( "poot ", 1000 ) ) local newFileName = path.tmpname() .. "Xx" local err, msg = dir.movefile( fileName, newFileName ) -- Make sure the move is successful assert( err, msg ) -- Check to make sure the original file is gone asserteq( path.exists( fileName ), false ) -- Check to make sure the new file is there asserteq( path.exists( newFileName ) , newFileName ) -- Test existence again, but explicitly check for correct casing local files = dir.getfiles(path.dirname(newFileName)) local found = false for i, filename in ipairs(files) do if filename == newFileName then found = true break end end assert(found, "file was not found in directory, check casing: " .. newFileName) -- Try to move the original file again (which should fail) local newFileName2 = path.tmpname() local err, msg = dir.movefile( fileName, newFileName2 ) asserteq( err, false ) -- Clean up file.delete( newFileName ) -- Test copy files ----------------------------------------- -- Create a dummy file local fileName = path.tmpname() file.write( fileName, string.rep( "poot ", 1000 ) ) local newFileName = path.tmpname() .. "xX" local err, msg = dir.copyfile( fileName, newFileName ) -- Make sure the move is successful assert( err, msg ) -- Check to make sure the new file is there asserteq( path.exists( newFileName ) , newFileName ) -- Test existence again, but explicitly check for correct casing local files = dir.getfiles(path.dirname(newFileName)) local found = false for i, filename in ipairs(files) do if filename == newFileName then found = true break end end assert(found, "file was not found in directory, check casing: " .. newFileName) -- Try to move a non-existant file (which should fail) local fileName2 = 'blub' local newFileName2 = 'snortsh' local err, msg = dir.copyfile( fileName2, newFileName2 ) asserteq( err, false ) -- Clean up the files file.delete( fileName ) file.delete( newFileName ) -- Test make directory ----------------------------------------- -- Create a dummy file local dirName = path.tmpname() .. "xX" local fullPath = dirName .. "/and/one/more" if path.is_windows then fullPath = fullPath:gsub("/", "\\") end local err, msg = dir.makepath(fullPath) -- Make sure the move is successful assert( err, msg ) -- Check to make sure the new file is there assert(path.isdir(dirName)) assert(path.isdir(fullPath)) -- Test existence again, but explicitly check for correct casing local files = dir.getdirectories(path.dirname(path.tmpname())) local found = false for i, filename in ipairs(files) do if filename == dirName then found = true break end end assert(found, "dir was not found in directory, check casing: " .. newFileName) -- Try to move a non-existant file (which should fail) local fileName2 = 'blub' local newFileName2 = 'snortsh' local err, msg = dir.copyfile( fileName2, newFileName2 ) asserteq( err, false ) -- Clean up the files file.delete( fileName ) file.delete( newFileName ) -- Test rmtree ----------------------------------------- do local dirName = path.tmpname() os.remove(dirName) assert(dir.makepath(dirName)) assert(file.write(path.normpath(dirName .. "/file_base.txt"), "hello world")) assert(dir.makepath(path.normpath(dirName .. "/sub1"))) assert(file.write(path.normpath(dirName .. "/sub1/file_sub1.txt"), "hello world")) assert(dir.makepath(path.normpath(dirName .. "/sub2"))) assert(file.write(path.normpath(dirName .. "/sub2/file_sub2.txt"), "hello world")) local linkTarget = path.tmpname() os.remove(linkTarget) assert(dir.makepath(linkTarget)) local linkFile = path.normpath(linkTarget .. "/file.txt") assert(file.write(linkFile, "hello world")) local linkSource = path.normpath(dirName .. "/link1") assert(lfs.link(linkTarget, linkSource, true)) -- test: rmtree will not follow symlinks local ok, err = dir.rmtree(linkSource) asserteq(ok, false) asserteq(err, "will not follow symlink") -- test: rmtree removes a tree without following symlinks in that tree local ok, err = dir.rmtree(dirName) asserteq(err, nil) asserteq(ok, true) asserteq(path.exists(dirName), false) -- tree is gone, including symlink assert(path.exists(linkFile), "expected linked-to file to still exist") -- symlink target file is still there -- cleanup assert(dir.rmtree(linkTarget)) end -- have NO idea why forcing the return code is necessary here (Windows 7 64-bit) os.exit(0) Penlight-1.12.0/tests/test-func.lua000066400000000000000000000055601416703176500171420ustar00rootroot00000000000000local utils = require 'pl.utils' local List = require 'pl.List' local tablex = require 'pl.tablex' asserteq = require('pl.test').asserteq utils.import('pl.func') -- _DEBUG = true function pprint (t) print(pretty.write(t)) end function test (e) local v = {} print('test',collect_values(e,v)) if #v > 0 then pprint(v) end local rep = repr(e) print(rep) end function teste (e,rs,ve) local v = {} collect_values(e,v) if #v > 0 then asserteq(v,ve,nil,1) end local rep = repr(e) asserteq(rep,rs, nil, 1) end teste(_1+_2('hello'),'_1 + _2(_C1)',{"hello"}) teste(_1:method(),'_1[_C1](_1)',{"method"}) teste(Not(_1),'not _1') asserteq(instantiate(_1+_2)(10,20),30) asserteq(instantiate(_1+20)(10),30) asserteq(instantiate(Or(Not(_1),_2))(true,true),true) teste(_1() + _2() + _3(),'_1() + _2() + _3()',30) asserteq(I(_1+_2)(10,20),30) teste(_1() - -_2() % _3(), '_1() - - _2() % _3()') teste((_1() - -_2()) % _3(), '(_1() - - _2()) % _3()') teste(_1() - _2() + _3(), '_1() - _2() + _3()') teste(_1() - (_2() + _3()), '_1() - (_2() + _3())') teste((_1() - _2()) + _3(), '_1() - _2() + _3()') teste(_1() .. _2() .. _3(), '_1() .. _2() .. _3()') teste(_1() .. (_2() .. _3()), '_1() .. _2() .. _3()') teste((_1() .. _2()) .. _3(), '(_1() .. _2()) .. _3()') teste(_1() ^ _2() ^ _3(), '_1() ^ _2() ^ _3()') teste(_1() ^ (_2() ^ _3()), '_1() ^ _2() ^ _3()') teste((_1() ^ _2()) ^ _3(), '(_1() ^ _2()) ^ _3()') teste(-_1() * _2(), '- _1() * _2()') teste(-(_1() * _2()), '- (_1() * _2())') teste((-_1()) * _2(), '- _1() * _2()') teste(-_1() ^ _2(), '- _1() ^ _2()') teste(-(_1() ^ _2()), '- _1() ^ _2()') teste((-_1()) ^ _2(), '(- _1()) ^ _2()') asserteq(instantiate(_1+_2)(10,20),30) ls = List {1,2,3,4} res = ls:map(10*_1 - 1) asserteq(res,List {9,19,29,39}) -- note that relational operators can't be overloaded for _different_ types ls = List {10,20,30,40} asserteq(ls:filter(Gt(_1,20)),List {30,40}) local map,map2 = tablex.map,tablex.map2 --~ test(Len(_1)) -- methods can be applied to all items in a table with map asserteq (map(_1:sub(1,2),{'one','four'}),{'on','fo'}) --~ -- or you can do this using List:map asserteq( List({'one','four'}):map(_1:sub(1,2)), List{'on','fo'}) --~ -- note that Len can't be represented generally by #, since this can only be overriden by userdata asserteq( map(Len(_1),{'one','four'}), {3,4} ) --~ -- simularly, 'and' and 'or' are not really operators in Lua, so we need a function notation for them asserteq (map2(Or(_1,_2),{false,'b'},{'.lua',false}),{'.lua','b'}) --~ -- binary operators: + - * / % ^ .. asserteq (map2(_1.._2,{'a','b'},{'.lua','.c'}),{'a.lua','b.c'}) t1 = {alice=23,fred=34} t2 = {bob=25,fred=34} intersection = bind(tablex.merge,_1,_2,false) asserteq(intersection(t1,t2),{fred=34}) union = bind(tablex.merge,_1,_2,true) asserteq(union(t1,t2),{bob=25,fred=34,alice=23}) asserteq(repr(_1+_2),"_1 + _2") Penlight-1.12.0/tests/test-import_into.lua000066400000000000000000000017121416703176500205450ustar00rootroot00000000000000local test = require 'pl.test' local utils = require 'pl.utils' require 'pl.app'.require_here 'lua' if not utils.lua51 then --- look at lua/mod52.lua local m = require 'mod52' test.asserteq(m.answer(),'{"10","20","30"}') assert(m.utils) -- !! implementation is leaky! -- that's a bugger. However, if 'pl.import_into' is passed true, -- then the returned module will _only_ contain the newly defined functions -- So reload after setting the global STRICT package.loaded.mod52 = nil STRICT = true m = require 'mod52' assert (m.answer) -- as before assert (not m.utils) -- cool! No underwear showing end local pl = require 'pl.import_into' () assert(pl.utils) assert(pl.tablex) assert(pl.data) assert(not _G.utils) assert(not _G.tablex) assert(not _G.data) require 'pl.import_into'(_G) assert(_G.utils) assert(_G.tablex) assert(_G.data) require 'pl.import_into'(_G) assert(_G.utils) assert(_G.tablex) assert(_G.data) Penlight-1.12.0/tests/test-lapp.lua000066400000000000000000000132701416703176500171400ustar00rootroot00000000000000 local test = require 'pl.test' local lapp = require 'pl.lapp' local utils = require 'pl.utils' local tablex = require 'pl.tablex' local path = require 'pl.path' local normpath = path.normpath local k = 1 function check (spec,args,match) local args = lapp(spec,args) for k,v in pairs(args) do if type(v) == 'userdata' then args[k]:close(); args[k] = '' end end test.asserteq(args,match,nil,1) end -- force Lapp to throw an error, rather than just calling os.exit() lapp.show_usage_error = 'throw' function check_error(spec,args,msg) arg = args local ok,err = pcall(lapp,spec) test.assertmatch(err,msg) end local parmtest = [[ Testing 'array' parameter handling -o,--output... (string) -v... ]] check (parmtest,{'-o','one'},{output={'one'},v={false}}) check (parmtest,{'-o','one','-v'},{output={'one'},v={true}}) check (parmtest,{'-o','one','-vv'},{output={'one'},v={true,true}}) check (parmtest,{'-o','one','-o','two'},{output={'one','two'},v={false}}) local simple = [[ Various flags and option types -p A simple optional flag, defaults to false -q,--quiet A simple flag with long name -o (string) A required option with argument (default stdin) Optional input file parameter... ]] check(simple, {'-o','in'}, {quiet=false,p=false,o='in',input=''}) ---- value of flag may be separated by '=' or ':' check(simple, {'-o=in'}, {quiet=false,p=false,o='in',input=''}) check(simple, {'-o:in'}, {quiet=false,p=false,o='in',input=''}) -- Check lapp.callback. local calls = {} function lapp.callback(param, arg) table.insert(calls, {param, arg}) end check(simple, {'-o','help','-q',normpath 'tests/test-lapp.lua'}, {quiet=true,p=false,o='help',input='',input_name=normpath 'tests/test-lapp.lua'}) test.asserteq(calls, { {'o', 'help'}, {'quiet', '-q'}, {'input', normpath 'tests/test-lapp.lua'} }) lapp.callback = nil local longs = [[ --open (string) ]] check(longs,{'--open','folder'},{open='folder'}) local long_file = [[ --open (default stdin) ]] check(long_file,{'--open',normpath 'tests/test-lapp.lua'},{open='',open_name=normpath 'tests/test-lapp.lua'}) local extras1 = [[ (string) A bunch of files ]] check(extras1,{'one','two'},{files={'one','two'}}) -- any extra parameters go into the array part of the result local extras2 = [[ (string) A file ]] check(extras2,{'one','two'},{file='one','two'}) local extended = [[ --foo (string default 1) -s,--speed (slow|medium|fast default medium) -n (1..10 default 1) -p print -v verbose ]] check(extended,{},{foo='1',speed='medium',n=1,p=false,v=false}) check(extended,{'-pv'},{foo='1',speed='medium',n=1,p=true,v=true}) check(extended,{'--foo','2','-s','fast'},{foo='2',speed='fast',n=1,p=false,v=false}) check(extended,{'--foo=2','-s=fast','-n2'},{foo='2',speed='fast',n=2,p=false,v=false}) check_error(extended,{'--speed','massive'},"value 'massive' not in slow|medium|fast") check_error(extended,{'-n','x'},"unable to convert to number: x") check_error(extended,{'-n','12'},"n out of range") local with_advanced_enum = [[ -s (test1|test2()|%a) -c (1-2|2-3|cool[]) ]] check(with_advanced_enum,{"-s", "test2()", "-c", "1-2"},{s='test2()',c='1-2'}) check(with_advanced_enum,{"-s", "test2()", "-c", "2-3"},{s='test2()',c='2-3'}) check(with_advanced_enum,{"-s", "%a", "-c", "2-3"},{s='%a',c='2-3'}) local with_dashes = [[ --first-dash dash --second-dash dash also ]] check(with_dashes,{'--first-dash'},{first_dash=true,second_dash=false}) -- optional parameters don't have to be set local optional = [[ -p (optional string) ]] check(optional,{'-p', 'test'},{p='test'}) check(optional,{},{}) -- boolean flags may have a true default... local false_flag = [[ -g group results -f (default true) force result ]] check (false_flag,{},{f=true,g=false}) check (false_flag,{'-g','-f'},{f=false,g=true}) -- '--' indicates end of parameter parsing check (false_flag,{'-g','--'},{f=true,g=true}) check (false_flag,{'-g','--','-a','frodo'},{f=true,g=true; '-a','frodo'}) local addtype = [[ -l (intlist) List of items ]] -- defining a custom type lapp.add_type('intlist', function(x) return tablex.imap(tonumber, utils.split(x, '%s*,%s*')) end, function(x) for _,v in ipairs(x) do lapp.assert(math.ceil(v) == v,'not an integer!') end end) check(addtype,{'-l', '1,2,3'},{l={1,2,3}}) check_error(addtype,{'-l', '1.5,2,3'},"not an integer!") -- short flags may be immediately followed by their value -- (previously only true for numerical values) local short_args = [[ -n (default 10) -I,--include (string) ]] check(short_args,{'-Ifrodo','-n5'},{include='frodo',n=5}) check(short_args,{'-I/usr/local/lua/5.1'},{include='/usr/local/lua/5.1',n=10}) -- ok, introducing _slack_ mode ;) -- 'short' flags may have multiple characters! (this is otherwise an error) -- Note that in _any case_ flags may contain hyphens, but these are turned -- into underscores for convenience. lapp.slack = true local spec = [[ Does some calculations -vs,--video-set (string) Use the German road sign dataset -w,--width (default 256) Width of the video -h,--height (default 144) Height of the video -t,--time (default 10) Seconds of video to process -sk,--seek (default 0) Seek number of seconds -dbg Debug! ]] test.asserteq(lapp(spec,{'-vs',200,'-sk',1}),{ video_set = 200, time = 10, height = 144, seek = 1, dbg = false, width = 256 }) Penlight-1.12.0/tests/test-lexer.lua000066400000000000000000000070311416703176500173210ustar00rootroot00000000000000local asserteq = require('pl.test').asserteq local lexer = require 'pl.lexer' local seq = require 'pl.seq' local List = require('pl.List') local open = require('pl.stringio').open local copy2 = seq.copy2 local function test_scan(str, filter, options, expected_tokens, lang) local matches if lang then matches, filter = filter, options else lang = 'scan' end asserteq(copy2(lexer[lang](str, matches, filter, options)), expected_tokens) if lang == 'scan' then asserteq(copy2(lexer[lang](open(str), matches, filter, options)), expected_tokens) end end local s = '20 = hello' test_scan(s, {space=false}, {number=false}, { {'number', '20'}, {'space', ' '}, {'=', '='}, {'space', ' '}, {'iden', 'hello'} }) test_scan(s, {space=true}, {number=true}, { {'number', 20}, {'=', '='}, {'iden', 'hello'} }) s = [[ 'help' "help" "dolly you're fine" "a \"quote\" here"]] test_scan(s, nil, nil, { {'string', 'help'}, {'string', 'help'}, {'string', "dolly you're fine"}, {'string', 'a \\\"quote\\\" here'} -- Escapes are preserved literally. }) test_scan([[\abc\]], nil, nil, { {'\\', '\\'}, {'iden', 'abc'}, {'\\', '\\'} }) test_scan([["" ""]], nil, nil, { {'string', ''}, {'string', ''} }) test_scan([["abc" "def\\"]], nil, nil, { {'string', 'abc'}, {'string', 'def\\\\'} }) test_scan([["abc\\" "def"]], nil, nil, { {'string', 'abc\\\\'}, {'string', 'def'} }) test_scan([["abc\\\" "]], nil, nil, { {'string', 'abc\\\\\\" '} }) local function test_roundtrip(str) test_scan(str, {}, {string=false}, {{'string', str}}) end test_roundtrip [["hello\\"]] test_roundtrip [["hello\"dolly"]] test_roundtrip [['hello\'dolly']] test_roundtrip [['']] test_roundtrip [[""]] test_scan('test(20 and a > b)', nil, nil, { {'iden', 'test'}, {'(', '('}, {'number', 20}, {'keyword', 'and'}, {'iden', 'a'}, {'>', '>'}, {'iden', 'b'}, {')', ')'} }, 'lua') test_scan('10+2.3', nil, nil, { {'number', 10}, {'+', '+'}, {'number', 2.3} }, 'lua') local txt = [==[ -- comment --[[ block comment ]][[ hello dammit ]][[hello]] ]==] test_scan(txt, {}, nil, { {'comment', '-- comment\n'}, {'comment', '--[[\nblock\ncomment\n]]'}, {'string', 'hello dammit\n'}, {'string', 'hello'}, {'space', '\n'} }, 'lua') local lines = [[ for k,v in pairs(t) do if type(k) == 'number' then print(v) -- array-like case else print(k,v) end -- if end ]] local ls = List() for tp,val in lexer.lua(lines,{space=true,comments=true}) do assert(tp ~= 'space' and tp ~= 'comment') if tp == 'keyword' then ls:append(val) end end asserteq(ls,List{'for','in','do','if','then','else','end','end'}) txt = [[ // comment /* a long set of words */ // more ]] test_scan(txt, {}, nil, { {'comment', '// comment\n'}, {'comment', '/* a long\nset of words */'}, {'space', ' '}, {'comment', '// more\n'} }, 'cpp') test_scan([['' "" " \\" '\'' "'"]], nil, nil, { {'char', ''}, -- Char literals with no or more than one characters are not a lexing error. {'string', ''}, {'string', ' \\\\'}, {'char', "\\'"}, {'string', "'"} }, 'cpp') local iter = lexer.lua([[ foo bar ]]) asserteq(lexer.lineno(iter), 0) iter() asserteq(lexer.lineno(iter), 1) asserteq(lexer.lineno(iter), 1) iter() asserteq(lexer.lineno(iter), 2) iter() asserteq(lexer.lineno(iter), 3) iter() iter() asserteq(lexer.lineno(iter), 3) do -- numbers without leading zero; ".123" local s = 'hello = +.234' test_scan(s, {space=true}, {number=true}, { {'iden', 'hello'}, {'=', '='}, {'number', .234} }) end Penlight-1.12.0/tests/test-list.lua000066400000000000000000000034021416703176500171530ustar00rootroot00000000000000local List = require 'pl.List' local class = require 'pl.class' local test = require 'pl.test' local asserteq, T = test.asserteq, test.tuple -- note that a _plain table_ is made directly into a list local t = {10,20,30} local ls = List(t) asserteq(t,ls) asserteq(List({}):reverse(), {}) asserteq(List({1}):reverse(), {1}) asserteq(List({1,2}):reverse(), {2,1}) asserteq(List({1,2,3}):reverse(), {3,2,1}) asserteq(List({1,2,3,4}):reverse(), {4,3,2,1}) -- you may derive classes from pl.List, and the result is covariant. -- That is, slice() etc will return a list of the derived type, not List. local NA = class(List) local function mapm(a1,op,a2) local M = type(a2)=='table' and List.map2 or List.map return M(a1,op,a2) end --- elementwise arithmetric operations function NA.__unm(a) return a:map '|X|-X' end function NA.__pow(a,s) return a:map '|X,Y|X^Y' end function NA.__add(a1,a2) return mapm(a1,'|X,Y|X+Y',a2) end function NA.__sub(a1,a2) return mapm(a1,'|X,Y|X-Y',a2) end function NA.__div(a1,a2) return mapm(a1,'|X,Y|X/Y',a2) end function NA.__mul(a1,a2) return mapm(a2,'|X,Y|X*Y',a1) end function NA:minmax () local min,max = math.huge,-math.huge for i = 1,#self do local val = self[i] if val > max then max = val end if val < min then min = val end end return min,max end function NA:sum () local res = 0 for i = 1,#self do res = res + self[i] end return res end function NA:normalize () return self:transform('|X,Y|X/Y',self:sum()) end n1 = NA{10,20,30} n2 = NA{1,2,3} ns = n1 + 2*n2 asserteq(List:class_of(ns),true) asserteq(NA:class_of(ns),true) asserteq(ns:is_a(NA),true) asserteq(ns,{12,24,36}) min,max = ns:slice(1,2):minmax() asserteq(T(min,max),T(12,24)) asserteq(n1:normalize():sum(),1,1e-8) Penlight-1.12.0/tests/test-list2.lua000066400000000000000000000033311416703176500172360ustar00rootroot00000000000000local List = require 'pl.List' local asserteq = require 'pl.test' . asserteq local s = List{1,2,3,4,5} -- test using: lua pylist.lua local lst = List() lst:append(20) lst:extend{30,40,50} lst:put(10) asserteq (lst,List{10,20,30,40,50}) asserteq (lst:len(),5) lst:insert(3,11) lst:remove_value(40) asserteq (lst,List{10,20,11,30,50}) asserteq (lst:contains(11),true) asserteq (lst:contains(40),false) local _ = lst:pop() asserteq( lst:index(30),4 ) asserteq( lst:count(10),1 ) lst:sort() lst:reverse() asserteq (lst , List{30,20,11,10}) asserteq (lst[#lst] , 10) asserteq (lst[#lst-2] , 20) asserteq (tostring(lst) , '{30,20,11,10}') lst = List {10,20,30,40,50} asserteq (lst:slice(2),{20,30,40,50}) asserteq (lst:slice(-2),{40,50}) asserteq (lst:slice(nil,3),{10,20,30}) asserteq (lst:slice(2,4),{20,30,40}) asserteq (lst:slice(-4,-2),{20,30,40}) lst = List.range(0,9) local seq = List{0,1,2,3,4,5,6,7,8,9} asserteq(List.range(4),{1,2,3,4}) asserteq(List.range(0,8,2),{0,2,4,6,8}) asserteq(List.range(0,1,0.2),{0,0.2,0.4,0.6,0.8,1},1e-9) asserteq(lst, seq) asserteq(lst:reduce '+', 45) local part = seq:partition(function(v) return v % 2 end) asserteq (part[0], List{0,2,4,6,8}) asserteq (part[1], List{1,3,5,7,9}) asserteq (List('abcd'),List{'a','b','c','d'}) local caps = List() List('abcd'):foreach(function(v) caps:append(v:upper()) end) asserteq (caps,List{'A','B','C','D'}) local ls = List{10,20,30,40} ls:slice_assign(2,3,{21,31}) asserteq (ls , List{10,21,31,40}) asserteq (ls:remove(2), List{10,31,40}) asserteq (ls:clear(), List{}) asserteq (ls:len(), 0) s = 'here the dog is just a dog' assert (List.split(s) == List{'here', 'the', 'dog', 'is', 'just', 'a', 'dog'}) assert (List.split('foo;bar;baz', ';') == List{'foo', 'bar', 'baz'}) Penlight-1.12.0/tests/test-map.lua000066400000000000000000000043321416703176500167600ustar00rootroot00000000000000-- testing Map functionality local test = require 'pl.test' local Map = require 'pl.Map' local tablex = require 'pl.tablex' local Set = require 'pl.Set' local utils = require 'pl.utils' local asserteq = test.asserteq local cmp = tablex.compare_no_order -- construction, plain local m = Map{alpha=1,beta=2,gamma=3} assert(cmp( m:values(), {1, 2, 3} )) assert(cmp( m:keys(), {'alpha', 'beta', 'gamma'} )) asserteq( m:items(), { {'alpha', 1}, {'beta', 2}, {'gamma', 3}, } ) asserteq (m:getvalues {'alpha','gamma'}, {1,3}) -- construction, from a set local s = Set{'red','orange','green','blue'} m = Map(s) asserteq( m:items(), { {'blue', true}, {'green', true}, {'orange', true}, {'red', true}, } ) -- iter() m = Map{alpha=1,beta=2,gamma=3} local t = {alpha=1,beta=2,gamma=3} for k,v in m:iter() do asserteq(v, t[k]) t[k] = nil end assert(next(t) == nil, "expected the table to be empty by now") -- setdefault() m = Map{alpha=1,beta=2,gamma=3} local v = m:setdefault("charlie", 4) asserteq(v, 4) v = m:setdefault("alpha", 10) asserteq(v, 1) asserteq( m:items(), { {'alpha', 1}, {'beta', 2}, {'charlie', 4}, {'gamma', 3}, } ) v = m:set("alpha", false) v = m:setdefault("alpha", true) -- falsy value should not be altered asserteq(false, m:get("alpha")) -- len() m = Map{alpha=1,beta=2,gamma=3} asserteq(3, m:len()) m = Map{} asserteq(0, m:len()) m:set("charlie", 4) asserteq(1, m:len()) -- set() & get() m = Map{} m:set("charlie", 4) asserteq(4, m:get("charlie")) m:set("charlie", 5) asserteq(5, m:get("charlie")) m:set("charlie", nil) asserteq(nil, m:get("charlie")) -- getvalues() m = Map{alpha=1,beta=2,gamma=3} local x = m:getvalues{"gamma", "beta"} asserteq({3, 2}, x) -- __eq() -- equality local m1 = Map{alpha=1,beta=2,gamma=3} local m2 = Map{alpha=1,beta=2,gamma=3} assert(m1 == m2) m1 = Map() m2 = Map() assert(m1 == m2) -- __tostring() m = Map() asserteq("{}", tostring(m)) m = Map{alpha=1} asserteq("{alpha=1}", tostring(m)) m = Map{alpha=1,beta=2} assert(({ -- test 2 versions, since we cannot rely on order ["{alpha=1,beta=2}"] = true, ["{beta=2,alpha=1}"] = true, })[tostring(m)]) Penlight-1.12.0/tests/test-orderedmap.lua000066400000000000000000000046141416703176500203300ustar00rootroot00000000000000local List = require 'pl.List' local asserteq = require 'pl.test' . asserteq local asserteq2 = require 'pl.test' . asserteq2 local OrderedMap = require 'pl.OrderedMap' m = OrderedMap() m:set('one',1) m:set('two',2) m:set('three',3) asserteq(m:values(),List{1,2,3}) -- usually exercized like this: --for k,v in m:iter() do print(k,v) end local fn = m:iter() asserteq2 ('one',1,fn()) asserteq2 ('two',2,fn()) asserteq2 ('three',3,fn()) -- Keys overriding methods can be used. m:set('set', 4) asserteq(m:values(),List{1,2,3,4}) local o1 = OrderedMap {{z=2},{beta=1},{name='fred'}} asserteq(tostring(o1),'{z=2,beta=1,name="fred"}') -- order of keys is not preserved here! local o2 = OrderedMap {z=4,beta=1.1,name='alice',extra='dolly'} o1:update(o2) asserteq(tostring(o1),'{z=4,beta=1.1,name="alice",extra="dolly"}') o1:set('beta',nil) asserteq(o1,OrderedMap{{z=4},{name='alice'},{extra='dolly'}}) local o3 = OrderedMap() o3:set('dog',10) o3:set('cat',20) o3:set('mouse',30) asserteq(o3:keys(),{'dog','cat','mouse'}) o3:set('dog',nil) asserteq(o3:keys(),{'cat','mouse'}) -- Vadim found a problem when clearing a key which did not exist already. -- The keys list would then contain the key, although the map would not o3:set('lizard',nil) asserteq(o3:keys(),{'cat','mouse'}) asserteq(o3:values(), {20,30}) asserteq(tostring(o3),'{cat=20,mouse=30}') -- copy constructor local o4 = OrderedMap(o3) asserteq(o4,o3) -- constructor throws an error if the argument is bad -- (errors same as OrderedMap:update) asserteq(false,pcall(function() m = OrderedMap('string') end)) ---- changing order of key/value pairs ---- o3 = OrderedMap{{cat=20},{mouse=30}} o3:insert(1,'bird',5) -- adds key/value before specified position o3:insert(1,'mouse') -- moves key keeping old value asserteq(o3:keys(),{'mouse','bird','cat'}) asserteq(tostring(o3),'{mouse=30,bird=5,cat=20}') o3:insert(2,'cat',21) -- moves key and sets new value asserteq(tostring(o3),'{mouse=30,cat=21,bird=5}') -- if you don't specify a value for an unknown key, nothing happens to the map o3:insert(3,'alligator') asserteq(tostring(o3),'{mouse=30,cat=21,bird=5}') ---- short-cut notation local o5 = OrderedMap() o5.alpha = 1 o5.beta = 2 o5.gamma = 3 asserteq(o5,OrderedMap{{alpha=1},{beta=2},{gamma=3}}) o5.alpha = 10 o5.beta = 20 o5.gamma = 30 o5.delta = 40 o5.checked = false asserteq(o5,OrderedMap{{alpha=10},{beta=20},{gamma=30},{delta=40},{checked=false}}) Penlight-1.12.0/tests/test-path.lua000066400000000000000000000135621416703176500171440ustar00rootroot00000000000000local path = require 'pl.path' asserteq = require 'pl.test'.asserteq function quote(s) return '"'..s..'"' end function print2(s1,s2) print(quote(s1),quote(s2)) end function slash (p) return (p:gsub('\\','/')) end -- path.currentdir do local cp = path.currentdir() path.chdir("docs") asserteq(path.currentdir(), cp .. path.sep .. "docs") path.chdir("..") asserteq(path.currentdir(), cp) end -- path.isdir asserteq( path.isdir( "docs" ), true ) asserteq( path.isdir( "docs/index.html" ), false ) -- path.isfile asserteq( path.isfile( "docs" ), false ) asserteq( path.isfile( "docs/index.html" ), true ) -- path.exists asserteq( path.exists( "docs"), "docs") asserteq( path.exists( "docs/index.html"), "docs/index.html") do -- path.splitpath & path.splitext function testpath(pth,p1,p2,p3) local dir,rest = path.splitpath(pth) local name,ext = path.splitext(rest) asserteq(dir,p1) asserteq(name,p2) asserteq(ext,p3) end testpath ([[/bonzo/dog_stuff/cat.txt]],[[/bonzo/dog_stuff]],'cat','.txt') testpath ([[/bonzo/dog/cat/fred.stuff]],'/bonzo/dog/cat','fred','.stuff') testpath ([[../../alice/jones]],'../../alice','jones','') testpath ([[alice]],'','alice','') testpath ([[/path-to/dog/]],[[/path-to/dog]],'','') asserteq({path.splitpath("some/dir/myfile.txt")}, {"some/dir", "myfile.txt"}) asserteq({path.splitpath("some/dir/")}, {"some/dir", ""}) asserteq({path.splitpath("some_dir")}, {"", "some_dir"}) asserteq({path.splitext("/bonzo/dog_stuff/cat.txt")}, {"/bonzo/dog_stuff/cat", ".txt"}) asserteq({path.splitext("cat.txt")}, {"cat", ".txt"}) asserteq({path.splitext("cat")}, {"cat", ""}) asserteq({path.splitext(".txt")}, {"", ".txt"}) asserteq({path.splitext("")}, {"", ""}) end -- TODO: path.abspath -- TODO: path.dirname -- TODO: path.basename -- TODO: path.extension do -- path.isabs asserteq(path.isabs("/hello/path"), true) asserteq(path.isabs("hello/path"), false) asserteq(path.isabs("./hello/path"), false) asserteq(path.isabs("../hello/path"), false) if path.is_windows then asserteq(path.isabs("c:/"), true) asserteq(path.isabs("c:/hello/path"), true) asserteq(path.isabs("c:"), false) asserteq(path.isabs("c:hello/path"), false) asserteq(path.isabs("c:./hello/path"), false) asserteq(path.isabs("c:../hello/path"), false) end end do -- path.join assert(path.join("somepath",".") == "somepath"..path.sep..".") assert(path.join(".","readme.txt") == "."..path.sep.."readme.txt") assert(path.join("/a_dir", "abs_path/") == "/a_dir"..path.sep.."abs_path/") assert(path.join("a_dir", "/abs_path/") == "/abs_path/") assert(path.join("a_dir", "/abs_path/", "/abs_path2/") == "/abs_path2/") assert(path.join("a_dir", "/abs_path/", "not_abs_path2/") == "/abs_path/not_abs_path2/") assert(path.join("a_dir", "/abs_path/", "not_abs_path2/", "/abs_path3/", "not_abs_path4/") == "/abs_path3/not_abs_path4/") assert(path.join("first","second","third") == "first"..path.sep.."second"..path.sep.."third") assert(path.join("first","second","") == "first"..path.sep.."second"..path.sep) assert(path.join("first","","third") == "first"..path.sep.."third") assert(path.join("","second","third") == "second"..path.sep.."third") assert(path.join("","") == "") end do -- path.normcase if path.iswindows then asserteq('c:\\hello\\world', 'c:\\hello\\world') asserteq('C:\\Hello\\wORLD', 'c:\\hello\\world') asserteq('c:/hello/world', 'c:\\hello\\world') else asserteq('/Hello/wORLD', '/Hello/wORLD') end end do -- path.normpath local norm = path.normpath local p = norm '/a/b' asserteq(norm '/a/fred/../b',p) asserteq(norm '/a//b',p) function testnorm(p1,p2) asserteq(norm(p1):gsub('\\','/'), p2) end testnorm('a/b/..','a') testnorm('a/b/../..','.') testnorm('a/b/../c/../../d','d') testnorm('a/.','a') testnorm('a/./','a') testnorm('a/b/.././..','.') testnorm('../../a/b','../../a/b') testnorm('../../a/b/../../','../..') testnorm('../../a/b/../c','../../a/c') testnorm('./../../a/b/../c','../../a/c') testnorm('a/..b', 'a/..b') testnorm('./a', 'a') testnorm('a/.', 'a') testnorm('a/', 'a') testnorm('/a', '/a') testnorm('', ".") if path.is_windows then testnorm('C://a', 'C:/a') testnorm('C:/../a', 'C:/../a') asserteq(norm [[\a\.\b]], p) -- UNC paths asserteq(norm [[\\bonzo\..\dog]], [[\\dog]]) asserteq(norm [[\\?\c:\bonzo\dog\.\]], [[\\?\c:\bonzo\dog]]) else testnorm('//a', '//a') testnorm('///a', '/a') end asserteq(norm '1/2/../3/4/../5',norm '1/3/5') asserteq(norm '1/hello/../3/hello/../HELLO',norm '1/3/HELLO') end do -- path.relpath local testpath = '/a/B/c' function try (p,r) asserteq(slash(path.relpath(p,testpath)),r) end try('/a/B/c/one.lua','one.lua') try('/a/B/c/bonZO/two.lua','bonZO/two.lua') try('/a/B/three.lua','../three.lua') try('/a/four.lua','../../four.lua') try('one.lua','one.lua') try('../two.lua','../two.lua') end -- TODO: path.expanduser -- TODO: path.tmpname do -- path.common_prefix asserteq(slash(path.common_prefix("../anything","../anything/goes")),"../anything") asserteq(slash(path.common_prefix("../anything/goes","../anything")),"../anything") asserteq(slash(path.common_prefix("../anything/goes","../anything/goes")),"../anything") asserteq(slash(path.common_prefix("../anything/","../anything/")),"../anything") asserteq(slash(path.common_prefix("../anything","../anything")),"..") asserteq(slash(path.common_prefix("/hello/world","/hello/world/filename.doc")),"/hello/world") asserteq(slash(path.common_prefix("/hello/filename.doc","/hello/filename.doc")),"/hello") if path.is_windows then asserteq(path.common_prefix("c:\\hey\\there","c:\\hey"),"c:\\hey") asserteq(path.common_prefix("c:/HEy/there","c:/hEy"),"c:\\hEy") -- normalized separators, original casing end end -- TODO: path.package_path Penlight-1.12.0/tests/test-pretty.lua000066400000000000000000000052621416703176500175350ustar00rootroot00000000000000local pretty = require 'pl.pretty' local utils = require 'pl.utils' local test = require 'pl.test' local asserteq, assertmatch = test.asserteq, test.assertmatch t1 = { 'one','two','three',{1,2,3}, alpha=1,beta=2,gamma=3,['&']=true,[0]=false, _fred = {true,true}, s = [[ hello dolly you're so fine ]] } s = pretty.write(t1) --,' ',true) t2,err = pretty.read(s) if err then return print(err) end asserteq(t1,t2) res,err = pretty.read [[ { ['function'] = true, ['do'] = true, } ]] assert(res) res,err = pretty.read [[ { ['function'] = true, ['do'] = "no function here...", } ]] assert(res) res,err = pretty.read [[ { ['function'] = true, ['do'] = function() return end } ]] assertmatch(err,'cannot have functions in table definition') res,err = pretty.load([[ -- comments are ok a = 2 bonzo = 'dog' t = {1,2,3} ]]) asserteq(res,{a=2,bonzo='dog',t={1,2,3}}) --- another potential problem is string functions called implicitly as methods-- res,err = pretty.read [[ {s = ('woo'):gsub('w','wwwwww'):gsub('w','wwwwww')} ]] assertmatch(err,(_VERSION ~= "Lua 5.2") and 'attempt to index a string value' or "attempt to index constant 'woo'") ---- pretty.load has a _paranoid_ option res,err = pretty.load([[ k = 0 for i = 1,1e12 do k = k + 1 end ]],{},true) assertmatch(err,'looping not allowed') -- Check to make sure that no spaces exist when write is told not to local tbl = { "a", 2, "c", false, 23, 453, "poot", 34 } asserteq( pretty.write( tbl, "" ), [[{"a",2,"c",false,23,453,"poot",34}]] ) -- Check that write correctly prevents cycles local t1,t2 = {},{} t1[1] = t1 asserteq( pretty.write(t1,""), [[{}]] ) t1[1],t1[2],t2[1] = 42,t2,t1 asserteq( pretty.write(t1,""), [[{42,{}}]] ) -- Check false positives in write's cycles prevention t2 = {} t1[1],t1[2] = t2,t2 asserteq( pretty.write(t1,""), [[{{},{}}]] ) -- Check that write correctly print table with non number or string as keys t1 = { [true] = "boolean", a = "a", b = "b", [1] = 1, [0] = 0 } asserteq( pretty.write(t1,""), [[{1,["true"]="boolean",a="a",b="b",[0]=0}]] ) -- Check number formatting asserteq(pretty.write({1/0, -1/0, 0/0, 1, 1/2}, ""), "{Inf,-Inf,NaN,1,0.5}") if _VERSION == "Lua 5.3" then asserteq(pretty.write({1.0}, ""), "{1.0}") else asserteq(pretty.write({1.0}, ""), "{1}") end do -- issue #203, item 3 local t = {}; t[t] = 1 pretty.write(t) -- should not crash end -- pretty.write fails if an __index metatable raises an error #257 -- only applies to 5.3+ where iterators respect metamethods do local t = setmetatable({},{ __index = function(self, key) error("oops... couldn't find " .. tostring(key)) end }) asserteq(pretty.write(t), "{\n}") end Penlight-1.12.0/tests/test-seq.lua000066400000000000000000000100721416703176500167710ustar00rootroot00000000000000local input = require 'pl.input' local seq = require 'pl.seq' local asserteq = require('pl.test').asserteq local utils = require 'pl.utils' local stringio = require 'pl.stringio' local unpack = utils.unpack local L = utils.string_lambda local S = seq.list local C = seq.copy local C2 = seq.copy2 asserteq (seq.sum(input.numbers '10 20 30 40 50'),150) local x,y = unpack(C(input.numbers('10 20'))) assert (x == 10 and y == 20) local test = {{1,10},{2,20},{3,30}} asserteq(C2(ipairs{10,20,30}),test) local res = C2(input.fields({1,2},',','1,10\n2,20\n3,30\n')) asserteq(res,test) asserteq( seq.copy(seq.filter(seq.list{10,20,5,15},seq.greater_than(10))), {20,15} ) asserteq( seq.copy(seq.filter(seq.list{10,20,5,15},seq.less_than(15))), {10,5} ) asserteq( #C(seq.filter(seq.list{10,20,5,10,15},seq.equal_to(10))), 2 ) asserteq( #seq{'green','yellow','red','blue','red'}:filter(seq.equal_to'red'):copy(), 2 ) asserteq( seq{'apple','orange','pineapple'}:filter(seq.matching'apple'):copy(), {'apple','pineapple'} ) asserteq( C(seq.sort(seq.keys{[11] = true, [17]= true, [23] = true})), {11,17,23} ) asserteq( C(seq.range(2,5)), {2,3,4,5} ) asserteq(seq.reduce('-',{1,2,3,4,5}),-13) asserteq(seq.count(S{10,20,30,40},L'|x| x > 20'), 2) asserteq(C2(seq.zip({1,2,3},{10,20,30})),test) asserteq(C(seq.splice({10,20},{30,40})),{10,20,30,40}) asserteq(C(seq.map(L'#_',{'one','tw'})),{3,2}) --for l1,l2 in seq.last{10,20,30} do print(l1,l2) end asserteq( C2(seq.last{10,20,30}),{{20,10},{30,20}} ) asserteq( C2(seq.last{40}),{} ) asserteq( C2(seq.last{}),{} ) asserteq( seq{10,20,30}:map(L'_+1'):copy(), {11,21,31} ) asserteq( seq {1,2,3,4,5}:reduce ('*'), 120 ) -- test reduce with an initial value asserteq( seq {1,2,3,4,5}:reduce ('+', 42), 57 ) -- test reduce with a short sequence asserteq( seq {7}:reduce ('+'), 7 ) asserteq( seq {5}:reduce ('/', 40), 8 ) asserteq( seq {}:reduce ('+', 42), 42 ) asserteq( seq {}:reduce ('-'), nil ) asserteq( seq{'one','two'}:upper():copy(), {'ONE','TWO'} ) asserteq( seq{'one','two','three'}:skip(1):copy(), {'two','three'} ) -- test skipping pass sequence asserteq( seq{'one','two','three'}:skip(4):copy(), {} ) asserteq( seq{7,8,9,10}:take(3):copy(), {7,8,9} ) asserteq( seq{7,8,9,10}:take(6):copy(), {7,8,9,10} ) asserteq( seq{7,8,9,10}:take(0):copy(), {} ) asserteq( seq{7,8,9,10}:take(-1):copy(), {} ) local l, u = 50, 100 local rand_seq = seq(seq.random(7, l, u)) asserteq( #rand_seq:filter(seq.less_than(u+1)):filter(seq.greater_than(l-1)):copy(), 7 ) rand_seq = seq(seq.random(7, u)) asserteq( #rand_seq:filter(seq.less_than(u+1)):filter(seq.greater_than(0)):copy(), 7 ) rand_seq = seq(seq.random(7)) asserteq( #rand_seq:filter(seq.less_than(1)):filter(seq.greater_than(0)):copy(), 7 ) test = {275,127,286,590,961,687,802,453,705,182} asserteq( C(seq.sort{seq(test):minmax()}), {127,961} ) asserteq( seq(test):take(5):enum():copy_tuples(), {{1,275},{2,127},{3,286},{4,590},{5,961}} ) asserteq( C(seq.unique(seq.list{1,2,3,2,1})), {1,2,3} ) local actualstr = {} local expectedstr = "275.00 127.00 286.00 590.00 961.00 687.00 802.00\n".. "453.00 705.00 182.00 \n" local function proxywrite_printall(head, ...) table.insert(actualstr, tostring(head)) if select('#', ...) == 0 then return true end return proxywrite_printall(...) end local iowrite = io.write io.write = proxywrite_printall seq(test):printall(nil,nil,'%.2f') io.write = iowrite asserteq( table.concat(actualstr), expectedstr ) local f = stringio.open '1 2 3 4' -- seq.lines may take format specifiers if using Lua 5.2, or a 5.2-compatible -- file object like that returned by stringio. asserteq( seq.lines(f,'*n'):copy(), {1,2,3,4} ) -- the seq() constructor can now take an iterator which consists of two parts, -- a function and an object - as returned e.g. by lfs.dir() local function my_iter(T) local idx = 0 return function(self) idx = idx + 1 return self[idx] end, T end asserteq( seq(my_iter{10,20,30}):copy(), {10,20,30} ) Penlight-1.12.0/tests/test-sip.lua000066400000000000000000000055551416703176500170060ustar00rootroot00000000000000local sip = require 'pl.sip' local tablex = require 'pl.tablex' local test = require 'pl.test' local function check(pat,line,tbl) local parms = {} if type(pat) == 'string' then pat = sip.compile(pat) end if pat(line,parms) then test.asserteq(parms,tbl) else -- only should happen if we're passed a nil! assert(tbl == nil) end end local c = sip.compile('ref=$S{file}:$d{line}') check(c,'ref=bonzo:23',{file='bonzo',line=23}) check(c,'here we go ref=c:\\bonzo\\dog.txt:53',{file='c:\\bonzo\\dog.txt',line=53}) check(c,'here is a line ref=xxxx:xx',nil) c = sip.compile('($i{x},$i{y},$i{z})') check(c,'(10,20,30)',{x=10,y=20,z=30}) check(c,' (+233,+99,-40) ',{x=233,y=99,z=-40}) local pat = '$v{name} = $q{str}' --assert(sip.create_pattern(pat) == [[([%a_][%w_]*)%s*=%s*(["'])(.-)%2]]) local m = sip.compile(pat) check(m,'a = "hello"',{name='a',str='hello'}) check(m,"a = 'hello'",{name='a',str='hello'}) check(m,'_fred="some text"',{name='_fred',str='some text'}) -- some cases broken in 0.6b release check('$v is $v','bonzo is dog for sure',{'bonzo','dog'}) check('$v is $','bonzo is dog for sure',{'bonzo','dog for sure'}) -- spaces check('$v $d','age 23',{'age',23}) check('$v $d','age 23',{'age',23}) check('$v $d','age23') -- the space is 'imcompressible' check('a b c $r', 'a bc d') check('a b c $r', 'a b c d',{'d'}) -- the spaces in this pattern, however, are compressible. check('$v = $d','age=23',{'age',23}) -- patterns without patterns check('just a string', 'just a string', {}) check('just a string', 'not that string') local months={"Jan","Feb","Mar","Apr","May","Jun","Jul","Aug","Sep","Oct","Nov","Dec"} local function adjust_year(res) if res.year < 100 then if res.year < 70 then res.year = res.year + 2000 else res.year = res.year + 1900 end end end local shortdate = sip.compile('$d{day}/$d{month}/$d{year}') local longdate = sip.compile('$d{day} $v{month} $d{year}') local isodate = sip.compile('$d{year}-$d{month}-$d{day}') local function dcheck (d1,d2) adjust_year(d1) test.asserteq(d1, d2) end local function dates(str,tbl) local res = {} if shortdate(str,res) then dcheck(res,tbl) elseif isodate(str,res) then dcheck(res,tbl) elseif longdate(str,res) then res.month = tablex.find(months,res.month) dcheck(res,tbl) else assert(tbl == nil) end end dates ('10/12/2007',{year=2007,month=12,day=10}) dates ('2006-03-01',{year=2006,month=3,day=1}) dates ('25/07/05',{year=2005,month=7,day=25}) dates ('20 Mar 1959',{year=1959,month=3,day=20}) local sio = require 'pl.stringio' local lines = [[ dodge much amazement kitteh cheezburger ]] sip.read(sio.open(lines),{ {'dodge $',function(rest) test.asserteq(rest,'much amazement') end}, {'kitteh $',function(rest) test.asserteq(rest,'cheezburger') end} }) Penlight-1.12.0/tests/test-strict.lua000066400000000000000000000071221416703176500175130ustar00rootroot00000000000000require 'pl.compat' -- require this one before loading strict local strict = require 'pl.strict' local test = require 'pl.test' local app = require 'pl.app' -- in strict mode, you must assign to a global first, even if just nil. test.assertraise(function() print(x) print 'ok?' end,"variable 'x' is not declared") -- can assign to globals in main (or from C extensions) but not anywhere else! test.assertraise(function() Boo = 3 end,"assign to undeclared global 'Boo'") Boo = true Boo2 = nil -- once declared, you can assign to globals from anywhere (function() Boo = 42; Boo2 = 6*7 end)() --- a module may use strict.module() to generate a simularly strict environment -- (see lua/mymod.lua) app.require_here 'lua' local M = require 'mymod' --- these are fine M.answer() M.question() -- spelling mistakes become errors... test.assertraise(function() print(M.Answer()) end,"variable 'Answer' is not declared in 'mymod'") --- for the extra paranoid, you can choose to make all global tables strict... strict.make_all_strict(_G) test.assertraise(function() print(math.sine(1.2)) end,"variable 'sine' is not declared in 'math'") -- module do local testmodule = { hello = function() return "supremacy" end } -- make strict and allow extra field "world" strict.module("my_test", testmodule, { world = true }) test.asserteq(testmodule.hello(), "supremacy") test.assertraise(function() print(testmodule.not_allowed_key) end, "variable 'not_allowed_key' is not declared in 'my_test'") test.asserteq(testmodule.world, nil) testmodule.world = "supremacy" test.asserteq(testmodule.world, "supremacy") -- table with a __newindex method local mod1 = strict.module("mod1", setmetatable( { hello = "world", }, { __newindex = function(self, key, value) if key == "Lua" then rawset(self, key, value) end end, } )) test.asserteq(mod1.hello, "world") mod1.Lua = "hello world" test.asserteq(mod1.Lua, "hello world") test.assertraise(function() print(mod1.not_allowed_key) end, "variable 'not_allowed_key' is not declared in 'mod1'") -- table with a __index method local mod1 = strict.module("mod1", setmetatable( { hello = "world", }, { __index = function(self, key) if key == "Lua" then return "rocks" end end, } )) test.asserteq(mod1.hello, "world") test.asserteq(mod1.Lua, "rocks") test.assertraise(function() print(mod1.not_allowed_key) end, "variable 'not_allowed_key' is not declared in 'mod1'") -- table with a __index table local mod1 = strict.module("mod1", setmetatable( { hello = "world", }, { __index = { Lua = "rocks!" } } )) test.asserteq(mod1.hello, "world") test.asserteq(mod1.Lua, "rocks!") test.assertraise(function() print(mod1.not_allowed_key) end, "variable 'not_allowed_key' is not declared in 'mod1'") end do -- closed_module -- what does this do? this does not seem a usefull function??? local testmodule = { hello = function() return "supremacy" end } local M = strict.closed_module(testmodule, "my_test") -- read acces to original is granted, but not to the new one test.asserteq(testmodule.hello(), "supremacy") test.assertraise(function() print(M.hello()) end, "variable 'hello' is not declared in 'my_test'") -- write access to both is granted testmodule.world = "domination" M.world = "domination" -- read acces to set field in original is granted, but not set test.asserteq(testmodule.world, nil) test.asserteq(M.world, "domination") end Penlight-1.12.0/tests/test-stringio.lua000066400000000000000000000027311416703176500200420ustar00rootroot00000000000000local stringio = require 'pl.stringio' local test = require 'pl.test' local asserteq = test.asserteq local T = test.tuple function fprintf(f,fmt,...) f:write(fmt:format(...)) end fs = stringio.create() for i = 1,100 do fs:write('hello','\n','dolly','\n') end asserteq(#fs:value(),1200) fs = stringio.create() fs:writef("%s %d",'answer',42) -- note writef() extension method asserteq(fs:value(),"answer 42") inf = stringio.open('10 20 30') asserteq(T(inf:read('*n','*n','*n')),T(10,20,30)) local txt = [[ Some lines here are they not for other english? ]] inf = stringio.open (txt) fs = stringio.create() for l in inf:lines() do fs:write(l,'\n') end asserteq(txt,fs:value()) inf = stringio.open '1234567890ABCDEF' asserteq(T(inf:read(3), inf:read(5), inf:read()),T('123','45678','90ABCDEF')) s = stringio.open 'one\ntwo' asserteq(s:read() , 'one') asserteq(s:read() , 'two') asserteq(s:read() , nil) s = stringio.open 'one\ntwo' iter = s:lines() asserteq(iter() , 'one') asserteq(iter() , 'two') asserteq(iter() , nil) s = stringio.open 'ABC' iter = s:lines(1) asserteq(iter() , 'A') asserteq(iter() , 'B') asserteq(iter() , 'C') asserteq(iter() , nil) s = stringio.open '20 5.2e-2 52.3' x,y,z = s:read('*n','*n','*n') out = stringio.create() fprintf(out,"%5.2f %5.2f %5.2f!",x,y,z) asserteq(out:value(),"20.00 0.05 52.30!") s = stringio.open 'one\ntwo\n\n' iter = s:lines '*L' asserteq(iter(),'one\n') asserteq(iter(),'two\n') asserteq(iter(),'\n') asserteq(iter(),nil) Penlight-1.12.0/tests/test-tablex.lua000066400000000000000000000204231416703176500174610ustar00rootroot00000000000000local tablex = require 'pl.tablex' local utils = require ('pl.utils') local L = utils.string_lambda local test = require('pl.test') -- bring tablex funtions into global namespace utils.import(tablex) local asserteq = test.asserteq local cmp = deepcompare function asserteq_no_order (x,y) if not compare_no_order(x,y) then test.complain(x,y,"these lists contained different elements") end end asserteq( copy {10,20,30}, {10,20,30} ) asserteq( deepcopy {10,20,{30,40}}, {10,20,{30,40}} ) local t = { a = "hello", b = { c = "world" } } t.b.d = t.b local tcopy = { a = "hello", b = { c = "world" } } tcopy.b.d = tcopy.b asserteq( deepcopy(t), tcopy ) asserteq( pairmap(function(i,v) return v end,{10,20,30}), {10,20,30} ) asserteq_no_order( pairmap(L'_',{fred=10,bonzo=20}), {'fred','bonzo'} ) asserteq_no_order( pairmap(function(k,v) return v end,{fred=10,bonzo=20}), {10,20} ) asserteq_no_order( pairmap(function(i,v) return v,i end,{10,20,30}), {10,20,30} ) asserteq( pairmap(function(k,v) return {k,v},k end,{one=1,two=2}), {one={'one',1},two={'two',2}} ) -- same as above, using string lambdas asserteq( pairmap(L'|k,v|{k,v},k',{one=1,two=2}), {one={'one',1},two={'two',2}} ) asserteq( map(function(v) return v*v end,{10,20,30}), {100,400,900} ) -- extra arguments to map() are passed to the function; can use -- the abbreviations provided by pl.operator asserteq( map('+',{10,20,30},1), {11,21,31} ) asserteq( map(L'_+1',{10,20,30}), {11,21,31} ) -- map2 generalizes for operations on two tables asserteq( map2(math.max,{1,2,3},{0,4,2}), {1,4,3} ) -- mapn operates over an arbitrary number of input tables (but use map2 for n=2) asserteq( mapn(function(x,y,z) return x+y+z end, {1,2,3},{10,20,30},{100,200,300}), {111,222,333} ) asserteq( mapn(math.max, {1,20,300},{10,2,3},{100,200,100}), {100,200,300} ) asserteq( count_map({"foo", "bar", "foo", "baz"}), {foo = 2, bar = 1, baz = 1} ) asserteq( zip({10,20,30},{100,200,300}), {{10,100},{20,200},{30,300}} ) assert(compare_no_order({1,2,3,4},{2,1,4,3})==true) assert(compare_no_order({1,2,3,4},{2,1,4,4})==false) asserteq(range(10,9),{}) asserteq(range(10,10),{10}) asserteq(range(10,11),{10,11}) -- update inserts key-value pairs from the second table t1 = {one=1,two=2} t2 = {three=3,two=20,four=4} asserteq(update(t1,t2),{one=1,three=3,two=20,four=4}) -- the difference between move and icopy is that the second removes -- any extra elements in the destination after end of copy -- 3rd arg is the index to start in the destination, defaults to 1 asserteq(move({1,2,3,4,5,6},{20,30}),{20,30,3,4,5,6}) asserteq(move({1,2,3,4,5,6},{20,30},2),{1,20,30,4,5,6}) asserteq(icopy({1,2,3,4,5,6},{20,30},2),{1,20,30}) -- 5th arg determines how many elements to copy (default size of source) asserteq(icopy({1,2,3,4,5,6},{20,30},2,1,1),{1,20}) -- 4th arg is where to stop copying from the source (default s to 1) asserteq(icopy({1,2,3,4,5,6},{20,30},2,2,1),{1,30}) -- whereas insertvalues works like table.insert, but inserts a range of values -- from the given table. asserteq(insertvalues({1,2,3,4,5,6},2,{20,30}),{1,20,30,2,3,4,5,6}) asserteq(insertvalues({1,2},{3,4}),{1,2,3,4}) -- the 4th arg of move and icopy gives the start index in the source table asserteq(move({1,2,3,4,5,6},{20,30},2,2),{1,30,3,4,5,6}) asserteq(icopy({1,2,3,4,5,6},{20,30},2,2),{1,30}) t = {1,2,3,4,5,6} move(t,{20,30},2,1,1) asserteq(t,{1,20,3,4,5,6}) set(t,0,2,3) asserteq(t,{1,0,0,4,5,6}) insertvalues(t,1,{10,20}) asserteq(t,{10,20,1,0,0,4,5,6}) asserteq(merge({10,20,30},{nil, nil, 30, 40}), {[3]=30}) asserteq(merge({10,20,30},{nil, nil, 30, 40}, true), {10,20,30,40}) -- Function to check that the order of elements returned by the iterator -- match the order of the elements in the list. function assert_iter_order(iter,l) local i = 0 for k,v in iter do i = i + 1 asserteq(k,l[i][1]) asserteq(v,l[i][2]) end end local t = {a=10,b=9,c=8,d=7,e=6,f=5,g=4,h=3,i=2,j=1} assert_iter_order( sort(t), {{'a',10},{'b',9},{'c',8},{'d',7},{'e',6},{'f',5},{'g',4},{'h',3},{'i',2},{'j',1}}) assert_iter_order( sortv(t), {{'j',1},{'i',2},{'h',3},{'g',4},{'f',5},{'e',6},{'d',7},{'c',8},{'b',9},{'a',10}}) asserteq(difference({a = true, b = true},{a = true, b = true}),{}) -- no longer confused by false values ;) asserteq(difference({v = false},{v = false}),{}) asserteq(difference({a = true},{b = true}),{a=true}) -- symmetric difference asserteq(difference({a = true},{b = true},true),{a=true,b=true}) --basic index_map test asserteq(index_map({10,20,30}), {[10]=1,[20]=2,[30]=3}) --test that repeated values return multiple indices asserteq(index_map({10,20,30,30,30}), {[10]=1,[20]=2,[30]={3,4,5}}) -- Reduce asserteq(tablex.reduce('-', {}, 2), 2) asserteq(tablex.reduce('-', {}), nil) asserteq(tablex.reduce('-', {1,2,3,4,5}), -13) asserteq(tablex.reduce('-', {1,2,3,4,5}, 1), -14) -- tablex.compare do asserteq(tablex.compare({},{}, "=="), true) asserteq(tablex.compare({1,2,3}, {1,2,3}, "=="), true) asserteq(tablex.compare({1,"hello",3}, {1,2,3}, "=="), false) asserteq(tablex.compare( {1,2,3, hello = "world"}, {1,2,3}, function(v1, v2) return v1 == v2 end), true) -- only compares the list part end -- tablex.rfind do local rfind = tablex.rfind local lst = { "Rudolph", "the", "red-nose", "raindeer" } asserteq(rfind(lst, "Santa"), nil) asserteq(rfind(lst, "raindeer", -2), nil) asserteq(rfind(lst, "raindeer"), 4) asserteq(rfind(lst, "Rudolph"), 1) asserteq(rfind(lst, "the", -3), 2) asserteq(rfind(lst, "the", -30), nil) asserteq(rfind({10,10,10},10), 3) end -- tablex.find_if do local fi = tablex.find_if local lst = { "Rudolph", true, false, 15 } asserteq({fi(lst, "==", "Rudolph")}, {1, true}) asserteq({fi(lst, "==", true)}, {2, true}) asserteq({fi(lst, "==", false)}, {3, true}) asserteq({fi(lst, "==", 15)}, {4, true}) local cmp = function(v1, v2) return v1 == v2 and v2 end asserteq({fi(lst, cmp, "Rudolph")}, {1, "Rudolph"}) asserteq({fi(lst, cmp, true)}, {2, true}) asserteq({fi(lst, cmp, false)}, {}) -- 'false' cannot be returned! asserteq({fi(lst, cmp, 15)}, {4, 15}) end -- tablex.map_named_method do local Car = {} Car.__index = Car function Car.new(car) return setmetatable(car or {}, Car) end Car.speed = 0 function Car:faster(increase) self.speed = self.speed + (increase or 1) return self.speed end function Car:slower(self, decrease) self.speed = self.speed - (decrease or 1) return self.speed end local ferrari = Car.new{ name = "Ferrari" } local lamborghini = Car.new{ name = "Lamborghini", speed = 50 } local cars = { ferrari, lamborghini } asserteq(ferrari.speed, 0) asserteq(lamborghini.speed, 50) asserteq(tablex.map_named_method("faster", cars, 10), {10, 60}) asserteq(ferrari.speed, 10) asserteq(lamborghini.speed, 60) end -- tablex.foreach do local lst = { "one", "two", "three", hello = "world" } tablex.foreach(lst, function(v, k, sep) lst[k] = tostring(k) .. sep .. v end, " = ") asserteq(lst, {"1 = one", "2 = two", "3 = three", hello = "hello = world"}) end -- tablex.foreachi do local lst = { "one", "two", "three", hello = "world" } tablex.foreachi(lst, function(v, k, sep) lst[k] = tostring(k) .. sep .. v end, " = ") asserteq(lst, {"1 = one", "2 = two", "3 = three", hello = "world"}) end -- tablex.new asserteq(tablex.new(3, "hi"), { "hi", "hi", "hi" }) -- tablex.search do local t = { penlight = { battery = { type = "AA", capacity = "1500mah", }, }, hello = { world = { also = "AA" } } } asserteq(tablex.search(t, "1500mah"), "penlight.battery.capacity") asserteq(tablex.search(t, "AA", {t.penlight} ), "hello.world.also") asserteq(tablex.search(t, "xxx"), nil) end -- tablex.readonly do local ro = tablex.readonly { 1,2,3, hello = "world" } asserteq(pcall(function() ro.hello = "hi there" end), false) asserteq(getmetatable(ro), false) if not utils.lua51 then asserteq(#ro, 3) local r = {} for k,v in pairs(ro) do r[k] = v end asserteq(r, { 1,2,3, hello = "world" }) r = {} for k,v in ipairs(ro) do r[k] = v end asserteq(r, { 1,2,3 }) end end Penlight-1.12.0/tests/test-tablex3.lua000066400000000000000000000004041416703176500175410ustar00rootroot00000000000000-- tablex.move when the tables are the same -- and there are overlapping ranges T = require 'pl.tablex' asserteq = require 'pl.test'.asserteq t1 = {1,2,3,4,5,6,7,8,9,10} t2 = T.copy(t1) t3 = T.copy(t1) T.move(t1,t2,4,1,4) T.move(t3,t3,4,1,4) asserteq(t1,t3) Penlight-1.12.0/tests/test-template.lua000066400000000000000000000113161416703176500200160ustar00rootroot00000000000000local template = require 'pl.template' local subst = template.substitute local List = require 'pl.List' local asserteq = require 'pl.test'.asserteq local utils = require 'pl.utils' asserteq(subst([[ # for i = 1,2 do

Hello $(tostring(i))

# end ]],_G),[[

Hello 1

Hello 2

]]) asserteq(subst([[
    # for name in ls:iter() do
  • $(name)
  • #end
]],{ls = List{'john','alice','jane'}}),[[
  • john
  • alice
  • jane
]]) -- can change the default escape from '#' so we can do C/C++ output. -- note that the environment can have a parent field. asserteq(subst([[ > for i,v in ipairs{'alpha','beta','gamma'} do cout << obj.${v} << endl; > end ]],{_parent=_G, _brackets='{}', _escape='>'}),[[ cout << obj.alpha << endl; cout << obj.beta << endl; cout << obj.gamma << endl; ]]) -- handle templates with a lot of substitutions asserteq(subst(("$(x)\n"):rep(300), {x = "y"}), ("y\n"):rep(300)) -------------------------------------------------- -- Test using no leading nor trailing linebreak local tmpl = [[
    # for i,val in ipairs(T) do
  • $(i) = $(val:upper())
  • # end
]] local my_env = { ipairs = ipairs, T = {'one','two','three'}, _debug = true, } local res, err = template.substitute(tmpl, my_env) --print(res, err) asserteq(res, [[
  • 1 = ONE
  • 2 = TWO
  • 3 = THREE
]]) -------------------------------------------------- -- Test using both leading and trailing linebreak local tmpl = [[
    # for i,val in ipairs(T) do
  • $(i) = $(val:upper())
  • # end
]] local my_env = { ipairs = ipairs, T = {'one','two','three'}, _debug = true, } local res, err = template.substitute(tmpl, my_env) --print(res, err) asserteq(res, [[
  • 1 = ONE
  • 2 = TWO
  • 3 = THREE
]]) -------------------------------------------------- -- Test reusing a compiled template local tmpl = [[
    # for i,val in ipairs(T) do
  • $(i) = $(val:upper())
  • # end
]] local my_env = { ipairs = ipairs, T = {'one','two','three'} } local t, err = template.compile(tmpl, { debug = true }) local res, err, code = t:render(my_env) --print(res, err, code) asserteq(res, [[
  • 1 = ONE
  • 2 = TWO
  • 3 = THREE
]]) -- now reuse with different env local my_env = { ipairs = ipairs, T = {'four','five','six'} } local t, err = template.compile(tmpl, { debug = true }) local res, err, code = t:render(my_env) --print(res, err, code) asserteq(res, [[
  • 1 = FOUR
  • 2 = FIVE
  • 3 = SIX
]]) -------------------------------------------------- -- Test the newline parameter local tmpl = [[ some list: $(T[1]:upper()) # for i = 2, #T do ,$(T[i]:upper()) # end ]] local my_env = { ipairs = ipairs, T = {'one','two','three'} } local t, err = template.compile(tmpl, { debug = true, newline = "" }) local res, err, code = t:render(my_env) --print(res, err, code) asserteq(res, [[some list: ONE,TWO,THREE]]) -------------------------------------------------- -- Test template run-time error local tmpl = [[ header: $("hello" * 10) ]] local t, err = template.compile(tmpl, { debug = true, newline = "" }) local res, err, code = t:render() --print(res, err, code) assert(res == nil, "expected nil here because of the runtime error") asserteq(type(err), "string") asserteq(type(utils.load(code)), "function") -------------------------------------------------- -- Test template run-time, doesn't fail on table value -- table.concat fails if we insert a non-string (table) value local tmpl = [[ header: $(myParam) ]] local t, err = template.compile(tmpl, { debug = true, newline = "" }) local myParam = {} local res, err, code = t:render( {myParam = myParam } ) -- insert a table --print(res, err, code) asserteq(res, "header: "..tostring(myParam)) asserteq(type(err), "nil") -------------------------------------------------- -- Test template compile-time error local tmpl = [[ header: $(this doesn't work) ]] local my_env = { ipairs = ipairs, T = {'one','two','three'} } local t, err, code = template.compile(tmpl, { debug = true, newline = "" }) --print(t, err, code) assert(t==nil, "expected t to be nil here because of the syntax error") asserteq(type(err), "string") asserteq(type(code), "string") -------------------------------------------------- -- Test using template being a single static string local tmpl = [[

    a paragraph

    a paragraph

]] local t, err = template.compile(tmpl, { debug = true }) local res, err, code = t:render(my_env) --print(res, err, code) asserteq(res, [[

    a paragraph

    a paragraph

]]) asserteq(code, [[return "
    \

    a paragraph

    \

    a paragraph

    \
\ "]]) print("template: success") Penlight-1.12.0/tests/test-template2.lua000066400000000000000000000026461416703176500201060ustar00rootroot00000000000000local T = require 'pl.text' local utils = require 'pl.utils' local Template = T.Template local asserteq = require 'pl.test'.asserteq local OrderedMap = require 'pl.OrderedMap' local template = require 'pl.template' local t = [[ # for i = 1,3 do print($(i+1)) # end ]] asserteq(template.substitute(t),[[ print(2) print(3) print(4) ]]) t = [[ > for i = 1,3 do print(${i+1}) > end ]] asserteq(template.substitute(t,{_brackets='{}',_escape='>'}),[[ print(2) print(3) print(4) ]]) t = [[ #@ for i = 1,3 do print(@{i+1}) #@ end ]] asserteq(template.substitute(t,{_brackets='{}',_escape='#@',_inline_escape='@'}),[[ print(2) print(3) print(4) ]]) --- iteration using pairs is usually unordered. But using OrderedMap --- we can get the exact original ordering. t = [[ # for k,v in pairs(T) do "$(k)", -- $(v) # end ]] if utils.lua51 then -- easy enough to define a general pairs in Lua 5.1 local rawpairs = pairs function pairs(t) local mt = getmetatable(t) local f = mt and mt.__pairs if f then return f(t) else return rawpairs(t) end end end local Tee = OrderedMap{{Dog = 'Bonzo'}, {Cat = 'Felix'}, {Lion = 'Leo'}} -- note that the template will also look up global functions using _parent asserteq(template.substitute(t,{T=Tee,_parent=_G}),[[ "Dog", -- Bonzo "Cat", -- Felix "Lion", -- Leo ]]) Penlight-1.12.0/tests/test-types.lua000066400000000000000000000073111416703176500173470ustar00rootroot00000000000000---- testing types local types = require 'pl.types' local asserteq = require 'pl.test'.asserteq local List = require 'pl.List' local list = List() local array = {10,20,30} local map = {one=1,two=2} -- extended type() function asserteq(types.type(array),'table') asserteq(types.type('hello'),'string') -- knows about Lua file objects asserteq(types.type(io.stdin),'file') local f = io.open("tests/test-types.lua") asserteq(types.type(f),'file') f:close() -- and class names asserteq(types.type(list),'List') -- tables with unknown metatable asserteq(types.type(setmetatable({},{})), "unknown table") -- userdata with unknown metatable if newproxy then asserteq(types.type(newproxy(true)), "unknown userdata") end asserteq(types.is_integer(10),true) asserteq(types.is_integer(10.1),false) asserteq(types.is_integer(-10),true) asserteq(types.is_integer(-10.1),false) -- do note that for Lua < 5.3, 10.0 is the same as 10; an integer. asserteq(types.is_callable(asserteq),true) asserteq(types.is_callable(List),true) asserteq(types.is_indexable(array),true) asserteq(types.is_indexable('hello'),nil) asserteq(types.is_indexable(10),nil) if newproxy then local v = newproxy(true) local mt = getmetatable(v) mt.__len = true mt.__index = true asserteq(types.is_indexable(v), true) end if newproxy then local v = newproxy(true) asserteq(types.is_indexable(v), nil) end asserteq(types.is_iterable(array),true) asserteq(types.is_iterable(true),nil) asserteq(types.is_iterable(42),nil) asserteq(types.is_iterable("array"),nil) if newproxy then local v = newproxy(true) local mt = getmetatable(v) mt.__pairs = true asserteq(types.is_iterable(v), true) end if newproxy then local v = newproxy(true) asserteq(types.is_iterable(v), nil) end asserteq(types.is_writeable(array),true) asserteq(types.is_writeable(true),nil) asserteq(types.is_writeable(42),nil) asserteq(types.is_writeable("array"),nil) if newproxy then local v = newproxy(true) local mt = getmetatable(v) mt.__newindex = true asserteq(types.is_writeable(v), true) end if newproxy then local v = newproxy(true) asserteq(types.is_writeable(v), nil) end asserteq(types.is_empty(nil),true) asserteq(types.is_empty({}),true) asserteq(types.is_empty({[false] = false}),false) asserteq(types.is_empty(""),true) asserteq(types.is_empty(" ",true),true) asserteq(types.is_empty(" "),false) asserteq(types.is_empty(true),true) -- Numbers asserteq(types.is_empty(0), true) asserteq(types.is_empty(20), true) -- Booleans asserteq(types.is_empty(false), true) asserteq(types.is_empty(true), true) -- Functions asserteq(types.is_empty(print), true) -- Userdata --asserteq(types.is_empty(newproxy()), true) --newproxy was removed in Lua 5.2 -- a more relaxed kind of truthiness.... asserteq(types.to_bool('yes'),true) asserteq(types.to_bool('true'),true) asserteq(types.to_bool('y'),true) asserteq(types.to_bool('t'),true) asserteq(types.to_bool('YES'),true) asserteq(types.to_bool('1'),true) asserteq(types.to_bool('no'),false) asserteq(types.to_bool('false'),false) asserteq(types.to_bool('n'),false) asserteq(types.to_bool('f'),false) asserteq(types.to_bool('NO'),false) asserteq(types.to_bool('0'),false) asserteq(types.to_bool(1),true) asserteq(types.to_bool(0),false) local de_fr = { 'ja', 'oui' } asserteq(types.to_bool('ja', de_fr),true) asserteq(types.to_bool('OUI', de_fr),true) local t_e = {} local t_ne = { "not empty" } asserteq(types.to_bool(t_e,nil,false),false) asserteq(types.to_bool(t_e,nil,true),false) asserteq(types.to_bool(t_ne,nil,false),false) asserteq(types.to_bool(t_ne,nil,true),true) asserteq(types.to_bool(coroutine.create(function() end),nil,true),true) asserteq(types.to_bool(coroutine.create(function() end),nil,false),false) Penlight-1.12.0/tests/test-url.lua000066400000000000000000000032311416703176500170020ustar00rootroot00000000000000local url = require 'pl.url' local asserteq = require 'pl.test' . asserteq asserteq(url.quote(''), '') asserteq(url.quote('ABCDEFGHIJKLMNOPQRSTUVWXYZ'), 'ABCDEFGHIJKLMNOPQRSTUVWXYZ') asserteq(url.quote('abcdefghijklmnopqrstuvwxyz'), 'abcdefghijklmnopqrstuvwxyz') asserteq(url.quote('0123456789'), '0123456789') asserteq(url.quote(' -_./'), '%20-_./') asserteq(url.quote('`~!@#$%^&*()'), '%60%7E%21%40%23%24%25%5E%26%2A%28%29') asserteq(url.quote('%2'), '%252') asserteq(url.quote('2R%1%%'), '2R%251%25%25') asserteq(url.quote('', true), '') asserteq(url.quote('ABCDEFGHIJKLMNOPQRSTUVWXYZ', true), 'ABCDEFGHIJKLMNOPQRSTUVWXYZ') asserteq(url.quote('abcdefghijklmnopqrstuvwxyz', true), 'abcdefghijklmnopqrstuvwxyz') asserteq(url.quote('0123456789'), '0123456789', true) asserteq(url.quote(' -_./', true), '+-_.%2F') asserteq(url.quote('`~!@#$%^&*()', true), '%60%7E%21%40%23%24%25%5E%26%2A%28%29') asserteq(url.quote('%2', true), '%252') asserteq(url.quote('2R%1%%', true), '2R%251%25%25') asserteq(url.unquote(''), '') asserteq(url.unquote('ABCDEFGHIJKLMNOPQRSTUVWXYZ'), 'ABCDEFGHIJKLMNOPQRSTUVWXYZ') asserteq(url.unquote('abcdefghijklmnopqrstuvwxyz'), 'abcdefghijklmnopqrstuvwxyz') asserteq(url.unquote('0123456789'), '0123456789') asserteq(url.unquote(' -_./'), ' -_./') asserteq(url.unquote('+-_.%2F'), ' -_./') asserteq(url.unquote('%20-_./'), ' -_./') asserteq(url.unquote('%60%7E%21%40%23%24%25%5E%26%2A%28%29'), '`~!@#$%^&*()') asserteq(url.unquote('%252'), '%2') asserteq(url.unquote('2%52%1%%'), '2R%1%%') asserteq(url.unquote('2R%251%25%25'), '2R%1%%') asserteq(url.quote(true), true) asserteq(url.quote(42), 42) asserteq(url.unquote(true), true) asserteq(url.unquote(42), 42) Penlight-1.12.0/tests/test-utils.lua000066400000000000000000000224561416703176500173520ustar00rootroot00000000000000local utils = require 'pl.utils' local path = require 'pl.path' local test = require 'pl.test' local asserteq, T = test.asserteq, test.tuple local function quote(s) if utils.is_windows then return '"'..s..'"' else return "'"..s.."'" end end -- construct command to run external lua, we need to to be able to run some -- tests on the same lua engine, but also need to pass on the LuaCov flag -- if it was used, to make sure we report the proper coverage. local cmd = "-e " do local i = 0 while arg[i-1] do local a = arg[i-1] if a:find("package%.path") and a:sub(1,1) ~= "'" then a = quote(a) end cmd = a .. " " .. cmd i = i - 1 end end --- quitting do local luacode = quote("require([[pl.utils]]).quit([[hello world]])") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, false) if utils.is_windows then asserteq(code, -1) else asserteq(code, 255) end asserteq(stdout, "") asserteq(stderr, "hello world\n") local luacode = quote("require([[pl.utils]]).quit(2, [[hello world]])") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, false) asserteq(code, 2) asserteq(stdout, "") asserteq(stderr, "hello world\n") local luacode = quote("require([[pl.utils]]).quit(2, [[hello %s]], 42)") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, false) asserteq(code, 2) asserteq(stdout, "") asserteq(stderr, "hello 42\n") local luacode = quote("require([[pl.utils]]).quit(2)") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, false) asserteq(code, 2) asserteq(stdout, "") asserteq(stderr, "") end ----- importing module tables wholesale --- utils.import(math) asserteq(type(sin),"function") asserteq(type(abs),"function") --- useful patterns local P = utils.patterns asserteq(("+0.1e10"):match(P.FLOAT) ~= nil, true) asserteq(("-23430"):match(P.INTEGER) ~= nil, true) asserteq(("my_little_pony99"):match(P.IDEN) ~= nil, true) --- escaping magic chars local escape = utils.escape asserteq(escape '[a]','%[a%]') asserteq(escape '$(bonzo)','%$%(bonzo%)') --- choose asserteq(utils.choose(true, 1, 2), 1) asserteq(utils.choose(false, 1, 2), 2) --- splitting strings --- local split = utils.split asserteq(split("hello dolly"),{"hello","dolly"}) asserteq(split("hello,dolly",","),{"hello","dolly"}) asserteq(split("hello,dolly,",","),{"hello","dolly"}) local first,second = utils.splitv("hello:dolly",":") asserteq(T(first,second),T("hello","dolly")) local first,second = utils.splitv("hello:dolly:parton",":", false, 2) asserteq(T(first,second),T("hello","dolly:parton")) local first,second,third = utils.splitv("hello=dolly:parton","[:=]") asserteq(T(first,second,third),T("hello","dolly","parton")) local first,second = utils.splitv("hello=dolly:parton","[:=]", false, 2) asserteq(T(first,second),T("hello","dolly:parton")) ----- table of values to table of strings asserteq(utils.array_tostring{1,2,3},{"1","2","3"}) -- writing into existing table local tmp = {} utils.array_tostring({1,2,3},tmp) asserteq(tmp,{"1","2","3"}) --- memoizing a function local kount = 0 local f = utils.memoize(function(x) kount = kount + 1 return x*x end) asserteq(f(2),4) asserteq(f(10),100) asserteq(f(2),4) -- actual function only called twice asserteq(kount,2) -- string lambdas local L = utils.string_lambda local g = L"|x| x:sub(1,1)" asserteq(g("hello"),"h") local f = L"|x,y| x - y" asserteq(f(10,2),8) -- alternative form for _one_ argument asserteq(L("2 * _")(4), 8) local List = require 'pl.List' local ls = List{10,20,30} -- string lambdas can be used throughout Penlight asserteq(ls:map"_+1", {11,21,31}) -- because they use this common function local function test_fn_arg(f) f = utils.function_arg(1,f) asserteq(f(10),11) end test_fn_arg (function (x) return x + 1 end) test_fn_arg '_ + 1' test.assertraise(function() test_fn_arg {} end, 'not a callable object') test.assertraise(function() test_fn_arg (0) end, 'must be callable') -- partial application local f1 = utils.bind1(f,10) asserteq(f1(2), 8) local f2 = utils.bind2(f,2) asserteq(f2(10), 8) --- extended type checking local is_type = utils.is_type -- anything without a metatable works as regular type() function asserteq(is_type("one","string"),true) asserteq(is_type({},"table"),true) -- but otherwise the type of an object is considered to be its metatable asserteq(is_type(ls,List),true) -- compatibility functions local chunk = utils.load 'return 42' asserteq(chunk(),42) chunk = utils.load 'a = 42' chunk() asserteq(a,42) local t = {} chunk = utils.load ('b = 42','','t',t) chunk() asserteq(t.b,42) chunk,err = utils.load ('a = ?','') assert(err,[[[string ""]:1: unexpected symbol near '?']]) asserteq(utils.quote_arg("foo"), [[foo]]) if path.is_windows then asserteq(utils.quote_arg(""), '^"^"') asserteq(utils.quote_arg('"'), '^"') asserteq(utils.quote_arg([[ \]]), [[^" \\^"]]) asserteq(utils.quote_arg([[foo\\ bar\\" baz\]]), [[^"foo\\ bar\\\\\^" baz\\^"]]) asserteq(utils.quote_arg("%path% ^^!()"), [[^"^%path^% ^^^^^!()^"]]) else asserteq(utils.quote_arg(""), "''") asserteq(utils.quote_arg("'"), [[''\''']]) asserteq(utils.quote_arg([['a\'b]]), [[''\''a\'\''b']]) end -- packing and unpacking arguments in a nil-safe way local t = utils.pack(nil, nil, "hello", nil) asserteq(t.n, 4) -- the last nil does count as an argument local arg1, arg2, arg3, arg4 = utils.unpack(t) assert(arg1 == nil) assert(arg2 == nil) asserteq("hello", arg3) assert(arg4 == nil) -- Assert arguments assert_arg local ok, err = pcall(function() utils.assert_arg(4,'!@#$%^&*','string',require("pl.path").isdir,'not a directory') end) asserteq(ok, false) asserteq(err:match("(argument .+)$"), "argument 4: '!@#$%^&*' not a directory") local ok, err = pcall(function() utils.assert_arg(1, "hello", "table") end) asserteq(ok, false) asserteq(err:match("(argument .+)$"), "argument 1 expected a 'table', got a 'string'") local ok, err = pcall(function() return utils.assert_arg(1, "hello", "string") end) asserteq(ok, true) asserteq(err, "hello") -- assert_string local success, err = pcall(utils.assert_string, 2, 5) asserteq(success, false) asserteq(err:match("(argument .+)$"), "argument 2 expected a 'string', got a 'number'") local x = utils.assert_string(2, "5") asserteq(x, "5") do -- printf -- without template local luacode = quote("require([[pl.utils]]).printf([[hello world]])") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, true) asserteq(code, 0) asserteq(stdout, "hello world") asserteq(stderr, "") -- printf -- with template local luacode = quote("require([[pl.utils]]).printf([[hello %s]], [[world]])") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, true) asserteq(code, 0) asserteq(stdout, "hello world") asserteq(stderr, "") -- printf -- with bad template local luacode = quote("require([[pl.utils]]).printf(42)") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, false) asserteq(code, 1) asserteq(stdout, "") assert(stderr:find("argument 1 expected a 'string', got a 'number'")) end do -- on_error, raise -- default utils.on_error("default") local ok, err = utils.raise("some error") asserteq(ok, nil) asserteq(err, "some error") local ok, err = pcall(utils.on_error, "bad one") asserteq(ok, false) asserteq(err, "Bad argument expected string; 'default', 'quit', or 'error'. Got 'bad one'") -- on_error, raise -- error utils.on_error("error") local ok, err = pcall(utils.raise, "some error") asserteq(ok, false) asserteq(err, "some error") local ok, err = pcall(utils.on_error, "bad one") asserteq(ok, false) assert(err:find("Bad argument expected string; 'default', 'quit', or 'error'. Got 'bad one'")) -- on_error, raise -- quit utils.on_error("quit") local luacode = quote("local u=require([[pl.utils]]) u.on_error([[quit]]) u.raise([[some error]])") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, false) if utils.is_windows then asserteq(code, -1) else asserteq(code, 255) end asserteq(stdout, "") asserteq(stderr, "some error\n") local luacode = quote("local u=require([[pl.utils]]) u.on_error([[quit]]) u.on_error([[bad one]])") local success, code, stdout, stderr = utils.executeex(cmd..luacode) asserteq(success, false) if utils.is_windows then asserteq(code, -1) else asserteq(code, 255) end asserteq(stdout, "") asserteq(stderr, "Bad argument expected string; 'default', 'quit', or 'error'. Got 'bad one'\n") utils.on_error("default") -- cleanup by restoring behaviour after on_error + raise tests end do -- readlines local f = utils.readlines("tests/test-utils.lua") asserteq(type(f), "table") local v = "some extraordinary string this is only in this file for test purposes so we can go and find it" local found = false for i, line in ipairs(f) do if line:find(v) then found = true break end end asserteq(found, true) end Penlight-1.12.0/tests/test-utils2.lua000066400000000000000000000044371416703176500174330ustar00rootroot00000000000000local path = require 'pl.path' local utils = require 'pl.utils' local asserteq = require 'pl.test'.asserteq local echo_lineending = "\n" if path.is_windows then echo_lineending = " \n" end local function test_executeex(cmd, expected_successful, expected_retcode, expected_stdout, expected_stderr) --print("\n"..cmd) --print(os.execute(cmd)) --print(utils.executeex(cmd)) local successful, retcode, stdout, stderr = utils.executeex(cmd) asserteq(successful, expected_successful) asserteq(retcode, expected_retcode) asserteq(stdout, expected_stdout) asserteq(stderr, expected_stderr) end -- Check the return codes if utils.is_windows then test_executeex("exit", true, 0, "", "") test_executeex("exit 0", true, 0, "", "") test_executeex("exit 1", false, 1, "", "") test_executeex("exit 13", false, 13, "", "") test_executeex("exit 255", false, 255, "", "") test_executeex("exit 256", false, 256, "", "") test_executeex("exit 257", false, 257, "", "") test_executeex("exit 3809", false, 3809, "", "") test_executeex("exit -1", false, -1, "", "") test_executeex("exit -13", false, -13, "", "") test_executeex("exit -255", false, -255, "", "") test_executeex("exit -256", false, -256, "", "") test_executeex("exit -257", false, -257, "", "") test_executeex("exit -3809", false, -3809, "", "") else test_executeex("exit", true, 0, "", "") test_executeex("exit 0", true, 0, "", "") test_executeex("exit 1", false, 1, "", "") test_executeex("exit 13", false, 13, "", "") test_executeex("exit 255", false, 255, "", "") -- on posix anything other than 0-255 is undefined test_executeex("exit 256", true, 0, "", "") test_executeex("exit 257", false, 1, "", "") test_executeex("exit 3809", false, 225, "", "") end -- Check output strings test_executeex("echo stdout", true, 0, "stdout" .. echo_lineending, "") test_executeex("(echo stderr 1>&2)", true, 0, "", "stderr" .. echo_lineending) test_executeex("(echo stdout && (echo stderr 1>&2))", true, 0, "stdout" .. echo_lineending, "stderr" .. echo_lineending) Penlight-1.12.0/tests/test-utils3.lua000066400000000000000000000034151416703176500174270ustar00rootroot00000000000000--- testing Lua 5.1/5.2 compatibility functions -- these are global side-effects of pl.utils local utils = require 'pl.utils' local test = require 'pl.test' local asserteq = test.asserteq local _,lua = require 'pl.app'. lua() local setfenv,getfenv = utils.setfenv, utils.getfenv -- utils.execute is a compromise between 5.1 and 5.2 for os.execute changes -- can we call Lua ? local ok,code = utils.execute(lua..' -v') asserteq(ok,true) asserteq(code,0) -- does it return false when it fails ? asserteq(utils.execute('most-likely-nonexistent-command'),false) -- table.pack is defined for 5.1 local t = table.pack(1,nil,'hello') asserteq(t.n,3) assert(t[1] == 1 and t[3] == 'hello') -- unpack is not globally available for 5.2 unless in compat mode. -- But utils.unpack is always defined. local a,b = utils.unpack{10,'wow'} assert(a == 10 and b == 'wow') -- utils.load() is Lua 5.2 style chunk = utils.load('return x+y','tmp','t',{x=1,y=2}) asserteq(chunk(),3) -- can only load a binary chunk if the mode permits! local f = string.dump(function() end) local res,err = utils.load(f,'tmp','t') test.assertmatch(err,'attempt to load') -- package.searchpath for Lua 5.1 -- nota bene: depends on ./?.lua being in the package.path! -- So we hack it if not found if not package.path:find '.[/\\]%?' then package.path = './?.lua;'..package.path end asserteq( package.searchpath('tests.test-utils3',package.path):gsub('\\','/'), './tests/test-utils3.lua' ) -- testing getfenv and setfenv for both interpreters function test() return X + Y + Z end t = {X = 1, Y = 2, Z = 3} setfenv(test,t) assert(test(),6) t.X = 10 assert(test(),15) local getfenv,_G = getfenv,_G function test2() local env = {x=2} setfenv(1,env) asserteq(getfenv(1),env) asserteq(x,2) end test2() Penlight-1.12.0/tests/test-xml.lua000066400000000000000000000254051416703176500170070ustar00rootroot00000000000000local xml = require 'pl.xml' local asserteq = require 'pl.test'.asserteq local dump = require 'pl.pretty'.dump -- Prosody stanza.lua style XML building d = xml.new 'top' : addtag 'child' : text 'alice' : up() : addtag 'child' : text 'bob' d = xml.new 'children' : addtag 'child' : addtag 'name' : text 'alice' : up() : addtag 'age' : text '5' : up() : addtag('toy',{type='fluffy'}) : up() : up() : addtag 'child': addtag 'name' : text 'bob' : up() : addtag 'age' : text '6' : up() : addtag('toy',{type='squeaky'}) asserteq(xml.tostring(d,'',' '), [[ alice 5 bob 6 ]]) -- Orbit-style 'xmlification' local children,child,toy,name,age = xml.tags 'children, child, toy, name, age' d1 = children { child {name 'alice', age '5', toy {type='fluffy'}}, child {name 'bob', age '6', toy {type='squeaky'}} } assert(xml.compare(d,d1)) -- or we can use a template document to convert Lua data to LOM templ = child {name '$name', age '$age', toy{type='$toy'}} d2 = children(templ:subst{ {name='alice',age='5',toy='fluffy'}, {name='bob',age='6',toy='squeaky'} }) assert(xml.compare(d1,d2)) -- Parsing Google Weather service results -- local joburg = [[ ]] -- we particularly want to test the built-in XML parser here, not lxp.lom local function parse (str) return xml.parse(str,false,true) end local d = parse(joburg) function match(t,xpect) local res,ret = d:match(t) asserteq(res,xpect,0,1) ---> note extra level, so we report on calls to this function! end t1 = [[ ]] match(t1,{ condition = "Clear", temp = "24", } ) t2 = [[ {{ }} ]] local conditions = { { low = "60", high = "89", day = "Sat", condition = "Clear", }, { low = "53", high = "86", day = "Sun", condition = "Clear", }, { low = "57", high = "87", day = "Mon", condition = "Clear", }, { low = "60", high = "84", day = "Tue", condition = "Clear", } } match(t2,conditions) config = [[ 1.3 10 bozo ]] d,err = parse(config) if not d then print(err); os.exit(1) end -- can match against wildcard tag names (end with -) -- can be names match([[ {{$value}} ]],{ {key="alpha", value = "1.3"}, {key="beta", value = "10"}, {key="name",value = "bozo"}, }) -- can be numerical indices match([[ {{<1->$2}} ]],{ {"alpha","1.3"}, {"beta","10"}, {"name","bozo"}, }) -- _ is special; means 'this value is key of captured table' match([[ {{<_->$1}} ]],{ alpha = {"1.3"}, beta = {"10"}, name = {"bozo"}, }) -- the numerical index 0 is special: a capture of {[0]=val} becomes simply the value val match([[ {{<_->$0}} ]],{ alpha = "1.3", name = "bozo", beta = "10" }) -- this can of course also work with attributes, but then we don't want to collapse! config = [[ 1.3 10 bozo ]] d,err = parse(config) if not d then print(err); os.exit(1) end match([[ {{<_- type='$1'>$2}} ]],{ alpha = {"number","1.3"}, beta = {"number","10"}, name = {"string","bozo"}, }) d,err = parse [[ ]] if not d then print(err); os.exit(1) end --xml.debug = true res,err = d:match [[ {{}} ]] asserteq(res,{ HOST = "windows-unknown-linux-gnu", COPYRIGHT = "Copyright (C) 1999-2009 ImageMagick Studio LLC", NAME = "ImageMagick", LIB_VERSION = "0x651", VERSION = "6.5.1", RELEASE_DATE = "2009-05-01", WEBSITE = "http://www.imagemagick.org", LIB_VERSION_NUMBER = "6,5,1,3", CC = "vs7", DELEGATES = "bzlib freetype jpeg jp2 lcms png tiff x11 xml wmf zlib" }) -- short excerpt from -- /usr/share/mobile-broadband-provider-info/serviceproviders.xml d = parse [[ Cell-c Cellcis 196.7.0.138 196.7.142.132 MTN 196.11.240.241 209.212.97.1 Vodacom 196.207.40.165 196.43.46.190 Unrestricted 196.207.32.69 196.43.45.190 Virgin Mobile 196.7.0.138 196.7.142.132 ]] res = d:match [[ {{ {{ $0 }} }} ]] asserteq(res,{ za = { "Cell-c", "MTN", "Vodacom", "Virgin Mobile" } }) res = d:match [[ $name 196.43.46.190 ]] asserteq(res,{ name = "Vodacom", country = "za", apn = "internet" }) d = parse[[ XXX YYY 1 ]] match([[ {{ $_ $0 }} ]],{XXX = '',YYY = '1'}) -- can always use xmlification to generate your templates... local SP, country, provider, gsm, apn, dns = xml.tags 'serviceprovider, country, provider, gsm, apn, dns' t = SP{country{code="$country",provider{ name '$name', gsm{apn {value="$apn",dns '196.43.46.190'}} }}} out = xml.tostring(t,' ',' ') asserteq(out,[[ $name 196.43.46.190 ]]) ----- HTML is a degenerate form of XML ;) -- attribute values don't need to be quoted, tags are case insensitive, -- and some are treated as self-closing doc = xml.parsehtml [[ Hello dolly
HTML is slack
]] asserteq(xml.tostring(doc),[[ Hello dolly
HTML is slack
]]) doc = xml.parsehtml [[ ]] asserteq(xml.tostring(doc),"") -- note that HTML mode currently barfs if there isn't whitespace around things -- like '<' and '>' in scripts. doc = xml.parsehtml [[

hello dammit

]] script = doc:get_elements_with_name 'script' asserteq(script[1]:get_text(), 'function less(a,b) { return a < b; }') -- test attribute order local test_attrlist = xml.new('AttrList',{ Attr3="Value3", ['Attr1'] = "Value1", ['Attr2'] = "Value2", [1] = 'Attr1', [2] = 'Attr2', [3] = 'Attr3' }) asserteq( xml.tostring(test_attrlist), "" ) -- commments str = [[ dolly ]] doc = parse(str) asserteq(xml.tostring(doc),[[ dolly ]]) -- underscores and dashes in attributes str = [[ dolly ]] doc = parse(str) print(doc) print(xml.tostring(doc)) asserteq(xml.tostring(doc),[[ dolly]])