diff-match-patch-20121119/0000750000175000017500000000000012254032721014766 5ustar lukeluke00000000000000diff-match-patch-20121119/python2/0000750000175000017500000000000012254032721016371 5ustar lukeluke00000000000000diff-match-patch-20121119/python2/diff_match_patch.egg-info/0000750000175000017500000000000012254032721023326 5ustar lukeluke00000000000000diff-match-patch-20121119/python2/diff_match_patch.egg-info/SOURCES.txt0000640000175000017500000000101712254032721025212 0ustar lukeluke00000000000000CHANGES.rst LICENCE MANIFEST.in README.original.txt README.rst setup.py python2/diff_match_patch/__init__.py python2/diff_match_patch/diff_match_patch.py python2/diff_match_patch/diff_match_patch_test.py python2/diff_match_patch.egg-info/PKG-INFO python2/diff_match_patch.egg-info/SOURCES.txt python2/diff_match_patch.egg-info/dependency_links.txt python2/diff_match_patch.egg-info/top_level.txt python3/diff_match_patch/__init__.py python3/diff_match_patch/diff_match_patch.py python3/diff_match_patch/diff_match_patch_test.pydiff-match-patch-20121119/python2/diff_match_patch.egg-info/PKG-INFO0000640000175000017500000000375412254032721024435 0ustar lukeluke00000000000000Metadata-Version: 1.1 Name: diff-match-patch Version: 20121119 Summary: The Diff Match and Patch libraries offer robust algorithms to perform the operations required for synchronizing plain text. Home-page: http://code.google.com/p/google-diff-match-patch/ Author: Neil Fraser Author-email: UNKNOWN License: Apache Description: ================ diff-match-patch ================ Documentation on the Google site: http://code.google.com/p/google-diff-match-patch/ See README.original.txt for the original README. This is an unofficial package of the Python versions for PyPI, by Luke Plant http://lukeplant.me.uk/ See https://bitbucket.org/spookylukey/diff-match-patch for packaging issues. Version 20121119 ---------------- Upstream release Version 20120106 ---------------- Upstream release Version 20111010 ---------------- Upstream release Version 20110725.1 ------------------ * Python 3 packaging bug (python3 directory excluded) * Other packaging bugs (extra files included). Version 20110725 ---------------- Initial PyPI package Platform: UNKNOWN Classifier: Topic :: Text Processing Classifier: Intended Audience :: Developers Classifier: Development Status :: 6 - Mature Classifier: License :: OSI Approved :: Apache Software License Classifier: Programming Language :: Python :: 2.4 Classifier: Programming Language :: Python :: 2.5 Classifier: Programming Language :: Python :: 2.6 Classifier: Programming Language :: Python :: 2.7 Classifier: Programming Language :: Python :: 3 Classifier: Programming Language :: Python :: 3.0 Classifier: Programming Language :: Python :: 3.1 Classifier: Programming Language :: Python :: 3.2 Classifier: Programming Language :: Python :: 3.3 diff-match-patch-20121119/python2/diff_match_patch.egg-info/top_level.txt0000640000175000017500000000002112254032721026052 0ustar lukeluke00000000000000diff_match_patch diff-match-patch-20121119/python2/diff_match_patch.egg-info/dependency_links.txt0000640000175000017500000000000112254032721027375 0ustar lukeluke00000000000000 diff-match-patch-20121119/python2/diff_match_patch/0000750000175000017500000000000012254032721021634 5ustar lukeluke00000000000000diff-match-patch-20121119/python2/diff_match_patch/diff_match_patch_test.py0000750000175000017500000012142012052423660026515 0ustar lukeluke00000000000000#!/usr/bin/python2.4 """Test harness for diff_match_patch.py Copyright 2006 Google Inc. http://code.google.com/p/google-diff-match-patch/ Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ import sys import time import unittest import diff_match_patch as dmp_module # Force a module reload. Allows one to edit the DMP module and rerun the tests # without leaving the Python interpreter. reload(dmp_module) class DiffMatchPatchTest(unittest.TestCase): def setUp(self): "Test harness for dmp_module." self.dmp = dmp_module.diff_match_patch() def diff_rebuildtexts(self, diffs): # Construct the two texts which made up the diff originally. text1 = "" text2 = "" for x in range(0, len(diffs)): if diffs[x][0] != dmp_module.diff_match_patch.DIFF_INSERT: text1 += diffs[x][1] if diffs[x][0] != dmp_module.diff_match_patch.DIFF_DELETE: text2 += diffs[x][1] return (text1, text2) class DiffTest(DiffMatchPatchTest): """DIFF TEST FUNCTIONS""" def testDiffCommonPrefix(self): # Detect any common prefix. # Null case. self.assertEquals(0, self.dmp.diff_commonPrefix("abc", "xyz")) # Non-null case. self.assertEquals(4, self.dmp.diff_commonPrefix("1234abcdef", "1234xyz")) # Whole case. self.assertEquals(4, self.dmp.diff_commonPrefix("1234", "1234xyz")) def testDiffCommonSuffix(self): # Detect any common suffix. # Null case. self.assertEquals(0, self.dmp.diff_commonSuffix("abc", "xyz")) # Non-null case. self.assertEquals(4, self.dmp.diff_commonSuffix("abcdef1234", "xyz1234")) # Whole case. self.assertEquals(4, self.dmp.diff_commonSuffix("1234", "xyz1234")) def testDiffCommonOverlap(self): # Null case. self.assertEquals(0, self.dmp.diff_commonOverlap("", "abcd")) # Whole case. self.assertEquals(3, self.dmp.diff_commonOverlap("abc", "abcd")) # No overlap. self.assertEquals(0, self.dmp.diff_commonOverlap("123456", "abcd")) # Overlap. self.assertEquals(3, self.dmp.diff_commonOverlap("123456xxx", "xxxabcd")) # Unicode. # Some overly clever languages (C#) may treat ligatures as equal to their # component letters. E.g. U+FB01 == 'fi' self.assertEquals(0, self.dmp.diff_commonOverlap("fi", u"\ufb01i")) def testDiffHalfMatch(self): # Detect a halfmatch. self.dmp.Diff_Timeout = 1 # No match. self.assertEquals(None, self.dmp.diff_halfMatch("1234567890", "abcdef")) self.assertEquals(None, self.dmp.diff_halfMatch("12345", "23")) # Single Match. self.assertEquals(("12", "90", "a", "z", "345678"), self.dmp.diff_halfMatch("1234567890", "a345678z")) self.assertEquals(("a", "z", "12", "90", "345678"), self.dmp.diff_halfMatch("a345678z", "1234567890")) self.assertEquals(("abc", "z", "1234", "0", "56789"), self.dmp.diff_halfMatch("abc56789z", "1234567890")) self.assertEquals(("a", "xyz", "1", "7890", "23456"), self.dmp.diff_halfMatch("a23456xyz", "1234567890")) # Multiple Matches. self.assertEquals(("12123", "123121", "a", "z", "1234123451234"), self.dmp.diff_halfMatch("121231234123451234123121", "a1234123451234z")) self.assertEquals(("", "-=-=-=-=-=", "x", "", "x-=-=-=-=-=-=-="), self.dmp.diff_halfMatch("x-=-=-=-=-=-=-=-=-=-=-=-=", "xx-=-=-=-=-=-=-=")) self.assertEquals(("-=-=-=-=-=", "", "", "y", "-=-=-=-=-=-=-=y"), self.dmp.diff_halfMatch("-=-=-=-=-=-=-=-=-=-=-=-=y", "-=-=-=-=-=-=-=yy")) # Non-optimal halfmatch. # Optimal diff would be -q+x=H-i+e=lloHe+Hu=llo-Hew+y not -qHillo+x=HelloHe-w+Hulloy self.assertEquals(("qHillo", "w", "x", "Hulloy", "HelloHe"), self.dmp.diff_halfMatch("qHilloHelloHew", "xHelloHeHulloy")) # Optimal no halfmatch. self.dmp.Diff_Timeout = 0 self.assertEquals(None, self.dmp.diff_halfMatch("qHilloHelloHew", "xHelloHeHulloy")) def testDiffLinesToChars(self): # Convert lines down to characters. self.assertEquals(("\x01\x02\x01", "\x02\x01\x02", ["", "alpha\n", "beta\n"]), self.dmp.diff_linesToChars("alpha\nbeta\nalpha\n", "beta\nalpha\nbeta\n")) self.assertEquals(("", "\x01\x02\x03\x03", ["", "alpha\r\n", "beta\r\n", "\r\n"]), self.dmp.diff_linesToChars("", "alpha\r\nbeta\r\n\r\n\r\n")) self.assertEquals(("\x01", "\x02", ["", "a", "b"]), self.dmp.diff_linesToChars("a", "b")) # More than 256 to reveal any 8-bit limitations. n = 300 lineList = [] charList = [] for x in range(1, n + 1): lineList.append(str(x) + "\n") charList.append(unichr(x)) self.assertEquals(n, len(lineList)) lines = "".join(lineList) chars = "".join(charList) self.assertEquals(n, len(chars)) lineList.insert(0, "") self.assertEquals((chars, "", lineList), self.dmp.diff_linesToChars(lines, "")) def testDiffCharsToLines(self): # Convert chars up to lines. diffs = [(self.dmp.DIFF_EQUAL, "\x01\x02\x01"), (self.dmp.DIFF_INSERT, "\x02\x01\x02")] self.dmp.diff_charsToLines(diffs, ["", "alpha\n", "beta\n"]) self.assertEquals([(self.dmp.DIFF_EQUAL, "alpha\nbeta\nalpha\n"), (self.dmp.DIFF_INSERT, "beta\nalpha\nbeta\n")], diffs) # More than 256 to reveal any 8-bit limitations. n = 300 lineList = [] charList = [] for x in range(1, n + 1): lineList.append(str(x) + "\n") charList.append(unichr(x)) self.assertEquals(n, len(lineList)) lines = "".join(lineList) chars = "".join(charList) self.assertEquals(n, len(chars)) lineList.insert(0, "") diffs = [(self.dmp.DIFF_DELETE, chars)] self.dmp.diff_charsToLines(diffs, lineList) self.assertEquals([(self.dmp.DIFF_DELETE, lines)], diffs) def testDiffCleanupMerge(self): # Cleanup a messy diff. # Null case. diffs = [] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([], diffs) # No change case. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_INSERT, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_INSERT, "c")], diffs) # Merge equalities. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_EQUAL, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "abc")], diffs) # Merge deletions. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_DELETE, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abc")], diffs) # Merge insertions. diffs = [(self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_INSERT, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_INSERT, "abc")], diffs) # Merge interweave. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_DELETE, "c"), (self.dmp.DIFF_INSERT, "d"), (self.dmp.DIFF_EQUAL, "e"), (self.dmp.DIFF_EQUAL, "f")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "ac"), (self.dmp.DIFF_INSERT, "bd"), (self.dmp.DIFF_EQUAL, "ef")], diffs) # Prefix and suffix detection. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "abc"), (self.dmp.DIFF_DELETE, "dc")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "d"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_EQUAL, "c")], diffs) # Prefix and suffix detection with equalities. diffs = [(self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "abc"), (self.dmp.DIFF_DELETE, "dc"), (self.dmp.DIFF_EQUAL, "y")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "xa"), (self.dmp.DIFF_DELETE, "d"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_EQUAL, "cy")], diffs) # Slide edit left. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_INSERT, "ba"), (self.dmp.DIFF_EQUAL, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_INSERT, "ab"), (self.dmp.DIFF_EQUAL, "ac")], diffs) # Slide edit right. diffs = [(self.dmp.DIFF_EQUAL, "c"), (self.dmp.DIFF_INSERT, "ab"), (self.dmp.DIFF_EQUAL, "a")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "ca"), (self.dmp.DIFF_INSERT, "ba")], diffs) # Slide edit left recursive. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_EQUAL, "c"), (self.dmp.DIFF_DELETE, "ac"), (self.dmp.DIFF_EQUAL, "x")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_EQUAL, "acx")], diffs) # Slide edit right recursive. diffs = [(self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "ca"), (self.dmp.DIFF_EQUAL, "c"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_EQUAL, "a")] self.dmp.diff_cleanupMerge(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "xca"), (self.dmp.DIFF_DELETE, "cba")], diffs) def testDiffCleanupSemanticLossless(self): # Slide diffs to match logical boundaries. # Null case. diffs = [] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([], diffs) # Blank lines. diffs = [(self.dmp.DIFF_EQUAL, "AAA\r\n\r\nBBB"), (self.dmp.DIFF_INSERT, "\r\nDDD\r\n\r\nBBB"), (self.dmp.DIFF_EQUAL, "\r\nEEE")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "AAA\r\n\r\n"), (self.dmp.DIFF_INSERT, "BBB\r\nDDD\r\n\r\n"), (self.dmp.DIFF_EQUAL, "BBB\r\nEEE")], diffs) # Line boundaries. diffs = [(self.dmp.DIFF_EQUAL, "AAA\r\nBBB"), (self.dmp.DIFF_INSERT, " DDD\r\nBBB"), (self.dmp.DIFF_EQUAL, " EEE")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "AAA\r\n"), (self.dmp.DIFF_INSERT, "BBB DDD\r\n"), (self.dmp.DIFF_EQUAL, "BBB EEE")], diffs) # Word boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The c"), (self.dmp.DIFF_INSERT, "ow and the c"), (self.dmp.DIFF_EQUAL, "at.")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "The "), (self.dmp.DIFF_INSERT, "cow and the "), (self.dmp.DIFF_EQUAL, "cat.")], diffs) # Alphanumeric boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The-c"), (self.dmp.DIFF_INSERT, "ow-and-the-c"), (self.dmp.DIFF_EQUAL, "at.")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "The-"), (self.dmp.DIFF_INSERT, "cow-and-the-"), (self.dmp.DIFF_EQUAL, "cat.")], diffs) # Hitting the start. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "ax")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "aax")], diffs) # Hitting the end. diffs = [(self.dmp.DIFF_EQUAL, "xa"), (self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "a")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "xaa"), (self.dmp.DIFF_DELETE, "a")], diffs) # Sentence boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The xxx. The "), (self.dmp.DIFF_INSERT, "zzz. The "), (self.dmp.DIFF_EQUAL, "yyy.")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "The xxx."), (self.dmp.DIFF_INSERT, " The zzz."), (self.dmp.DIFF_EQUAL, " The yyy.")], diffs) def testDiffCleanupSemantic(self): # Cleanup semantically trivial equalities. # Null case. diffs = [] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([], diffs) # No elimination #1. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "cd"), (self.dmp.DIFF_EQUAL, "12"), (self.dmp.DIFF_DELETE, "e")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "cd"), (self.dmp.DIFF_EQUAL, "12"), (self.dmp.DIFF_DELETE, "e")], diffs) # No elimination #2. diffs = [(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "ABC"), (self.dmp.DIFF_EQUAL, "1234"), (self.dmp.DIFF_DELETE, "wxyz")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "ABC"), (self.dmp.DIFF_EQUAL, "1234"), (self.dmp.DIFF_DELETE, "wxyz")], diffs) # Simple elimination. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_DELETE, "c")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "b")], diffs) # Backpass elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_EQUAL, "cd"), (self.dmp.DIFF_DELETE, "e"), (self.dmp.DIFF_EQUAL, "f"), (self.dmp.DIFF_INSERT, "g")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abcdef"), (self.dmp.DIFF_INSERT, "cdfg")], diffs) # Multiple eliminations. diffs = [(self.dmp.DIFF_INSERT, "1"), (self.dmp.DIFF_EQUAL, "A"), (self.dmp.DIFF_DELETE, "B"), (self.dmp.DIFF_INSERT, "2"), (self.dmp.DIFF_EQUAL, "_"), (self.dmp.DIFF_INSERT, "1"), (self.dmp.DIFF_EQUAL, "A"), (self.dmp.DIFF_DELETE, "B"), (self.dmp.DIFF_INSERT, "2")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "AB_AB"), (self.dmp.DIFF_INSERT, "1A2_1A2")], diffs) # Word boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The c"), (self.dmp.DIFF_DELETE, "ow and the c"), (self.dmp.DIFF_EQUAL, "at.")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_EQUAL, "The "), (self.dmp.DIFF_DELETE, "cow and the "), (self.dmp.DIFF_EQUAL, "cat.")], diffs) # No overlap elimination. diffs = [(self.dmp.DIFF_DELETE, "abcxx"), (self.dmp.DIFF_INSERT, "xxdef")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abcxx"), (self.dmp.DIFF_INSERT, "xxdef")], diffs) # Overlap elimination. diffs = [(self.dmp.DIFF_DELETE, "abcxxx"), (self.dmp.DIFF_INSERT, "xxxdef")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_EQUAL, "xxx"), (self.dmp.DIFF_INSERT, "def")], diffs) # Reverse overlap elimination. diffs = [(self.dmp.DIFF_DELETE, "xxxabc"), (self.dmp.DIFF_INSERT, "defxxx")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_INSERT, "def"), (self.dmp.DIFF_EQUAL, "xxx"), (self.dmp.DIFF_DELETE, "abc")], diffs) # Two overlap eliminations. diffs = [(self.dmp.DIFF_DELETE, "abcd1212"), (self.dmp.DIFF_INSERT, "1212efghi"), (self.dmp.DIFF_EQUAL, "----"), (self.dmp.DIFF_DELETE, "A3"), (self.dmp.DIFF_INSERT, "3BC")] self.dmp.diff_cleanupSemantic(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abcd"), (self.dmp.DIFF_EQUAL, "1212"), (self.dmp.DIFF_INSERT, "efghi"), (self.dmp.DIFF_EQUAL, "----"), (self.dmp.DIFF_DELETE, "A"), (self.dmp.DIFF_EQUAL, "3"), (self.dmp.DIFF_INSERT, "BC")], diffs) def testDiffCleanupEfficiency(self): # Cleanup operationally trivial equalities. self.dmp.Diff_EditCost = 4 # Null case. diffs = [] self.dmp.diff_cleanupEfficiency(diffs) self.assertEquals([], diffs) # No elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "wxyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "wxyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")], diffs) # Four-edit elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "xyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abxyzcd"), (self.dmp.DIFF_INSERT, "12xyz34")], diffs) # Three-edit elimination. diffs = [(self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "xcd"), (self.dmp.DIFF_INSERT, "12x34")], diffs) # Backpass elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "xy"), (self.dmp.DIFF_INSERT, "34"), (self.dmp.DIFF_EQUAL, "z"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "56")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abxyzcd"), (self.dmp.DIFF_INSERT, "12xy34z56")], diffs) # High cost elimination. self.dmp.Diff_EditCost = 5 diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "wxyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEquals([(self.dmp.DIFF_DELETE, "abwxyzcd"), (self.dmp.DIFF_INSERT, "12wxyz34")], diffs) self.dmp.Diff_EditCost = 4 def testDiffPrettyHtml(self): # Pretty print. diffs = [(self.dmp.DIFF_EQUAL, "a\n"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_INSERT, "c&d")] self.assertEquals("
<B>b</B>c&d", self.dmp.diff_prettyHtml(diffs)) def testDiffText(self): # Compute the source and destination texts. diffs = [(self.dmp.DIFF_EQUAL, "jump"), (self.dmp.DIFF_DELETE, "s"), (self.dmp.DIFF_INSERT, "ed"), (self.dmp.DIFF_EQUAL, " over "), (self.dmp.DIFF_DELETE, "the"), (self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_EQUAL, " lazy")] self.assertEquals("jumps over the lazy", self.dmp.diff_text1(diffs)) self.assertEquals("jumped over a lazy", self.dmp.diff_text2(diffs)) def testDiffDelta(self): # Convert a diff into delta string. diffs = [(self.dmp.DIFF_EQUAL, "jump"), (self.dmp.DIFF_DELETE, "s"), (self.dmp.DIFF_INSERT, "ed"), (self.dmp.DIFF_EQUAL, " over "), (self.dmp.DIFF_DELETE, "the"), (self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_EQUAL, " lazy"), (self.dmp.DIFF_INSERT, "old dog")] text1 = self.dmp.diff_text1(diffs) self.assertEquals("jumps over the lazy", text1) delta = self.dmp.diff_toDelta(diffs) self.assertEquals("=4\t-1\t+ed\t=6\t-3\t+a\t=5\t+old dog", delta) # Convert delta string into a diff. self.assertEquals(diffs, self.dmp.diff_fromDelta(text1, delta)) # Generates error (19 != 20). try: self.dmp.diff_fromDelta(text1 + "x", delta) self.assertFalse(True) except ValueError: # Exception expected. pass # Generates error (19 != 18). try: self.dmp.diff_fromDelta(text1[1:], delta) self.assertFalse(True) except ValueError: # Exception expected. pass # Generates error (%c3%xy invalid Unicode). try: self.dmp.diff_fromDelta("", "+%c3xy") self.assertFalse(True) except ValueError: # Exception expected. pass # Test deltas with special characters. diffs = [(self.dmp.DIFF_EQUAL, u"\u0680 \x00 \t %"), (self.dmp.DIFF_DELETE, u"\u0681 \x01 \n ^"), (self.dmp.DIFF_INSERT, u"\u0682 \x02 \\ |")] text1 = self.dmp.diff_text1(diffs) self.assertEquals(u"\u0680 \x00 \t %\u0681 \x01 \n ^", text1) delta = self.dmp.diff_toDelta(diffs) self.assertEquals("=7\t-7\t+%DA%82 %02 %5C %7C", delta) # Convert delta string into a diff. self.assertEquals(diffs, self.dmp.diff_fromDelta(text1, delta)) # Verify pool of unchanged characters. diffs = [(self.dmp.DIFF_INSERT, "A-Z a-z 0-9 - _ . ! ~ * ' ( ) ; / ? : @ & = + $ , # ")] text2 = self.dmp.diff_text2(diffs) self.assertEquals("A-Z a-z 0-9 - _ . ! ~ * \' ( ) ; / ? : @ & = + $ , # ", text2) delta = self.dmp.diff_toDelta(diffs) self.assertEquals("+A-Z a-z 0-9 - _ . ! ~ * \' ( ) ; / ? : @ & = + $ , # ", delta) # Convert delta string into a diff. self.assertEquals(diffs, self.dmp.diff_fromDelta("", delta)) def testDiffXIndex(self): # Translate a location in text1 to text2. self.assertEquals(5, self.dmp.diff_xIndex([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "1234"), (self.dmp.DIFF_EQUAL, "xyz")], 2)) # Translation on deletion. self.assertEquals(1, self.dmp.diff_xIndex([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "1234"), (self.dmp.DIFF_EQUAL, "xyz")], 3)) def testDiffLevenshtein(self): # Levenshtein with trailing equality. self.assertEquals(4, self.dmp.diff_levenshtein([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "1234"), (self.dmp.DIFF_EQUAL, "xyz")])) # Levenshtein with leading equality. self.assertEquals(4, self.dmp.diff_levenshtein([(self.dmp.DIFF_EQUAL, "xyz"), (self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "1234")])) # Levenshtein with middle equality. self.assertEquals(7, self.dmp.diff_levenshtein([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_EQUAL, "xyz"), (self.dmp.DIFF_INSERT, "1234")])) def testDiffBisect(self): # Normal. a = "cat" b = "map" # Since the resulting diff hasn't been normalized, it would be ok if # the insertion and deletion pairs are swapped. # If the order changes, tweak this test as required. self.assertEquals([(self.dmp.DIFF_DELETE, "c"), (self.dmp.DIFF_INSERT, "m"), (self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "t"), (self.dmp.DIFF_INSERT, "p")], self.dmp.diff_bisect(a, b, sys.maxint)) # Timeout. self.assertEquals([(self.dmp.DIFF_DELETE, "cat"), (self.dmp.DIFF_INSERT, "map")], self.dmp.diff_bisect(a, b, 0)) def testDiffMain(self): # Perform a trivial diff. # Null case. self.assertEquals([], self.dmp.diff_main("", "", False)) # Equality. self.assertEquals([(self.dmp.DIFF_EQUAL, "abc")], self.dmp.diff_main("abc", "abc", False)) # Simple insertion. self.assertEquals([(self.dmp.DIFF_EQUAL, "ab"), (self.dmp.DIFF_INSERT, "123"), (self.dmp.DIFF_EQUAL, "c")], self.dmp.diff_main("abc", "ab123c", False)) # Simple deletion. self.assertEquals([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "123"), (self.dmp.DIFF_EQUAL, "bc")], self.dmp.diff_main("a123bc", "abc", False)) # Two insertions. self.assertEquals([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_INSERT, "123"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_INSERT, "456"), (self.dmp.DIFF_EQUAL, "c")], self.dmp.diff_main("abc", "a123b456c", False)) # Two deletions. self.assertEquals([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "123"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_DELETE, "456"), (self.dmp.DIFF_EQUAL, "c")], self.dmp.diff_main("a123b456c", "abc", False)) # Perform a real diff. # Switch off the timeout. self.dmp.Diff_Timeout = 0 # Simple cases. self.assertEquals([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "b")], self.dmp.diff_main("a", "b", False)) self.assertEquals([(self.dmp.DIFF_DELETE, "Apple"), (self.dmp.DIFF_INSERT, "Banana"), (self.dmp.DIFF_EQUAL, "s are a"), (self.dmp.DIFF_INSERT, "lso"), (self.dmp.DIFF_EQUAL, " fruit.")], self.dmp.diff_main("Apples are a fruit.", "Bananas are also fruit.", False)) self.assertEquals([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, u"\u0680"), (self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "\t"), (self.dmp.DIFF_INSERT, "\x00")], self.dmp.diff_main("ax\t", u"\u0680x\x00", False)) # Overlaps. self.assertEquals([(self.dmp.DIFF_DELETE, "1"), (self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "y"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_DELETE, "2"), (self.dmp.DIFF_INSERT, "xab")], self.dmp.diff_main("1ayb2", "abxab", False)) self.assertEquals([(self.dmp.DIFF_INSERT, "xaxcx"), (self.dmp.DIFF_EQUAL, "abc"), (self.dmp.DIFF_DELETE, "y")], self.dmp.diff_main("abcy", "xaxcxabc", False)) self.assertEquals([(self.dmp.DIFF_DELETE, "ABCD"), (self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "="), (self.dmp.DIFF_INSERT, "-"), (self.dmp.DIFF_EQUAL, "bcd"), (self.dmp.DIFF_DELETE, "="), (self.dmp.DIFF_INSERT, "-"), (self.dmp.DIFF_EQUAL, "efghijklmnopqrs"), (self.dmp.DIFF_DELETE, "EFGHIJKLMNOefg")], self.dmp.diff_main("ABCDa=bcd=efghijklmnopqrsEFGHIJKLMNOefg", "a-bcd-efghijklmnopqrs", False)) # Large equality. self.assertEquals([(self.dmp.DIFF_INSERT, " "), (self.dmp.DIFF_EQUAL,"a"), (self.dmp.DIFF_INSERT,"nd"), (self.dmp.DIFF_EQUAL," [[Pennsylvania]]"), (self.dmp.DIFF_DELETE," and [[New")], self.dmp.diff_main("a [[Pennsylvania]] and [[New", " and [[Pennsylvania]]", False)) # Timeout. self.dmp.Diff_Timeout = 0.1 # 100ms a = "`Twas brillig, and the slithy toves\nDid gyre and gimble in the wabe:\nAll mimsy were the borogoves,\nAnd the mome raths outgrabe.\n" b = "I am the very model of a modern major general,\nI've information vegetable, animal, and mineral,\nI know the kings of England, and I quote the fights historical,\nFrom Marathon to Waterloo, in order categorical.\n" # Increase the text lengths by 1024 times to ensure a timeout. for x in range(10): a = a + a b = b + b startTime = time.time() self.dmp.diff_main(a, b) endTime = time.time() # Test that we took at least the timeout period. self.assertTrue(self.dmp.Diff_Timeout <= endTime - startTime) # Test that we didn't take forever (be forgiving). # Theoretically this test could fail very occasionally if the # OS task swaps or locks up for a second at the wrong moment. self.assertTrue(self.dmp.Diff_Timeout * 2 > endTime - startTime) self.dmp.Diff_Timeout = 0 # Test the linemode speedup. # Must be long to pass the 100 char cutoff. # Simple line-mode. a = "1234567890\n" * 13 b = "abcdefghij\n" * 13 self.assertEquals(self.dmp.diff_main(a, b, False), self.dmp.diff_main(a, b, True)) # Single line-mode. a = "1234567890" * 13 b = "abcdefghij" * 13 self.assertEquals(self.dmp.diff_main(a, b, False), self.dmp.diff_main(a, b, True)) # Overlap line-mode. a = "1234567890\n" * 13 b = "abcdefghij\n1234567890\n1234567890\n1234567890\nabcdefghij\n1234567890\n1234567890\n1234567890\nabcdefghij\n1234567890\n1234567890\n1234567890\nabcdefghij\n" texts_linemode = self.diff_rebuildtexts(self.dmp.diff_main(a, b, True)) texts_textmode = self.diff_rebuildtexts(self.dmp.diff_main(a, b, False)) self.assertEquals(texts_textmode, texts_linemode) # Test null inputs. try: self.dmp.diff_main(None, None) self.assertFalse(True) except ValueError: # Exception expected. pass class MatchTest(DiffMatchPatchTest): """MATCH TEST FUNCTIONS""" def testMatchAlphabet(self): # Initialise the bitmasks for Bitap. self.assertEquals({"a":4, "b":2, "c":1}, self.dmp.match_alphabet("abc")) self.assertEquals({"a":37, "b":18, "c":8}, self.dmp.match_alphabet("abcaba")) def testMatchBitap(self): self.dmp.Match_Distance = 100 self.dmp.Match_Threshold = 0.5 # Exact matches. self.assertEquals(5, self.dmp.match_bitap("abcdefghijk", "fgh", 5)) self.assertEquals(5, self.dmp.match_bitap("abcdefghijk", "fgh", 0)) # Fuzzy matches. self.assertEquals(4, self.dmp.match_bitap("abcdefghijk", "efxhi", 0)) self.assertEquals(2, self.dmp.match_bitap("abcdefghijk", "cdefxyhijk", 5)) self.assertEquals(-1, self.dmp.match_bitap("abcdefghijk", "bxy", 1)) # Overflow. self.assertEquals(2, self.dmp.match_bitap("123456789xx0", "3456789x0", 2)) self.assertEquals(0, self.dmp.match_bitap("abcdef", "xxabc", 4)) self.assertEquals(3, self.dmp.match_bitap("abcdef", "defyy", 4)) self.assertEquals(0, self.dmp.match_bitap("abcdef", "xabcdefy", 0)) # Threshold test. self.dmp.Match_Threshold = 0.4 self.assertEquals(4, self.dmp.match_bitap("abcdefghijk", "efxyhi", 1)) self.dmp.Match_Threshold = 0.3 self.assertEquals(-1, self.dmp.match_bitap("abcdefghijk", "efxyhi", 1)) self.dmp.Match_Threshold = 0.0 self.assertEquals(1, self.dmp.match_bitap("abcdefghijk", "bcdef", 1)) self.dmp.Match_Threshold = 0.5 # Multiple select. self.assertEquals(0, self.dmp.match_bitap("abcdexyzabcde", "abccde", 3)) self.assertEquals(8, self.dmp.match_bitap("abcdexyzabcde", "abccde", 5)) # Distance test. self.dmp.Match_Distance = 10 # Strict location. self.assertEquals(-1, self.dmp.match_bitap("abcdefghijklmnopqrstuvwxyz", "abcdefg", 24)) self.assertEquals(0, self.dmp.match_bitap("abcdefghijklmnopqrstuvwxyz", "abcdxxefg", 1)) self.dmp.Match_Distance = 1000 # Loose location. self.assertEquals(0, self.dmp.match_bitap("abcdefghijklmnopqrstuvwxyz", "abcdefg", 24)) def testMatchMain(self): # Full match. # Shortcut matches. self.assertEquals(0, self.dmp.match_main("abcdef", "abcdef", 1000)) self.assertEquals(-1, self.dmp.match_main("", "abcdef", 1)) self.assertEquals(3, self.dmp.match_main("abcdef", "", 3)) self.assertEquals(3, self.dmp.match_main("abcdef", "de", 3)) self.assertEquals(3, self.dmp.match_main("abcdef", "defy", 4)) self.assertEquals(0, self.dmp.match_main("abcdef", "abcdefy", 0)) # Complex match. self.dmp.Match_Threshold = 0.7 self.assertEquals(4, self.dmp.match_main("I am the very model of a modern major general.", " that berry ", 5)) self.dmp.Match_Threshold = 0.5 # Test null inputs. try: self.dmp.match_main(None, None, 0) self.assertFalse(True) except ValueError: # Exception expected. pass class PatchTest(DiffMatchPatchTest): """PATCH TEST FUNCTIONS""" def testPatchObj(self): # Patch Object. p = dmp_module.patch_obj() p.start1 = 20 p.start2 = 21 p.length1 = 18 p.length2 = 17 p.diffs = [(self.dmp.DIFF_EQUAL, "jump"), (self.dmp.DIFF_DELETE, "s"), (self.dmp.DIFF_INSERT, "ed"), (self.dmp.DIFF_EQUAL, " over "), (self.dmp.DIFF_DELETE, "the"), (self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_EQUAL, "\nlaz")] strp = str(p) self.assertEquals("@@ -21,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n %0Alaz\n", strp) def testPatchFromText(self): self.assertEquals([], self.dmp.patch_fromText("")) strp = "@@ -21,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n %0Alaz\n" self.assertEquals(strp, str(self.dmp.patch_fromText(strp)[0])) self.assertEquals("@@ -1 +1 @@\n-a\n+b\n", str(self.dmp.patch_fromText("@@ -1 +1 @@\n-a\n+b\n")[0])) self.assertEquals("@@ -1,3 +0,0 @@\n-abc\n", str(self.dmp.patch_fromText("@@ -1,3 +0,0 @@\n-abc\n")[0])) self.assertEquals("@@ -0,0 +1,3 @@\n+abc\n", str(self.dmp.patch_fromText("@@ -0,0 +1,3 @@\n+abc\n")[0])) # Generates error. try: self.dmp.patch_fromText("Bad\nPatch\n") self.assertFalse(True) except ValueError: # Exception expected. pass def testPatchToText(self): strp = "@@ -21,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n laz\n" p = self.dmp.patch_fromText(strp) self.assertEquals(strp, self.dmp.patch_toText(p)) strp = "@@ -1,9 +1,9 @@\n-f\n+F\n oo+fooba\n@@ -7,9 +7,9 @@\n obar\n-,\n+.\n tes\n" p = self.dmp.patch_fromText(strp) self.assertEquals(strp, self.dmp.patch_toText(p)) def testPatchAddContext(self): self.dmp.Patch_Margin = 4 p = self.dmp.patch_fromText("@@ -21,4 +21,10 @@\n-jump\n+somersault\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps over the lazy dog.") self.assertEquals("@@ -17,12 +17,18 @@\n fox \n-jump\n+somersault\n s ov\n", str(p)) # Same, but not enough trailing context. p = self.dmp.patch_fromText("@@ -21,4 +21,10 @@\n-jump\n+somersault\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps.") self.assertEquals("@@ -17,10 +17,16 @@\n fox \n-jump\n+somersault\n s.\n", str(p)) # Same, but not enough leading context. p = self.dmp.patch_fromText("@@ -3 +3,2 @@\n-e\n+at\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps.") self.assertEquals("@@ -1,7 +1,8 @@\n Th\n-e\n+at\n qui\n", str(p)) # Same, but with ambiguity. p = self.dmp.patch_fromText("@@ -3 +3,2 @@\n-e\n+at\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps. The quick brown fox crashes.") self.assertEquals("@@ -1,27 +1,28 @@\n Th\n-e\n+at\n quick brown fox jumps. \n", str(p)) def testPatchMake(self): # Null case. patches = self.dmp.patch_make("", "") self.assertEquals("", self.dmp.patch_toText(patches)) text1 = "The quick brown fox jumps over the lazy dog." text2 = "That quick brown fox jumped over a lazy dog." # Text2+Text1 inputs. expectedPatch = "@@ -1,8 +1,7 @@\n Th\n-at\n+e\n qui\n@@ -21,17 +21,18 @@\n jump\n-ed\n+s\n over \n-a\n+the\n laz\n" # The second patch must be "-21,17 +21,18", not "-22,17 +21,18" due to rolling context. patches = self.dmp.patch_make(text2, text1) self.assertEquals(expectedPatch, self.dmp.patch_toText(patches)) # Text1+Text2 inputs. expectedPatch = "@@ -1,11 +1,12 @@\n Th\n-e\n+at\n quick b\n@@ -22,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n laz\n" patches = self.dmp.patch_make(text1, text2) self.assertEquals(expectedPatch, self.dmp.patch_toText(patches)) # Diff input. diffs = self.dmp.diff_main(text1, text2, False) patches = self.dmp.patch_make(diffs) self.assertEquals(expectedPatch, self.dmp.patch_toText(patches)) # Text1+Diff inputs. patches = self.dmp.patch_make(text1, diffs) self.assertEquals(expectedPatch, self.dmp.patch_toText(patches)) # Text1+Text2+Diff inputs (deprecated). patches = self.dmp.patch_make(text1, text2, diffs) self.assertEquals(expectedPatch, self.dmp.patch_toText(patches)) # Character encoding. patches = self.dmp.patch_make("`1234567890-=[]\\;',./", "~!@#$%^&*()_+{}|:\"<>?") self.assertEquals("@@ -1,21 +1,21 @@\n-%601234567890-=%5B%5D%5C;',./\n+~!@#$%25%5E&*()_+%7B%7D%7C:%22%3C%3E?\n", self.dmp.patch_toText(patches)) # Character decoding. diffs = [(self.dmp.DIFF_DELETE, "`1234567890-=[]\\;',./"), (self.dmp.DIFF_INSERT, "~!@#$%^&*()_+{}|:\"<>?")] self.assertEquals(diffs, self.dmp.patch_fromText("@@ -1,21 +1,21 @@\n-%601234567890-=%5B%5D%5C;',./\n+~!@#$%25%5E&*()_+%7B%7D%7C:%22%3C%3E?\n")[0].diffs) # Long string with repeats. text1 = "" for x in range(100): text1 += "abcdef" text2 = text1 + "123" expectedPatch = "@@ -573,28 +573,31 @@\n cdefabcdefabcdefabcdefabcdef\n+123\n" patches = self.dmp.patch_make(text1, text2) self.assertEquals(expectedPatch, self.dmp.patch_toText(patches)) # Test null inputs. try: self.dmp.patch_make(None, None) self.assertFalse(True) except ValueError: # Exception expected. pass def testPatchSplitMax(self): # Assumes that Match_MaxBits is 32. patches = self.dmp.patch_make("abcdefghijklmnopqrstuvwxyz01234567890", "XabXcdXefXghXijXklXmnXopXqrXstXuvXwxXyzX01X23X45X67X89X0") self.dmp.patch_splitMax(patches) self.assertEquals("@@ -1,32 +1,46 @@\n+X\n ab\n+X\n cd\n+X\n ef\n+X\n gh\n+X\n ij\n+X\n kl\n+X\n mn\n+X\n op\n+X\n qr\n+X\n st\n+X\n uv\n+X\n wx\n+X\n yz\n+X\n 012345\n@@ -25,13 +39,18 @@\n zX01\n+X\n 23\n+X\n 45\n+X\n 67\n+X\n 89\n+X\n 0\n", self.dmp.patch_toText(patches)) patches = self.dmp.patch_make("abcdef1234567890123456789012345678901234567890123456789012345678901234567890uvwxyz", "abcdefuvwxyz") oldToText = self.dmp.patch_toText(patches) self.dmp.patch_splitMax(patches) self.assertEquals(oldToText, self.dmp.patch_toText(patches)) patches = self.dmp.patch_make("1234567890123456789012345678901234567890123456789012345678901234567890", "abc") self.dmp.patch_splitMax(patches) self.assertEquals("@@ -1,32 +1,4 @@\n-1234567890123456789012345678\n 9012\n@@ -29,32 +1,4 @@\n-9012345678901234567890123456\n 7890\n@@ -57,14 +1,3 @@\n-78901234567890\n+abc\n", self.dmp.patch_toText(patches)) patches = self.dmp.patch_make("abcdefghij , h : 0 , t : 1 abcdefghij , h : 0 , t : 1 abcdefghij , h : 0 , t : 1", "abcdefghij , h : 1 , t : 1 abcdefghij , h : 1 , t : 1 abcdefghij , h : 0 , t : 1") self.dmp.patch_splitMax(patches) self.assertEquals("@@ -2,32 +2,32 @@\n bcdefghij , h : \n-0\n+1\n , t : 1 abcdef\n@@ -29,32 +29,32 @@\n bcdefghij , h : \n-0\n+1\n , t : 1 abcdef\n", self.dmp.patch_toText(patches)) def testPatchAddPadding(self): # Both edges full. patches = self.dmp.patch_make("", "test") self.assertEquals("@@ -0,0 +1,4 @@\n+test\n", self.dmp.patch_toText(patches)) self.dmp.patch_addPadding(patches) self.assertEquals("@@ -1,8 +1,12 @@\n %01%02%03%04\n+test\n %01%02%03%04\n", self.dmp.patch_toText(patches)) # Both edges partial. patches = self.dmp.patch_make("XY", "XtestY") self.assertEquals("@@ -1,2 +1,6 @@\n X\n+test\n Y\n", self.dmp.patch_toText(patches)) self.dmp.patch_addPadding(patches) self.assertEquals("@@ -2,8 +2,12 @@\n %02%03%04X\n+test\n Y%01%02%03\n", self.dmp.patch_toText(patches)) # Both edges none. patches = self.dmp.patch_make("XXXXYYYY", "XXXXtestYYYY") self.assertEquals("@@ -1,8 +1,12 @@\n XXXX\n+test\n YYYY\n", self.dmp.patch_toText(patches)) self.dmp.patch_addPadding(patches) self.assertEquals("@@ -5,8 +5,12 @@\n XXXX\n+test\n YYYY\n", self.dmp.patch_toText(patches)) def testPatchApply(self): self.dmp.Match_Distance = 1000 self.dmp.Match_Threshold = 0.5 self.dmp.Patch_DeleteThreshold = 0.5 # Null case. patches = self.dmp.patch_make("", "") results = self.dmp.patch_apply(patches, "Hello world.") self.assertEquals(("Hello world.", []), results) # Exact match. patches = self.dmp.patch_make("The quick brown fox jumps over the lazy dog.", "That quick brown fox jumped over a lazy dog.") results = self.dmp.patch_apply(patches, "The quick brown fox jumps over the lazy dog.") self.assertEquals(("That quick brown fox jumped over a lazy dog.", [True, True]), results) # Partial match. results = self.dmp.patch_apply(patches, "The quick red rabbit jumps over the tired tiger.") self.assertEquals(("That quick red rabbit jumped over a tired tiger.", [True, True]), results) # Failed match. results = self.dmp.patch_apply(patches, "I am the very model of a modern major general.") self.assertEquals(("I am the very model of a modern major general.", [False, False]), results) # Big delete, small change. patches = self.dmp.patch_make("x1234567890123456789012345678901234567890123456789012345678901234567890y", "xabcy") results = self.dmp.patch_apply(patches, "x123456789012345678901234567890-----++++++++++-----123456789012345678901234567890y") self.assertEquals(("xabcy", [True, True]), results) # Big delete, big change 1. patches = self.dmp.patch_make("x1234567890123456789012345678901234567890123456789012345678901234567890y", "xabcy") results = self.dmp.patch_apply(patches, "x12345678901234567890---------------++++++++++---------------12345678901234567890y") self.assertEquals(("xabc12345678901234567890---------------++++++++++---------------12345678901234567890y", [False, True]), results) # Big delete, big change 2. self.dmp.Patch_DeleteThreshold = 0.6 patches = self.dmp.patch_make("x1234567890123456789012345678901234567890123456789012345678901234567890y", "xabcy") results = self.dmp.patch_apply(patches, "x12345678901234567890---------------++++++++++---------------12345678901234567890y") self.assertEquals(("xabcy", [True, True]), results) self.dmp.Patch_DeleteThreshold = 0.5 # Compensate for failed patch. self.dmp.Match_Threshold = 0.0 self.dmp.Match_Distance = 0 patches = self.dmp.patch_make("abcdefghijklmnopqrstuvwxyz--------------------1234567890", "abcXXXXXXXXXXdefghijklmnopqrstuvwxyz--------------------1234567YYYYYYYYYY890") results = self.dmp.patch_apply(patches, "ABCDEFGHIJKLMNOPQRSTUVWXYZ--------------------1234567890") self.assertEquals(("ABCDEFGHIJKLMNOPQRSTUVWXYZ--------------------1234567YYYYYYYYYY890", [False, True]), results) self.dmp.Match_Threshold = 0.5 self.dmp.Match_Distance = 1000 # No side effects. patches = self.dmp.patch_make("", "test") patchstr = self.dmp.patch_toText(patches) results = self.dmp.patch_apply(patches, "") self.assertEquals(patchstr, self.dmp.patch_toText(patches)) # No side effects with major delete. patches = self.dmp.patch_make("The quick brown fox jumps over the lazy dog.", "Woof") patchstr = self.dmp.patch_toText(patches) self.dmp.patch_apply(patches, "The quick brown fox jumps over the lazy dog.") self.assertEquals(patchstr, self.dmp.patch_toText(patches)) # Edge exact match. patches = self.dmp.patch_make("", "test") self.dmp.patch_apply(patches, "") self.assertEquals(("test", [True]), results) # Near edge exact match. patches = self.dmp.patch_make("XY", "XtestY") results = self.dmp.patch_apply(patches, "XY") self.assertEquals(("XtestY", [True]), results) # Edge partial match. patches = self.dmp.patch_make("y", "y123") results = self.dmp.patch_apply(patches, "x") self.assertEquals(("x123", [True]), results) if __name__ == "__main__": unittest.main() diff-match-patch-20121119/python2/diff_match_patch/__init__.py0000750000175000017500000000007312052423660023752 0ustar lukeluke00000000000000from .diff_match_patch import diff_match_patch, patch_obj diff-match-patch-20121119/python2/diff_match_patch/diff_match_patch.py0000750000175000017500000020453612052423660025470 0ustar lukeluke00000000000000#!/usr/bin/python2.4 from __future__ import division """Diff Match and Patch Copyright 2006 Google Inc. http://code.google.com/p/google-diff-match-patch/ Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ """Functions for diff, match and patch. Computes the difference between two texts to create a patch. Applies the patch onto another text, allowing for errors. """ __author__ = 'fraser@google.com (Neil Fraser)' import math import re import sys import time import urllib class diff_match_patch: """Class containing the diff, match and patch methods. Also contains the behaviour settings. """ def __init__(self): """Inits a diff_match_patch object with default settings. Redefine these in your program to override the defaults. """ # Number of seconds to map a diff before giving up (0 for infinity). self.Diff_Timeout = 1.0 # Cost of an empty edit operation in terms of edit characters. self.Diff_EditCost = 4 # At what point is no match declared (0.0 = perfection, 1.0 = very loose). self.Match_Threshold = 0.5 # How far to search for a match (0 = exact location, 1000+ = broad match). # A match this many characters away from the expected location will add # 1.0 to the score (0.0 is a perfect match). self.Match_Distance = 1000 # When deleting a large block of text (over ~64 characters), how close do # the contents have to be to match the expected contents. (0.0 = perfection, # 1.0 = very loose). Note that Match_Threshold controls how closely the # end points of a delete need to match. self.Patch_DeleteThreshold = 0.5 # Chunk size for context length. self.Patch_Margin = 4 # The number of bits in an int. # Python has no maximum, thus to disable patch splitting set to 0. # However to avoid long patches in certain pathological cases, use 32. # Multiple short patches (using native ints) are much faster than long ones. self.Match_MaxBits = 32 # DIFF FUNCTIONS # The data structure representing a diff is an array of tuples: # [(DIFF_DELETE, "Hello"), (DIFF_INSERT, "Goodbye"), (DIFF_EQUAL, " world.")] # which means: delete "Hello", add "Goodbye" and keep " world." DIFF_DELETE = -1 DIFF_INSERT = 1 DIFF_EQUAL = 0 def diff_main(self, text1, text2, checklines=True, deadline=None): """Find the differences between two texts. Simplifies the problem by stripping any common prefix or suffix off the texts before diffing. Args: text1: Old string to be diffed. text2: New string to be diffed. checklines: Optional speedup flag. If present and false, then don't run a line-level diff first to identify the changed areas. Defaults to true, which does a faster, slightly less optimal diff. deadline: Optional time when the diff should be complete by. Used internally for recursive calls. Users should set DiffTimeout instead. Returns: Array of changes. """ # Set a deadline by which time the diff must be complete. if deadline == None: # Unlike in most languages, Python counts time in seconds. if self.Diff_Timeout <= 0: deadline = sys.maxint else: deadline = time.time() + self.Diff_Timeout # Check for null inputs. if text1 == None or text2 == None: raise ValueError("Null inputs. (diff_main)") # Check for equality (speedup). if text1 == text2: if text1: return [(self.DIFF_EQUAL, text1)] return [] # Trim off common prefix (speedup). commonlength = self.diff_commonPrefix(text1, text2) commonprefix = text1[:commonlength] text1 = text1[commonlength:] text2 = text2[commonlength:] # Trim off common suffix (speedup). commonlength = self.diff_commonSuffix(text1, text2) if commonlength == 0: commonsuffix = '' else: commonsuffix = text1[-commonlength:] text1 = text1[:-commonlength] text2 = text2[:-commonlength] # Compute the diff on the middle block. diffs = self.diff_compute(text1, text2, checklines, deadline) # Restore the prefix and suffix. if commonprefix: diffs[:0] = [(self.DIFF_EQUAL, commonprefix)] if commonsuffix: diffs.append((self.DIFF_EQUAL, commonsuffix)) self.diff_cleanupMerge(diffs) return diffs def diff_compute(self, text1, text2, checklines, deadline): """Find the differences between two texts. Assumes that the texts do not have any common prefix or suffix. Args: text1: Old string to be diffed. text2: New string to be diffed. checklines: Speedup flag. If false, then don't run a line-level diff first to identify the changed areas. If true, then run a faster, slightly less optimal diff. deadline: Time when the diff should be complete by. Returns: Array of changes. """ if not text1: # Just add some text (speedup). return [(self.DIFF_INSERT, text2)] if not text2: # Just delete some text (speedup). return [(self.DIFF_DELETE, text1)] if len(text1) > len(text2): (longtext, shorttext) = (text1, text2) else: (shorttext, longtext) = (text1, text2) i = longtext.find(shorttext) if i != -1: # Shorter text is inside the longer text (speedup). diffs = [(self.DIFF_INSERT, longtext[:i]), (self.DIFF_EQUAL, shorttext), (self.DIFF_INSERT, longtext[i + len(shorttext):])] # Swap insertions for deletions if diff is reversed. if len(text1) > len(text2): diffs[0] = (self.DIFF_DELETE, diffs[0][1]) diffs[2] = (self.DIFF_DELETE, diffs[2][1]) return diffs if len(shorttext) == 1: # Single character string. # After the previous speedup, the character can't be an equality. return [(self.DIFF_DELETE, text1), (self.DIFF_INSERT, text2)] # Check to see if the problem can be split in two. hm = self.diff_halfMatch(text1, text2) if hm: # A half-match was found, sort out the return data. (text1_a, text1_b, text2_a, text2_b, mid_common) = hm # Send both pairs off for separate processing. diffs_a = self.diff_main(text1_a, text2_a, checklines, deadline) diffs_b = self.diff_main(text1_b, text2_b, checklines, deadline) # Merge the results. return diffs_a + [(self.DIFF_EQUAL, mid_common)] + diffs_b if checklines and len(text1) > 100 and len(text2) > 100: return self.diff_lineMode(text1, text2, deadline) return self.diff_bisect(text1, text2, deadline) def diff_lineMode(self, text1, text2, deadline): """Do a quick line-level diff on both strings, then rediff the parts for greater accuracy. This speedup can produce non-minimal diffs. Args: text1: Old string to be diffed. text2: New string to be diffed. deadline: Time when the diff should be complete by. Returns: Array of changes. """ # Scan the text on a line-by-line basis first. (text1, text2, linearray) = self.diff_linesToChars(text1, text2) diffs = self.diff_main(text1, text2, False, deadline) # Convert the diff back to original text. self.diff_charsToLines(diffs, linearray) # Eliminate freak matches (e.g. blank lines) self.diff_cleanupSemantic(diffs) # Rediff any replacement blocks, this time character-by-character. # Add a dummy entry at the end. diffs.append((self.DIFF_EQUAL, '')) pointer = 0 count_delete = 0 count_insert = 0 text_delete = '' text_insert = '' while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_INSERT: count_insert += 1 text_insert += diffs[pointer][1] elif diffs[pointer][0] == self.DIFF_DELETE: count_delete += 1 text_delete += diffs[pointer][1] elif diffs[pointer][0] == self.DIFF_EQUAL: # Upon reaching an equality, check for prior redundancies. if count_delete >= 1 and count_insert >= 1: # Delete the offending records and add the merged ones. a = self.diff_main(text_delete, text_insert, False, deadline) diffs[pointer - count_delete - count_insert : pointer] = a pointer = pointer - count_delete - count_insert + len(a) count_insert = 0 count_delete = 0 text_delete = '' text_insert = '' pointer += 1 diffs.pop() # Remove the dummy entry at the end. return diffs def diff_bisect(self, text1, text2, deadline): """Find the 'middle snake' of a diff, split the problem in two and return the recursively constructed diff. See Myers 1986 paper: An O(ND) Difference Algorithm and Its Variations. Args: text1: Old string to be diffed. text2: New string to be diffed. deadline: Time at which to bail if not yet complete. Returns: Array of diff tuples. """ # Cache the text lengths to prevent multiple calls. text1_length = len(text1) text2_length = len(text2) max_d = (text1_length + text2_length + 1) // 2 v_offset = max_d v_length = 2 * max_d v1 = [-1] * v_length v1[v_offset + 1] = 0 v2 = v1[:] delta = text1_length - text2_length # If the total number of characters is odd, then the front path will # collide with the reverse path. front = (delta % 2 != 0) # Offsets for start and end of k loop. # Prevents mapping of space beyond the grid. k1start = 0 k1end = 0 k2start = 0 k2end = 0 for d in xrange(max_d): # Bail out if deadline is reached. if time.time() > deadline: break # Walk the front path one step. for k1 in xrange(-d + k1start, d + 1 - k1end, 2): k1_offset = v_offset + k1 if k1 == -d or (k1 != d and v1[k1_offset - 1] < v1[k1_offset + 1]): x1 = v1[k1_offset + 1] else: x1 = v1[k1_offset - 1] + 1 y1 = x1 - k1 while (x1 < text1_length and y1 < text2_length and text1[x1] == text2[y1]): x1 += 1 y1 += 1 v1[k1_offset] = x1 if x1 > text1_length: # Ran off the right of the graph. k1end += 2 elif y1 > text2_length: # Ran off the bottom of the graph. k1start += 2 elif front: k2_offset = v_offset + delta - k1 if k2_offset >= 0 and k2_offset < v_length and v2[k2_offset] != -1: # Mirror x2 onto top-left coordinate system. x2 = text1_length - v2[k2_offset] if x1 >= x2: # Overlap detected. return self.diff_bisectSplit(text1, text2, x1, y1, deadline) # Walk the reverse path one step. for k2 in xrange(-d + k2start, d + 1 - k2end, 2): k2_offset = v_offset + k2 if k2 == -d or (k2 != d and v2[k2_offset - 1] < v2[k2_offset + 1]): x2 = v2[k2_offset + 1] else: x2 = v2[k2_offset - 1] + 1 y2 = x2 - k2 while (x2 < text1_length and y2 < text2_length and text1[-x2 - 1] == text2[-y2 - 1]): x2 += 1 y2 += 1 v2[k2_offset] = x2 if x2 > text1_length: # Ran off the left of the graph. k2end += 2 elif y2 > text2_length: # Ran off the top of the graph. k2start += 2 elif not front: k1_offset = v_offset + delta - k2 if k1_offset >= 0 and k1_offset < v_length and v1[k1_offset] != -1: x1 = v1[k1_offset] y1 = v_offset + x1 - k1_offset # Mirror x2 onto top-left coordinate system. x2 = text1_length - x2 if x1 >= x2: # Overlap detected. return self.diff_bisectSplit(text1, text2, x1, y1, deadline) # Diff took too long and hit the deadline or # number of diffs equals number of characters, no commonality at all. return [(self.DIFF_DELETE, text1), (self.DIFF_INSERT, text2)] def diff_bisectSplit(self, text1, text2, x, y, deadline): """Given the location of the 'middle snake', split the diff in two parts and recurse. Args: text1: Old string to be diffed. text2: New string to be diffed. x: Index of split point in text1. y: Index of split point in text2. deadline: Time at which to bail if not yet complete. Returns: Array of diff tuples. """ text1a = text1[:x] text2a = text2[:y] text1b = text1[x:] text2b = text2[y:] # Compute both diffs serially. diffs = self.diff_main(text1a, text2a, False, deadline) diffsb = self.diff_main(text1b, text2b, False, deadline) return diffs + diffsb def diff_linesToChars(self, text1, text2): """Split two texts into an array of strings. Reduce the texts to a string of hashes where each Unicode character represents one line. Args: text1: First string. text2: Second string. Returns: Three element tuple, containing the encoded text1, the encoded text2 and the array of unique strings. The zeroth element of the array of unique strings is intentionally blank. """ lineArray = [] # e.g. lineArray[4] == "Hello\n" lineHash = {} # e.g. lineHash["Hello\n"] == 4 # "\x00" is a valid character, but various debuggers don't like it. # So we'll insert a junk entry to avoid generating a null character. lineArray.append('') def diff_linesToCharsMunge(text): """Split a text into an array of strings. Reduce the texts to a string of hashes where each Unicode character represents one line. Modifies linearray and linehash through being a closure. Args: text: String to encode. Returns: Encoded string. """ chars = [] # Walk the text, pulling out a substring for each line. # text.split('\n') would would temporarily double our memory footprint. # Modifying text would create many large strings to garbage collect. lineStart = 0 lineEnd = -1 while lineEnd < len(text) - 1: lineEnd = text.find('\n', lineStart) if lineEnd == -1: lineEnd = len(text) - 1 line = text[lineStart:lineEnd + 1] lineStart = lineEnd + 1 if line in lineHash: chars.append(unichr(lineHash[line])) else: lineArray.append(line) lineHash[line] = len(lineArray) - 1 chars.append(unichr(len(lineArray) - 1)) return "".join(chars) chars1 = diff_linesToCharsMunge(text1) chars2 = diff_linesToCharsMunge(text2) return (chars1, chars2, lineArray) def diff_charsToLines(self, diffs, lineArray): """Rehydrate the text in a diff from a string of line hashes to real lines of text. Args: diffs: Array of diff tuples. lineArray: Array of unique strings. """ for x in xrange(len(diffs)): text = [] for char in diffs[x][1]: text.append(lineArray[ord(char)]) diffs[x] = (diffs[x][0], "".join(text)) def diff_commonPrefix(self, text1, text2): """Determine the common prefix of two strings. Args: text1: First string. text2: Second string. Returns: The number of characters common to the start of each string. """ # Quick check for common null cases. if not text1 or not text2 or text1[0] != text2[0]: return 0 # Binary search. # Performance analysis: http://neil.fraser.name/news/2007/10/09/ pointermin = 0 pointermax = min(len(text1), len(text2)) pointermid = pointermax pointerstart = 0 while pointermin < pointermid: if text1[pointerstart:pointermid] == text2[pointerstart:pointermid]: pointermin = pointermid pointerstart = pointermin else: pointermax = pointermid pointermid = (pointermax - pointermin) // 2 + pointermin return pointermid def diff_commonSuffix(self, text1, text2): """Determine the common suffix of two strings. Args: text1: First string. text2: Second string. Returns: The number of characters common to the end of each string. """ # Quick check for common null cases. if not text1 or not text2 or text1[-1] != text2[-1]: return 0 # Binary search. # Performance analysis: http://neil.fraser.name/news/2007/10/09/ pointermin = 0 pointermax = min(len(text1), len(text2)) pointermid = pointermax pointerend = 0 while pointermin < pointermid: if (text1[-pointermid:len(text1) - pointerend] == text2[-pointermid:len(text2) - pointerend]): pointermin = pointermid pointerend = pointermin else: pointermax = pointermid pointermid = (pointermax - pointermin) // 2 + pointermin return pointermid def diff_commonOverlap(self, text1, text2): """Determine if the suffix of one string is the prefix of another. Args: text1 First string. text2 Second string. Returns: The number of characters common to the end of the first string and the start of the second string. """ # Cache the text lengths to prevent multiple calls. text1_length = len(text1) text2_length = len(text2) # Eliminate the null case. if text1_length == 0 or text2_length == 0: return 0 # Truncate the longer string. if text1_length > text2_length: text1 = text1[-text2_length:] elif text1_length < text2_length: text2 = text2[:text1_length] text_length = min(text1_length, text2_length) # Quick check for the worst case. if text1 == text2: return text_length # Start by looking for a single character match # and increase length until no match is found. # Performance analysis: http://neil.fraser.name/news/2010/11/04/ best = 0 length = 1 while True: pattern = text1[-length:] found = text2.find(pattern) if found == -1: return best length += found if found == 0 or text1[-length:] == text2[:length]: best = length length += 1 def diff_halfMatch(self, text1, text2): """Do the two texts share a substring which is at least half the length of the longer text? This speedup can produce non-minimal diffs. Args: text1: First string. text2: Second string. Returns: Five element Array, containing the prefix of text1, the suffix of text1, the prefix of text2, the suffix of text2 and the common middle. Or None if there was no match. """ if self.Diff_Timeout <= 0: # Don't risk returning a non-optimal diff if we have unlimited time. return None if len(text1) > len(text2): (longtext, shorttext) = (text1, text2) else: (shorttext, longtext) = (text1, text2) if len(longtext) < 4 or len(shorttext) * 2 < len(longtext): return None # Pointless. def diff_halfMatchI(longtext, shorttext, i): """Does a substring of shorttext exist within longtext such that the substring is at least half the length of longtext? Closure, but does not reference any external variables. Args: longtext: Longer string. shorttext: Shorter string. i: Start index of quarter length substring within longtext. Returns: Five element Array, containing the prefix of longtext, the suffix of longtext, the prefix of shorttext, the suffix of shorttext and the common middle. Or None if there was no match. """ seed = longtext[i:i + len(longtext) // 4] best_common = '' j = shorttext.find(seed) while j != -1: prefixLength = self.diff_commonPrefix(longtext[i:], shorttext[j:]) suffixLength = self.diff_commonSuffix(longtext[:i], shorttext[:j]) if len(best_common) < suffixLength + prefixLength: best_common = (shorttext[j - suffixLength:j] + shorttext[j:j + prefixLength]) best_longtext_a = longtext[:i - suffixLength] best_longtext_b = longtext[i + prefixLength:] best_shorttext_a = shorttext[:j - suffixLength] best_shorttext_b = shorttext[j + prefixLength:] j = shorttext.find(seed, j + 1) if len(best_common) * 2 >= len(longtext): return (best_longtext_a, best_longtext_b, best_shorttext_a, best_shorttext_b, best_common) else: return None # First check if the second quarter is the seed for a half-match. hm1 = diff_halfMatchI(longtext, shorttext, (len(longtext) + 3) // 4) # Check again based on the third quarter. hm2 = diff_halfMatchI(longtext, shorttext, (len(longtext) + 1) // 2) if not hm1 and not hm2: return None elif not hm2: hm = hm1 elif not hm1: hm = hm2 else: # Both matched. Select the longest. if len(hm1[4]) > len(hm2[4]): hm = hm1 else: hm = hm2 # A half-match was found, sort out the return data. if len(text1) > len(text2): (text1_a, text1_b, text2_a, text2_b, mid_common) = hm else: (text2_a, text2_b, text1_a, text1_b, mid_common) = hm return (text1_a, text1_b, text2_a, text2_b, mid_common) def diff_cleanupSemantic(self, diffs): """Reduce the number of edits by eliminating semantically trivial equalities. Args: diffs: Array of diff tuples. """ changes = False equalities = [] # Stack of indices where equalities are found. lastequality = None # Always equal to diffs[equalities[-1]][1] pointer = 0 # Index of current position. # Number of chars that changed prior to the equality. length_insertions1, length_deletions1 = 0, 0 # Number of chars that changed after the equality. length_insertions2, length_deletions2 = 0, 0 while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_EQUAL: # Equality found. equalities.append(pointer) length_insertions1, length_insertions2 = length_insertions2, 0 length_deletions1, length_deletions2 = length_deletions2, 0 lastequality = diffs[pointer][1] else: # An insertion or deletion. if diffs[pointer][0] == self.DIFF_INSERT: length_insertions2 += len(diffs[pointer][1]) else: length_deletions2 += len(diffs[pointer][1]) # Eliminate an equality that is smaller or equal to the edits on both # sides of it. if (lastequality and (len(lastequality) <= max(length_insertions1, length_deletions1)) and (len(lastequality) <= max(length_insertions2, length_deletions2))): # Duplicate record. diffs.insert(equalities[-1], (self.DIFF_DELETE, lastequality)) # Change second copy to insert. diffs[equalities[-1] + 1] = (self.DIFF_INSERT, diffs[equalities[-1] + 1][1]) # Throw away the equality we just deleted. equalities.pop() # Throw away the previous equality (it needs to be reevaluated). if len(equalities): equalities.pop() if len(equalities): pointer = equalities[-1] else: pointer = -1 # Reset the counters. length_insertions1, length_deletions1 = 0, 0 length_insertions2, length_deletions2 = 0, 0 lastequality = None changes = True pointer += 1 # Normalize the diff. if changes: self.diff_cleanupMerge(diffs) self.diff_cleanupSemanticLossless(diffs) # Find any overlaps between deletions and insertions. # e.g: abcxxxxxxdef # -> abcxxxdef # e.g: xxxabcdefxxx # -> defxxxabc # Only extract an overlap if it is as big as the edit ahead or behind it. pointer = 1 while pointer < len(diffs): if (diffs[pointer - 1][0] == self.DIFF_DELETE and diffs[pointer][0] == self.DIFF_INSERT): deletion = diffs[pointer - 1][1] insertion = diffs[pointer][1] overlap_length1 = self.diff_commonOverlap(deletion, insertion) overlap_length2 = self.diff_commonOverlap(insertion, deletion) if overlap_length1 >= overlap_length2: if (overlap_length1 >= len(deletion) / 2.0 or overlap_length1 >= len(insertion) / 2.0): # Overlap found. Insert an equality and trim the surrounding edits. diffs.insert(pointer, (self.DIFF_EQUAL, insertion[:overlap_length1])) diffs[pointer - 1] = (self.DIFF_DELETE, deletion[:len(deletion) - overlap_length1]) diffs[pointer + 1] = (self.DIFF_INSERT, insertion[overlap_length1:]) pointer += 1 else: if (overlap_length2 >= len(deletion) / 2.0 or overlap_length2 >= len(insertion) / 2.0): # Reverse overlap found. # Insert an equality and swap and trim the surrounding edits. diffs.insert(pointer, (self.DIFF_EQUAL, deletion[:overlap_length2])) diffs[pointer - 1] = (self.DIFF_INSERT, insertion[:len(insertion) - overlap_length2]) diffs[pointer + 1] = (self.DIFF_DELETE, deletion[overlap_length2:]) pointer += 1 pointer += 1 pointer += 1 def diff_cleanupSemanticLossless(self, diffs): """Look for single edits surrounded on both sides by equalities which can be shifted sideways to align the edit to a word boundary. e.g: The cat came. -> The cat came. Args: diffs: Array of diff tuples. """ def diff_cleanupSemanticScore(one, two): """Given two strings, compute a score representing whether the internal boundary falls on logical boundaries. Scores range from 6 (best) to 0 (worst). Closure, but does not reference any external variables. Args: one: First string. two: Second string. Returns: The score. """ if not one or not two: # Edges are the best. return 6 # Each port of this function behaves slightly differently due to # subtle differences in each language's definition of things like # 'whitespace'. Since this function's purpose is largely cosmetic, # the choice has been made to use each language's native features # rather than force total conformity. char1 = one[-1] char2 = two[0] nonAlphaNumeric1 = not char1.isalnum() nonAlphaNumeric2 = not char2.isalnum() whitespace1 = nonAlphaNumeric1 and char1.isspace() whitespace2 = nonAlphaNumeric2 and char2.isspace() lineBreak1 = whitespace1 and (char1 == "\r" or char1 == "\n") lineBreak2 = whitespace2 and (char2 == "\r" or char2 == "\n") blankLine1 = lineBreak1 and self.BLANKLINEEND.search(one) blankLine2 = lineBreak2 and self.BLANKLINESTART.match(two) if blankLine1 or blankLine2: # Five points for blank lines. return 5 elif lineBreak1 or lineBreak2: # Four points for line breaks. return 4 elif nonAlphaNumeric1 and not whitespace1 and whitespace2: # Three points for end of sentences. return 3 elif whitespace1 or whitespace2: # Two points for whitespace. return 2 elif nonAlphaNumeric1 or nonAlphaNumeric2: # One point for non-alphanumeric. return 1 return 0 pointer = 1 # Intentionally ignore the first and last element (don't need checking). while pointer < len(diffs) - 1: if (diffs[pointer - 1][0] == self.DIFF_EQUAL and diffs[pointer + 1][0] == self.DIFF_EQUAL): # This is a single edit surrounded by equalities. equality1 = diffs[pointer - 1][1] edit = diffs[pointer][1] equality2 = diffs[pointer + 1][1] # First, shift the edit as far left as possible. commonOffset = self.diff_commonSuffix(equality1, edit) if commonOffset: commonString = edit[-commonOffset:] equality1 = equality1[:-commonOffset] edit = commonString + edit[:-commonOffset] equality2 = commonString + equality2 # Second, step character by character right, looking for the best fit. bestEquality1 = equality1 bestEdit = edit bestEquality2 = equality2 bestScore = (diff_cleanupSemanticScore(equality1, edit) + diff_cleanupSemanticScore(edit, equality2)) while edit and equality2 and edit[0] == equality2[0]: equality1 += edit[0] edit = edit[1:] + equality2[0] equality2 = equality2[1:] score = (diff_cleanupSemanticScore(equality1, edit) + diff_cleanupSemanticScore(edit, equality2)) # The >= encourages trailing rather than leading whitespace on edits. if score >= bestScore: bestScore = score bestEquality1 = equality1 bestEdit = edit bestEquality2 = equality2 if diffs[pointer - 1][1] != bestEquality1: # We have an improvement, save it back to the diff. if bestEquality1: diffs[pointer - 1] = (diffs[pointer - 1][0], bestEquality1) else: del diffs[pointer - 1] pointer -= 1 diffs[pointer] = (diffs[pointer][0], bestEdit) if bestEquality2: diffs[pointer + 1] = (diffs[pointer + 1][0], bestEquality2) else: del diffs[pointer + 1] pointer -= 1 pointer += 1 # Define some regex patterns for matching boundaries. BLANKLINEEND = re.compile(r"\n\r?\n$"); BLANKLINESTART = re.compile(r"^\r?\n\r?\n"); def diff_cleanupEfficiency(self, diffs): """Reduce the number of edits by eliminating operationally trivial equalities. Args: diffs: Array of diff tuples. """ changes = False equalities = [] # Stack of indices where equalities are found. lastequality = None # Always equal to diffs[equalities[-1]][1] pointer = 0 # Index of current position. pre_ins = False # Is there an insertion operation before the last equality. pre_del = False # Is there a deletion operation before the last equality. post_ins = False # Is there an insertion operation after the last equality. post_del = False # Is there a deletion operation after the last equality. while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_EQUAL: # Equality found. if (len(diffs[pointer][1]) < self.Diff_EditCost and (post_ins or post_del)): # Candidate found. equalities.append(pointer) pre_ins = post_ins pre_del = post_del lastequality = diffs[pointer][1] else: # Not a candidate, and can never become one. equalities = [] lastequality = None post_ins = post_del = False else: # An insertion or deletion. if diffs[pointer][0] == self.DIFF_DELETE: post_del = True else: post_ins = True # Five types to be split: # ABXYCD # AXCD # ABXC # AXCD # ABXC if lastequality and ((pre_ins and pre_del and post_ins and post_del) or ((len(lastequality) < self.Diff_EditCost / 2) and (pre_ins + pre_del + post_ins + post_del) == 3)): # Duplicate record. diffs.insert(equalities[-1], (self.DIFF_DELETE, lastequality)) # Change second copy to insert. diffs[equalities[-1] + 1] = (self.DIFF_INSERT, diffs[equalities[-1] + 1][1]) equalities.pop() # Throw away the equality we just deleted. lastequality = None if pre_ins and pre_del: # No changes made which could affect previous entry, keep going. post_ins = post_del = True equalities = [] else: if len(equalities): equalities.pop() # Throw away the previous equality. if len(equalities): pointer = equalities[-1] else: pointer = -1 post_ins = post_del = False changes = True pointer += 1 if changes: self.diff_cleanupMerge(diffs) def diff_cleanupMerge(self, diffs): """Reorder and merge like edit sections. Merge equalities. Any edit section can move as long as it doesn't cross an equality. Args: diffs: Array of diff tuples. """ diffs.append((self.DIFF_EQUAL, '')) # Add a dummy entry at the end. pointer = 0 count_delete = 0 count_insert = 0 text_delete = '' text_insert = '' while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_INSERT: count_insert += 1 text_insert += diffs[pointer][1] pointer += 1 elif diffs[pointer][0] == self.DIFF_DELETE: count_delete += 1 text_delete += diffs[pointer][1] pointer += 1 elif diffs[pointer][0] == self.DIFF_EQUAL: # Upon reaching an equality, check for prior redundancies. if count_delete + count_insert > 1: if count_delete != 0 and count_insert != 0: # Factor out any common prefixies. commonlength = self.diff_commonPrefix(text_insert, text_delete) if commonlength != 0: x = pointer - count_delete - count_insert - 1 if x >= 0 and diffs[x][0] == self.DIFF_EQUAL: diffs[x] = (diffs[x][0], diffs[x][1] + text_insert[:commonlength]) else: diffs.insert(0, (self.DIFF_EQUAL, text_insert[:commonlength])) pointer += 1 text_insert = text_insert[commonlength:] text_delete = text_delete[commonlength:] # Factor out any common suffixies. commonlength = self.diff_commonSuffix(text_insert, text_delete) if commonlength != 0: diffs[pointer] = (diffs[pointer][0], text_insert[-commonlength:] + diffs[pointer][1]) text_insert = text_insert[:-commonlength] text_delete = text_delete[:-commonlength] # Delete the offending records and add the merged ones. if count_delete == 0: diffs[pointer - count_insert : pointer] = [ (self.DIFF_INSERT, text_insert)] elif count_insert == 0: diffs[pointer - count_delete : pointer] = [ (self.DIFF_DELETE, text_delete)] else: diffs[pointer - count_delete - count_insert : pointer] = [ (self.DIFF_DELETE, text_delete), (self.DIFF_INSERT, text_insert)] pointer = pointer - count_delete - count_insert + 1 if count_delete != 0: pointer += 1 if count_insert != 0: pointer += 1 elif pointer != 0 and diffs[pointer - 1][0] == self.DIFF_EQUAL: # Merge this equality with the previous one. diffs[pointer - 1] = (diffs[pointer - 1][0], diffs[pointer - 1][1] + diffs[pointer][1]) del diffs[pointer] else: pointer += 1 count_insert = 0 count_delete = 0 text_delete = '' text_insert = '' if diffs[-1][1] == '': diffs.pop() # Remove the dummy entry at the end. # Second pass: look for single edits surrounded on both sides by equalities # which can be shifted sideways to eliminate an equality. # e.g: ABAC -> ABAC changes = False pointer = 1 # Intentionally ignore the first and last element (don't need checking). while pointer < len(diffs) - 1: if (diffs[pointer - 1][0] == self.DIFF_EQUAL and diffs[pointer + 1][0] == self.DIFF_EQUAL): # This is a single edit surrounded by equalities. if diffs[pointer][1].endswith(diffs[pointer - 1][1]): # Shift the edit over the previous equality. diffs[pointer] = (diffs[pointer][0], diffs[pointer - 1][1] + diffs[pointer][1][:-len(diffs[pointer - 1][1])]) diffs[pointer + 1] = (diffs[pointer + 1][0], diffs[pointer - 1][1] + diffs[pointer + 1][1]) del diffs[pointer - 1] changes = True elif diffs[pointer][1].startswith(diffs[pointer + 1][1]): # Shift the edit over the next equality. diffs[pointer - 1] = (diffs[pointer - 1][0], diffs[pointer - 1][1] + diffs[pointer + 1][1]) diffs[pointer] = (diffs[pointer][0], diffs[pointer][1][len(diffs[pointer + 1][1]):] + diffs[pointer + 1][1]) del diffs[pointer + 1] changes = True pointer += 1 # If shifts were made, the diff needs reordering and another shift sweep. if changes: self.diff_cleanupMerge(diffs) def diff_xIndex(self, diffs, loc): """loc is a location in text1, compute and return the equivalent location in text2. e.g. "The cat" vs "The big cat", 1->1, 5->8 Args: diffs: Array of diff tuples. loc: Location within text1. Returns: Location within text2. """ chars1 = 0 chars2 = 0 last_chars1 = 0 last_chars2 = 0 for x in xrange(len(diffs)): (op, text) = diffs[x] if op != self.DIFF_INSERT: # Equality or deletion. chars1 += len(text) if op != self.DIFF_DELETE: # Equality or insertion. chars2 += len(text) if chars1 > loc: # Overshot the location. break last_chars1 = chars1 last_chars2 = chars2 if len(diffs) != x and diffs[x][0] == self.DIFF_DELETE: # The location was deleted. return last_chars2 # Add the remaining len(character). return last_chars2 + (loc - last_chars1) def diff_prettyHtml(self, diffs): """Convert a diff array into a pretty HTML report. Args: diffs: Array of diff tuples. Returns: HTML representation. """ html = [] for (op, data) in diffs: text = (data.replace("&", "&").replace("<", "<") .replace(">", ">").replace("\n", "¶
")) if op == self.DIFF_INSERT: html.append("%s" % text) elif op == self.DIFF_DELETE: html.append("%s" % text) elif op == self.DIFF_EQUAL: html.append("%s" % text) return "".join(html) def diff_text1(self, diffs): """Compute and return the source text (all equalities and deletions). Args: diffs: Array of diff tuples. Returns: Source text. """ text = [] for (op, data) in diffs: if op != self.DIFF_INSERT: text.append(data) return "".join(text) def diff_text2(self, diffs): """Compute and return the destination text (all equalities and insertions). Args: diffs: Array of diff tuples. Returns: Destination text. """ text = [] for (op, data) in diffs: if op != self.DIFF_DELETE: text.append(data) return "".join(text) def diff_levenshtein(self, diffs): """Compute the Levenshtein distance; the number of inserted, deleted or substituted characters. Args: diffs: Array of diff tuples. Returns: Number of changes. """ levenshtein = 0 insertions = 0 deletions = 0 for (op, data) in diffs: if op == self.DIFF_INSERT: insertions += len(data) elif op == self.DIFF_DELETE: deletions += len(data) elif op == self.DIFF_EQUAL: # A deletion and an insertion is one substitution. levenshtein += max(insertions, deletions) insertions = 0 deletions = 0 levenshtein += max(insertions, deletions) return levenshtein def diff_toDelta(self, diffs): """Crush the diff into an encoded string which describes the operations required to transform text1 into text2. E.g. =3\t-2\t+ing -> Keep 3 chars, delete 2 chars, insert 'ing'. Operations are tab-separated. Inserted text is escaped using %xx notation. Args: diffs: Array of diff tuples. Returns: Delta text. """ text = [] for (op, data) in diffs: if op == self.DIFF_INSERT: # High ascii will raise UnicodeDecodeError. Use Unicode instead. data = data.encode("utf-8") text.append("+" + urllib.quote(data, "!~*'();/?:@&=+$,# ")) elif op == self.DIFF_DELETE: text.append("-%d" % len(data)) elif op == self.DIFF_EQUAL: text.append("=%d" % len(data)) return "\t".join(text) def diff_fromDelta(self, text1, delta): """Given the original text1, and an encoded string which describes the operations required to transform text1 into text2, compute the full diff. Args: text1: Source string for the diff. delta: Delta text. Returns: Array of diff tuples. Raises: ValueError: If invalid input. """ if type(delta) == unicode: # Deltas should be composed of a subset of ascii chars, Unicode not # required. If this encode raises UnicodeEncodeError, delta is invalid. delta = delta.encode("ascii") diffs = [] pointer = 0 # Cursor in text1 tokens = delta.split("\t") for token in tokens: if token == "": # Blank tokens are ok (from a trailing \t). continue # Each token begins with a one character parameter which specifies the # operation of this token (delete, insert, equality). param = token[1:] if token[0] == "+": param = urllib.unquote(param).decode("utf-8") diffs.append((self.DIFF_INSERT, param)) elif token[0] == "-" or token[0] == "=": try: n = int(param) except ValueError: raise ValueError("Invalid number in diff_fromDelta: " + param) if n < 0: raise ValueError("Negative number in diff_fromDelta: " + param) text = text1[pointer : pointer + n] pointer += n if token[0] == "=": diffs.append((self.DIFF_EQUAL, text)) else: diffs.append((self.DIFF_DELETE, text)) else: # Anything else is an error. raise ValueError("Invalid diff operation in diff_fromDelta: " + token[0]) if pointer != len(text1): raise ValueError( "Delta length (%d) does not equal source text length (%d)." % (pointer, len(text1))) return diffs # MATCH FUNCTIONS def match_main(self, text, pattern, loc): """Locate the best instance of 'pattern' in 'text' near 'loc'. Args: text: The text to search. pattern: The pattern to search for. loc: The location to search around. Returns: Best match index or -1. """ # Check for null inputs. if text == None or pattern == None: raise ValueError("Null inputs. (match_main)") loc = max(0, min(loc, len(text))) if text == pattern: # Shortcut (potentially not guaranteed by the algorithm) return 0 elif not text: # Nothing to match. return -1 elif text[loc:loc + len(pattern)] == pattern: # Perfect match at the perfect spot! (Includes case of null pattern) return loc else: # Do a fuzzy compare. match = self.match_bitap(text, pattern, loc) return match def match_bitap(self, text, pattern, loc): """Locate the best instance of 'pattern' in 'text' near 'loc' using the Bitap algorithm. Args: text: The text to search. pattern: The pattern to search for. loc: The location to search around. Returns: Best match index or -1. """ # Python doesn't have a maxint limit, so ignore this check. #if self.Match_MaxBits != 0 and len(pattern) > self.Match_MaxBits: # raise ValueError("Pattern too long for this application.") # Initialise the alphabet. s = self.match_alphabet(pattern) def match_bitapScore(e, x): """Compute and return the score for a match with e errors and x location. Accesses loc and pattern through being a closure. Args: e: Number of errors in match. x: Location of match. Returns: Overall score for match (0.0 = good, 1.0 = bad). """ accuracy = float(e) / len(pattern) proximity = abs(loc - x) if not self.Match_Distance: # Dodge divide by zero error. return proximity and 1.0 or accuracy return accuracy + (proximity / float(self.Match_Distance)) # Highest score beyond which we give up. score_threshold = self.Match_Threshold # Is there a nearby exact match? (speedup) best_loc = text.find(pattern, loc) if best_loc != -1: score_threshold = min(match_bitapScore(0, best_loc), score_threshold) # What about in the other direction? (speedup) best_loc = text.rfind(pattern, loc + len(pattern)) if best_loc != -1: score_threshold = min(match_bitapScore(0, best_loc), score_threshold) # Initialise the bit arrays. matchmask = 1 << (len(pattern) - 1) best_loc = -1 bin_max = len(pattern) + len(text) # Empty initialization added to appease pychecker. last_rd = None for d in xrange(len(pattern)): # Scan for the best match each iteration allows for one more error. # Run a binary search to determine how far from 'loc' we can stray at # this error level. bin_min = 0 bin_mid = bin_max while bin_min < bin_mid: if match_bitapScore(d, loc + bin_mid) <= score_threshold: bin_min = bin_mid else: bin_max = bin_mid bin_mid = (bin_max - bin_min) // 2 + bin_min # Use the result from this iteration as the maximum for the next. bin_max = bin_mid start = max(1, loc - bin_mid + 1) finish = min(loc + bin_mid, len(text)) + len(pattern) rd = [0] * (finish + 2) rd[finish + 1] = (1 << d) - 1 for j in xrange(finish, start - 1, -1): if len(text) <= j - 1: # Out of range. charMatch = 0 else: charMatch = s.get(text[j - 1], 0) if d == 0: # First pass: exact match. rd[j] = ((rd[j + 1] << 1) | 1) & charMatch else: # Subsequent passes: fuzzy match. rd[j] = (((rd[j + 1] << 1) | 1) & charMatch) | ( ((last_rd[j + 1] | last_rd[j]) << 1) | 1) | last_rd[j + 1] if rd[j] & matchmask: score = match_bitapScore(d, j - 1) # This match will almost certainly be better than any existing match. # But check anyway. if score <= score_threshold: # Told you so. score_threshold = score best_loc = j - 1 if best_loc > loc: # When passing loc, don't exceed our current distance from loc. start = max(1, 2 * loc - best_loc) else: # Already passed loc, downhill from here on in. break # No hope for a (better) match at greater error levels. if match_bitapScore(d + 1, loc) > score_threshold: break last_rd = rd return best_loc def match_alphabet(self, pattern): """Initialise the alphabet for the Bitap algorithm. Args: pattern: The text to encode. Returns: Hash of character locations. """ s = {} for char in pattern: s[char] = 0 for i in xrange(len(pattern)): s[pattern[i]] |= 1 << (len(pattern) - i - 1) return s # PATCH FUNCTIONS def patch_addContext(self, patch, text): """Increase the context until it is unique, but don't let the pattern expand beyond Match_MaxBits. Args: patch: The patch to grow. text: Source text. """ if len(text) == 0: return pattern = text[patch.start2 : patch.start2 + patch.length1] padding = 0 # Look for the first and last matches of pattern in text. If two different # matches are found, increase the pattern length. while (text.find(pattern) != text.rfind(pattern) and (self.Match_MaxBits == 0 or len(pattern) < self.Match_MaxBits - self.Patch_Margin - self.Patch_Margin)): padding += self.Patch_Margin pattern = text[max(0, patch.start2 - padding) : patch.start2 + patch.length1 + padding] # Add one chunk for good luck. padding += self.Patch_Margin # Add the prefix. prefix = text[max(0, patch.start2 - padding) : patch.start2] if prefix: patch.diffs[:0] = [(self.DIFF_EQUAL, prefix)] # Add the suffix. suffix = text[patch.start2 + patch.length1 : patch.start2 + patch.length1 + padding] if suffix: patch.diffs.append((self.DIFF_EQUAL, suffix)) # Roll back the start points. patch.start1 -= len(prefix) patch.start2 -= len(prefix) # Extend lengths. patch.length1 += len(prefix) + len(suffix) patch.length2 += len(prefix) + len(suffix) def patch_make(self, a, b=None, c=None): """Compute a list of patches to turn text1 into text2. Use diffs if provided, otherwise compute it ourselves. There are four ways to call this function, depending on what data is available to the caller: Method 1: a = text1, b = text2 Method 2: a = diffs Method 3 (optimal): a = text1, b = diffs Method 4 (deprecated, use method 3): a = text1, b = text2, c = diffs Args: a: text1 (methods 1,3,4) or Array of diff tuples for text1 to text2 (method 2). b: text2 (methods 1,4) or Array of diff tuples for text1 to text2 (method 3) or undefined (method 2). c: Array of diff tuples for text1 to text2 (method 4) or undefined (methods 1,2,3). Returns: Array of Patch objects. """ text1 = None diffs = None # Note that texts may arrive as 'str' or 'unicode'. if isinstance(a, basestring) and isinstance(b, basestring) and c is None: # Method 1: text1, text2 # Compute diffs from text1 and text2. text1 = a diffs = self.diff_main(text1, b, True) if len(diffs) > 2: self.diff_cleanupSemantic(diffs) self.diff_cleanupEfficiency(diffs) elif isinstance(a, list) and b is None and c is None: # Method 2: diffs # Compute text1 from diffs. diffs = a text1 = self.diff_text1(diffs) elif isinstance(a, basestring) and isinstance(b, list) and c is None: # Method 3: text1, diffs text1 = a diffs = b elif (isinstance(a, basestring) and isinstance(b, basestring) and isinstance(c, list)): # Method 4: text1, text2, diffs # text2 is not used. text1 = a diffs = c else: raise ValueError("Unknown call format to patch_make.") if not diffs: return [] # Get rid of the None case. patches = [] patch = patch_obj() char_count1 = 0 # Number of characters into the text1 string. char_count2 = 0 # Number of characters into the text2 string. prepatch_text = text1 # Recreate the patches to determine context info. postpatch_text = text1 for x in xrange(len(diffs)): (diff_type, diff_text) = diffs[x] if len(patch.diffs) == 0 and diff_type != self.DIFF_EQUAL: # A new patch starts here. patch.start1 = char_count1 patch.start2 = char_count2 if diff_type == self.DIFF_INSERT: # Insertion patch.diffs.append(diffs[x]) patch.length2 += len(diff_text) postpatch_text = (postpatch_text[:char_count2] + diff_text + postpatch_text[char_count2:]) elif diff_type == self.DIFF_DELETE: # Deletion. patch.length1 += len(diff_text) patch.diffs.append(diffs[x]) postpatch_text = (postpatch_text[:char_count2] + postpatch_text[char_count2 + len(diff_text):]) elif (diff_type == self.DIFF_EQUAL and len(diff_text) <= 2 * self.Patch_Margin and len(patch.diffs) != 0 and len(diffs) != x + 1): # Small equality inside a patch. patch.diffs.append(diffs[x]) patch.length1 += len(diff_text) patch.length2 += len(diff_text) if (diff_type == self.DIFF_EQUAL and len(diff_text) >= 2 * self.Patch_Margin): # Time for a new patch. if len(patch.diffs) != 0: self.patch_addContext(patch, prepatch_text) patches.append(patch) patch = patch_obj() # Unlike Unidiff, our patch lists have a rolling context. # http://code.google.com/p/google-diff-match-patch/wiki/Unidiff # Update prepatch text & pos to reflect the application of the # just completed patch. prepatch_text = postpatch_text char_count1 = char_count2 # Update the current character count. if diff_type != self.DIFF_INSERT: char_count1 += len(diff_text) if diff_type != self.DIFF_DELETE: char_count2 += len(diff_text) # Pick up the leftover patch if not empty. if len(patch.diffs) != 0: self.patch_addContext(patch, prepatch_text) patches.append(patch) return patches def patch_deepCopy(self, patches): """Given an array of patches, return another array that is identical. Args: patches: Array of Patch objects. Returns: Array of Patch objects. """ patchesCopy = [] for patch in patches: patchCopy = patch_obj() # No need to deep copy the tuples since they are immutable. patchCopy.diffs = patch.diffs[:] patchCopy.start1 = patch.start1 patchCopy.start2 = patch.start2 patchCopy.length1 = patch.length1 patchCopy.length2 = patch.length2 patchesCopy.append(patchCopy) return patchesCopy def patch_apply(self, patches, text): """Merge a set of patches onto the text. Return a patched text, as well as a list of true/false values indicating which patches were applied. Args: patches: Array of Patch objects. text: Old text. Returns: Two element Array, containing the new text and an array of boolean values. """ if not patches: return (text, []) # Deep copy the patches so that no changes are made to originals. patches = self.patch_deepCopy(patches) nullPadding = self.patch_addPadding(patches) text = nullPadding + text + nullPadding self.patch_splitMax(patches) # delta keeps track of the offset between the expected and actual location # of the previous patch. If there are patches expected at positions 10 and # 20, but the first patch was found at 12, delta is 2 and the second patch # has an effective expected position of 22. delta = 0 results = [] for patch in patches: expected_loc = patch.start2 + delta text1 = self.diff_text1(patch.diffs) end_loc = -1 if len(text1) > self.Match_MaxBits: # patch_splitMax will only provide an oversized pattern in the case of # a monster delete. start_loc = self.match_main(text, text1[:self.Match_MaxBits], expected_loc) if start_loc != -1: end_loc = self.match_main(text, text1[-self.Match_MaxBits:], expected_loc + len(text1) - self.Match_MaxBits) if end_loc == -1 or start_loc >= end_loc: # Can't find valid trailing context. Drop this patch. start_loc = -1 else: start_loc = self.match_main(text, text1, expected_loc) if start_loc == -1: # No match found. :( results.append(False) # Subtract the delta for this failed patch from subsequent patches. delta -= patch.length2 - patch.length1 else: # Found a match. :) results.append(True) delta = start_loc - expected_loc if end_loc == -1: text2 = text[start_loc : start_loc + len(text1)] else: text2 = text[start_loc : end_loc + self.Match_MaxBits] if text1 == text2: # Perfect match, just shove the replacement text in. text = (text[:start_loc] + self.diff_text2(patch.diffs) + text[start_loc + len(text1):]) else: # Imperfect match. # Run a diff to get a framework of equivalent indices. diffs = self.diff_main(text1, text2, False) if (len(text1) > self.Match_MaxBits and self.diff_levenshtein(diffs) / float(len(text1)) > self.Patch_DeleteThreshold): # The end points match, but the content is unacceptably bad. results[-1] = False else: self.diff_cleanupSemanticLossless(diffs) index1 = 0 for (op, data) in patch.diffs: if op != self.DIFF_EQUAL: index2 = self.diff_xIndex(diffs, index1) if op == self.DIFF_INSERT: # Insertion text = text[:start_loc + index2] + data + text[start_loc + index2:] elif op == self.DIFF_DELETE: # Deletion text = text[:start_loc + index2] + text[start_loc + self.diff_xIndex(diffs, index1 + len(data)):] if op != self.DIFF_DELETE: index1 += len(data) # Strip the padding off. text = text[len(nullPadding):-len(nullPadding)] return (text, results) def patch_addPadding(self, patches): """Add some padding on text start and end so that edges can match something. Intended to be called only from within patch_apply. Args: patches: Array of Patch objects. Returns: The padding string added to each side. """ paddingLength = self.Patch_Margin nullPadding = "" for x in xrange(1, paddingLength + 1): nullPadding += chr(x) # Bump all the patches forward. for patch in patches: patch.start1 += paddingLength patch.start2 += paddingLength # Add some padding on start of first diff. patch = patches[0] diffs = patch.diffs if not diffs or diffs[0][0] != self.DIFF_EQUAL: # Add nullPadding equality. diffs.insert(0, (self.DIFF_EQUAL, nullPadding)) patch.start1 -= paddingLength # Should be 0. patch.start2 -= paddingLength # Should be 0. patch.length1 += paddingLength patch.length2 += paddingLength elif paddingLength > len(diffs[0][1]): # Grow first equality. extraLength = paddingLength - len(diffs[0][1]) newText = nullPadding[len(diffs[0][1]):] + diffs[0][1] diffs[0] = (diffs[0][0], newText) patch.start1 -= extraLength patch.start2 -= extraLength patch.length1 += extraLength patch.length2 += extraLength # Add some padding on end of last diff. patch = patches[-1] diffs = patch.diffs if not diffs or diffs[-1][0] != self.DIFF_EQUAL: # Add nullPadding equality. diffs.append((self.DIFF_EQUAL, nullPadding)) patch.length1 += paddingLength patch.length2 += paddingLength elif paddingLength > len(diffs[-1][1]): # Grow last equality. extraLength = paddingLength - len(diffs[-1][1]) newText = diffs[-1][1] + nullPadding[:extraLength] diffs[-1] = (diffs[-1][0], newText) patch.length1 += extraLength patch.length2 += extraLength return nullPadding def patch_splitMax(self, patches): """Look through the patches and break up any which are longer than the maximum limit of the match algorithm. Intended to be called only from within patch_apply. Args: patches: Array of Patch objects. """ patch_size = self.Match_MaxBits if patch_size == 0: # Python has the option of not splitting strings due to its ability # to handle integers of arbitrary precision. return for x in xrange(len(patches)): if patches[x].length1 <= patch_size: continue bigpatch = patches[x] # Remove the big old patch. del patches[x] x -= 1 start1 = bigpatch.start1 start2 = bigpatch.start2 precontext = '' while len(bigpatch.diffs) != 0: # Create one of several smaller patches. patch = patch_obj() empty = True patch.start1 = start1 - len(precontext) patch.start2 = start2 - len(precontext) if precontext: patch.length1 = patch.length2 = len(precontext) patch.diffs.append((self.DIFF_EQUAL, precontext)) while (len(bigpatch.diffs) != 0 and patch.length1 < patch_size - self.Patch_Margin): (diff_type, diff_text) = bigpatch.diffs[0] if diff_type == self.DIFF_INSERT: # Insertions are harmless. patch.length2 += len(diff_text) start2 += len(diff_text) patch.diffs.append(bigpatch.diffs.pop(0)) empty = False elif (diff_type == self.DIFF_DELETE and len(patch.diffs) == 1 and patch.diffs[0][0] == self.DIFF_EQUAL and len(diff_text) > 2 * patch_size): # This is a large deletion. Let it pass in one chunk. patch.length1 += len(diff_text) start1 += len(diff_text) empty = False patch.diffs.append((diff_type, diff_text)) del bigpatch.diffs[0] else: # Deletion or equality. Only take as much as we can stomach. diff_text = diff_text[:patch_size - patch.length1 - self.Patch_Margin] patch.length1 += len(diff_text) start1 += len(diff_text) if diff_type == self.DIFF_EQUAL: patch.length2 += len(diff_text) start2 += len(diff_text) else: empty = False patch.diffs.append((diff_type, diff_text)) if diff_text == bigpatch.diffs[0][1]: del bigpatch.diffs[0] else: bigpatch.diffs[0] = (bigpatch.diffs[0][0], bigpatch.diffs[0][1][len(diff_text):]) # Compute the head context for the next patch. precontext = self.diff_text2(patch.diffs) precontext = precontext[-self.Patch_Margin:] # Append the end context for this patch. postcontext = self.diff_text1(bigpatch.diffs)[:self.Patch_Margin] if postcontext: patch.length1 += len(postcontext) patch.length2 += len(postcontext) if len(patch.diffs) != 0 and patch.diffs[-1][0] == self.DIFF_EQUAL: patch.diffs[-1] = (self.DIFF_EQUAL, patch.diffs[-1][1] + postcontext) else: patch.diffs.append((self.DIFF_EQUAL, postcontext)) if not empty: x += 1 patches.insert(x, patch) def patch_toText(self, patches): """Take a list of patches and return a textual representation. Args: patches: Array of Patch objects. Returns: Text representation of patches. """ text = [] for patch in patches: text.append(str(patch)) return "".join(text) def patch_fromText(self, textline): """Parse a textual representation of patches and return a list of patch objects. Args: textline: Text representation of patches. Returns: Array of Patch objects. Raises: ValueError: If invalid input. """ if type(textline) == unicode: # Patches should be composed of a subset of ascii chars, Unicode not # required. If this encode raises UnicodeEncodeError, patch is invalid. textline = textline.encode("ascii") patches = [] if not textline: return patches text = textline.split('\n') while len(text) != 0: m = re.match("^@@ -(\d+),?(\d*) \+(\d+),?(\d*) @@$", text[0]) if not m: raise ValueError("Invalid patch string: " + text[0]) patch = patch_obj() patches.append(patch) patch.start1 = int(m.group(1)) if m.group(2) == '': patch.start1 -= 1 patch.length1 = 1 elif m.group(2) == '0': patch.length1 = 0 else: patch.start1 -= 1 patch.length1 = int(m.group(2)) patch.start2 = int(m.group(3)) if m.group(4) == '': patch.start2 -= 1 patch.length2 = 1 elif m.group(4) == '0': patch.length2 = 0 else: patch.start2 -= 1 patch.length2 = int(m.group(4)) del text[0] while len(text) != 0: if text[0]: sign = text[0][0] else: sign = '' line = urllib.unquote(text[0][1:]) line = line.decode("utf-8") if sign == '+': # Insertion. patch.diffs.append((self.DIFF_INSERT, line)) elif sign == '-': # Deletion. patch.diffs.append((self.DIFF_DELETE, line)) elif sign == ' ': # Minor equality. patch.diffs.append((self.DIFF_EQUAL, line)) elif sign == '@': # Start of next patch. break elif sign == '': # Blank line? Whatever. pass else: # WTF? raise ValueError("Invalid patch mode: '%s'\n%s" % (sign, line)) del text[0] return patches class patch_obj: """Class representing one patch operation. """ def __init__(self): """Initializes with an empty list of diffs. """ self.diffs = [] self.start1 = None self.start2 = None self.length1 = 0 self.length2 = 0 def __str__(self): """Emmulate GNU diff's format. Header: @@ -382,8 +481,9 @@ Indicies are printed as 1-based, not 0-based. Returns: The GNU diff string. """ if self.length1 == 0: coords1 = str(self.start1) + ",0" elif self.length1 == 1: coords1 = str(self.start1 + 1) else: coords1 = str(self.start1 + 1) + "," + str(self.length1) if self.length2 == 0: coords2 = str(self.start2) + ",0" elif self.length2 == 1: coords2 = str(self.start2 + 1) else: coords2 = str(self.start2 + 1) + "," + str(self.length2) text = ["@@ -", coords1, " +", coords2, " @@\n"] # Escape the body of the patch with %xx notation. for (op, data) in self.diffs: if op == diff_match_patch.DIFF_INSERT: text.append("+") elif op == diff_match_patch.DIFF_DELETE: text.append("-") elif op == diff_match_patch.DIFF_EQUAL: text.append(" ") # High ascii will raise UnicodeDecodeError. Use Unicode instead. data = data.encode("utf-8") text.append(urllib.quote(data, "!~*'();/?:@&=+$,# ") + "\n") return "".join(text) diff-match-patch-20121119/LICENCE0000640000175000017500000002367611613340077015776 0ustar lukeluke00000000000000 Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 1. Definitions. "License" shall mean the terms and conditions for use, reproduction, and distribution as defined by Sections 1 through 9 of this document. "Licensor" shall mean the copyright owner or entity authorized by the copyright owner that is granting the License. "Legal Entity" shall mean the union of the acting entity and all other entities that control, are controlled by, or are under common control with that entity. For the purposes of this definition, "control" means (i) the power, direct or indirect, to cause the direction or management of such entity, whether by contract or otherwise, or (ii) ownership of fifty percent (50%) or more of the outstanding shares, or (iii) beneficial ownership of such entity. "You" (or "Your") shall mean an individual or Legal Entity exercising permissions granted by this License. "Source" form shall mean the preferred form for making modifications, including but not limited to software source code, documentation source, and configuration files. "Object" form shall mean any form resulting from mechanical transformation or translation of a Source form, including but not limited to compiled object code, generated documentation, and conversions to other media types. "Work" shall mean the work of authorship, whether in Source or Object form, made available under the License, as indicated by a copyright notice that is included in or attached to the work (an example is provided in the Appendix below). "Derivative Works" shall mean any work, whether in Source or Object form, that is based on (or derived from) the Work and for which the editorial revisions, annotations, elaborations, or other modifications represent, as a whole, an original work of authorship. For the purposes of this License, Derivative Works shall not include works that remain separable from, or merely link (or bind by name) to the interfaces of, the Work and Derivative Works thereof. "Contribution" shall mean any work of authorship, including the original version of the Work and any modifications or additions to that Work or Derivative Works thereof, that is intentionally submitted to Licensor for inclusion in the Work by the copyright owner or by an individual or Legal Entity authorized to submit on behalf of the copyright owner. For the purposes of this definition, "submitted" means any form of electronic, verbal, or written communication sent to the Licensor or its representatives, including but not limited to communication on electronic mailing lists, source code control systems, and issue tracking systems that are managed by, or on behalf of, the Licensor for the purpose of discussing and improving the Work, but excluding communication that is conspicuously marked or otherwise designated in writing by the copyright owner as "Not a Contribution." "Contributor" shall mean Licensor and any individual or Legal Entity on behalf of whom a Contribution has been received by Licensor and subsequently incorporated within the Work. 2. Grant of Copyright License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable copyright license to reproduce, prepare Derivative Works of, publicly display, publicly perform, sublicense, and distribute the Work and such Derivative Works in Source or Object form. 3. Grant of Patent License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this section) patent license to make, have made, use, offer to sell, sell, import, and otherwise transfer the Work, where such license applies only to those patent claims licensable by such Contributor that are necessarily infringed by their Contribution(s) alone or by combination of their Contribution(s) with the Work to which such Contribution(s) was submitted. If You institute patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed. 4. Redistribution. You may reproduce and distribute copies of the Work or Derivative Works thereof in any medium, with or without modifications, and in Source or Object form, provided that You meet the following conditions: (a) You must give any other recipients of the Work or Derivative Works a copy of this License; and (b) You must cause any modified files to carry prominent notices stating that You changed the files; and (c) You must retain, in the Source form of any Derivative Works that You distribute, all copyright, patent, trademark, and attribution notices from the Source form of the Work, excluding those notices that do not pertain to any part of the Derivative Works; and (d) If the Work includes a "NOTICE" text file as part of its distribution, then any Derivative Works that You distribute must include a readable copy of the attribution notices contained within such NOTICE file, excluding those notices that do not pertain to any part of the Derivative Works, in at least one of the following places: within a NOTICE text file distributed as part of the Derivative Works; within the Source form or documentation, if provided along with the Derivative Works; or, within a display generated by the Derivative Works, if and wherever such third-party notices normally appear. The contents of the NOTICE file are for informational purposes only and do not modify the License. You may add Your own attribution notices within Derivative Works that You distribute, alongside or as an addendum to the NOTICE text from the Work, provided that such additional attribution notices cannot be construed as modifying the License. You may add Your own copyright statement to Your modifications and may provide additional or different license terms and conditions for use, reproduction, or distribution of Your modifications, or for any such Derivative Works as a whole, provided Your use, reproduction, and distribution of the Work otherwise complies with the conditions stated in this License. 5. Submission of Contributions. Unless You explicitly state otherwise, any Contribution intentionally submitted for inclusion in the Work by You to the Licensor shall be under the terms and conditions of this License, without any additional terms or conditions. Notwithstanding the above, nothing herein shall supersede or modify the terms of any separate license agreement you may have executed with Licensor regarding such Contributions. 6. Trademarks. This License does not grant permission to use the trade names, trademarks, service marks, or product names of the Licensor, except as required for reasonable and customary use in describing the origin of the Work and reproducing the content of the NOTICE file. 7. Disclaimer of Warranty. Unless required by applicable law or agreed to in writing, Licensor provides the Work (and each Contributor provides its Contributions) on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied, including, without limitation, any warranties or conditions of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A PARTICULAR PURPOSE. You are solely responsible for determining the appropriateness of using or redistributing the Work and assume any risks associated with Your exercise of permissions under this License. 8. Limitation of Liability. In no event and under no legal theory, whether in tort (including negligence), contract, or otherwise, unless required by applicable law (such as deliberate and grossly negligent acts) or agreed to in writing, shall any Contributor be liable to You for damages, including any direct, indirect, special, incidental, or consequential damages of any character arising as a result of this License or out of the use or inability to use the Work (including but not limited to damages for loss of goodwill, work stoppage, computer failure or malfunction, or any and all other commercial damages or losses), even if such Contributor has been advised of the possibility of such damages. 9. Accepting Warranty or Additional Liability. While redistributing the Work or Derivative Works thereof, You may choose to offer, and charge a fee for, acceptance of support, warranty, indemnity, or other liability obligations and/or rights consistent with this License. However, in accepting such obligations, You may act only on Your own behalf and on Your sole responsibility, not on behalf of any other Contributor, and only if You agree to indemnify, defend, and hold each Contributor harmless for any liability incurred by, or claims asserted against, such Contributor by reason of your accepting any such warranty or additional liability. END OF TERMS AND CONDITIONS diff-match-patch-20121119/MANIFEST.in0000640000175000017500000000027411632173147016537 0ustar lukeluke00000000000000include README.rst include CHANGES.rst include README.original.txt include LICENCE # Need to include the other versions stuff recursive-include python2 *.py recursive-include python3 *.py diff-match-patch-20121119/setup.cfg0000640000175000017500000000007312254032721016610 0ustar lukeluke00000000000000[egg_info] tag_build = tag_date = 0 tag_svn_revision = 0 diff-match-patch-20121119/CHANGES.rst0000640000175000017500000000054012254031310016561 0ustar lukeluke00000000000000Version 20121119 ---------------- Upstream release Version 20120106 ---------------- Upstream release Version 20111010 ---------------- Upstream release Version 20110725.1 ------------------ * Python 3 packaging bug (python3 directory excluded) * Other packaging bugs (extra files included). Version 20110725 ---------------- Initial PyPI package diff-match-patch-20121119/PKG-INFO0000640000175000017500000000375412254032721016075 0ustar lukeluke00000000000000Metadata-Version: 1.1 Name: diff-match-patch Version: 20121119 Summary: The Diff Match and Patch libraries offer robust algorithms to perform the operations required for synchronizing plain text. Home-page: http://code.google.com/p/google-diff-match-patch/ Author: Neil Fraser Author-email: UNKNOWN License: Apache Description: ================ diff-match-patch ================ Documentation on the Google site: http://code.google.com/p/google-diff-match-patch/ See README.original.txt for the original README. This is an unofficial package of the Python versions for PyPI, by Luke Plant http://lukeplant.me.uk/ See https://bitbucket.org/spookylukey/diff-match-patch for packaging issues. Version 20121119 ---------------- Upstream release Version 20120106 ---------------- Upstream release Version 20111010 ---------------- Upstream release Version 20110725.1 ------------------ * Python 3 packaging bug (python3 directory excluded) * Other packaging bugs (extra files included). Version 20110725 ---------------- Initial PyPI package Platform: UNKNOWN Classifier: Topic :: Text Processing Classifier: Intended Audience :: Developers Classifier: Development Status :: 6 - Mature Classifier: License :: OSI Approved :: Apache Software License Classifier: Programming Language :: Python :: 2.4 Classifier: Programming Language :: Python :: 2.5 Classifier: Programming Language :: Python :: 2.6 Classifier: Programming Language :: Python :: 2.7 Classifier: Programming Language :: Python :: 3 Classifier: Programming Language :: Python :: 3.0 Classifier: Programming Language :: Python :: 3.1 Classifier: Programming Language :: Python :: 3.2 Classifier: Programming Language :: Python :: 3.3 diff-match-patch-20121119/setup.py0000750000175000017500000000263512254032662016515 0ustar lukeluke00000000000000#!/usr/bin/env python from setuptools import setup, find_packages import os import sys def read(*rnames): return open(os.path.join(os.path.dirname(__file__), *rnames)).read() if sys.version_info >= (3,): subdir = 'python3' else: subdir = 'python2' setup( name='diff-match-patch', version='20121119', description = "The Diff Match and Patch libraries offer robust algorithms to perform the operations required for synchronizing plain text.", long_description = read('README.rst') + "\n\n" + read("CHANGES.rst"), packages = ['diff_match_patch'], package_dir = {'': subdir}, author='Neil Fraser', url='http://code.google.com/p/google-diff-match-patch/', test_suite='diff_match_patch.diff_match_patch_test', license = "Apache", classifiers = [ "Topic :: Text Processing", "Intended Audience :: Developers", "Development Status :: 6 - Mature", "License :: OSI Approved :: Apache Software License", "Programming Language :: Python :: 2.4", "Programming Language :: Python :: 2.5", "Programming Language :: Python :: 2.6", "Programming Language :: Python :: 2.7", "Programming Language :: Python :: 3", "Programming Language :: Python :: 3.0", "Programming Language :: Python :: 3.1", "Programming Language :: Python :: 3.2", "Programming Language :: Python :: 3.3", ] ) diff-match-patch-20121119/README.original.txt0000640000175000017500000000174311613340077020301 0ustar lukeluke00000000000000Diff, Match and Patch Library http://code.google.com/p/google-diff-match-patch/ Neil Fraser This library is currently available in seven different ports, all using the same API. Every version includes a full set of unit tests. C++: * Ported by Mike Slemmer. * Currently requires the Qt library. C#: * Ported by Matthaeus G. Chajdas. Java: * Included is both the source and a Maven package. JavaScript: * diff_match_patch_uncompressed.js is the human-readable version. * diff_match_patch.js has been compressed using Google's internal JavaScript compressor. External hackers are recommended to use http://dean.edwards.name/packer/ Lua: * Ported by Duncan Cross. * Does not support line-mode speedup. Objective C: * Ported by Jan Weiss. * Includes speed test (this is a separate bundle for other languages). Python: * Two versions, one for Python 2.x, the other for Python 3.x. * Runs 10x faster under PyPy than CPython. Demos: * Separate demos for Diff, Match and Patch in JavaScript. diff-match-patch-20121119/python3/0000750000175000017500000000000012254032721016372 5ustar lukeluke00000000000000diff-match-patch-20121119/python3/diff_match_patch/0000750000175000017500000000000012254032721021635 5ustar lukeluke00000000000000diff-match-patch-20121119/python3/diff_match_patch/diff_match_patch_test.py0000750000175000017500000012121712052423660026522 0ustar lukeluke00000000000000#!/usr/bin/python3 """Test harness for diff_match_patch.py Copyright 2006 Google Inc. http://code.google.com/p/google-diff-match-patch/ Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ import imp import sys import time import unittest import diff_match_patch as dmp_module # Force a module reload. Allows one to edit the DMP module and rerun the tests # without leaving the Python interpreter. imp.reload(dmp_module) class DiffMatchPatchTest(unittest.TestCase): def setUp(self): "Test harness for dmp_module." self.dmp = dmp_module.diff_match_patch() def diff_rebuildtexts(self, diffs): # Construct the two texts which made up the diff originally. text1 = "" text2 = "" for x in range(0, len(diffs)): if diffs[x][0] != dmp_module.diff_match_patch.DIFF_INSERT: text1 += diffs[x][1] if diffs[x][0] != dmp_module.diff_match_patch.DIFF_DELETE: text2 += diffs[x][1] return (text1, text2) class DiffTest(DiffMatchPatchTest): """DIFF TEST FUNCTIONS""" def testDiffCommonPrefix(self): # Detect any common prefix. # Null case. self.assertEqual(0, self.dmp.diff_commonPrefix("abc", "xyz")) # Non-null case. self.assertEqual(4, self.dmp.diff_commonPrefix("1234abcdef", "1234xyz")) # Whole case. self.assertEqual(4, self.dmp.diff_commonPrefix("1234", "1234xyz")) def testDiffCommonSuffix(self): # Detect any common suffix. # Null case. self.assertEqual(0, self.dmp.diff_commonSuffix("abc", "xyz")) # Non-null case. self.assertEqual(4, self.dmp.diff_commonSuffix("abcdef1234", "xyz1234")) # Whole case. self.assertEqual(4, self.dmp.diff_commonSuffix("1234", "xyz1234")) def testDiffCommonOverlap(self): # Null case. self.assertEqual(0, self.dmp.diff_commonOverlap("", "abcd")) # Whole case. self.assertEqual(3, self.dmp.diff_commonOverlap("abc", "abcd")) # No overlap. self.assertEqual(0, self.dmp.diff_commonOverlap("123456", "abcd")) # Overlap. self.assertEqual(3, self.dmp.diff_commonOverlap("123456xxx", "xxxabcd")) # Unicode. # Some overly clever languages (C#) may treat ligatures as equal to their # component letters. E.g. U+FB01 == 'fi' self.assertEqual(0, self.dmp.diff_commonOverlap("fi", "\ufb01i")) def testDiffHalfMatch(self): # Detect a halfmatch. self.dmp.Diff_Timeout = 1 # No match. self.assertEqual(None, self.dmp.diff_halfMatch("1234567890", "abcdef")) self.assertEqual(None, self.dmp.diff_halfMatch("12345", "23")) # Single Match. self.assertEqual(("12", "90", "a", "z", "345678"), self.dmp.diff_halfMatch("1234567890", "a345678z")) self.assertEqual(("a", "z", "12", "90", "345678"), self.dmp.diff_halfMatch("a345678z", "1234567890")) self.assertEqual(("abc", "z", "1234", "0", "56789"), self.dmp.diff_halfMatch("abc56789z", "1234567890")) self.assertEqual(("a", "xyz", "1", "7890", "23456"), self.dmp.diff_halfMatch("a23456xyz", "1234567890")) # Multiple Matches. self.assertEqual(("12123", "123121", "a", "z", "1234123451234"), self.dmp.diff_halfMatch("121231234123451234123121", "a1234123451234z")) self.assertEqual(("", "-=-=-=-=-=", "x", "", "x-=-=-=-=-=-=-="), self.dmp.diff_halfMatch("x-=-=-=-=-=-=-=-=-=-=-=-=", "xx-=-=-=-=-=-=-=")) self.assertEqual(("-=-=-=-=-=", "", "", "y", "-=-=-=-=-=-=-=y"), self.dmp.diff_halfMatch("-=-=-=-=-=-=-=-=-=-=-=-=y", "-=-=-=-=-=-=-=yy")) # Non-optimal halfmatch. # Optimal diff would be -q+x=H-i+e=lloHe+Hu=llo-Hew+y not -qHillo+x=HelloHe-w+Hulloy self.assertEqual(("qHillo", "w", "x", "Hulloy", "HelloHe"), self.dmp.diff_halfMatch("qHilloHelloHew", "xHelloHeHulloy")) # Optimal no halfmatch. self.dmp.Diff_Timeout = 0 self.assertEqual(None, self.dmp.diff_halfMatch("qHilloHelloHew", "xHelloHeHulloy")) def testDiffLinesToChars(self): # Convert lines down to characters. self.assertEqual(("\x01\x02\x01", "\x02\x01\x02", ["", "alpha\n", "beta\n"]), self.dmp.diff_linesToChars("alpha\nbeta\nalpha\n", "beta\nalpha\nbeta\n")) self.assertEqual(("", "\x01\x02\x03\x03", ["", "alpha\r\n", "beta\r\n", "\r\n"]), self.dmp.diff_linesToChars("", "alpha\r\nbeta\r\n\r\n\r\n")) self.assertEqual(("\x01", "\x02", ["", "a", "b"]), self.dmp.diff_linesToChars("a", "b")) # More than 256 to reveal any 8-bit limitations. n = 300 lineList = [] charList = [] for x in range(1, n + 1): lineList.append(str(x) + "\n") charList.append(chr(x)) self.assertEqual(n, len(lineList)) lines = "".join(lineList) chars = "".join(charList) self.assertEqual(n, len(chars)) lineList.insert(0, "") self.assertEqual((chars, "", lineList), self.dmp.diff_linesToChars(lines, "")) def testDiffCharsToLines(self): # Convert chars up to lines. diffs = [(self.dmp.DIFF_EQUAL, "\x01\x02\x01"), (self.dmp.DIFF_INSERT, "\x02\x01\x02")] self.dmp.diff_charsToLines(diffs, ["", "alpha\n", "beta\n"]) self.assertEqual([(self.dmp.DIFF_EQUAL, "alpha\nbeta\nalpha\n"), (self.dmp.DIFF_INSERT, "beta\nalpha\nbeta\n")], diffs) # More than 256 to reveal any 8-bit limitations. n = 300 lineList = [] charList = [] for x in range(1, n + 1): lineList.append(str(x) + "\n") charList.append(chr(x)) self.assertEqual(n, len(lineList)) lines = "".join(lineList) chars = "".join(charList) self.assertEqual(n, len(chars)) lineList.insert(0, "") diffs = [(self.dmp.DIFF_DELETE, chars)] self.dmp.diff_charsToLines(diffs, lineList) self.assertEqual([(self.dmp.DIFF_DELETE, lines)], diffs) def testDiffCleanupMerge(self): # Cleanup a messy diff. # Null case. diffs = [] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([], diffs) # No change case. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_INSERT, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_INSERT, "c")], diffs) # Merge equalities. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_EQUAL, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "abc")], diffs) # Merge deletions. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_DELETE, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abc")], diffs) # Merge insertions. diffs = [(self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_INSERT, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_INSERT, "abc")], diffs) # Merge interweave. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_DELETE, "c"), (self.dmp.DIFF_INSERT, "d"), (self.dmp.DIFF_EQUAL, "e"), (self.dmp.DIFF_EQUAL, "f")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "ac"), (self.dmp.DIFF_INSERT, "bd"), (self.dmp.DIFF_EQUAL, "ef")], diffs) # Prefix and suffix detection. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "abc"), (self.dmp.DIFF_DELETE, "dc")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "d"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_EQUAL, "c")], diffs) # Prefix and suffix detection with equalities. diffs = [(self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "abc"), (self.dmp.DIFF_DELETE, "dc"), (self.dmp.DIFF_EQUAL, "y")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "xa"), (self.dmp.DIFF_DELETE, "d"), (self.dmp.DIFF_INSERT, "b"), (self.dmp.DIFF_EQUAL, "cy")], diffs) # Slide edit left. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_INSERT, "ba"), (self.dmp.DIFF_EQUAL, "c")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_INSERT, "ab"), (self.dmp.DIFF_EQUAL, "ac")], diffs) # Slide edit right. diffs = [(self.dmp.DIFF_EQUAL, "c"), (self.dmp.DIFF_INSERT, "ab"), (self.dmp.DIFF_EQUAL, "a")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "ca"), (self.dmp.DIFF_INSERT, "ba")], diffs) # Slide edit left recursive. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_EQUAL, "c"), (self.dmp.DIFF_DELETE, "ac"), (self.dmp.DIFF_EQUAL, "x")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_EQUAL, "acx")], diffs) # Slide edit right recursive. diffs = [(self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "ca"), (self.dmp.DIFF_EQUAL, "c"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_EQUAL, "a")] self.dmp.diff_cleanupMerge(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "xca"), (self.dmp.DIFF_DELETE, "cba")], diffs) def testDiffCleanupSemanticLossless(self): # Slide diffs to match logical boundaries. # Null case. diffs = [] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([], diffs) # Blank lines. diffs = [(self.dmp.DIFF_EQUAL, "AAA\r\n\r\nBBB"), (self.dmp.DIFF_INSERT, "\r\nDDD\r\n\r\nBBB"), (self.dmp.DIFF_EQUAL, "\r\nEEE")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "AAA\r\n\r\n"), (self.dmp.DIFF_INSERT, "BBB\r\nDDD\r\n\r\n"), (self.dmp.DIFF_EQUAL, "BBB\r\nEEE")], diffs) # Line boundaries. diffs = [(self.dmp.DIFF_EQUAL, "AAA\r\nBBB"), (self.dmp.DIFF_INSERT, " DDD\r\nBBB"), (self.dmp.DIFF_EQUAL, " EEE")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "AAA\r\n"), (self.dmp.DIFF_INSERT, "BBB DDD\r\n"), (self.dmp.DIFF_EQUAL, "BBB EEE")], diffs) # Word boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The c"), (self.dmp.DIFF_INSERT, "ow and the c"), (self.dmp.DIFF_EQUAL, "at.")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "The "), (self.dmp.DIFF_INSERT, "cow and the "), (self.dmp.DIFF_EQUAL, "cat.")], diffs) # Alphanumeric boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The-c"), (self.dmp.DIFF_INSERT, "ow-and-the-c"), (self.dmp.DIFF_EQUAL, "at.")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "The-"), (self.dmp.DIFF_INSERT, "cow-and-the-"), (self.dmp.DIFF_EQUAL, "cat.")], diffs) # Hitting the start. diffs = [(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "ax")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "aax")], diffs) # Hitting the end. diffs = [(self.dmp.DIFF_EQUAL, "xa"), (self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "a")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "xaa"), (self.dmp.DIFF_DELETE, "a")], diffs) # Sentence boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The xxx. The "), (self.dmp.DIFF_INSERT, "zzz. The "), (self.dmp.DIFF_EQUAL, "yyy.")] self.dmp.diff_cleanupSemanticLossless(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "The xxx."), (self.dmp.DIFF_INSERT, " The zzz."), (self.dmp.DIFF_EQUAL, " The yyy.")], diffs) def testDiffCleanupSemantic(self): # Cleanup semantically trivial equalities. # Null case. diffs = [] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([], diffs) # No elimination #1. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "cd"), (self.dmp.DIFF_EQUAL, "12"), (self.dmp.DIFF_DELETE, "e")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "cd"), (self.dmp.DIFF_EQUAL, "12"), (self.dmp.DIFF_DELETE, "e")], diffs) # No elimination #2. diffs = [(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "ABC"), (self.dmp.DIFF_EQUAL, "1234"), (self.dmp.DIFF_DELETE, "wxyz")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "ABC"), (self.dmp.DIFF_EQUAL, "1234"), (self.dmp.DIFF_DELETE, "wxyz")], diffs) # Simple elimination. diffs = [(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_DELETE, "c")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "b")], diffs) # Backpass elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_EQUAL, "cd"), (self.dmp.DIFF_DELETE, "e"), (self.dmp.DIFF_EQUAL, "f"), (self.dmp.DIFF_INSERT, "g")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abcdef"), (self.dmp.DIFF_INSERT, "cdfg")], diffs) # Multiple eliminations. diffs = [(self.dmp.DIFF_INSERT, "1"), (self.dmp.DIFF_EQUAL, "A"), (self.dmp.DIFF_DELETE, "B"), (self.dmp.DIFF_INSERT, "2"), (self.dmp.DIFF_EQUAL, "_"), (self.dmp.DIFF_INSERT, "1"), (self.dmp.DIFF_EQUAL, "A"), (self.dmp.DIFF_DELETE, "B"), (self.dmp.DIFF_INSERT, "2")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "AB_AB"), (self.dmp.DIFF_INSERT, "1A2_1A2")], diffs) # Word boundaries. diffs = [(self.dmp.DIFF_EQUAL, "The c"), (self.dmp.DIFF_DELETE, "ow and the c"), (self.dmp.DIFF_EQUAL, "at.")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_EQUAL, "The "), (self.dmp.DIFF_DELETE, "cow and the "), (self.dmp.DIFF_EQUAL, "cat.")], diffs) # No overlap elimination. diffs = [(self.dmp.DIFF_DELETE, "abcxx"), (self.dmp.DIFF_INSERT, "xxdef")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abcxx"), (self.dmp.DIFF_INSERT, "xxdef")], diffs) # Overlap elimination. diffs = [(self.dmp.DIFF_DELETE, "abcxxx"), (self.dmp.DIFF_INSERT, "xxxdef")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_EQUAL, "xxx"), (self.dmp.DIFF_INSERT, "def")], diffs) # Reverse overlap elimination. diffs = [(self.dmp.DIFF_DELETE, "xxxabc"), (self.dmp.DIFF_INSERT, "defxxx")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_INSERT, "def"), (self.dmp.DIFF_EQUAL, "xxx"), (self.dmp.DIFF_DELETE, "abc")], diffs) # Two overlap eliminations. diffs = [(self.dmp.DIFF_DELETE, "abcd1212"), (self.dmp.DIFF_INSERT, "1212efghi"), (self.dmp.DIFF_EQUAL, "----"), (self.dmp.DIFF_DELETE, "A3"), (self.dmp.DIFF_INSERT, "3BC")] self.dmp.diff_cleanupSemantic(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abcd"), (self.dmp.DIFF_EQUAL, "1212"), (self.dmp.DIFF_INSERT, "efghi"), (self.dmp.DIFF_EQUAL, "----"), (self.dmp.DIFF_DELETE, "A"), (self.dmp.DIFF_EQUAL, "3"), (self.dmp.DIFF_INSERT, "BC")], diffs) def testDiffCleanupEfficiency(self): # Cleanup operationally trivial equalities. self.dmp.Diff_EditCost = 4 # Null case. diffs = [] self.dmp.diff_cleanupEfficiency(diffs) self.assertEqual([], diffs) # No elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "wxyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "wxyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")], diffs) # Four-edit elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "xyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abxyzcd"), (self.dmp.DIFF_INSERT, "12xyz34")], diffs) # Three-edit elimination. diffs = [(self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "xcd"), (self.dmp.DIFF_INSERT, "12x34")], diffs) # Backpass elimination. diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "xy"), (self.dmp.DIFF_INSERT, "34"), (self.dmp.DIFF_EQUAL, "z"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "56")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abxyzcd"), (self.dmp.DIFF_INSERT, "12xy34z56")], diffs) # High cost elimination. self.dmp.Diff_EditCost = 5 diffs = [(self.dmp.DIFF_DELETE, "ab"), (self.dmp.DIFF_INSERT, "12"), (self.dmp.DIFF_EQUAL, "wxyz"), (self.dmp.DIFF_DELETE, "cd"), (self.dmp.DIFF_INSERT, "34")] self.dmp.diff_cleanupEfficiency(diffs) self.assertEqual([(self.dmp.DIFF_DELETE, "abwxyzcd"), (self.dmp.DIFF_INSERT, "12wxyz34")], diffs) self.dmp.Diff_EditCost = 4 def testDiffPrettyHtml(self): # Pretty print. diffs = [(self.dmp.DIFF_EQUAL, "a\n"), (self.dmp.DIFF_DELETE, "b"), (self.dmp.DIFF_INSERT, "c&d")] self.assertEqual("
<B>b</B>c&d", self.dmp.diff_prettyHtml(diffs)) def testDiffText(self): # Compute the source and destination texts. diffs = [(self.dmp.DIFF_EQUAL, "jump"), (self.dmp.DIFF_DELETE, "s"), (self.dmp.DIFF_INSERT, "ed"), (self.dmp.DIFF_EQUAL, " over "), (self.dmp.DIFF_DELETE, "the"), (self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_EQUAL, " lazy")] self.assertEqual("jumps over the lazy", self.dmp.diff_text1(diffs)) self.assertEqual("jumped over a lazy", self.dmp.diff_text2(diffs)) def testDiffDelta(self): # Convert a diff into delta string. diffs = [(self.dmp.DIFF_EQUAL, "jump"), (self.dmp.DIFF_DELETE, "s"), (self.dmp.DIFF_INSERT, "ed"), (self.dmp.DIFF_EQUAL, " over "), (self.dmp.DIFF_DELETE, "the"), (self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_EQUAL, " lazy"), (self.dmp.DIFF_INSERT, "old dog")] text1 = self.dmp.diff_text1(diffs) self.assertEqual("jumps over the lazy", text1) delta = self.dmp.diff_toDelta(diffs) self.assertEqual("=4\t-1\t+ed\t=6\t-3\t+a\t=5\t+old dog", delta) # Convert delta string into a diff. self.assertEqual(diffs, self.dmp.diff_fromDelta(text1, delta)) # Generates error (19 != 20). try: self.dmp.diff_fromDelta(text1 + "x", delta) self.assertFalse(True) except ValueError: # Exception expected. pass # Generates error (19 != 18). try: self.dmp.diff_fromDelta(text1[1:], delta) self.assertFalse(True) except ValueError: # Exception expected. pass # Generates error (%c3%xy invalid Unicode). # Note: Python 3 can decode this. #try: # self.dmp.diff_fromDelta("", "+%c3xy") # self.assertFalse(True) #except ValueError: # # Exception expected. # pass # Test deltas with special characters. diffs = [(self.dmp.DIFF_EQUAL, "\u0680 \x00 \t %"), (self.dmp.DIFF_DELETE, "\u0681 \x01 \n ^"), (self.dmp.DIFF_INSERT, "\u0682 \x02 \\ |")] text1 = self.dmp.diff_text1(diffs) self.assertEqual("\u0680 \x00 \t %\u0681 \x01 \n ^", text1) delta = self.dmp.diff_toDelta(diffs) self.assertEqual("=7\t-7\t+%DA%82 %02 %5C %7C", delta) # Convert delta string into a diff. self.assertEqual(diffs, self.dmp.diff_fromDelta(text1, delta)) # Verify pool of unchanged characters. diffs = [(self.dmp.DIFF_INSERT, "A-Z a-z 0-9 - _ . ! ~ * ' ( ) ; / ? : @ & = + $ , # ")] text2 = self.dmp.diff_text2(diffs) self.assertEqual("A-Z a-z 0-9 - _ . ! ~ * \' ( ) ; / ? : @ & = + $ , # ", text2) delta = self.dmp.diff_toDelta(diffs) self.assertEqual("+A-Z a-z 0-9 - _ . ! ~ * \' ( ) ; / ? : @ & = + $ , # ", delta) # Convert delta string into a diff. self.assertEqual(diffs, self.dmp.diff_fromDelta("", delta)) def testDiffXIndex(self): # Translate a location in text1 to text2. self.assertEqual(5, self.dmp.diff_xIndex([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "1234"), (self.dmp.DIFF_EQUAL, "xyz")], 2)) # Translation on deletion. self.assertEqual(1, self.dmp.diff_xIndex([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "1234"), (self.dmp.DIFF_EQUAL, "xyz")], 3)) def testDiffLevenshtein(self): # Levenshtein with trailing equality. self.assertEqual(4, self.dmp.diff_levenshtein([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "1234"), (self.dmp.DIFF_EQUAL, "xyz")])) # Levenshtein with leading equality. self.assertEqual(4, self.dmp.diff_levenshtein([(self.dmp.DIFF_EQUAL, "xyz"), (self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_INSERT, "1234")])) # Levenshtein with middle equality. self.assertEqual(7, self.dmp.diff_levenshtein([(self.dmp.DIFF_DELETE, "abc"), (self.dmp.DIFF_EQUAL, "xyz"), (self.dmp.DIFF_INSERT, "1234")])) def testDiffBisect(self): # Normal. a = "cat" b = "map" # Since the resulting diff hasn't been normalized, it would be ok if # the insertion and deletion pairs are swapped. # If the order changes, tweak this test as required. self.assertEqual([(self.dmp.DIFF_DELETE, "c"), (self.dmp.DIFF_INSERT, "m"), (self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "t"), (self.dmp.DIFF_INSERT, "p")], self.dmp.diff_bisect(a, b, sys.maxsize)) # Timeout. self.assertEqual([(self.dmp.DIFF_DELETE, "cat"), (self.dmp.DIFF_INSERT, "map")], self.dmp.diff_bisect(a, b, 0)) def testDiffMain(self): # Perform a trivial diff. # Null case. self.assertEqual([], self.dmp.diff_main("", "", False)) # Equality. self.assertEqual([(self.dmp.DIFF_EQUAL, "abc")], self.dmp.diff_main("abc", "abc", False)) # Simple insertion. self.assertEqual([(self.dmp.DIFF_EQUAL, "ab"), (self.dmp.DIFF_INSERT, "123"), (self.dmp.DIFF_EQUAL, "c")], self.dmp.diff_main("abc", "ab123c", False)) # Simple deletion. self.assertEqual([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "123"), (self.dmp.DIFF_EQUAL, "bc")], self.dmp.diff_main("a123bc", "abc", False)) # Two insertions. self.assertEqual([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_INSERT, "123"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_INSERT, "456"), (self.dmp.DIFF_EQUAL, "c")], self.dmp.diff_main("abc", "a123b456c", False)) # Two deletions. self.assertEqual([(self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "123"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_DELETE, "456"), (self.dmp.DIFF_EQUAL, "c")], self.dmp.diff_main("a123b456c", "abc", False)) # Perform a real diff. # Switch off the timeout. self.dmp.Diff_Timeout = 0 # Simple cases. self.assertEqual([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "b")], self.dmp.diff_main("a", "b", False)) self.assertEqual([(self.dmp.DIFF_DELETE, "Apple"), (self.dmp.DIFF_INSERT, "Banana"), (self.dmp.DIFF_EQUAL, "s are a"), (self.dmp.DIFF_INSERT, "lso"), (self.dmp.DIFF_EQUAL, " fruit.")], self.dmp.diff_main("Apples are a fruit.", "Bananas are also fruit.", False)) self.assertEqual([(self.dmp.DIFF_DELETE, "a"), (self.dmp.DIFF_INSERT, "\u0680"), (self.dmp.DIFF_EQUAL, "x"), (self.dmp.DIFF_DELETE, "\t"), (self.dmp.DIFF_INSERT, "\x00")], self.dmp.diff_main("ax\t", "\u0680x\x00", False)) # Overlaps. self.assertEqual([(self.dmp.DIFF_DELETE, "1"), (self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "y"), (self.dmp.DIFF_EQUAL, "b"), (self.dmp.DIFF_DELETE, "2"), (self.dmp.DIFF_INSERT, "xab")], self.dmp.diff_main("1ayb2", "abxab", False)) self.assertEqual([(self.dmp.DIFF_INSERT, "xaxcx"), (self.dmp.DIFF_EQUAL, "abc"), (self.dmp.DIFF_DELETE, "y")], self.dmp.diff_main("abcy", "xaxcxabc", False)) self.assertEqual([(self.dmp.DIFF_DELETE, "ABCD"), (self.dmp.DIFF_EQUAL, "a"), (self.dmp.DIFF_DELETE, "="), (self.dmp.DIFF_INSERT, "-"), (self.dmp.DIFF_EQUAL, "bcd"), (self.dmp.DIFF_DELETE, "="), (self.dmp.DIFF_INSERT, "-"), (self.dmp.DIFF_EQUAL, "efghijklmnopqrs"), (self.dmp.DIFF_DELETE, "EFGHIJKLMNOefg")], self.dmp.diff_main("ABCDa=bcd=efghijklmnopqrsEFGHIJKLMNOefg", "a-bcd-efghijklmnopqrs", False)) # Large equality. self.assertEqual([(self.dmp.DIFF_INSERT, " "), (self.dmp.DIFF_EQUAL,"a"), (self.dmp.DIFF_INSERT,"nd"), (self.dmp.DIFF_EQUAL," [[Pennsylvania]]"), (self.dmp.DIFF_DELETE," and [[New")], self.dmp.diff_main("a [[Pennsylvania]] and [[New", " and [[Pennsylvania]]", False)) # Timeout. self.dmp.Diff_Timeout = 0.1 # 100ms a = "`Twas brillig, and the slithy toves\nDid gyre and gimble in the wabe:\nAll mimsy were the borogoves,\nAnd the mome raths outgrabe.\n" b = "I am the very model of a modern major general,\nI've information vegetable, animal, and mineral,\nI know the kings of England, and I quote the fights historical,\nFrom Marathon to Waterloo, in order categorical.\n" # Increase the text lengths by 1024 times to ensure a timeout. for x in range(10): a = a + a b = b + b startTime = time.time() self.dmp.diff_main(a, b) endTime = time.time() # Test that we took at least the timeout period. self.assertTrue(self.dmp.Diff_Timeout <= endTime - startTime) # Test that we didn't take forever (be forgiving). # Theoretically this test could fail very occasionally if the # OS task swaps or locks up for a second at the wrong moment. self.assertTrue(self.dmp.Diff_Timeout * 2 > endTime - startTime) self.dmp.Diff_Timeout = 0 # Test the linemode speedup. # Must be long to pass the 100 char cutoff. # Simple line-mode. a = "1234567890\n" * 13 b = "abcdefghij\n" * 13 self.assertEqual(self.dmp.diff_main(a, b, False), self.dmp.diff_main(a, b, True)) # Single line-mode. a = "1234567890" * 13 b = "abcdefghij" * 13 self.assertEqual(self.dmp.diff_main(a, b, False), self.dmp.diff_main(a, b, True)) # Overlap line-mode. a = "1234567890\n" * 13 b = "abcdefghij\n1234567890\n1234567890\n1234567890\nabcdefghij\n1234567890\n1234567890\n1234567890\nabcdefghij\n1234567890\n1234567890\n1234567890\nabcdefghij\n" texts_linemode = self.diff_rebuildtexts(self.dmp.diff_main(a, b, True)) texts_textmode = self.diff_rebuildtexts(self.dmp.diff_main(a, b, False)) self.assertEqual(texts_textmode, texts_linemode) # Test null inputs. try: self.dmp.diff_main(None, None) self.assertFalse(True) except ValueError: # Exception expected. pass class MatchTest(DiffMatchPatchTest): """MATCH TEST FUNCTIONS""" def testMatchAlphabet(self): # Initialise the bitmasks for Bitap. self.assertEqual({"a":4, "b":2, "c":1}, self.dmp.match_alphabet("abc")) self.assertEqual({"a":37, "b":18, "c":8}, self.dmp.match_alphabet("abcaba")) def testMatchBitap(self): self.dmp.Match_Distance = 100 self.dmp.Match_Threshold = 0.5 # Exact matches. self.assertEqual(5, self.dmp.match_bitap("abcdefghijk", "fgh", 5)) self.assertEqual(5, self.dmp.match_bitap("abcdefghijk", "fgh", 0)) # Fuzzy matches. self.assertEqual(4, self.dmp.match_bitap("abcdefghijk", "efxhi", 0)) self.assertEqual(2, self.dmp.match_bitap("abcdefghijk", "cdefxyhijk", 5)) self.assertEqual(-1, self.dmp.match_bitap("abcdefghijk", "bxy", 1)) # Overflow. self.assertEqual(2, self.dmp.match_bitap("123456789xx0", "3456789x0", 2)) self.assertEqual(0, self.dmp.match_bitap("abcdef", "xxabc", 4)) self.assertEqual(3, self.dmp.match_bitap("abcdef", "defyy", 4)) self.assertEqual(0, self.dmp.match_bitap("abcdef", "xabcdefy", 0)) # Threshold test. self.dmp.Match_Threshold = 0.4 self.assertEqual(4, self.dmp.match_bitap("abcdefghijk", "efxyhi", 1)) self.dmp.Match_Threshold = 0.3 self.assertEqual(-1, self.dmp.match_bitap("abcdefghijk", "efxyhi", 1)) self.dmp.Match_Threshold = 0.0 self.assertEqual(1, self.dmp.match_bitap("abcdefghijk", "bcdef", 1)) self.dmp.Match_Threshold = 0.5 # Multiple select. self.assertEqual(0, self.dmp.match_bitap("abcdexyzabcde", "abccde", 3)) self.assertEqual(8, self.dmp.match_bitap("abcdexyzabcde", "abccde", 5)) # Distance test. self.dmp.Match_Distance = 10 # Strict location. self.assertEqual(-1, self.dmp.match_bitap("abcdefghijklmnopqrstuvwxyz", "abcdefg", 24)) self.assertEqual(0, self.dmp.match_bitap("abcdefghijklmnopqrstuvwxyz", "abcdxxefg", 1)) self.dmp.Match_Distance = 1000 # Loose location. self.assertEqual(0, self.dmp.match_bitap("abcdefghijklmnopqrstuvwxyz", "abcdefg", 24)) def testMatchMain(self): # Full match. # Shortcut matches. self.assertEqual(0, self.dmp.match_main("abcdef", "abcdef", 1000)) self.assertEqual(-1, self.dmp.match_main("", "abcdef", 1)) self.assertEqual(3, self.dmp.match_main("abcdef", "", 3)) self.assertEqual(3, self.dmp.match_main("abcdef", "de", 3)) self.assertEqual(3, self.dmp.match_main("abcdef", "defy", 4)) self.assertEqual(0, self.dmp.match_main("abcdef", "abcdefy", 0)) # Complex match. self.dmp.Match_Threshold = 0.7 self.assertEqual(4, self.dmp.match_main("I am the very model of a modern major general.", " that berry ", 5)) self.dmp.Match_Threshold = 0.5 # Test null inputs. try: self.dmp.match_main(None, None, 0) self.assertFalse(True) except ValueError: # Exception expected. pass class PatchTest(DiffMatchPatchTest): """PATCH TEST FUNCTIONS""" def testPatchObj(self): # Patch Object. p = dmp_module.patch_obj() p.start1 = 20 p.start2 = 21 p.length1 = 18 p.length2 = 17 p.diffs = [(self.dmp.DIFF_EQUAL, "jump"), (self.dmp.DIFF_DELETE, "s"), (self.dmp.DIFF_INSERT, "ed"), (self.dmp.DIFF_EQUAL, " over "), (self.dmp.DIFF_DELETE, "the"), (self.dmp.DIFF_INSERT, "a"), (self.dmp.DIFF_EQUAL, "\nlaz")] strp = str(p) self.assertEqual("@@ -21,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n %0Alaz\n", strp) def testPatchFromText(self): self.assertEqual([], self.dmp.patch_fromText("")) strp = "@@ -21,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n %0Alaz\n" self.assertEqual(strp, str(self.dmp.patch_fromText(strp)[0])) self.assertEqual("@@ -1 +1 @@\n-a\n+b\n", str(self.dmp.patch_fromText("@@ -1 +1 @@\n-a\n+b\n")[0])) self.assertEqual("@@ -1,3 +0,0 @@\n-abc\n", str(self.dmp.patch_fromText("@@ -1,3 +0,0 @@\n-abc\n")[0])) self.assertEqual("@@ -0,0 +1,3 @@\n+abc\n", str(self.dmp.patch_fromText("@@ -0,0 +1,3 @@\n+abc\n")[0])) # Generates error. try: self.dmp.patch_fromText("Bad\nPatch\n") self.assertFalse(True) except ValueError: # Exception expected. pass def testPatchToText(self): strp = "@@ -21,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n laz\n" p = self.dmp.patch_fromText(strp) self.assertEqual(strp, self.dmp.patch_toText(p)) strp = "@@ -1,9 +1,9 @@\n-f\n+F\n oo+fooba\n@@ -7,9 +7,9 @@\n obar\n-,\n+.\n tes\n" p = self.dmp.patch_fromText(strp) self.assertEqual(strp, self.dmp.patch_toText(p)) def testPatchAddContext(self): self.dmp.Patch_Margin = 4 p = self.dmp.patch_fromText("@@ -21,4 +21,10 @@\n-jump\n+somersault\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps over the lazy dog.") self.assertEqual("@@ -17,12 +17,18 @@\n fox \n-jump\n+somersault\n s ov\n", str(p)) # Same, but not enough trailing context. p = self.dmp.patch_fromText("@@ -21,4 +21,10 @@\n-jump\n+somersault\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps.") self.assertEqual("@@ -17,10 +17,16 @@\n fox \n-jump\n+somersault\n s.\n", str(p)) # Same, but not enough leading context. p = self.dmp.patch_fromText("@@ -3 +3,2 @@\n-e\n+at\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps.") self.assertEqual("@@ -1,7 +1,8 @@\n Th\n-e\n+at\n qui\n", str(p)) # Same, but with ambiguity. p = self.dmp.patch_fromText("@@ -3 +3,2 @@\n-e\n+at\n")[0] self.dmp.patch_addContext(p, "The quick brown fox jumps. The quick brown fox crashes.") self.assertEqual("@@ -1,27 +1,28 @@\n Th\n-e\n+at\n quick brown fox jumps. \n", str(p)) def testPatchMake(self): # Null case. patches = self.dmp.patch_make("", "") self.assertEqual("", self.dmp.patch_toText(patches)) text1 = "The quick brown fox jumps over the lazy dog." text2 = "That quick brown fox jumped over a lazy dog." # Text2+Text1 inputs. expectedPatch = "@@ -1,8 +1,7 @@\n Th\n-at\n+e\n qui\n@@ -21,17 +21,18 @@\n jump\n-ed\n+s\n over \n-a\n+the\n laz\n" # The second patch must be "-21,17 +21,18", not "-22,17 +21,18" due to rolling context. patches = self.dmp.patch_make(text2, text1) self.assertEqual(expectedPatch, self.dmp.patch_toText(patches)) # Text1+Text2 inputs. expectedPatch = "@@ -1,11 +1,12 @@\n Th\n-e\n+at\n quick b\n@@ -22,18 +22,17 @@\n jump\n-s\n+ed\n over \n-the\n+a\n laz\n" patches = self.dmp.patch_make(text1, text2) self.assertEqual(expectedPatch, self.dmp.patch_toText(patches)) # Diff input. diffs = self.dmp.diff_main(text1, text2, False) patches = self.dmp.patch_make(diffs) self.assertEqual(expectedPatch, self.dmp.patch_toText(patches)) # Text1+Diff inputs. patches = self.dmp.patch_make(text1, diffs) self.assertEqual(expectedPatch, self.dmp.patch_toText(patches)) # Text1+Text2+Diff inputs (deprecated). patches = self.dmp.patch_make(text1, text2, diffs) self.assertEqual(expectedPatch, self.dmp.patch_toText(patches)) # Character encoding. patches = self.dmp.patch_make("`1234567890-=[]\\;',./", "~!@#$%^&*()_+{}|:\"<>?") self.assertEqual("@@ -1,21 +1,21 @@\n-%601234567890-=%5B%5D%5C;',./\n+~!@#$%25%5E&*()_+%7B%7D%7C:%22%3C%3E?\n", self.dmp.patch_toText(patches)) # Character decoding. diffs = [(self.dmp.DIFF_DELETE, "`1234567890-=[]\\;',./"), (self.dmp.DIFF_INSERT, "~!@#$%^&*()_+{}|:\"<>?")] self.assertEqual(diffs, self.dmp.patch_fromText("@@ -1,21 +1,21 @@\n-%601234567890-=%5B%5D%5C;',./\n+~!@#$%25%5E&*()_+%7B%7D%7C:%22%3C%3E?\n")[0].diffs) # Long string with repeats. text1 = "" for x in range(100): text1 += "abcdef" text2 = text1 + "123" expectedPatch = "@@ -573,28 +573,31 @@\n cdefabcdefabcdefabcdefabcdef\n+123\n" patches = self.dmp.patch_make(text1, text2) self.assertEqual(expectedPatch, self.dmp.patch_toText(patches)) # Test null inputs. try: self.dmp.patch_make(None, None) self.assertFalse(True) except ValueError: # Exception expected. pass def testPatchSplitMax(self): # Assumes that Match_MaxBits is 32. patches = self.dmp.patch_make("abcdefghijklmnopqrstuvwxyz01234567890", "XabXcdXefXghXijXklXmnXopXqrXstXuvXwxXyzX01X23X45X67X89X0") self.dmp.patch_splitMax(patches) self.assertEqual("@@ -1,32 +1,46 @@\n+X\n ab\n+X\n cd\n+X\n ef\n+X\n gh\n+X\n ij\n+X\n kl\n+X\n mn\n+X\n op\n+X\n qr\n+X\n st\n+X\n uv\n+X\n wx\n+X\n yz\n+X\n 012345\n@@ -25,13 +39,18 @@\n zX01\n+X\n 23\n+X\n 45\n+X\n 67\n+X\n 89\n+X\n 0\n", self.dmp.patch_toText(patches)) patches = self.dmp.patch_make("abcdef1234567890123456789012345678901234567890123456789012345678901234567890uvwxyz", "abcdefuvwxyz") oldToText = self.dmp.patch_toText(patches) self.dmp.patch_splitMax(patches) self.assertEqual(oldToText, self.dmp.patch_toText(patches)) patches = self.dmp.patch_make("1234567890123456789012345678901234567890123456789012345678901234567890", "abc") self.dmp.patch_splitMax(patches) self.assertEqual("@@ -1,32 +1,4 @@\n-1234567890123456789012345678\n 9012\n@@ -29,32 +1,4 @@\n-9012345678901234567890123456\n 7890\n@@ -57,14 +1,3 @@\n-78901234567890\n+abc\n", self.dmp.patch_toText(patches)) patches = self.dmp.patch_make("abcdefghij , h : 0 , t : 1 abcdefghij , h : 0 , t : 1 abcdefghij , h : 0 , t : 1", "abcdefghij , h : 1 , t : 1 abcdefghij , h : 1 , t : 1 abcdefghij , h : 0 , t : 1") self.dmp.patch_splitMax(patches) self.assertEqual("@@ -2,32 +2,32 @@\n bcdefghij , h : \n-0\n+1\n , t : 1 abcdef\n@@ -29,32 +29,32 @@\n bcdefghij , h : \n-0\n+1\n , t : 1 abcdef\n", self.dmp.patch_toText(patches)) def testPatchAddPadding(self): # Both edges full. patches = self.dmp.patch_make("", "test") self.assertEqual("@@ -0,0 +1,4 @@\n+test\n", self.dmp.patch_toText(patches)) self.dmp.patch_addPadding(patches) self.assertEqual("@@ -1,8 +1,12 @@\n %01%02%03%04\n+test\n %01%02%03%04\n", self.dmp.patch_toText(patches)) # Both edges partial. patches = self.dmp.patch_make("XY", "XtestY") self.assertEqual("@@ -1,2 +1,6 @@\n X\n+test\n Y\n", self.dmp.patch_toText(patches)) self.dmp.patch_addPadding(patches) self.assertEqual("@@ -2,8 +2,12 @@\n %02%03%04X\n+test\n Y%01%02%03\n", self.dmp.patch_toText(patches)) # Both edges none. patches = self.dmp.patch_make("XXXXYYYY", "XXXXtestYYYY") self.assertEqual("@@ -1,8 +1,12 @@\n XXXX\n+test\n YYYY\n", self.dmp.patch_toText(patches)) self.dmp.patch_addPadding(patches) self.assertEqual("@@ -5,8 +5,12 @@\n XXXX\n+test\n YYYY\n", self.dmp.patch_toText(patches)) def testPatchApply(self): self.dmp.Match_Distance = 1000 self.dmp.Match_Threshold = 0.5 self.dmp.Patch_DeleteThreshold = 0.5 # Null case. patches = self.dmp.patch_make("", "") results = self.dmp.patch_apply(patches, "Hello world.") self.assertEqual(("Hello world.", []), results) # Exact match. patches = self.dmp.patch_make("The quick brown fox jumps over the lazy dog.", "That quick brown fox jumped over a lazy dog.") results = self.dmp.patch_apply(patches, "The quick brown fox jumps over the lazy dog.") self.assertEqual(("That quick brown fox jumped over a lazy dog.", [True, True]), results) # Partial match. results = self.dmp.patch_apply(patches, "The quick red rabbit jumps over the tired tiger.") self.assertEqual(("That quick red rabbit jumped over a tired tiger.", [True, True]), results) # Failed match. results = self.dmp.patch_apply(patches, "I am the very model of a modern major general.") self.assertEqual(("I am the very model of a modern major general.", [False, False]), results) # Big delete, small change. patches = self.dmp.patch_make("x1234567890123456789012345678901234567890123456789012345678901234567890y", "xabcy") results = self.dmp.patch_apply(patches, "x123456789012345678901234567890-----++++++++++-----123456789012345678901234567890y") self.assertEqual(("xabcy", [True, True]), results) # Big delete, big change 1. patches = self.dmp.patch_make("x1234567890123456789012345678901234567890123456789012345678901234567890y", "xabcy") results = self.dmp.patch_apply(patches, "x12345678901234567890---------------++++++++++---------------12345678901234567890y") self.assertEqual(("xabc12345678901234567890---------------++++++++++---------------12345678901234567890y", [False, True]), results) # Big delete, big change 2. self.dmp.Patch_DeleteThreshold = 0.6 patches = self.dmp.patch_make("x1234567890123456789012345678901234567890123456789012345678901234567890y", "xabcy") results = self.dmp.patch_apply(patches, "x12345678901234567890---------------++++++++++---------------12345678901234567890y") self.assertEqual(("xabcy", [True, True]), results) self.dmp.Patch_DeleteThreshold = 0.5 # Compensate for failed patch. self.dmp.Match_Threshold = 0.0 self.dmp.Match_Distance = 0 patches = self.dmp.patch_make("abcdefghijklmnopqrstuvwxyz--------------------1234567890", "abcXXXXXXXXXXdefghijklmnopqrstuvwxyz--------------------1234567YYYYYYYYYY890") results = self.dmp.patch_apply(patches, "ABCDEFGHIJKLMNOPQRSTUVWXYZ--------------------1234567890") self.assertEqual(("ABCDEFGHIJKLMNOPQRSTUVWXYZ--------------------1234567YYYYYYYYYY890", [False, True]), results) self.dmp.Match_Threshold = 0.5 self.dmp.Match_Distance = 1000 # No side effects. patches = self.dmp.patch_make("", "test") patchstr = self.dmp.patch_toText(patches) results = self.dmp.patch_apply(patches, "") self.assertEqual(patchstr, self.dmp.patch_toText(patches)) # No side effects with major delete. patches = self.dmp.patch_make("The quick brown fox jumps over the lazy dog.", "Woof") patchstr = self.dmp.patch_toText(patches) self.dmp.patch_apply(patches, "The quick brown fox jumps over the lazy dog.") self.assertEqual(patchstr, self.dmp.patch_toText(patches)) # Edge exact match. patches = self.dmp.patch_make("", "test") self.dmp.patch_apply(patches, "") self.assertEqual(("test", [True]), results) # Near edge exact match. patches = self.dmp.patch_make("XY", "XtestY") results = self.dmp.patch_apply(patches, "XY") self.assertEqual(("XtestY", [True]), results) # Edge partial match. patches = self.dmp.patch_make("y", "y123") results = self.dmp.patch_apply(patches, "x") self.assertEqual(("x123", [True]), results) if __name__ == "__main__": unittest.main() diff-match-patch-20121119/python3/diff_match_patch/__init__.py0000750000175000017500000000007312052423660023753 0ustar lukeluke00000000000000from .diff_match_patch import diff_match_patch, patch_obj diff-match-patch-20121119/python3/diff_match_patch/diff_match_patch.py0000750000175000017500000020337012052423660025464 0ustar lukeluke00000000000000#!/usr/bin/python3 """Diff Match and Patch Copyright 2006 Google Inc. http://code.google.com/p/google-diff-match-patch/ Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ """Functions for diff, match and patch. Computes the difference between two texts to create a patch. Applies the patch onto another text, allowing for errors. """ __author__ = 'fraser@google.com (Neil Fraser)' import math import re import sys import time import urllib.parse class diff_match_patch: """Class containing the diff, match and patch methods. Also contains the behaviour settings. """ def __init__(self): """Inits a diff_match_patch object with default settings. Redefine these in your program to override the defaults. """ # Number of seconds to map a diff before giving up (0 for infinity). self.Diff_Timeout = 1.0 # Cost of an empty edit operation in terms of edit characters. self.Diff_EditCost = 4 # At what point is no match declared (0.0 = perfection, 1.0 = very loose). self.Match_Threshold = 0.5 # How far to search for a match (0 = exact location, 1000+ = broad match). # A match this many characters away from the expected location will add # 1.0 to the score (0.0 is a perfect match). self.Match_Distance = 1000 # When deleting a large block of text (over ~64 characters), how close do # the contents have to be to match the expected contents. (0.0 = perfection, # 1.0 = very loose). Note that Match_Threshold controls how closely the # end points of a delete need to match. self.Patch_DeleteThreshold = 0.5 # Chunk size for context length. self.Patch_Margin = 4 # The number of bits in an int. # Python has no maximum, thus to disable patch splitting set to 0. # However to avoid long patches in certain pathological cases, use 32. # Multiple short patches (using native ints) are much faster than long ones. self.Match_MaxBits = 32 # DIFF FUNCTIONS # The data structure representing a diff is an array of tuples: # [(DIFF_DELETE, "Hello"), (DIFF_INSERT, "Goodbye"), (DIFF_EQUAL, " world.")] # which means: delete "Hello", add "Goodbye" and keep " world." DIFF_DELETE = -1 DIFF_INSERT = 1 DIFF_EQUAL = 0 def diff_main(self, text1, text2, checklines=True, deadline=None): """Find the differences between two texts. Simplifies the problem by stripping any common prefix or suffix off the texts before diffing. Args: text1: Old string to be diffed. text2: New string to be diffed. checklines: Optional speedup flag. If present and false, then don't run a line-level diff first to identify the changed areas. Defaults to true, which does a faster, slightly less optimal diff. deadline: Optional time when the diff should be complete by. Used internally for recursive calls. Users should set DiffTimeout instead. Returns: Array of changes. """ # Set a deadline by which time the diff must be complete. if deadline == None: # Unlike in most languages, Python counts time in seconds. if self.Diff_Timeout <= 0: deadline = sys.maxsize else: deadline = time.time() + self.Diff_Timeout # Check for null inputs. if text1 == None or text2 == None: raise ValueError("Null inputs. (diff_main)") # Check for equality (speedup). if text1 == text2: if text1: return [(self.DIFF_EQUAL, text1)] return [] # Trim off common prefix (speedup). commonlength = self.diff_commonPrefix(text1, text2) commonprefix = text1[:commonlength] text1 = text1[commonlength:] text2 = text2[commonlength:] # Trim off common suffix (speedup). commonlength = self.diff_commonSuffix(text1, text2) if commonlength == 0: commonsuffix = '' else: commonsuffix = text1[-commonlength:] text1 = text1[:-commonlength] text2 = text2[:-commonlength] # Compute the diff on the middle block. diffs = self.diff_compute(text1, text2, checklines, deadline) # Restore the prefix and suffix. if commonprefix: diffs[:0] = [(self.DIFF_EQUAL, commonprefix)] if commonsuffix: diffs.append((self.DIFF_EQUAL, commonsuffix)) self.diff_cleanupMerge(diffs) return diffs def diff_compute(self, text1, text2, checklines, deadline): """Find the differences between two texts. Assumes that the texts do not have any common prefix or suffix. Args: text1: Old string to be diffed. text2: New string to be diffed. checklines: Speedup flag. If false, then don't run a line-level diff first to identify the changed areas. If true, then run a faster, slightly less optimal diff. deadline: Time when the diff should be complete by. Returns: Array of changes. """ if not text1: # Just add some text (speedup). return [(self.DIFF_INSERT, text2)] if not text2: # Just delete some text (speedup). return [(self.DIFF_DELETE, text1)] if len(text1) > len(text2): (longtext, shorttext) = (text1, text2) else: (shorttext, longtext) = (text1, text2) i = longtext.find(shorttext) if i != -1: # Shorter text is inside the longer text (speedup). diffs = [(self.DIFF_INSERT, longtext[:i]), (self.DIFF_EQUAL, shorttext), (self.DIFF_INSERT, longtext[i + len(shorttext):])] # Swap insertions for deletions if diff is reversed. if len(text1) > len(text2): diffs[0] = (self.DIFF_DELETE, diffs[0][1]) diffs[2] = (self.DIFF_DELETE, diffs[2][1]) return diffs if len(shorttext) == 1: # Single character string. # After the previous speedup, the character can't be an equality. return [(self.DIFF_DELETE, text1), (self.DIFF_INSERT, text2)] # Check to see if the problem can be split in two. hm = self.diff_halfMatch(text1, text2) if hm: # A half-match was found, sort out the return data. (text1_a, text1_b, text2_a, text2_b, mid_common) = hm # Send both pairs off for separate processing. diffs_a = self.diff_main(text1_a, text2_a, checklines, deadline) diffs_b = self.diff_main(text1_b, text2_b, checklines, deadline) # Merge the results. return diffs_a + [(self.DIFF_EQUAL, mid_common)] + diffs_b if checklines and len(text1) > 100 and len(text2) > 100: return self.diff_lineMode(text1, text2, deadline) return self.diff_bisect(text1, text2, deadline) def diff_lineMode(self, text1, text2, deadline): """Do a quick line-level diff on both strings, then rediff the parts for greater accuracy. This speedup can produce non-minimal diffs. Args: text1: Old string to be diffed. text2: New string to be diffed. deadline: Time when the diff should be complete by. Returns: Array of changes. """ # Scan the text on a line-by-line basis first. (text1, text2, linearray) = self.diff_linesToChars(text1, text2) diffs = self.diff_main(text1, text2, False, deadline) # Convert the diff back to original text. self.diff_charsToLines(diffs, linearray) # Eliminate freak matches (e.g. blank lines) self.diff_cleanupSemantic(diffs) # Rediff any replacement blocks, this time character-by-character. # Add a dummy entry at the end. diffs.append((self.DIFF_EQUAL, '')) pointer = 0 count_delete = 0 count_insert = 0 text_delete = '' text_insert = '' while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_INSERT: count_insert += 1 text_insert += diffs[pointer][1] elif diffs[pointer][0] == self.DIFF_DELETE: count_delete += 1 text_delete += diffs[pointer][1] elif diffs[pointer][0] == self.DIFF_EQUAL: # Upon reaching an equality, check for prior redundancies. if count_delete >= 1 and count_insert >= 1: # Delete the offending records and add the merged ones. a = self.diff_main(text_delete, text_insert, False, deadline) diffs[pointer - count_delete - count_insert : pointer] = a pointer = pointer - count_delete - count_insert + len(a) count_insert = 0 count_delete = 0 text_delete = '' text_insert = '' pointer += 1 diffs.pop() # Remove the dummy entry at the end. return diffs def diff_bisect(self, text1, text2, deadline): """Find the 'middle snake' of a diff, split the problem in two and return the recursively constructed diff. See Myers 1986 paper: An O(ND) Difference Algorithm and Its Variations. Args: text1: Old string to be diffed. text2: New string to be diffed. deadline: Time at which to bail if not yet complete. Returns: Array of diff tuples. """ # Cache the text lengths to prevent multiple calls. text1_length = len(text1) text2_length = len(text2) max_d = (text1_length + text2_length + 1) // 2 v_offset = max_d v_length = 2 * max_d v1 = [-1] * v_length v1[v_offset + 1] = 0 v2 = v1[:] delta = text1_length - text2_length # If the total number of characters is odd, then the front path will # collide with the reverse path. front = (delta % 2 != 0) # Offsets for start and end of k loop. # Prevents mapping of space beyond the grid. k1start = 0 k1end = 0 k2start = 0 k2end = 0 for d in range(max_d): # Bail out if deadline is reached. if time.time() > deadline: break # Walk the front path one step. for k1 in range(-d + k1start, d + 1 - k1end, 2): k1_offset = v_offset + k1 if k1 == -d or (k1 != d and v1[k1_offset - 1] < v1[k1_offset + 1]): x1 = v1[k1_offset + 1] else: x1 = v1[k1_offset - 1] + 1 y1 = x1 - k1 while (x1 < text1_length and y1 < text2_length and text1[x1] == text2[y1]): x1 += 1 y1 += 1 v1[k1_offset] = x1 if x1 > text1_length: # Ran off the right of the graph. k1end += 2 elif y1 > text2_length: # Ran off the bottom of the graph. k1start += 2 elif front: k2_offset = v_offset + delta - k1 if k2_offset >= 0 and k2_offset < v_length and v2[k2_offset] != -1: # Mirror x2 onto top-left coordinate system. x2 = text1_length - v2[k2_offset] if x1 >= x2: # Overlap detected. return self.diff_bisectSplit(text1, text2, x1, y1, deadline) # Walk the reverse path one step. for k2 in range(-d + k2start, d + 1 - k2end, 2): k2_offset = v_offset + k2 if k2 == -d or (k2 != d and v2[k2_offset - 1] < v2[k2_offset + 1]): x2 = v2[k2_offset + 1] else: x2 = v2[k2_offset - 1] + 1 y2 = x2 - k2 while (x2 < text1_length and y2 < text2_length and text1[-x2 - 1] == text2[-y2 - 1]): x2 += 1 y2 += 1 v2[k2_offset] = x2 if x2 > text1_length: # Ran off the left of the graph. k2end += 2 elif y2 > text2_length: # Ran off the top of the graph. k2start += 2 elif not front: k1_offset = v_offset + delta - k2 if k1_offset >= 0 and k1_offset < v_length and v1[k1_offset] != -1: x1 = v1[k1_offset] y1 = v_offset + x1 - k1_offset # Mirror x2 onto top-left coordinate system. x2 = text1_length - x2 if x1 >= x2: # Overlap detected. return self.diff_bisectSplit(text1, text2, x1, y1, deadline) # Diff took too long and hit the deadline or # number of diffs equals number of characters, no commonality at all. return [(self.DIFF_DELETE, text1), (self.DIFF_INSERT, text2)] def diff_bisectSplit(self, text1, text2, x, y, deadline): """Given the location of the 'middle snake', split the diff in two parts and recurse. Args: text1: Old string to be diffed. text2: New string to be diffed. x: Index of split point in text1. y: Index of split point in text2. deadline: Time at which to bail if not yet complete. Returns: Array of diff tuples. """ text1a = text1[:x] text2a = text2[:y] text1b = text1[x:] text2b = text2[y:] # Compute both diffs serially. diffs = self.diff_main(text1a, text2a, False, deadline) diffsb = self.diff_main(text1b, text2b, False, deadline) return diffs + diffsb def diff_linesToChars(self, text1, text2): """Split two texts into an array of strings. Reduce the texts to a string of hashes where each Unicode character represents one line. Args: text1: First string. text2: Second string. Returns: Three element tuple, containing the encoded text1, the encoded text2 and the array of unique strings. The zeroth element of the array of unique strings is intentionally blank. """ lineArray = [] # e.g. lineArray[4] == "Hello\n" lineHash = {} # e.g. lineHash["Hello\n"] == 4 # "\x00" is a valid character, but various debuggers don't like it. # So we'll insert a junk entry to avoid generating a null character. lineArray.append('') def diff_linesToCharsMunge(text): """Split a text into an array of strings. Reduce the texts to a string of hashes where each Unicode character represents one line. Modifies linearray and linehash through being a closure. Args: text: String to encode. Returns: Encoded string. """ chars = [] # Walk the text, pulling out a substring for each line. # text.split('\n') would would temporarily double our memory footprint. # Modifying text would create many large strings to garbage collect. lineStart = 0 lineEnd = -1 while lineEnd < len(text) - 1: lineEnd = text.find('\n', lineStart) if lineEnd == -1: lineEnd = len(text) - 1 line = text[lineStart:lineEnd + 1] lineStart = lineEnd + 1 if line in lineHash: chars.append(chr(lineHash[line])) else: lineArray.append(line) lineHash[line] = len(lineArray) - 1 chars.append(chr(len(lineArray) - 1)) return "".join(chars) chars1 = diff_linesToCharsMunge(text1) chars2 = diff_linesToCharsMunge(text2) return (chars1, chars2, lineArray) def diff_charsToLines(self, diffs, lineArray): """Rehydrate the text in a diff from a string of line hashes to real lines of text. Args: diffs: Array of diff tuples. lineArray: Array of unique strings. """ for x in range(len(diffs)): text = [] for char in diffs[x][1]: text.append(lineArray[ord(char)]) diffs[x] = (diffs[x][0], "".join(text)) def diff_commonPrefix(self, text1, text2): """Determine the common prefix of two strings. Args: text1: First string. text2: Second string. Returns: The number of characters common to the start of each string. """ # Quick check for common null cases. if not text1 or not text2 or text1[0] != text2[0]: return 0 # Binary search. # Performance analysis: http://neil.fraser.name/news/2007/10/09/ pointermin = 0 pointermax = min(len(text1), len(text2)) pointermid = pointermax pointerstart = 0 while pointermin < pointermid: if text1[pointerstart:pointermid] == text2[pointerstart:pointermid]: pointermin = pointermid pointerstart = pointermin else: pointermax = pointermid pointermid = (pointermax - pointermin) // 2 + pointermin return pointermid def diff_commonSuffix(self, text1, text2): """Determine the common suffix of two strings. Args: text1: First string. text2: Second string. Returns: The number of characters common to the end of each string. """ # Quick check for common null cases. if not text1 or not text2 or text1[-1] != text2[-1]: return 0 # Binary search. # Performance analysis: http://neil.fraser.name/news/2007/10/09/ pointermin = 0 pointermax = min(len(text1), len(text2)) pointermid = pointermax pointerend = 0 while pointermin < pointermid: if (text1[-pointermid:len(text1) - pointerend] == text2[-pointermid:len(text2) - pointerend]): pointermin = pointermid pointerend = pointermin else: pointermax = pointermid pointermid = (pointermax - pointermin) // 2 + pointermin return pointermid def diff_commonOverlap(self, text1, text2): """Determine if the suffix of one string is the prefix of another. Args: text1 First string. text2 Second string. Returns: The number of characters common to the end of the first string and the start of the second string. """ # Cache the text lengths to prevent multiple calls. text1_length = len(text1) text2_length = len(text2) # Eliminate the null case. if text1_length == 0 or text2_length == 0: return 0 # Truncate the longer string. if text1_length > text2_length: text1 = text1[-text2_length:] elif text1_length < text2_length: text2 = text2[:text1_length] text_length = min(text1_length, text2_length) # Quick check for the worst case. if text1 == text2: return text_length # Start by looking for a single character match # and increase length until no match is found. # Performance analysis: http://neil.fraser.name/news/2010/11/04/ best = 0 length = 1 while True: pattern = text1[-length:] found = text2.find(pattern) if found == -1: return best length += found if found == 0 or text1[-length:] == text2[:length]: best = length length += 1 def diff_halfMatch(self, text1, text2): """Do the two texts share a substring which is at least half the length of the longer text? This speedup can produce non-minimal diffs. Args: text1: First string. text2: Second string. Returns: Five element Array, containing the prefix of text1, the suffix of text1, the prefix of text2, the suffix of text2 and the common middle. Or None if there was no match. """ if self.Diff_Timeout <= 0: # Don't risk returning a non-optimal diff if we have unlimited time. return None if len(text1) > len(text2): (longtext, shorttext) = (text1, text2) else: (shorttext, longtext) = (text1, text2) if len(longtext) < 4 or len(shorttext) * 2 < len(longtext): return None # Pointless. def diff_halfMatchI(longtext, shorttext, i): """Does a substring of shorttext exist within longtext such that the substring is at least half the length of longtext? Closure, but does not reference any external variables. Args: longtext: Longer string. shorttext: Shorter string. i: Start index of quarter length substring within longtext. Returns: Five element Array, containing the prefix of longtext, the suffix of longtext, the prefix of shorttext, the suffix of shorttext and the common middle. Or None if there was no match. """ seed = longtext[i:i + len(longtext) // 4] best_common = '' j = shorttext.find(seed) while j != -1: prefixLength = self.diff_commonPrefix(longtext[i:], shorttext[j:]) suffixLength = self.diff_commonSuffix(longtext[:i], shorttext[:j]) if len(best_common) < suffixLength + prefixLength: best_common = (shorttext[j - suffixLength:j] + shorttext[j:j + prefixLength]) best_longtext_a = longtext[:i - suffixLength] best_longtext_b = longtext[i + prefixLength:] best_shorttext_a = shorttext[:j - suffixLength] best_shorttext_b = shorttext[j + prefixLength:] j = shorttext.find(seed, j + 1) if len(best_common) * 2 >= len(longtext): return (best_longtext_a, best_longtext_b, best_shorttext_a, best_shorttext_b, best_common) else: return None # First check if the second quarter is the seed for a half-match. hm1 = diff_halfMatchI(longtext, shorttext, (len(longtext) + 3) // 4) # Check again based on the third quarter. hm2 = diff_halfMatchI(longtext, shorttext, (len(longtext) + 1) // 2) if not hm1 and not hm2: return None elif not hm2: hm = hm1 elif not hm1: hm = hm2 else: # Both matched. Select the longest. if len(hm1[4]) > len(hm2[4]): hm = hm1 else: hm = hm2 # A half-match was found, sort out the return data. if len(text1) > len(text2): (text1_a, text1_b, text2_a, text2_b, mid_common) = hm else: (text2_a, text2_b, text1_a, text1_b, mid_common) = hm return (text1_a, text1_b, text2_a, text2_b, mid_common) def diff_cleanupSemantic(self, diffs): """Reduce the number of edits by eliminating semantically trivial equalities. Args: diffs: Array of diff tuples. """ changes = False equalities = [] # Stack of indices where equalities are found. lastequality = None # Always equal to diffs[equalities[-1]][1] pointer = 0 # Index of current position. # Number of chars that changed prior to the equality. length_insertions1, length_deletions1 = 0, 0 # Number of chars that changed after the equality. length_insertions2, length_deletions2 = 0, 0 while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_EQUAL: # Equality found. equalities.append(pointer) length_insertions1, length_insertions2 = length_insertions2, 0 length_deletions1, length_deletions2 = length_deletions2, 0 lastequality = diffs[pointer][1] else: # An insertion or deletion. if diffs[pointer][0] == self.DIFF_INSERT: length_insertions2 += len(diffs[pointer][1]) else: length_deletions2 += len(diffs[pointer][1]) # Eliminate an equality that is smaller or equal to the edits on both # sides of it. if (lastequality and (len(lastequality) <= max(length_insertions1, length_deletions1)) and (len(lastequality) <= max(length_insertions2, length_deletions2))): # Duplicate record. diffs.insert(equalities[-1], (self.DIFF_DELETE, lastequality)) # Change second copy to insert. diffs[equalities[-1] + 1] = (self.DIFF_INSERT, diffs[equalities[-1] + 1][1]) # Throw away the equality we just deleted. equalities.pop() # Throw away the previous equality (it needs to be reevaluated). if len(equalities): equalities.pop() if len(equalities): pointer = equalities[-1] else: pointer = -1 # Reset the counters. length_insertions1, length_deletions1 = 0, 0 length_insertions2, length_deletions2 = 0, 0 lastequality = None changes = True pointer += 1 # Normalize the diff. if changes: self.diff_cleanupMerge(diffs) self.diff_cleanupSemanticLossless(diffs) # Find any overlaps between deletions and insertions. # e.g: abcxxxxxxdef # -> abcxxxdef # e.g: xxxabcdefxxx # -> defxxxabc # Only extract an overlap if it is as big as the edit ahead or behind it. pointer = 1 while pointer < len(diffs): if (diffs[pointer - 1][0] == self.DIFF_DELETE and diffs[pointer][0] == self.DIFF_INSERT): deletion = diffs[pointer - 1][1] insertion = diffs[pointer][1] overlap_length1 = self.diff_commonOverlap(deletion, insertion) overlap_length2 = self.diff_commonOverlap(insertion, deletion) if overlap_length1 >= overlap_length2: if (overlap_length1 >= len(deletion) / 2.0 or overlap_length1 >= len(insertion) / 2.0): # Overlap found. Insert an equality and trim the surrounding edits. diffs.insert(pointer, (self.DIFF_EQUAL, insertion[:overlap_length1])) diffs[pointer - 1] = (self.DIFF_DELETE, deletion[:len(deletion) - overlap_length1]) diffs[pointer + 1] = (self.DIFF_INSERT, insertion[overlap_length1:]) pointer += 1 else: if (overlap_length2 >= len(deletion) / 2.0 or overlap_length2 >= len(insertion) / 2.0): # Reverse overlap found. # Insert an equality and swap and trim the surrounding edits. diffs.insert(pointer, (self.DIFF_EQUAL, deletion[:overlap_length2])) diffs[pointer - 1] = (self.DIFF_INSERT, insertion[:len(insertion) - overlap_length2]) diffs[pointer + 1] = (self.DIFF_DELETE, deletion[overlap_length2:]) pointer += 1 pointer += 1 pointer += 1 def diff_cleanupSemanticLossless(self, diffs): """Look for single edits surrounded on both sides by equalities which can be shifted sideways to align the edit to a word boundary. e.g: The cat came. -> The cat came. Args: diffs: Array of diff tuples. """ def diff_cleanupSemanticScore(one, two): """Given two strings, compute a score representing whether the internal boundary falls on logical boundaries. Scores range from 6 (best) to 0 (worst). Closure, but does not reference any external variables. Args: one: First string. two: Second string. Returns: The score. """ if not one or not two: # Edges are the best. return 6 # Each port of this function behaves slightly differently due to # subtle differences in each language's definition of things like # 'whitespace'. Since this function's purpose is largely cosmetic, # the choice has been made to use each language's native features # rather than force total conformity. char1 = one[-1] char2 = two[0] nonAlphaNumeric1 = not char1.isalnum() nonAlphaNumeric2 = not char2.isalnum() whitespace1 = nonAlphaNumeric1 and char1.isspace() whitespace2 = nonAlphaNumeric2 and char2.isspace() lineBreak1 = whitespace1 and (char1 == "\r" or char1 == "\n") lineBreak2 = whitespace2 and (char2 == "\r" or char2 == "\n") blankLine1 = lineBreak1 and self.BLANKLINEEND.search(one) blankLine2 = lineBreak2 and self.BLANKLINESTART.match(two) if blankLine1 or blankLine2: # Five points for blank lines. return 5 elif lineBreak1 or lineBreak2: # Four points for line breaks. return 4 elif nonAlphaNumeric1 and not whitespace1 and whitespace2: # Three points for end of sentences. return 3 elif whitespace1 or whitespace2: # Two points for whitespace. return 2 elif nonAlphaNumeric1 or nonAlphaNumeric2: # One point for non-alphanumeric. return 1 return 0 pointer = 1 # Intentionally ignore the first and last element (don't need checking). while pointer < len(diffs) - 1: if (diffs[pointer - 1][0] == self.DIFF_EQUAL and diffs[pointer + 1][0] == self.DIFF_EQUAL): # This is a single edit surrounded by equalities. equality1 = diffs[pointer - 1][1] edit = diffs[pointer][1] equality2 = diffs[pointer + 1][1] # First, shift the edit as far left as possible. commonOffset = self.diff_commonSuffix(equality1, edit) if commonOffset: commonString = edit[-commonOffset:] equality1 = equality1[:-commonOffset] edit = commonString + edit[:-commonOffset] equality2 = commonString + equality2 # Second, step character by character right, looking for the best fit. bestEquality1 = equality1 bestEdit = edit bestEquality2 = equality2 bestScore = (diff_cleanupSemanticScore(equality1, edit) + diff_cleanupSemanticScore(edit, equality2)) while edit and equality2 and edit[0] == equality2[0]: equality1 += edit[0] edit = edit[1:] + equality2[0] equality2 = equality2[1:] score = (diff_cleanupSemanticScore(equality1, edit) + diff_cleanupSemanticScore(edit, equality2)) # The >= encourages trailing rather than leading whitespace on edits. if score >= bestScore: bestScore = score bestEquality1 = equality1 bestEdit = edit bestEquality2 = equality2 if diffs[pointer - 1][1] != bestEquality1: # We have an improvement, save it back to the diff. if bestEquality1: diffs[pointer - 1] = (diffs[pointer - 1][0], bestEquality1) else: del diffs[pointer - 1] pointer -= 1 diffs[pointer] = (diffs[pointer][0], bestEdit) if bestEquality2: diffs[pointer + 1] = (diffs[pointer + 1][0], bestEquality2) else: del diffs[pointer + 1] pointer -= 1 pointer += 1 # Define some regex patterns for matching boundaries. BLANKLINEEND = re.compile(r"\n\r?\n$"); BLANKLINESTART = re.compile(r"^\r?\n\r?\n"); def diff_cleanupEfficiency(self, diffs): """Reduce the number of edits by eliminating operationally trivial equalities. Args: diffs: Array of diff tuples. """ changes = False equalities = [] # Stack of indices where equalities are found. lastequality = None # Always equal to diffs[equalities[-1]][1] pointer = 0 # Index of current position. pre_ins = False # Is there an insertion operation before the last equality. pre_del = False # Is there a deletion operation before the last equality. post_ins = False # Is there an insertion operation after the last equality. post_del = False # Is there a deletion operation after the last equality. while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_EQUAL: # Equality found. if (len(diffs[pointer][1]) < self.Diff_EditCost and (post_ins or post_del)): # Candidate found. equalities.append(pointer) pre_ins = post_ins pre_del = post_del lastequality = diffs[pointer][1] else: # Not a candidate, and can never become one. equalities = [] lastequality = None post_ins = post_del = False else: # An insertion or deletion. if diffs[pointer][0] == self.DIFF_DELETE: post_del = True else: post_ins = True # Five types to be split: # ABXYCD # AXCD # ABXC # AXCD # ABXC if lastequality and ((pre_ins and pre_del and post_ins and post_del) or ((len(lastequality) < self.Diff_EditCost / 2) and (pre_ins + pre_del + post_ins + post_del) == 3)): # Duplicate record. diffs.insert(equalities[-1], (self.DIFF_DELETE, lastequality)) # Change second copy to insert. diffs[equalities[-1] + 1] = (self.DIFF_INSERT, diffs[equalities[-1] + 1][1]) equalities.pop() # Throw away the equality we just deleted. lastequality = None if pre_ins and pre_del: # No changes made which could affect previous entry, keep going. post_ins = post_del = True equalities = [] else: if len(equalities): equalities.pop() # Throw away the previous equality. if len(equalities): pointer = equalities[-1] else: pointer = -1 post_ins = post_del = False changes = True pointer += 1 if changes: self.diff_cleanupMerge(diffs) def diff_cleanupMerge(self, diffs): """Reorder and merge like edit sections. Merge equalities. Any edit section can move as long as it doesn't cross an equality. Args: diffs: Array of diff tuples. """ diffs.append((self.DIFF_EQUAL, '')) # Add a dummy entry at the end. pointer = 0 count_delete = 0 count_insert = 0 text_delete = '' text_insert = '' while pointer < len(diffs): if diffs[pointer][0] == self.DIFF_INSERT: count_insert += 1 text_insert += diffs[pointer][1] pointer += 1 elif diffs[pointer][0] == self.DIFF_DELETE: count_delete += 1 text_delete += diffs[pointer][1] pointer += 1 elif diffs[pointer][0] == self.DIFF_EQUAL: # Upon reaching an equality, check for prior redundancies. if count_delete + count_insert > 1: if count_delete != 0 and count_insert != 0: # Factor out any common prefixies. commonlength = self.diff_commonPrefix(text_insert, text_delete) if commonlength != 0: x = pointer - count_delete - count_insert - 1 if x >= 0 and diffs[x][0] == self.DIFF_EQUAL: diffs[x] = (diffs[x][0], diffs[x][1] + text_insert[:commonlength]) else: diffs.insert(0, (self.DIFF_EQUAL, text_insert[:commonlength])) pointer += 1 text_insert = text_insert[commonlength:] text_delete = text_delete[commonlength:] # Factor out any common suffixies. commonlength = self.diff_commonSuffix(text_insert, text_delete) if commonlength != 0: diffs[pointer] = (diffs[pointer][0], text_insert[-commonlength:] + diffs[pointer][1]) text_insert = text_insert[:-commonlength] text_delete = text_delete[:-commonlength] # Delete the offending records and add the merged ones. if count_delete == 0: diffs[pointer - count_insert : pointer] = [ (self.DIFF_INSERT, text_insert)] elif count_insert == 0: diffs[pointer - count_delete : pointer] = [ (self.DIFF_DELETE, text_delete)] else: diffs[pointer - count_delete - count_insert : pointer] = [ (self.DIFF_DELETE, text_delete), (self.DIFF_INSERT, text_insert)] pointer = pointer - count_delete - count_insert + 1 if count_delete != 0: pointer += 1 if count_insert != 0: pointer += 1 elif pointer != 0 and diffs[pointer - 1][0] == self.DIFF_EQUAL: # Merge this equality with the previous one. diffs[pointer - 1] = (diffs[pointer - 1][0], diffs[pointer - 1][1] + diffs[pointer][1]) del diffs[pointer] else: pointer += 1 count_insert = 0 count_delete = 0 text_delete = '' text_insert = '' if diffs[-1][1] == '': diffs.pop() # Remove the dummy entry at the end. # Second pass: look for single edits surrounded on both sides by equalities # which can be shifted sideways to eliminate an equality. # e.g: ABAC -> ABAC changes = False pointer = 1 # Intentionally ignore the first and last element (don't need checking). while pointer < len(diffs) - 1: if (diffs[pointer - 1][0] == self.DIFF_EQUAL and diffs[pointer + 1][0] == self.DIFF_EQUAL): # This is a single edit surrounded by equalities. if diffs[pointer][1].endswith(diffs[pointer - 1][1]): # Shift the edit over the previous equality. diffs[pointer] = (diffs[pointer][0], diffs[pointer - 1][1] + diffs[pointer][1][:-len(diffs[pointer - 1][1])]) diffs[pointer + 1] = (diffs[pointer + 1][0], diffs[pointer - 1][1] + diffs[pointer + 1][1]) del diffs[pointer - 1] changes = True elif diffs[pointer][1].startswith(diffs[pointer + 1][1]): # Shift the edit over the next equality. diffs[pointer - 1] = (diffs[pointer - 1][0], diffs[pointer - 1][1] + diffs[pointer + 1][1]) diffs[pointer] = (diffs[pointer][0], diffs[pointer][1][len(diffs[pointer + 1][1]):] + diffs[pointer + 1][1]) del diffs[pointer + 1] changes = True pointer += 1 # If shifts were made, the diff needs reordering and another shift sweep. if changes: self.diff_cleanupMerge(diffs) def diff_xIndex(self, diffs, loc): """loc is a location in text1, compute and return the equivalent location in text2. e.g. "The cat" vs "The big cat", 1->1, 5->8 Args: diffs: Array of diff tuples. loc: Location within text1. Returns: Location within text2. """ chars1 = 0 chars2 = 0 last_chars1 = 0 last_chars2 = 0 for x in range(len(diffs)): (op, text) = diffs[x] if op != self.DIFF_INSERT: # Equality or deletion. chars1 += len(text) if op != self.DIFF_DELETE: # Equality or insertion. chars2 += len(text) if chars1 > loc: # Overshot the location. break last_chars1 = chars1 last_chars2 = chars2 if len(diffs) != x and diffs[x][0] == self.DIFF_DELETE: # The location was deleted. return last_chars2 # Add the remaining len(character). return last_chars2 + (loc - last_chars1) def diff_prettyHtml(self, diffs): """Convert a diff array into a pretty HTML report. Args: diffs: Array of diff tuples. Returns: HTML representation. """ html = [] for (op, data) in diffs: text = (data.replace("&", "&").replace("<", "<") .replace(">", ">").replace("\n", "¶
")) if op == self.DIFF_INSERT: html.append("%s" % text) elif op == self.DIFF_DELETE: html.append("%s" % text) elif op == self.DIFF_EQUAL: html.append("%s" % text) return "".join(html) def diff_text1(self, diffs): """Compute and return the source text (all equalities and deletions). Args: diffs: Array of diff tuples. Returns: Source text. """ text = [] for (op, data) in diffs: if op != self.DIFF_INSERT: text.append(data) return "".join(text) def diff_text2(self, diffs): """Compute and return the destination text (all equalities and insertions). Args: diffs: Array of diff tuples. Returns: Destination text. """ text = [] for (op, data) in diffs: if op != self.DIFF_DELETE: text.append(data) return "".join(text) def diff_levenshtein(self, diffs): """Compute the Levenshtein distance; the number of inserted, deleted or substituted characters. Args: diffs: Array of diff tuples. Returns: Number of changes. """ levenshtein = 0 insertions = 0 deletions = 0 for (op, data) in diffs: if op == self.DIFF_INSERT: insertions += len(data) elif op == self.DIFF_DELETE: deletions += len(data) elif op == self.DIFF_EQUAL: # A deletion and an insertion is one substitution. levenshtein += max(insertions, deletions) insertions = 0 deletions = 0 levenshtein += max(insertions, deletions) return levenshtein def diff_toDelta(self, diffs): """Crush the diff into an encoded string which describes the operations required to transform text1 into text2. E.g. =3\t-2\t+ing -> Keep 3 chars, delete 2 chars, insert 'ing'. Operations are tab-separated. Inserted text is escaped using %xx notation. Args: diffs: Array of diff tuples. Returns: Delta text. """ text = [] for (op, data) in diffs: if op == self.DIFF_INSERT: # High ascii will raise UnicodeDecodeError. Use Unicode instead. data = data.encode("utf-8") text.append("+" + urllib.parse.quote(data, "!~*'();/?:@&=+$,# ")) elif op == self.DIFF_DELETE: text.append("-%d" % len(data)) elif op == self.DIFF_EQUAL: text.append("=%d" % len(data)) return "\t".join(text) def diff_fromDelta(self, text1, delta): """Given the original text1, and an encoded string which describes the operations required to transform text1 into text2, compute the full diff. Args: text1: Source string for the diff. delta: Delta text. Returns: Array of diff tuples. Raises: ValueError: If invalid input. """ diffs = [] pointer = 0 # Cursor in text1 tokens = delta.split("\t") for token in tokens: if token == "": # Blank tokens are ok (from a trailing \t). continue # Each token begins with a one character parameter which specifies the # operation of this token (delete, insert, equality). param = token[1:] if token[0] == "+": param = urllib.parse.unquote(param) diffs.append((self.DIFF_INSERT, param)) elif token[0] == "-" or token[0] == "=": try: n = int(param) except ValueError: raise ValueError("Invalid number in diff_fromDelta: " + param) if n < 0: raise ValueError("Negative number in diff_fromDelta: " + param) text = text1[pointer : pointer + n] pointer += n if token[0] == "=": diffs.append((self.DIFF_EQUAL, text)) else: diffs.append((self.DIFF_DELETE, text)) else: # Anything else is an error. raise ValueError("Invalid diff operation in diff_fromDelta: " + token[0]) if pointer != len(text1): raise ValueError( "Delta length (%d) does not equal source text length (%d)." % (pointer, len(text1))) return diffs # MATCH FUNCTIONS def match_main(self, text, pattern, loc): """Locate the best instance of 'pattern' in 'text' near 'loc'. Args: text: The text to search. pattern: The pattern to search for. loc: The location to search around. Returns: Best match index or -1. """ # Check for null inputs. if text == None or pattern == None: raise ValueError("Null inputs. (match_main)") loc = max(0, min(loc, len(text))) if text == pattern: # Shortcut (potentially not guaranteed by the algorithm) return 0 elif not text: # Nothing to match. return -1 elif text[loc:loc + len(pattern)] == pattern: # Perfect match at the perfect spot! (Includes case of null pattern) return loc else: # Do a fuzzy compare. match = self.match_bitap(text, pattern, loc) return match def match_bitap(self, text, pattern, loc): """Locate the best instance of 'pattern' in 'text' near 'loc' using the Bitap algorithm. Args: text: The text to search. pattern: The pattern to search for. loc: The location to search around. Returns: Best match index or -1. """ # Python doesn't have a maxint limit, so ignore this check. #if self.Match_MaxBits != 0 and len(pattern) > self.Match_MaxBits: # raise ValueError("Pattern too long for this application.") # Initialise the alphabet. s = self.match_alphabet(pattern) def match_bitapScore(e, x): """Compute and return the score for a match with e errors and x location. Accesses loc and pattern through being a closure. Args: e: Number of errors in match. x: Location of match. Returns: Overall score for match (0.0 = good, 1.0 = bad). """ accuracy = float(e) / len(pattern) proximity = abs(loc - x) if not self.Match_Distance: # Dodge divide by zero error. return proximity and 1.0 or accuracy return accuracy + (proximity / float(self.Match_Distance)) # Highest score beyond which we give up. score_threshold = self.Match_Threshold # Is there a nearby exact match? (speedup) best_loc = text.find(pattern, loc) if best_loc != -1: score_threshold = min(match_bitapScore(0, best_loc), score_threshold) # What about in the other direction? (speedup) best_loc = text.rfind(pattern, loc + len(pattern)) if best_loc != -1: score_threshold = min(match_bitapScore(0, best_loc), score_threshold) # Initialise the bit arrays. matchmask = 1 << (len(pattern) - 1) best_loc = -1 bin_max = len(pattern) + len(text) # Empty initialization added to appease pychecker. last_rd = None for d in range(len(pattern)): # Scan for the best match each iteration allows for one more error. # Run a binary search to determine how far from 'loc' we can stray at # this error level. bin_min = 0 bin_mid = bin_max while bin_min < bin_mid: if match_bitapScore(d, loc + bin_mid) <= score_threshold: bin_min = bin_mid else: bin_max = bin_mid bin_mid = (bin_max - bin_min) // 2 + bin_min # Use the result from this iteration as the maximum for the next. bin_max = bin_mid start = max(1, loc - bin_mid + 1) finish = min(loc + bin_mid, len(text)) + len(pattern) rd = [0] * (finish + 2) rd[finish + 1] = (1 << d) - 1 for j in range(finish, start - 1, -1): if len(text) <= j - 1: # Out of range. charMatch = 0 else: charMatch = s.get(text[j - 1], 0) if d == 0: # First pass: exact match. rd[j] = ((rd[j + 1] << 1) | 1) & charMatch else: # Subsequent passes: fuzzy match. rd[j] = (((rd[j + 1] << 1) | 1) & charMatch) | ( ((last_rd[j + 1] | last_rd[j]) << 1) | 1) | last_rd[j + 1] if rd[j] & matchmask: score = match_bitapScore(d, j - 1) # This match will almost certainly be better than any existing match. # But check anyway. if score <= score_threshold: # Told you so. score_threshold = score best_loc = j - 1 if best_loc > loc: # When passing loc, don't exceed our current distance from loc. start = max(1, 2 * loc - best_loc) else: # Already passed loc, downhill from here on in. break # No hope for a (better) match at greater error levels. if match_bitapScore(d + 1, loc) > score_threshold: break last_rd = rd return best_loc def match_alphabet(self, pattern): """Initialise the alphabet for the Bitap algorithm. Args: pattern: The text to encode. Returns: Hash of character locations. """ s = {} for char in pattern: s[char] = 0 for i in range(len(pattern)): s[pattern[i]] |= 1 << (len(pattern) - i - 1) return s # PATCH FUNCTIONS def patch_addContext(self, patch, text): """Increase the context until it is unique, but don't let the pattern expand beyond Match_MaxBits. Args: patch: The patch to grow. text: Source text. """ if len(text) == 0: return pattern = text[patch.start2 : patch.start2 + patch.length1] padding = 0 # Look for the first and last matches of pattern in text. If two different # matches are found, increase the pattern length. while (text.find(pattern) != text.rfind(pattern) and (self.Match_MaxBits == 0 or len(pattern) < self.Match_MaxBits - self.Patch_Margin - self.Patch_Margin)): padding += self.Patch_Margin pattern = text[max(0, patch.start2 - padding) : patch.start2 + patch.length1 + padding] # Add one chunk for good luck. padding += self.Patch_Margin # Add the prefix. prefix = text[max(0, patch.start2 - padding) : patch.start2] if prefix: patch.diffs[:0] = [(self.DIFF_EQUAL, prefix)] # Add the suffix. suffix = text[patch.start2 + patch.length1 : patch.start2 + patch.length1 + padding] if suffix: patch.diffs.append((self.DIFF_EQUAL, suffix)) # Roll back the start points. patch.start1 -= len(prefix) patch.start2 -= len(prefix) # Extend lengths. patch.length1 += len(prefix) + len(suffix) patch.length2 += len(prefix) + len(suffix) def patch_make(self, a, b=None, c=None): """Compute a list of patches to turn text1 into text2. Use diffs if provided, otherwise compute it ourselves. There are four ways to call this function, depending on what data is available to the caller: Method 1: a = text1, b = text2 Method 2: a = diffs Method 3 (optimal): a = text1, b = diffs Method 4 (deprecated, use method 3): a = text1, b = text2, c = diffs Args: a: text1 (methods 1,3,4) or Array of diff tuples for text1 to text2 (method 2). b: text2 (methods 1,4) or Array of diff tuples for text1 to text2 (method 3) or undefined (method 2). c: Array of diff tuples for text1 to text2 (method 4) or undefined (methods 1,2,3). Returns: Array of Patch objects. """ text1 = None diffs = None if isinstance(a, str) and isinstance(b, str) and c is None: # Method 1: text1, text2 # Compute diffs from text1 and text2. text1 = a diffs = self.diff_main(text1, b, True) if len(diffs) > 2: self.diff_cleanupSemantic(diffs) self.diff_cleanupEfficiency(diffs) elif isinstance(a, list) and b is None and c is None: # Method 2: diffs # Compute text1 from diffs. diffs = a text1 = self.diff_text1(diffs) elif isinstance(a, str) and isinstance(b, list) and c is None: # Method 3: text1, diffs text1 = a diffs = b elif (isinstance(a, str) and isinstance(b, str) and isinstance(c, list)): # Method 4: text1, text2, diffs # text2 is not used. text1 = a diffs = c else: raise ValueError("Unknown call format to patch_make.") if not diffs: return [] # Get rid of the None case. patches = [] patch = patch_obj() char_count1 = 0 # Number of characters into the text1 string. char_count2 = 0 # Number of characters into the text2 string. prepatch_text = text1 # Recreate the patches to determine context info. postpatch_text = text1 for x in range(len(diffs)): (diff_type, diff_text) = diffs[x] if len(patch.diffs) == 0 and diff_type != self.DIFF_EQUAL: # A new patch starts here. patch.start1 = char_count1 patch.start2 = char_count2 if diff_type == self.DIFF_INSERT: # Insertion patch.diffs.append(diffs[x]) patch.length2 += len(diff_text) postpatch_text = (postpatch_text[:char_count2] + diff_text + postpatch_text[char_count2:]) elif diff_type == self.DIFF_DELETE: # Deletion. patch.length1 += len(diff_text) patch.diffs.append(diffs[x]) postpatch_text = (postpatch_text[:char_count2] + postpatch_text[char_count2 + len(diff_text):]) elif (diff_type == self.DIFF_EQUAL and len(diff_text) <= 2 * self.Patch_Margin and len(patch.diffs) != 0 and len(diffs) != x + 1): # Small equality inside a patch. patch.diffs.append(diffs[x]) patch.length1 += len(diff_text) patch.length2 += len(diff_text) if (diff_type == self.DIFF_EQUAL and len(diff_text) >= 2 * self.Patch_Margin): # Time for a new patch. if len(patch.diffs) != 0: self.patch_addContext(patch, prepatch_text) patches.append(patch) patch = patch_obj() # Unlike Unidiff, our patch lists have a rolling context. # http://code.google.com/p/google-diff-match-patch/wiki/Unidiff # Update prepatch text & pos to reflect the application of the # just completed patch. prepatch_text = postpatch_text char_count1 = char_count2 # Update the current character count. if diff_type != self.DIFF_INSERT: char_count1 += len(diff_text) if diff_type != self.DIFF_DELETE: char_count2 += len(diff_text) # Pick up the leftover patch if not empty. if len(patch.diffs) != 0: self.patch_addContext(patch, prepatch_text) patches.append(patch) return patches def patch_deepCopy(self, patches): """Given an array of patches, return another array that is identical. Args: patches: Array of Patch objects. Returns: Array of Patch objects. """ patchesCopy = [] for patch in patches: patchCopy = patch_obj() # No need to deep copy the tuples since they are immutable. patchCopy.diffs = patch.diffs[:] patchCopy.start1 = patch.start1 patchCopy.start2 = patch.start2 patchCopy.length1 = patch.length1 patchCopy.length2 = patch.length2 patchesCopy.append(patchCopy) return patchesCopy def patch_apply(self, patches, text): """Merge a set of patches onto the text. Return a patched text, as well as a list of true/false values indicating which patches were applied. Args: patches: Array of Patch objects. text: Old text. Returns: Two element Array, containing the new text and an array of boolean values. """ if not patches: return (text, []) # Deep copy the patches so that no changes are made to originals. patches = self.patch_deepCopy(patches) nullPadding = self.patch_addPadding(patches) text = nullPadding + text + nullPadding self.patch_splitMax(patches) # delta keeps track of the offset between the expected and actual location # of the previous patch. If there are patches expected at positions 10 and # 20, but the first patch was found at 12, delta is 2 and the second patch # has an effective expected position of 22. delta = 0 results = [] for patch in patches: expected_loc = patch.start2 + delta text1 = self.diff_text1(patch.diffs) end_loc = -1 if len(text1) > self.Match_MaxBits: # patch_splitMax will only provide an oversized pattern in the case of # a monster delete. start_loc = self.match_main(text, text1[:self.Match_MaxBits], expected_loc) if start_loc != -1: end_loc = self.match_main(text, text1[-self.Match_MaxBits:], expected_loc + len(text1) - self.Match_MaxBits) if end_loc == -1 or start_loc >= end_loc: # Can't find valid trailing context. Drop this patch. start_loc = -1 else: start_loc = self.match_main(text, text1, expected_loc) if start_loc == -1: # No match found. :( results.append(False) # Subtract the delta for this failed patch from subsequent patches. delta -= patch.length2 - patch.length1 else: # Found a match. :) results.append(True) delta = start_loc - expected_loc if end_loc == -1: text2 = text[start_loc : start_loc + len(text1)] else: text2 = text[start_loc : end_loc + self.Match_MaxBits] if text1 == text2: # Perfect match, just shove the replacement text in. text = (text[:start_loc] + self.diff_text2(patch.diffs) + text[start_loc + len(text1):]) else: # Imperfect match. # Run a diff to get a framework of equivalent indices. diffs = self.diff_main(text1, text2, False) if (len(text1) > self.Match_MaxBits and self.diff_levenshtein(diffs) / float(len(text1)) > self.Patch_DeleteThreshold): # The end points match, but the content is unacceptably bad. results[-1] = False else: self.diff_cleanupSemanticLossless(diffs) index1 = 0 for (op, data) in patch.diffs: if op != self.DIFF_EQUAL: index2 = self.diff_xIndex(diffs, index1) if op == self.DIFF_INSERT: # Insertion text = text[:start_loc + index2] + data + text[start_loc + index2:] elif op == self.DIFF_DELETE: # Deletion text = text[:start_loc + index2] + text[start_loc + self.diff_xIndex(diffs, index1 + len(data)):] if op != self.DIFF_DELETE: index1 += len(data) # Strip the padding off. text = text[len(nullPadding):-len(nullPadding)] return (text, results) def patch_addPadding(self, patches): """Add some padding on text start and end so that edges can match something. Intended to be called only from within patch_apply. Args: patches: Array of Patch objects. Returns: The padding string added to each side. """ paddingLength = self.Patch_Margin nullPadding = "" for x in range(1, paddingLength + 1): nullPadding += chr(x) # Bump all the patches forward. for patch in patches: patch.start1 += paddingLength patch.start2 += paddingLength # Add some padding on start of first diff. patch = patches[0] diffs = patch.diffs if not diffs or diffs[0][0] != self.DIFF_EQUAL: # Add nullPadding equality. diffs.insert(0, (self.DIFF_EQUAL, nullPadding)) patch.start1 -= paddingLength # Should be 0. patch.start2 -= paddingLength # Should be 0. patch.length1 += paddingLength patch.length2 += paddingLength elif paddingLength > len(diffs[0][1]): # Grow first equality. extraLength = paddingLength - len(diffs[0][1]) newText = nullPadding[len(diffs[0][1]):] + diffs[0][1] diffs[0] = (diffs[0][0], newText) patch.start1 -= extraLength patch.start2 -= extraLength patch.length1 += extraLength patch.length2 += extraLength # Add some padding on end of last diff. patch = patches[-1] diffs = patch.diffs if not diffs or diffs[-1][0] != self.DIFF_EQUAL: # Add nullPadding equality. diffs.append((self.DIFF_EQUAL, nullPadding)) patch.length1 += paddingLength patch.length2 += paddingLength elif paddingLength > len(diffs[-1][1]): # Grow last equality. extraLength = paddingLength - len(diffs[-1][1]) newText = diffs[-1][1] + nullPadding[:extraLength] diffs[-1] = (diffs[-1][0], newText) patch.length1 += extraLength patch.length2 += extraLength return nullPadding def patch_splitMax(self, patches): """Look through the patches and break up any which are longer than the maximum limit of the match algorithm. Intended to be called only from within patch_apply. Args: patches: Array of Patch objects. """ patch_size = self.Match_MaxBits if patch_size == 0: # Python has the option of not splitting strings due to its ability # to handle integers of arbitrary precision. return for x in range(len(patches)): if patches[x].length1 <= patch_size: continue bigpatch = patches[x] # Remove the big old patch. del patches[x] x -= 1 start1 = bigpatch.start1 start2 = bigpatch.start2 precontext = '' while len(bigpatch.diffs) != 0: # Create one of several smaller patches. patch = patch_obj() empty = True patch.start1 = start1 - len(precontext) patch.start2 = start2 - len(precontext) if precontext: patch.length1 = patch.length2 = len(precontext) patch.diffs.append((self.DIFF_EQUAL, precontext)) while (len(bigpatch.diffs) != 0 and patch.length1 < patch_size - self.Patch_Margin): (diff_type, diff_text) = bigpatch.diffs[0] if diff_type == self.DIFF_INSERT: # Insertions are harmless. patch.length2 += len(diff_text) start2 += len(diff_text) patch.diffs.append(bigpatch.diffs.pop(0)) empty = False elif (diff_type == self.DIFF_DELETE and len(patch.diffs) == 1 and patch.diffs[0][0] == self.DIFF_EQUAL and len(diff_text) > 2 * patch_size): # This is a large deletion. Let it pass in one chunk. patch.length1 += len(diff_text) start1 += len(diff_text) empty = False patch.diffs.append((diff_type, diff_text)) del bigpatch.diffs[0] else: # Deletion or equality. Only take as much as we can stomach. diff_text = diff_text[:patch_size - patch.length1 - self.Patch_Margin] patch.length1 += len(diff_text) start1 += len(diff_text) if diff_type == self.DIFF_EQUAL: patch.length2 += len(diff_text) start2 += len(diff_text) else: empty = False patch.diffs.append((diff_type, diff_text)) if diff_text == bigpatch.diffs[0][1]: del bigpatch.diffs[0] else: bigpatch.diffs[0] = (bigpatch.diffs[0][0], bigpatch.diffs[0][1][len(diff_text):]) # Compute the head context for the next patch. precontext = self.diff_text2(patch.diffs) precontext = precontext[-self.Patch_Margin:] # Append the end context for this patch. postcontext = self.diff_text1(bigpatch.diffs)[:self.Patch_Margin] if postcontext: patch.length1 += len(postcontext) patch.length2 += len(postcontext) if len(patch.diffs) != 0 and patch.diffs[-1][0] == self.DIFF_EQUAL: patch.diffs[-1] = (self.DIFF_EQUAL, patch.diffs[-1][1] + postcontext) else: patch.diffs.append((self.DIFF_EQUAL, postcontext)) if not empty: x += 1 patches.insert(x, patch) def patch_toText(self, patches): """Take a list of patches and return a textual representation. Args: patches: Array of Patch objects. Returns: Text representation of patches. """ text = [] for patch in patches: text.append(str(patch)) return "".join(text) def patch_fromText(self, textline): """Parse a textual representation of patches and return a list of patch objects. Args: textline: Text representation of patches. Returns: Array of Patch objects. Raises: ValueError: If invalid input. """ patches = [] if not textline: return patches text = textline.split('\n') while len(text) != 0: m = re.match("^@@ -(\d+),?(\d*) \+(\d+),?(\d*) @@$", text[0]) if not m: raise ValueError("Invalid patch string: " + text[0]) patch = patch_obj() patches.append(patch) patch.start1 = int(m.group(1)) if m.group(2) == '': patch.start1 -= 1 patch.length1 = 1 elif m.group(2) == '0': patch.length1 = 0 else: patch.start1 -= 1 patch.length1 = int(m.group(2)) patch.start2 = int(m.group(3)) if m.group(4) == '': patch.start2 -= 1 patch.length2 = 1 elif m.group(4) == '0': patch.length2 = 0 else: patch.start2 -= 1 patch.length2 = int(m.group(4)) del text[0] while len(text) != 0: if text[0]: sign = text[0][0] else: sign = '' line = urllib.parse.unquote(text[0][1:]) if sign == '+': # Insertion. patch.diffs.append((self.DIFF_INSERT, line)) elif sign == '-': # Deletion. patch.diffs.append((self.DIFF_DELETE, line)) elif sign == ' ': # Minor equality. patch.diffs.append((self.DIFF_EQUAL, line)) elif sign == '@': # Start of next patch. break elif sign == '': # Blank line? Whatever. pass else: # WTF? raise ValueError("Invalid patch mode: '%s'\n%s" % (sign, line)) del text[0] return patches class patch_obj: """Class representing one patch operation. """ def __init__(self): """Initializes with an empty list of diffs. """ self.diffs = [] self.start1 = None self.start2 = None self.length1 = 0 self.length2 = 0 def __str__(self): """Emmulate GNU diff's format. Header: @@ -382,8 +481,9 @@ Indicies are printed as 1-based, not 0-based. Returns: The GNU diff string. """ if self.length1 == 0: coords1 = str(self.start1) + ",0" elif self.length1 == 1: coords1 = str(self.start1 + 1) else: coords1 = str(self.start1 + 1) + "," + str(self.length1) if self.length2 == 0: coords2 = str(self.start2) + ",0" elif self.length2 == 1: coords2 = str(self.start2 + 1) else: coords2 = str(self.start2 + 1) + "," + str(self.length2) text = ["@@ -", coords1, " +", coords2, " @@\n"] # Escape the body of the patch with %xx notation. for (op, data) in self.diffs: if op == diff_match_patch.DIFF_INSERT: text.append("+") elif op == diff_match_patch.DIFF_DELETE: text.append("-") elif op == diff_match_patch.DIFF_EQUAL: text.append(" ") # High ascii will raise UnicodeDecodeError. Use Unicode instead. data = data.encode("utf-8") text.append(urllib.parse.quote(data, "!~*'();/?:@&=+$,# ") + "\n") return "".join(text) diff-match-patch-20121119/README.rst0000640000175000017500000000056211632165424016467 0ustar lukeluke00000000000000================ diff-match-patch ================ Documentation on the Google site: http://code.google.com/p/google-diff-match-patch/ See README.original.txt for the original README. This is an unofficial package of the Python versions for PyPI, by Luke Plant http://lukeplant.me.uk/ See https://bitbucket.org/spookylukey/diff-match-patch for packaging issues.