pax_global_header 0000666 0000000 0000000 00000000064 13700665644 0014525 g ustar 00root root 0000000 0000000 52 comment=bcb54d096f2303ce80960e125b5991ac4d6325c3
ristretto-0.0.3/ 0000775 0000000 0000000 00000000000 13700665644 0013564 5 ustar 00root root 0000000 0000000 ristretto-0.0.3/.deepsource.toml 0000664 0000000 0000000 00000000275 13700665644 0016701 0 ustar 00root root 0000000 0000000 version = 1
test_patterns = [
'**/*_test.go'
]
exclude_patterns = [
]
[[analyzers]]
name = 'go'
enabled = true
[analyzers.meta]
import_path = 'github.com/dgraph-io/ristretto'
ristretto-0.0.3/.github/ 0000775 0000000 0000000 00000000000 13700665644 0015124 5 ustar 00root root 0000000 0000000 ristretto-0.0.3/.github/CODEOWNERS 0000664 0000000 0000000 00000000325 13700665644 0016517 0 ustar 00root root 0000000 0000000 # CODEOWNERS info: https://help.github.com/en/articles/about-code-owners
# Owners are automatically requested for review for PRs that changes code
# that they own.
* @manishrjain @karlmcguire @martinmr @anurags92
ristretto-0.0.3/.travis.yml 0000664 0000000 0000000 00000000203 13700665644 0015670 0 ustar 00root root 0000000 0000000 language: go
go:
- 1.12
- 1.13
env:
- GO111MODULE=on # Enable go module for all versions.
script:
- go test -race ./...
ristretto-0.0.3/CHANGELOG.md 0000664 0000000 0000000 00000002664 13700665644 0015405 0 ustar 00root root 0000000 0000000 # Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/)
and this project will adhere to [Semantic Versioning](http://semver.org/spec/v2.0.0.html) starting v1.0.0.
## Unreleased
### Changed
### Added
### Fixed
## [0.0.3] - 2020-07-06
[0.0.3]: https://github.com/dgraph-io/ristretto/compare/v0.0.2..v0.0.3
### Changed
### Added
### Fixed
- z: use MemHashString and xxhash.Sum64String ([#153][])
- Check conflict key before updating expiration map. ([#154][])
- Fix race condition in Cache.Clear ([#133][])
- Improve handling of updated items ([#168][])
- Fix droppedSets count while updating the item ([#171][])
## [0.0.2] - 2020-02-24
[0.0.2]: https://github.com/dgraph-io/ristretto/compare/v0.0.1..v0.0.2
### Added
- Sets with TTL. ([#122][])
### Fixed
- Fix the way metrics are handled for deletions. ([#111][])
- Support nil `*Cache` values in `Clear` and `Close`. ([#119][])
- Delete item immediately. ([#113][])
- Remove key from policy after TTL eviction. ([#130][])
[#111]: https://github.com/dgraph-io/ristretto/issues/111
[#113]: https://github.com/dgraph-io/ristretto/issues/113
[#119]: https://github.com/dgraph-io/ristretto/issues/119
[#122]: https://github.com/dgraph-io/ristretto/issues/122
[#130]: https://github.com/dgraph-io/ristretto/issues/130
## 0.0.1
First release. Basic cache functionality based on a LFU policy.
ristretto-0.0.3/LICENSE 0000664 0000000 0000000 00000023675 13700665644 0014606 0 ustar 00root root 0000000 0000000 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
ristretto-0.0.3/README.md 0000664 0000000 0000000 00000020354 13700665644 0015047 0 ustar 00root root 0000000 0000000 # Ristretto
[](http://godoc.org/github.com/dgraph-io/ristretto)
[](https://goreportcard.com/report/github.com/dgraph-io/ristretto)
[](https://gocover.io/github.com/dgraph-io/ristretto)
Ristretto is a fast, concurrent cache library built with a focus on performance and correctness.
The motivation to build Ristretto comes from the need for a contention-free
cache in [Dgraph][].
[Dgraph]: https://github.com/dgraph-io/dgraph
## Features
* **High Hit Ratios** - with our unique admission/eviction policy pairing, Ristretto's performance is best in class.
* **Eviction: SampledLFU** - on par with exact LRU and better performance on Search and Database traces.
* **Admission: TinyLFU** - extra performance with little memory overhead (12 bits per counter).
* **Fast Throughput** - we use a variety of techniques for managing contention and the result is excellent throughput.
* **Cost-Based Eviction** - any large new item deemed valuable can evict multiple smaller items (cost could be anything).
* **Fully Concurrent** - you can use as many goroutines as you want with little throughput degradation.
* **Metrics** - optional performance metrics for throughput, hit ratios, and other stats.
* **Simple API** - just figure out your ideal `Config` values and you're off and running.
## Status
Ristretto is usable but still under active development. We expect it to be production ready in the near future.
## Table of Contents
* [Usage](#Usage)
* [Example](#Example)
* [Config](#Config)
* [NumCounters](#Config)
* [MaxCost](#Config)
* [BufferItems](#Config)
* [Metrics](#Config)
* [OnEvict](#Config)
* [KeyToHash](#Config)
* [Cost](#Config)
* [Benchmarks](#Benchmarks)
* [Hit Ratios](#Hit-Ratios)
* [Search](#Search)
* [Database](#Database)
* [Looping](#Looping)
* [CODASYL](#CODASYL)
* [Throughput](#Throughput)
* [Mixed](#Mixed)
* [Read](#Read)
* [Write](#Write)
* [FAQ](#FAQ)
## Usage
### Example
```go
func main() {
cache, err := ristretto.NewCache(&ristretto.Config{
NumCounters: 1e7, // number of keys to track frequency of (10M).
MaxCost: 1 << 30, // maximum cost of cache (1GB).
BufferItems: 64, // number of keys per Get buffer.
})
if err != nil {
panic(err)
}
// set a value with a cost of 1
cache.Set("key", "value", 1)
// wait for value to pass through buffers
time.Sleep(10 * time.Millisecond)
value, found := cache.Get("key")
if !found {
panic("missing value")
}
fmt.Println(value)
cache.Del("key")
}
```
### Config
The `Config` struct is passed to `NewCache` when creating Ristretto instances (see the example above).
**NumCounters** `int64`
NumCounters is the number of 4-bit access counters to keep for admission and eviction. We've seen good performance in setting this to 10x the number of items you expect to keep in the cache when full.
For example, if you expect each item to have a cost of 1 and MaxCost is 100, set NumCounters to 1,000. Or, if you use variable cost values but expect the cache to hold around 10,000 items when full, set NumCounters to 100,000. The important thing is the *number of unique items* in the full cache, not necessarily the MaxCost value.
**MaxCost** `int64`
MaxCost is how eviction decisions are made. For example, if MaxCost is 100 and a new item with a cost of 1 increases total cache cost to 101, 1 item will be evicted.
MaxCost can also be used to denote the max size in bytes. For example, if MaxCost is 1,000,000 (1MB) and the cache is full with 1,000 1KB items, a new item (that's accepted) would cause 5 1KB items to be evicted.
MaxCost could be anything as long as it matches how you're using the cost values when calling Set.
**BufferItems** `int64`
BufferItems is the size of the Get buffers. The best value we've found for this is 64.
If for some reason you see Get performance decreasing with lots of contention (you shouldn't), try increasing this value in increments of 64. This is a fine-tuning mechanism and you probably won't have to touch this.
**Metrics** `bool`
Metrics is true when you want real-time logging of a variety of stats. The reason this is a Config flag is because there's a 10% throughput performance overhead.
**OnEvict** `func(hashes [2]uint64, value interface{}, cost int64)`
OnEvict is called for every eviction.
**KeyToHash** `func(key interface{}) [2]uint64`
KeyToHash is the hashing algorithm used for every key. If this is nil, Ristretto has a variety of [defaults depending on the underlying interface type](https://github.com/dgraph-io/ristretto/blob/master/z/z.go#L19-L41).
Note that if you want 128bit hashes you should use the full `[2]uint64`,
otherwise just fill the `uint64` at the `0` position and it will behave like
any 64bit hash.
**Cost** `func(value interface{}) int64`
Cost is an optional function you can pass to the Config in order to evaluate
item cost at runtime, and only for the Set calls that aren't dropped (this is
useful if calculating item cost is particularly expensive and you don't want to
waste time on items that will be dropped anyways).
To signal to Ristretto that you'd like to use this Cost function:
1. Set the Cost field to a non-nil function.
2. When calling Set for new items or item updates, use a `cost` of 0.
## Benchmarks
The benchmarks can be found in https://github.com/dgraph-io/benchmarks/tree/master/cachebench/ristretto.
### Hit Ratios
#### Search
This trace is described as "disk read accesses initiated by a large commercial
search engine in response to various web search requests."
#### Database
This trace is described as "a database server running at a commercial site
running an ERP application on top of a commercial database."
#### Looping
This trace demonstrates a looping access pattern.
#### CODASYL
This trace is described as "references to a CODASYL database for a one hour
period."
### Throughput
All throughput benchmarks were ran on an Intel Core i7-8700K (3.7GHz) with 16gb
of RAM.
#### Mixed
#### Read
#### Write
## FAQ
### How are you achieving this performance? What shortcuts are you taking?
We go into detail in the [Ristretto blog post](https://blog.dgraph.io/post/introducing-ristretto-high-perf-go-cache/), but in short: our throughput performance can be attributed to a mix of batching and eventual consistency. Our hit ratio performance is mostly due to an excellent [admission policy](https://arxiv.org/abs/1512.00727) and SampledLFU eviction policy.
As for "shortcuts," the only thing Ristretto does that could be construed as one is dropping some Set calls. That means a Set call for a new item (updates are guaranteed) isn't guaranteed to make it into the cache. The new item could be dropped at two points: when passing through the Set buffer or when passing through the admission policy. However, this doesn't affect hit ratios much at all as we expect the most popular items to be Set multiple times and eventually make it in the cache.
### Is Ristretto distributed?
No, it's just like any other Go library that you can import into your project and use in a single process.
ristretto-0.0.3/benchmarks/ 0000775 0000000 0000000 00000000000 13700665644 0015701 5 ustar 00root root 0000000 0000000 ristretto-0.0.3/benchmarks/Hit Ratios - CODASYL (ARC-OLTP).svg 0000664 0000000 0000000 00000266123 13700665644 0023061 0 ustar 00root root 0000000 0000000 Ù=çgÛ·Ÿn?Þ~¾ýh—GÊ#å‘¥z®V½³æ¹î¹<ÞòxËã-·<ÞòxËã-·<^tÙ=W oy¼åñ–Ç[ oy¼åñ–Ç[ o´ã=W oy¼åñ–Ç[ oy¼åñ–Ç[ ïÒ·÷\y¼åñ–Ç[ oy¼åñ–Ç[ oy¼[ƒßsåñ–Ç[ oy¼åñ–Ç[ oy¼åñ^&ž+·<ÞòxËã-·<ÞòxËã-÷62ô\y¼åñ–Ç[ oy¼åñ–Ç[ oy¼Ù¢çÊã5|˜>Œ æˆ Äb™Òsåñ–Ç[ oy¼åñ–Ç[ oy¼åñ~¦ãJóJy|åñ•ÇW _y|åñ•ÇW _y|˜kz®<¾òøÊã+¯<¾òøÊã+¯<¾€z®<¾òøÊã+¯<¾òøÊã+¯<¾eRê¹òøÊã+¯<¾òøÊã+¯<¾òø¶‘ªçÊã+¯<¾òøÊã+¯<¾òøÊã»Ì^=W _y|åñ•ÇW _y|åñ•ÇW ßmHë¹òøÊã+¯<¾òøÊã+¯<¾òø Ó\Ï•ÇW _y|åñ•ÇW _y|åñ•Ç÷ûzÎXh.4š †fCáéÐx8ùp¢ q"âdÄ ‰“'&NNœ 8IѨx2™Ò'L‹§qñ4/žÆÓÄxO3ãihaœ<Í“§ò4QžFÊÓLy*OSåi¬<¯ÉÇ>a²<–§Ùò4\ž¦ËÓxyš/OæiÂ<ï‰Ô>aÈaÞ<
œ§‰ó4ržfÎÓÐyš:Ocçiî<ß î>aô<Íž§áó4}žÆÏÓüy@Oèi=¿ÉúöMû2Ÿ^`Ši¦˜n`Êiþê¿~À'¦!˜Š`:‚) ¦%˜š`z™OS0UÁtSL[0uÁôSLc0•ÁtSLk0µÁôS Ls`uðŸ•Í{h:ÿQýªœ_•ó«r~UίÊùU9¿*ç_«rþ»‹œóWåüÏT9Ìôžñ=ó{øLðá3ÃgˆÏŸ1>s|ùLòå3Ëg˜Ï4Ÿq>ó|úLôé3Óg¨ÏTŸ±>s}ûLöí3Ûg¸ÏtŸñ>ó}üLøñ3ãgÈÏ”Ÿ1?s~ýLúõ3ëgØÏ´Ÿq?ó~þLüù3ógèÏÔŸ±?sÿL~G?Î~ þ8ýqüãüG€ - z 4èÈüB©/Ó @+€^ Í º2ölüÙŸAó‰±hãÑƤK›6>M£¦3@k€Þ ͺ´èÐ C@‹€ Mº´ èÐ( S@«€^ͺ´ èÐ0 c@Ë€žMº´
èÐ8 s@ë€Þͺ´è Ð@ ƒ@ Mº´è#ÐH “@+^ͺ ´è'ÐP £@KžMº
´è+ÐX ³@kÞͺ ´è/Ð` Ã@‹ Mº ´è3Ðh Ó@«Ášß¥e®Û@»~
:´ è9Ðt ë@Û¾:´ è=XcÎÇ=ÿóç>1},úxô1éãÒe®A#‚Nz4#èFÐŽ AC‚Ž- z4%èJЖ /Ac‚Î z4'èNО ?Aƒ‚-
z4)èRЦ OA£‚N
z4+èVЮ _AÂŽ- z4-èZж oAã‚Î z4/è^о Aƒ- z41èbÐÆ A#ƒN z43èfÐÎ ŸaÏ6‚̵4èiÐÔ «A[ƒ¾
:´6èmÐÜ »A{ƒþ
:œ_
þ¥à_
þ¥à_
þ¥à_
þ¥à_
þWR°JA© VP,¨” êÞ ?#~fü ù¿)ï3çgÐϤŸQï¬W<¨ ”ê„
B ¡†PD¨"”ê…„JB)¡–PL¨&”ê …ŠBI¡¦PT¨*”ê
……ÊBi¡¶P\¨.”ê †
C‰¡ÆPd¨2”ê …†JC©¡ÖPl¨6”ê
‡ŠCÉ¡æPt¨:” ê…‡ÊCé¡ö¸f¡W÷¦üP(@T J5ˆ"D¢ Q‡(DT"JµˆbD5¢ Q(HT$J5‰¢DM¢(Q•(KÔ%
•‰ÒDm¢8¹ÆßÁ‡7o<ÞŸÉó‰±yãóÆèÉ\¡¢RQª¨U+ªåŠzEÁ¢bQ²¨Y-ªe‹ºEá¢rQº¨]/ªå‹úE£‚Q¨a1ªeŒ:F!£’Qʨe3ªåŒzFA£¢QÒ¨i5ªeºFa£²û(¿}”ß>Êo å·ò_ÚGÑ¡B+„^Í÷¼ôâ߯~
:"´Dè‰Ð¡+B[„¾:#Æ ¹Ý‰ë¸ß‰ ž¸á‰+ž¸ã‰Kž¸å‰kž¸ç‰‹ž¸é‰«ž¸ë‰Ëž¸í‰ëž¸ï‰ Ÿ¸ñ‰+Ÿ¸ó‰KŸ¸õ‰kŸ¸÷‰‹ŸÜó÷넼§ ™2dÚ©C¦™Bä¯ñ‰éD¦qBºŠ{ ¸Š› ¸
Š» ¸ ŠÛ ¸Šû ¸Š¡¸Š;¡¸Š[¡¸Š{¡¸Š›¡¸Š»¡¸ ŠÛ¡¸ Šû¡¸ Š¢¸"Š;¢¸$Š[¢¸&Š{¢¸(Š›¢¸*Š»¢¸,ŠÛ¢¸.Šû¢¸0Š£¸2Š;£<óª“Ì]ŽQ\ ÅÍQ\ ÅÝQ\ ÅíQ\ ÅýQ\ Å
R\!Å R\"Å-R\#Å=R\$ÅMR\%Å]R\&Åéãç7pœà8Âq†ãÇ)Žc ç8rœä8Êq–ã0ÇiŽã ç9tœè8Òq¦ãPÇ©Žc ç:vži¦
›.lÊ°iæ›>ì¯ó‰©ÄdîÇ)c ç<zœô8êqÖã°Çiã ç=|œø8òqæãÐÇ©c ç>~œü8úqöãðÇéã ç? t hÐ @€6 } ÞyÁMæz4èÐ @C€Ž -z4è
Ð /@c€Î z4èÐ ?@ƒ€-z4 èÐ& O@£€Nz4 èÐ. _@ÀŽ-z4
èÐ6 o@ã€Îz4è Ð> @-į&ÿÕä¿šüW“ÿjò_Mþ«É5ù¯&ÿWjr ld°’ÁN†w¦åŒË™—30gbÎÈœ™9Cóojú„sÓ~ lh°¢ÁŽKli°¦Áž‹lj°ªÁ®Ëlk°®Á¾ ll°²ÁÎKlm°¶ÁÞ‹ln°ºá›?#B—c{ƒõ
ö7Xà`ƒƒv8Xâ`‹ƒ5ö8Xä`“ƒUv9Xæ`›ƒuö9Xè`£ƒ•v:Xê`«ƒµö:Xì`³ƒÕv;Xî`»ƒõö;Xð`ÃvXú`냵ö>Xü`óƒÕv?Xþ`ûƒõö?X aÄ7^iÌÒ¸¥±Kã—Æ0cË4žéÏ4kÒ6oã4Îi¬Óx§1OãžÆ>éŸ,ƒbË XÅ2(–A± ŠeP,ƒrΟ 2ÖöÿÆ6\,µb©KXjÅR+–Z±ÔŠ¥V,µb©KXjÅR+–Z±ÔŠ¥V,µb©KXjÅR+–Z±ÔŠ¥V,µb©KXjÅR+–Z±ÔŠ¥V,µb©KXjÅR+¾d_²Š/YÅ—¬âKVñ%«ø’U|É*¾d_²Š/YÅ—¬âKVñ%«ø’U|É*¾d_²Š/YÅ—¬rÎ
é&}É*¾d_²Š/YÅ—¬âKVñ%«ø’U|É*¾d_²Š/YÅ—¬âKVñ%«ø’U|É*¾d_²Š/YÅ3X a1…ÅsX b1‰Å(Æûÿ™Ÿû?þÀ$€‰ “&øÙeþ\›L^ð ™Íb6‹á,¦³Ïb>‹-&´ÑbF‹!-)íßlã.æÆÌ[_&Ç cvŒá1¦Ç c~Œ2&È!c†Œ!2¦È#cŽŒA2&É%c–Œa2¦É'cžŒ2&Ê)c¦Œ¡2¦Ê+c®ŒÁ2&Ë-c¶Œá2¦Ë/c¾Œ3&Ì1cÆŒ!3¦Ì3cÎŒA3&Í5cÖŒa3¦Í7cÞŒ3&Î9c挡3¦ÎLìœÜ9Ás’çDÏÉž>'}Nüœü9tèDÐÉ B'…N :At’è_,úF}bâèäÑ ¤¾¨þûï÷ßï¾ß|¿ ø~?ðý~àûwúïà n¿F{*m=ì”ÚÓjO=½öÛÓlOµ=Ýö”ÛÓnO½=ýöÜÓpOÅ= ÷”ÜÓrOÍ==÷ÝÓtOÕ=]÷”ÝÓvOÝ=}÷ÞÓxOå=÷”ÞÓzOí=½÷ßÓ|Oõ=Ý÷”ßÓ~Oý=ý÷àÓ€O> ø$»iÁ§Ÿ |Šði§
Ÿ.|ÊðiçŸ>|
ñiħŸN|JñiŧŸ^|Šq›ñXÇn<–ã± õxìÇcA òX‘ÇŽ<–ä±%5yìÉcQ ›òX•Ç®<–å±-uyì˳æÏäÔ2Z™ÇÎ<–æ±5µyÖüŽ1?dÌ/?eøÄü˜1¿fÌÏó{Æü !s+ôØ¡Ç=¶è±F=z,Òc“ «ôØ¥Ç2=¶é±N}z,Ôc£ +õØ©ÇR=¶ê±V½z,Öc³ «õØÇr=¶ë±^ýz,Øcà +öرÇÔSWL]1uÅÔSWL]1uÅÔSWL]1uÅÔSWL]1uÅÔSWL]1uÅÔSWL]1uÅÔSWL]1ueR—
|¬àc KøØÂÇ>öð±ˆÏ¼çô[øüV~+¿ßŠÀoEà·"ð[øüVþŸ®DÕEÙEÝEáEåEéEíEñEõEùEýeÏÿ/ƒJ F
FfÏ_Ô15êBŒJŒRŒZŒbŒjŒrŒzÌž5kFÖ̬¿¡å3¶fnÍàšÉåèR›QœQQžQŸQ Q¡Q¢Q£Q¤Q¥Q¦Q§Q¨Q©QªQ«Q¬QQ®Q¯Q°Q±Q²Q³Q´QµQ¶Q·Q¸Q¹QºQ»Q¼Q½Q¾Q¿QÀQÁQÂQÃQÄQÅQÆQÇQÈQÉQÊQËQÌQÍQÎQÏQÐQÑQÒQÓQÔQÕQÖQ×QØQÙQÚQÛQÜQÝQÞQßQàQáQâQãQä¹æÿCæê<
=*ýø·²]„æc ristretto-0.0.3/sim/sim.go 0000664 0000000 0000000 00000012424 13700665644 0015476 0 ustar 00root root 0000000 0000000 /*
* Copyright 2019 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
package sim
import (
"bufio"
"errors"
"fmt"
"io"
"math/rand"
"strconv"
"strings"
"time"
)
var (
// ErrDone is returned when the underlying file has ran out of lines.
ErrDone = errors.New("no more values in the Simulator")
// ErrBadLine is returned when the trace file line is unrecognizable to
// the Parser.
ErrBadLine = errors.New("bad line for trace format")
)
// Simulator is the central type of the `sim` package. It is a function
// returning a key from some source (composed from the other functions in this
// package, either generated or parsed). You can use these Simulators to
// approximate access distributions.
type Simulator func() (uint64, error)
// NewZipfian creates a Simulator returning numbers following a Zipfian [1]
// distribution infinitely. Zipfian distributions are useful for simulating real
// workloads.
//
// [1]: https://en.wikipedia.org/wiki/Zipf%27s_law
func NewZipfian(s, v float64, n uint64) Simulator {
z := rand.NewZipf(rand.New(rand.NewSource(time.Now().UnixNano())), s, v, n)
return func() (uint64, error) {
return z.Uint64(), nil
}
}
// NewUniform creates a Simulator returning uniformly distributed [1] (random)
// numbers [0, max) infinitely.
//
// [1]: https://en.wikipedia.org/wiki/Uniform_distribution_(continuous)
func NewUniform(max uint64) Simulator {
m := int64(max)
r := rand.New(rand.NewSource(time.Now().UnixNano()))
return func() (uint64, error) {
return uint64(r.Int63n(m)), nil
}
}
// Parser is used as a parameter to NewReader so we can create Simulators from
// varying trace file formats easily.
type Parser func(string, error) ([]uint64, error)
// NewReader creates a Simulator from two components: the Parser, which is a
// filetype specific function for parsing lines, and the file itself, which will
// be read from.
//
// When every line in the file has been read, ErrDone will be returned. For some
// trace formats (LIRS) there is one item per line. For others (ARC) there is a
// range of items on each line. Thus, the true number of items in each file
// is hard to determine, so it's up to the user to handle ErrDone accordingly.
func NewReader(parser Parser, file io.Reader) Simulator {
b := bufio.NewReader(file)
s := make([]uint64, 0)
i := -1
var err error
return func() (uint64, error) {
// only parse a new line when we've run out of items
if i++; i == len(s) {
// parse sequence from line
if s, err = parser(b.ReadString('\n')); err != nil {
s = []uint64{0}
}
i = 0
}
return s[i], err
}
}
// ParseLIRS takes a single line of input from a LIRS trace file as described in
// multiple papers [1] and returns a slice containing one number. A nice
// collection of LIRS trace files can be found in Ben Manes' repo [2].
//
// [1]: https://en.wikipedia.org/wiki/LIRS_caching_algorithm
// [2]: https://git.io/fj9gU
func ParseLIRS(line string, err error) ([]uint64, error) {
if line = strings.TrimSpace(line); line != "" {
// example: "1\r\n"
key, err := strconv.ParseUint(line, 10, 64)
return []uint64{key}, err
}
return nil, ErrDone
}
// ParseARC takes a single line of input from an ARC trace file as described in
// "ARC: a self-tuning, low overhead replacement cache" [1] by Nimrod Megiddo
// and Dharmendra S. Modha [1] and returns a sequence of numbers generated from
// the line and any error. For use with NewReader.
//
// [1]: https://scinapse.io/papers/1860107648
func ParseARC(line string, err error) ([]uint64, error) {
if line != "" {
// example: "0 5 0 0\n"
//
// - first block: starting number in sequence
// - second block: number of items in sequence
// - third block: ignore
// - fourth block: global line number (not used)
cols := strings.Fields(line)
if len(cols) != 4 {
return nil, ErrBadLine
}
start, err := strconv.ParseUint(cols[0], 10, 64)
if err != nil {
return nil, err
}
count, err := strconv.ParseUint(cols[1], 10, 64)
if err != nil {
return nil, err
}
// populate sequence from start to start + count
seq := make([]uint64, count)
for i := range seq {
seq[i] = start + uint64(i)
}
return seq, nil
}
return nil, ErrDone
}
// Collection evaluates the Simulator size times and saves each item to the
// returned slice.
func Collection(simulator Simulator, size uint64) []uint64 {
collection := make([]uint64, size)
for i := range collection {
collection[i], _ = simulator()
}
return collection
}
// StringCollection evaluates the Simulator size times and saves each item to
// the returned slice, after converting it to a string.
func StringCollection(simulator Simulator, size uint64) []string {
collection := make([]string, size)
for i := range collection {
n, _ := simulator()
collection[i] = fmt.Sprintf("%d", n)
}
return collection
}
ristretto-0.0.3/sim/sim_test.go 0000664 0000000 0000000 00000004440 13700665644 0016534 0 ustar 00root root 0000000 0000000 /*
* Copyright 2019 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
package sim
import (
"bytes"
"compress/gzip"
"os"
"testing"
)
func TestZipfian(t *testing.T) {
s := NewZipfian(1.5, 1, 100)
m := make(map[uint64]uint64, 100)
for i := 0; i < 100; i++ {
k, err := s()
if err != nil {
t.Fatal(err)
}
m[k]++
}
if len(m) == 0 || len(m) == 100 {
t.Fatal("zipfian not skewed")
}
}
func TestUniform(t *testing.T) {
s := NewUniform(100)
for i := 0; i < 100; i++ {
if _, err := s(); err != nil {
t.Fatal(err)
}
}
}
func TestParseLIRS(t *testing.T) {
s := NewReader(ParseLIRS, bytes.NewReader([]byte{
'0', '\n',
'1', '\r', '\n',
'2', '\r', '\n',
}))
for i := uint64(0); i < 3; i++ {
v, err := s()
if err != nil {
t.Fatal(err)
}
if v != i {
t.Fatal("value mismatch")
}
}
}
func TestReadLIRS(t *testing.T) {
f, err := os.Open("./gli.lirs.gz")
if err != nil {
t.Fatal(err)
}
r, err := gzip.NewReader(f)
if err != nil {
t.Fatal(err)
}
s := NewReader(ParseLIRS, r)
for i := uint64(0); i < 100; i++ {
if _, err = s(); err != nil {
t.Fatal(err)
}
}
}
func TestParseARC(t *testing.T) {
s := NewReader(ParseARC, bytes.NewReader([]byte{
'1', '2', '7', ' ', '6', '4', ' ', '0', ' ', '0', '\r', '\n',
'1', '9', '1', ' ', '3', '6', ' ', '0', ' ', '0', '\r', '\n',
}))
for i := uint64(0); i < 100; i++ {
v, err := s()
if err != nil {
t.Fatal(err)
}
if v != 127+i {
t.Fatal("value mismatch")
}
}
}
func TestCollection(t *testing.T) {
s := NewUniform(100)
c := Collection(s, 100)
if len(c) != 100 {
t.Fatal("collection not full")
}
}
func TestStringCollection(t *testing.T) {
s := NewUniform(100)
c := StringCollection(s, 100)
if len(c) != 100 {
t.Fatal("string collection not full")
}
}
ristretto-0.0.3/sketch.go 0000664 0000000 0000000 00000007305 13700665644 0015401 0 ustar 00root root 0000000 0000000 /*
* Copyright 2019 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
// This package includes multiple probabalistic data structures needed for
// admission/eviction metadata. Most are Counting Bloom Filter variations, but
// a caching-specific feature that is also required is a "freshness" mechanism,
// which basically serves as a "lifetime" process. This freshness mechanism
// was described in the original TinyLFU paper [1], but other mechanisms may
// be better suited for certain data distributions.
//
// [1]: https://arxiv.org/abs/1512.00727
package ristretto
import (
"fmt"
"math/rand"
"time"
)
// cmSketch is a Count-Min sketch implementation with 4-bit counters, heavily
// based on Damian Gryski's CM4 [1].
//
// [1]: https://github.com/dgryski/go-tinylfu/blob/master/cm4.go
type cmSketch struct {
rows [cmDepth]cmRow
seed [cmDepth]uint64
mask uint64
}
const (
// cmDepth is the number of counter copies to store (think of it as rows).
cmDepth = 4
)
func newCmSketch(numCounters int64) *cmSketch {
if numCounters == 0 {
panic("cmSketch: bad numCounters")
}
// Get the next power of 2 for better cache performance.
numCounters = next2Power(numCounters)
sketch := &cmSketch{mask: uint64(numCounters - 1)}
// Initialize rows of counters and seeds.
source := rand.New(rand.NewSource(time.Now().UnixNano()))
for i := 0; i < cmDepth; i++ {
sketch.seed[i] = source.Uint64()
sketch.rows[i] = newCmRow(numCounters)
}
return sketch
}
// Increment increments the count(ers) for the specified key.
func (s *cmSketch) Increment(hashed uint64) {
for i := range s.rows {
s.rows[i].increment((hashed ^ s.seed[i]) & s.mask)
}
}
// Estimate returns the value of the specified key.
func (s *cmSketch) Estimate(hashed uint64) int64 {
min := byte(255)
for i := range s.rows {
val := s.rows[i].get((hashed ^ s.seed[i]) & s.mask)
if val < min {
min = val
}
}
return int64(min)
}
// Reset halves all counter values.
func (s *cmSketch) Reset() {
for _, r := range s.rows {
r.reset()
}
}
// Clear zeroes all counters.
func (s *cmSketch) Clear() {
for _, r := range s.rows {
r.clear()
}
}
// cmRow is a row of bytes, with each byte holding two counters.
type cmRow []byte
func newCmRow(numCounters int64) cmRow {
return make(cmRow, numCounters/2)
}
func (r cmRow) get(n uint64) byte {
return byte(r[n/2]>>((n&1)*4)) & 0x0f
}
func (r cmRow) increment(n uint64) {
// Index of the counter.
i := n / 2
// Shift distance (even 0, odd 4).
s := (n & 1) * 4
// Counter value.
v := (r[i] >> s) & 0x0f
// Only increment if not max value (overflow wrap is bad for LFU).
if v < 15 {
r[i] += 1 << s
}
}
func (r cmRow) reset() {
// Halve each counter.
for i := range r {
r[i] = (r[i] >> 1) & 0x77
}
}
func (r cmRow) clear() {
// Zero each counter.
for i := range r {
r[i] = 0
}
}
func (r cmRow) string() string {
s := ""
for i := uint64(0); i < uint64(len(r)*2); i++ {
s += fmt.Sprintf("%02d ", (r[(i/2)]>>((i&1)*4))&0x0f)
}
s = s[:len(s)-1]
return s
}
// next2Power rounds x up to the next power of 2, if it's not already one.
func next2Power(x int64) int64 {
x--
x |= x >> 1
x |= x >> 2
x |= x >> 4
x |= x >> 8
x |= x >> 16
x |= x >> 32
x++
return x
}
ristretto-0.0.3/sketch_test.go 0000664 0000000 0000000 00000002536 13700665644 0016441 0 ustar 00root root 0000000 0000000 package ristretto
import (
"testing"
"github.com/stretchr/testify/require"
)
func TestSketch(t *testing.T) {
defer func() {
require.NotNil(t, recover())
}()
s := newCmSketch(5)
require.Equal(t, uint64(7), s.mask)
newCmSketch(0)
}
func TestSketchIncrement(t *testing.T) {
s := newCmSketch(16)
s.Increment(1)
s.Increment(5)
s.Increment(9)
for i := 0; i < cmDepth; i++ {
if s.rows[i].string() != s.rows[0].string() {
break
}
require.False(t, i == cmDepth-1, "identical rows, bad seeding")
}
}
func TestSketchEstimate(t *testing.T) {
s := newCmSketch(16)
s.Increment(1)
s.Increment(1)
require.Equal(t, int64(2), s.Estimate(1))
require.Equal(t, int64(0), s.Estimate(0))
}
func TestSketchReset(t *testing.T) {
s := newCmSketch(16)
s.Increment(1)
s.Increment(1)
s.Increment(1)
s.Increment(1)
s.Reset()
require.Equal(t, int64(2), s.Estimate(1))
}
func TestSketchClear(t *testing.T) {
s := newCmSketch(16)
for i := 0; i < 16; i++ {
s.Increment(uint64(i))
}
s.Clear()
for i := 0; i < 16; i++ {
require.Equal(t, int64(0), s.Estimate(uint64(i)))
}
}
func BenchmarkSketchIncrement(b *testing.B) {
s := newCmSketch(16)
b.SetBytes(1)
for n := 0; n < b.N; n++ {
s.Increment(1)
}
}
func BenchmarkSketchEstimate(b *testing.B) {
s := newCmSketch(16)
s.Increment(1)
b.SetBytes(1)
for n := 0; n < b.N; n++ {
s.Estimate(1)
}
}
ristretto-0.0.3/store.go 0000664 0000000 0000000 00000012623 13700665644 0015253 0 ustar 00root root 0000000 0000000 /*
* Copyright 2019 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
package ristretto
import (
"sync"
"time"
)
type storeItem struct {
key uint64
conflict uint64
value interface{}
expiration time.Time
}
// store is the interface fulfilled by all hash map implementations in this
// file. Some hash map implementations are better suited for certain data
// distributions than others, so this allows us to abstract that out for use
// in Ristretto.
//
// Every store is safe for concurrent usage.
type store interface {
// Get returns the value associated with the key parameter.
Get(uint64, uint64) (interface{}, bool)
// Expiration returns the expiration time for this key.
Expiration(uint64) time.Time
// Set adds the key-value pair to the Map or updates the value if it's
// already present. The key-value pair is passed as a pointer to an
// item object.
Set(*item)
// Del deletes the key-value pair from the Map.
Del(uint64, uint64) (uint64, interface{})
// Update attempts to update the key with a new value and returns true if
// successful.
Update(*item) bool
// Cleanup removes items that have an expired TTL.
Cleanup(policy policy, onEvict onEvictFunc)
// Clear clears all contents of the store.
Clear()
}
// newStore returns the default store implementation.
func newStore() store {
return newShardedMap()
}
const numShards uint64 = 256
type shardedMap struct {
shards []*lockedMap
expiryMap *expirationMap
}
func newShardedMap() *shardedMap {
sm := &shardedMap{
shards: make([]*lockedMap, int(numShards)),
expiryMap: newExpirationMap(),
}
for i := range sm.shards {
sm.shards[i] = newLockedMap(sm.expiryMap)
}
return sm
}
func (sm *shardedMap) Get(key, conflict uint64) (interface{}, bool) {
return sm.shards[key%numShards].get(key, conflict)
}
func (sm *shardedMap) Expiration(key uint64) time.Time {
return sm.shards[key%numShards].Expiration(key)
}
func (sm *shardedMap) Set(i *item) {
if i == nil {
// If item is nil make this Set a no-op.
return
}
sm.shards[i.key%numShards].Set(i)
}
func (sm *shardedMap) Del(key, conflict uint64) (uint64, interface{}) {
return sm.shards[key%numShards].Del(key, conflict)
}
func (sm *shardedMap) Update(newItem *item) bool {
return sm.shards[newItem.key%numShards].Update(newItem)
}
func (sm *shardedMap) Cleanup(policy policy, onEvict onEvictFunc) {
sm.expiryMap.cleanup(sm, policy, onEvict)
}
func (sm *shardedMap) Clear() {
for i := uint64(0); i < numShards; i++ {
sm.shards[i].Clear()
}
}
type lockedMap struct {
sync.RWMutex
data map[uint64]storeItem
em *expirationMap
}
func newLockedMap(em *expirationMap) *lockedMap {
return &lockedMap{
data: make(map[uint64]storeItem),
em: em,
}
}
func (m *lockedMap) get(key, conflict uint64) (interface{}, bool) {
m.RLock()
item, ok := m.data[key]
m.RUnlock()
if !ok {
return nil, false
}
if conflict != 0 && (conflict != item.conflict) {
return nil, false
}
// Handle expired items.
if !item.expiration.IsZero() && time.Now().After(item.expiration) {
return nil, false
}
return item.value, true
}
func (m *lockedMap) Expiration(key uint64) time.Time {
m.RLock()
defer m.RUnlock()
return m.data[key].expiration
}
func (m *lockedMap) Set(i *item) {
if i == nil {
// If the item is nil make this Set a no-op.
return
}
m.Lock()
defer m.Unlock()
item, ok := m.data[i.key]
if ok {
// The item existed already. We need to check the conflict key and reject the
// update if they do not match. Only after that the expiration map is updated.
if i.conflict != 0 && (i.conflict != item.conflict) {
return
}
m.em.update(i.key, i.conflict, item.expiration, i.expiration)
} else {
// The value is not in the map already. There's no need to return anything.
// Simply add the expiration map.
m.em.add(i.key, i.conflict, i.expiration)
}
m.data[i.key] = storeItem{
key: i.key,
conflict: i.conflict,
value: i.value,
expiration: i.expiration,
}
}
func (m *lockedMap) Del(key, conflict uint64) (uint64, interface{}) {
m.Lock()
item, ok := m.data[key]
if !ok {
m.Unlock()
return 0, nil
}
if conflict != 0 && (conflict != item.conflict) {
m.Unlock()
return 0, nil
}
if !item.expiration.IsZero() {
m.em.del(key, item.expiration)
}
delete(m.data, key)
m.Unlock()
return item.conflict, item.value
}
func (m *lockedMap) Update(newItem *item) bool {
m.Lock()
item, ok := m.data[newItem.key]
if !ok {
m.Unlock()
return false
}
if newItem.conflict != 0 && (newItem.conflict != item.conflict) {
m.Unlock()
return false
}
m.em.update(newItem.key, newItem.conflict, item.expiration, newItem.expiration)
m.data[newItem.key] = storeItem{
key: newItem.key,
conflict: newItem.conflict,
value: newItem.value,
expiration: newItem.expiration,
}
m.Unlock()
return true
}
func (m *lockedMap) Clear() {
m.Lock()
m.data = make(map[uint64]storeItem)
m.Unlock()
}
ristretto-0.0.3/store_test.go 0000664 0000000 0000000 00000006721 13700665644 0016314 0 ustar 00root root 0000000 0000000 package ristretto
import (
"testing"
"github.com/dgraph-io/ristretto/z"
"github.com/stretchr/testify/require"
)
func TestStoreSetGet(t *testing.T) {
s := newStore()
key, conflict := z.KeyToHash(1)
i := item{
key: key,
conflict: conflict,
value: 2,
}
s.Set(&i)
val, ok := s.Get(key, conflict)
require.True(t, ok)
require.Equal(t, 2, val.(int))
i.value = 3
s.Set(&i)
val, ok = s.Get(key, conflict)
require.True(t, ok)
require.Equal(t, 3, val.(int))
key, conflict = z.KeyToHash(2)
i = item{
key: key,
conflict: conflict,
value: 2,
}
s.Set(&i)
val, ok = s.Get(key, conflict)
require.True(t, ok)
require.Equal(t, 2, val.(int))
}
func TestStoreDel(t *testing.T) {
s := newStore()
key, conflict := z.KeyToHash(1)
i := item{
key: key,
conflict: conflict,
value: 1,
}
s.Set(&i)
s.Del(key, conflict)
val, ok := s.Get(key, conflict)
require.False(t, ok)
require.Nil(t, val)
s.Del(2, 0)
}
func TestStoreClear(t *testing.T) {
s := newStore()
for i := uint64(0); i < 1000; i++ {
key, conflict := z.KeyToHash(i)
it := item{
key: key,
conflict: conflict,
value: i,
}
s.Set(&it)
}
s.Clear()
for i := uint64(0); i < 1000; i++ {
key, conflict := z.KeyToHash(i)
val, ok := s.Get(key, conflict)
require.False(t, ok)
require.Nil(t, val)
}
}
func TestStoreUpdate(t *testing.T) {
s := newStore()
key, conflict := z.KeyToHash(1)
i := item{
key: key,
conflict: conflict,
value: 1,
}
s.Set(&i)
i.value = 2
require.True(t, s.Update(&i))
val, ok := s.Get(key, conflict)
require.True(t, ok)
require.NotNil(t, val)
val, ok = s.Get(key, conflict)
require.True(t, ok)
require.Equal(t, 2, val.(int))
i.value = 3
require.True(t, s.Update(&i))
val, ok = s.Get(key, conflict)
require.True(t, ok)
require.Equal(t, 3, val.(int))
key, conflict = z.KeyToHash(2)
i = item{
key: key,
conflict: conflict,
value: 2,
}
require.False(t, s.Update(&i))
val, ok = s.Get(key, conflict)
require.False(t, ok)
require.Nil(t, val)
}
func TestStoreCollision(t *testing.T) {
s := newShardedMap()
s.shards[1].Lock()
s.shards[1].data[1] = storeItem{
key: 1,
conflict: 0,
value: 1,
}
s.shards[1].Unlock()
val, ok := s.Get(1, 1)
require.False(t, ok)
require.Nil(t, val)
i := item{
key: 1,
conflict: 1,
value: 2,
}
s.Set(&i)
val, ok = s.Get(1, 0)
require.True(t, ok)
require.NotEqual(t, 2, val.(int))
require.False(t, s.Update(&i))
val, ok = s.Get(1, 0)
require.True(t, ok)
require.NotEqual(t, 2, val.(int))
s.Del(1, 1)
val, ok = s.Get(1, 0)
require.True(t, ok)
require.NotNil(t, val)
}
func BenchmarkStoreGet(b *testing.B) {
s := newStore()
key, conflict := z.KeyToHash(1)
i := item{
key: key,
conflict: conflict,
value: 1,
}
s.Set(&i)
b.SetBytes(1)
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
s.Get(key, conflict)
}
})
}
func BenchmarkStoreSet(b *testing.B) {
s := newStore()
key, conflict := z.KeyToHash(1)
b.SetBytes(1)
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
i := item{
key: key,
conflict: conflict,
value: 1,
}
s.Set(&i)
}
})
}
func BenchmarkStoreUpdate(b *testing.B) {
s := newStore()
key, conflict := z.KeyToHash(1)
i := item{
key: key,
conflict: conflict,
value: 1,
}
s.Set(&i)
b.SetBytes(1)
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
s.Update(&item{
key: key,
conflict: conflict,
value: 2,
})
}
})
}
ristretto-0.0.3/stress_test.go 0000664 0000000 0000000 00000006647 13700665644 0016512 0 ustar 00root root 0000000 0000000 package ristretto
import (
"container/heap"
"fmt"
"math/rand"
"runtime"
"sync"
"testing"
"time"
"github.com/dgraph-io/ristretto/sim"
"github.com/stretchr/testify/require"
)
func TestStressSetGet(t *testing.T) {
c, err := NewCache(&Config{
NumCounters: 1000,
MaxCost: 100,
BufferItems: 64,
Metrics: true,
})
require.NoError(t, err)
for i := 0; i < 100; i++ {
c.Set(i, i, 1)
}
time.Sleep(wait)
wg := &sync.WaitGroup{}
for i := 0; i < runtime.GOMAXPROCS(0); i++ {
wg.Add(1)
go func() {
r := rand.New(rand.NewSource(time.Now().UnixNano()))
for a := 0; a < 1000; a++ {
k := r.Int() % 10
if val, ok := c.Get(k); val == nil || !ok {
err = fmt.Errorf("expected %d but got nil", k)
break
} else if val != nil && val.(int) != k {
err = fmt.Errorf("expected %d but got %d", k, val.(int))
break
}
}
wg.Done()
}()
}
wg.Wait()
require.NoError(t, err)
require.Equal(t, 1.0, c.Metrics.Ratio())
}
func TestStressHitRatio(t *testing.T) {
key := sim.NewZipfian(1.0001, 1, 1000)
c, err := NewCache(&Config{
NumCounters: 1000,
MaxCost: 100,
BufferItems: 64,
Metrics: true,
})
require.NoError(t, err)
o := NewClairvoyant(100)
for i := 0; i < 10000; i++ {
k, err := key()
require.NoError(t, err)
if _, ok := o.Get(k); !ok {
o.Set(k, k, 1)
}
if _, ok := c.Get(k); !ok {
c.Set(k, k, 1)
}
}
t.Logf("actual: %.2f, optimal: %.2f", c.Metrics.Ratio(), o.Metrics().Ratio())
}
// Clairvoyant is a mock cache providing us with optimal hit ratios to compare
// with Ristretto's. It looks ahead and evicts the absolute least valuable item,
// which we try to approximate in a real cache.
type Clairvoyant struct {
capacity uint64
hits map[uint64]uint64
access []uint64
}
func NewClairvoyant(capacity uint64) *Clairvoyant {
return &Clairvoyant{
capacity: capacity,
hits: make(map[uint64]uint64),
access: make([]uint64, 0),
}
}
// Get just records the cache access so that we can later take this event into
// consideration when calculating the absolute least valuable item to evict.
func (c *Clairvoyant) Get(key interface{}) (interface{}, bool) {
c.hits[key.(uint64)]++
c.access = append(c.access, key.(uint64))
return nil, false
}
// Set isn't important because it is only called after a Get (in the case of our
// hit ratio benchmarks, at least).
func (c *Clairvoyant) Set(key, value interface{}, cost int64) bool {
return false
}
func (c *Clairvoyant) Metrics() *Metrics {
stat := newMetrics()
look := make(map[uint64]struct{}, c.capacity)
data := &clairvoyantHeap{}
heap.Init(data)
for _, key := range c.access {
if _, has := look[key]; has {
stat.add(hit, 0, 1)
continue
}
if uint64(data.Len()) >= c.capacity {
victim := heap.Pop(data)
delete(look, victim.(*clairvoyantItem).key)
}
stat.add(miss, 0, 1)
look[key] = struct{}{}
heap.Push(data, &clairvoyantItem{key, c.hits[key]})
}
return stat
}
type clairvoyantItem struct {
key uint64
hits uint64
}
type clairvoyantHeap []*clairvoyantItem
func (h clairvoyantHeap) Len() int { return len(h) }
func (h clairvoyantHeap) Less(i, j int) bool { return h[i].hits < h[j].hits }
func (h clairvoyantHeap) Swap(i, j int) { h[i], h[j] = h[j], h[i] }
func (h *clairvoyantHeap) Push(x interface{}) {
*h = append(*h, x.(*clairvoyantItem))
}
func (h *clairvoyantHeap) Pop() interface{} {
old := *h
n := len(old)
x := old[n-1]
*h = old[0 : n-1]
return x
}
ristretto-0.0.3/ttl.go 0000664 0000000 0000000 00000006327 13700665644 0014726 0 ustar 00root root 0000000 0000000 /*
* Copyright 2020 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
package ristretto
import (
"sync"
"time"
)
var (
// TODO: find the optimal value or make it configurable.
bucketDurationSecs = int64(5)
)
func storageBucket(t time.Time) int64 {
return (t.Unix() / bucketDurationSecs) + 1
}
func cleanupBucket(t time.Time) int64 {
// The bucket to cleanup is always behind the storage bucket by one so that
// no elements in that bucket (which might not have expired yet) are deleted.
return storageBucket(t) - 1
}
// bucket type is a map of key to conflict.
type bucket map[uint64]uint64
// expirationMap is a map of bucket number to the corresponding bucket.
type expirationMap struct {
sync.RWMutex
buckets map[int64]bucket
}
func newExpirationMap() *expirationMap {
return &expirationMap{
buckets: make(map[int64]bucket),
}
}
func (m *expirationMap) add(key, conflict uint64, expiration time.Time) {
if m == nil {
return
}
// Items that don't expire don't need to be in the expiration map.
if expiration.IsZero() {
return
}
bucketNum := storageBucket(expiration)
m.Lock()
defer m.Unlock()
b, ok := m.buckets[bucketNum]
if !ok {
b = make(bucket)
m.buckets[bucketNum] = b
}
b[key] = conflict
}
func (m *expirationMap) update(key, conflict uint64, oldExpTime, newExpTime time.Time) {
if m == nil {
return
}
m.Lock()
defer m.Unlock()
oldBucketNum := storageBucket(oldExpTime)
oldBucket, ok := m.buckets[oldBucketNum]
if ok {
delete(oldBucket, key)
}
newBucketNum := storageBucket(newExpTime)
newBucket, ok := m.buckets[newBucketNum]
if !ok {
newBucket = make(bucket)
m.buckets[newBucketNum] = newBucket
}
newBucket[key] = conflict
}
func (m *expirationMap) del(key uint64, expiration time.Time) {
if m == nil {
return
}
bucketNum := storageBucket(expiration)
m.Lock()
defer m.Unlock()
_, ok := m.buckets[bucketNum]
if !ok {
return
}
delete(m.buckets[bucketNum], key)
}
// cleanup removes all the items in the bucket that was just completed. It deletes
// those items from the store, and calls the onEvict function on those items.
// This function is meant to be called periodically.
func (m *expirationMap) cleanup(store store, policy policy, onEvict onEvictFunc) {
if m == nil {
return
}
m.Lock()
now := time.Now()
bucketNum := cleanupBucket(now)
keys := m.buckets[bucketNum]
delete(m.buckets, bucketNum)
m.Unlock()
for key, conflict := range keys {
// Sanity check. Verify that the store agrees that this key is expired.
if store.Expiration(key).After(now) {
continue
}
cost := policy.Cost(key)
policy.Del(key)
_, value := store.Del(key, conflict)
if onEvict != nil {
onEvict(key, conflict, value, cost)
}
}
}
ristretto-0.0.3/z/ 0000775 0000000 0000000 00000000000 13700665644 0014035 5 ustar 00root root 0000000 0000000 ristretto-0.0.3/z/LICENSE 0000664 0000000 0000000 00000005571 13700665644 0015052 0 ustar 00root root 0000000 0000000 bbloom.go
// The MIT License (MIT)
// Copyright (c) 2014 Andreas Briese, eduToolbox@Bri-C GmbH, Sarstedt
// Permission is hereby granted, free of charge, to any person obtaining a copy of
// this software and associated documentation files (the "Software"), to deal in
// the Software without restriction, including without limitation the rights to
// use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
// the Software, and to permit persons to whom the Software is furnished to do so,
// subject to the following conditions:
// The above copyright notice and this permission notice shall be included in all
// copies or substantial portions of the Software.
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
// FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
// COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
// IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
// CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
rtutil.go
// MIT License
// Copyright (c) 2019 Ewan Chou
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
// in the Software without restriction, including without limitation the rights
// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
// copies of the Software, and to permit persons to whom the Software is
// furnished to do so, subject to the following conditions:
// The above copyright notice and this permission notice shall be included in all
// copies or substantial portions of the Software.
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
// SOFTWARE.
Modifications:
/*
* Copyright 2019 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
ristretto-0.0.3/z/README.md 0000664 0000000 0000000 00000012510 13700665644 0015313 0 ustar 00root root 0000000 0000000 ## bbloom: a bitset Bloom filter for go/golang
===
package implements a fast bloom filter with real 'bitset' and JSONMarshal/JSONUnmarshal to store/reload the Bloom filter.
NOTE: the package uses unsafe.Pointer to set and read the bits from the bitset. If you're uncomfortable with using the unsafe package, please consider using my bloom filter package at github.com/AndreasBriese/bloom
===
changelog 11/2015: new thread safe methods AddTS(), HasTS(), AddIfNotHasTS() following a suggestion from Srdjan Marinovic (github @a-little-srdjan), who used this to code a bloomfilter cache.
This bloom filter was developed to strengthen a website-log database and was tested and optimized for this log-entry mask: "2014/%02i/%02i %02i:%02i:%02i /info.html".
Nonetheless bbloom should work with any other form of entries.
~~Hash function is a modified Berkeley DB sdbm hash (to optimize for smaller strings). sdbm http://www.cse.yorku.ca/~oz/hash.html~~
Found sipHash (SipHash-2-4, a fast short-input PRF created by Jean-Philippe Aumasson and Daniel J. Bernstein.) to be about as fast. sipHash had been ported by Dimtry Chestnyk to Go (github.com/dchest/siphash )
Minimum hashset size is: 512 ([4]uint64; will be set automatically).
###install
```sh
go get github.com/AndreasBriese/bbloom
```
###test
+ change to folder ../bbloom
+ create wordlist in file "words.txt" (you might use `python permut.py`)
+ run 'go test -bench=.' within the folder
```go
go test -bench=.
```
~~If you've installed the GOCONVEY TDD-framework http://goconvey.co/ you can run the tests automatically.~~
using go's testing framework now (have in mind that the op timing is related to 65536 operations of Add, Has, AddIfNotHas respectively)
### usage
after installation add
```go
import (
...
"github.com/AndreasBriese/bbloom"
...
)
```
at your header. In the program use
```go
// create a bloom filter for 65536 items and 1 % wrong-positive ratio
bf := bbloom.New(float64(1<<16), float64(0.01))
// or
// create a bloom filter with 650000 for 65536 items and 7 locs per hash explicitly
// bf = bbloom.New(float64(650000), float64(7))
// or
bf = bbloom.New(650000.0, 7.0)
// add one item
bf.Add([]byte("butter"))
// Number of elements added is exposed now
// Note: ElemNum will not be included in JSON export (for compatability to older version)
nOfElementsInFilter := bf.ElemNum
// check if item is in the filter
isIn := bf.Has([]byte("butter")) // should be true
isNotIn := bf.Has([]byte("Butter")) // should be false
// 'add only if item is new' to the bloomfilter
added := bf.AddIfNotHas([]byte("butter")) // should be false because 'butter' is already in the set
added = bf.AddIfNotHas([]byte("buTTer")) // should be true because 'buTTer' is new
// thread safe versions for concurrent use: AddTS, HasTS, AddIfNotHasTS
// add one item
bf.AddTS([]byte("peanutbutter"))
// check if item is in the filter
isIn = bf.HasTS([]byte("peanutbutter")) // should be true
isNotIn = bf.HasTS([]byte("peanutButter")) // should be false
// 'add only if item is new' to the bloomfilter
added = bf.AddIfNotHasTS([]byte("butter")) // should be false because 'peanutbutter' is already in the set
added = bf.AddIfNotHasTS([]byte("peanutbuTTer")) // should be true because 'penutbuTTer' is new
// convert to JSON ([]byte)
Json := bf.JSONMarshal()
// bloomfilters Mutex is exposed for external un-/locking
// i.e. mutex lock while doing JSON conversion
bf.Mtx.Lock()
Json = bf.JSONMarshal()
bf.Mtx.Unlock()
// restore a bloom filter from storage
bfNew := bbloom.JSONUnmarshal(Json)
isInNew := bfNew.Has([]byte("butter")) // should be true
isNotInNew := bfNew.Has([]byte("Butter")) // should be false
```
to work with the bloom filter.
### why 'fast'?
It's about 3 times faster than William Fitzgeralds bitset bloom filter https://github.com/willf/bloom . And it is about so fast as my []bool set variant for Boom filters (see https://github.com/AndreasBriese/bloom ) but having a 8times smaller memory footprint:
Bloom filter (filter size 524288, 7 hashlocs)
github.com/AndreasBriese/bbloom 'Add' 65536 items (10 repetitions): 6595800 ns (100 ns/op)
github.com/AndreasBriese/bbloom 'Has' 65536 items (10 repetitions): 5986600 ns (91 ns/op)
github.com/AndreasBriese/bloom 'Add' 65536 items (10 repetitions): 6304684 ns (96 ns/op)
github.com/AndreasBriese/bloom 'Has' 65536 items (10 repetitions): 6568663 ns (100 ns/op)
github.com/willf/bloom 'Add' 65536 items (10 repetitions): 24367224 ns (371 ns/op)
github.com/willf/bloom 'Test' 65536 items (10 repetitions): 21881142 ns (333 ns/op)
github.com/dataence/bloom/standard 'Add' 65536 items (10 repetitions): 23041644 ns (351 ns/op)
github.com/dataence/bloom/standard 'Check' 65536 items (10 repetitions): 19153133 ns (292 ns/op)
github.com/cabello/bloom 'Add' 65536 items (10 repetitions): 131921507 ns (2012 ns/op)
github.com/cabello/bloom 'Contains' 65536 items (10 repetitions): 131108962 ns (2000 ns/op)
(on MBPro15 OSX10.8.5 i7 4Core 2.4Ghz)
With 32bit bloom filters (bloom32) using modified sdbm, bloom32 does hashing with only 2 bit shifts, one xor and one substraction per byte. smdb is about as fast as fnv64a but gives less collisions with the dataset (see mask above). bloom.New(float64(10 * 1<<16),float64(7)) populated with 1<<16 random items from the dataset (see above) and tested against the rest results in less than 0.05% collisions.
ristretto-0.0.3/z/bbloom.go 0000664 0000000 0000000 00000013664 13700665644 0015650 0 ustar 00root root 0000000 0000000 // The MIT License (MIT)
// Copyright (c) 2014 Andreas Briese, eduToolbox@Bri-C GmbH, Sarstedt
// Permission is hereby granted, free of charge, to any person obtaining a copy of
// this software and associated documentation files (the "Software"), to deal in
// the Software without restriction, including without limitation the rights to
// use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
// the Software, and to permit persons to whom the Software is furnished to do so,
// subject to the following conditions:
// The above copyright notice and this permission notice shall be included in all
// copies or substantial portions of the Software.
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
// FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
// COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
// IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
// CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
package z
import (
"bytes"
"encoding/json"
"log"
"math"
"unsafe"
)
// helper
var mask = []uint8{1, 2, 4, 8, 16, 32, 64, 128}
func getSize(ui64 uint64) (size uint64, exponent uint64) {
if ui64 < uint64(512) {
ui64 = uint64(512)
}
size = uint64(1)
for size < ui64 {
size <<= 1
exponent++
}
return size, exponent
}
func calcSizeByWrongPositives(numEntries, wrongs float64) (uint64, uint64) {
size := -1 * numEntries * math.Log(wrongs) / math.Pow(float64(0.69314718056), 2)
locs := math.Ceil(float64(0.69314718056) * size / numEntries)
return uint64(size), uint64(locs)
}
// NewBloomFilter returns a new bloomfilter.
func NewBloomFilter(params ...float64) (bloomfilter *Bloom) {
var entries, locs uint64
if len(params) == 2 {
if params[1] < 1 {
entries, locs = calcSizeByWrongPositives(params[0], params[1])
} else {
entries, locs = uint64(params[0]), uint64(params[1])
}
} else {
log.Fatal("usage: New(float64(number_of_entries), float64(number_of_hashlocations))" +
" i.e. New(float64(1000), float64(3)) or New(float64(number_of_entries)," +
" float64(number_of_hashlocations)) i.e. New(float64(1000), float64(0.03))")
}
size, exponent := getSize(entries)
bloomfilter = &Bloom{
sizeExp: exponent,
size: size - 1,
setLocs: locs,
shift: 64 - exponent,
}
bloomfilter.Size(size)
return bloomfilter
}
// Bloom filter
type Bloom struct {
bitset []uint64
ElemNum uint64
sizeExp uint64
size uint64
setLocs uint64
shift uint64
}
// <--- http://www.cse.yorku.ca/~oz/hash.html
// modified Berkeley DB Hash (32bit)
// hash is casted to l, h = 16bit fragments
// func (bl Bloom) absdbm(b *[]byte) (l, h uint64) {
// hash := uint64(len(*b))
// for _, c := range *b {
// hash = uint64(c) + (hash << 6) + (hash << bl.sizeExp) - hash
// }
// h = hash >> bl.shift
// l = hash << bl.shift >> bl.shift
// return l, h
// }
// Add adds hash of a key to the bloomfilter.
func (bl *Bloom) Add(hash uint64) {
h := hash >> bl.shift
l := hash << bl.shift >> bl.shift
for i := uint64(0); i < bl.setLocs; i++ {
bl.Set((h + i*l) & bl.size)
bl.ElemNum++
}
}
// Has checks if bit(s) for entry hash is/are set,
// returns true if the hash was added to the Bloom Filter.
func (bl Bloom) Has(hash uint64) bool {
h := hash >> bl.shift
l := hash << bl.shift >> bl.shift
for i := uint64(0); i < bl.setLocs; i++ {
if !bl.IsSet((h + i*l) & bl.size) {
return false
}
}
return true
}
// AddIfNotHas only Adds hash, if it's not present in the bloomfilter.
// Returns true if hash was added.
// Returns false if hash was already registered in the bloomfilter.
func (bl *Bloom) AddIfNotHas(hash uint64) bool {
if bl.Has(hash) {
return false
}
bl.Add(hash)
return true
}
// Size makes Bloom filter with as bitset of size sz.
func (bl *Bloom) Size(sz uint64) {
bl.bitset = make([]uint64, sz>>6)
}
// Clear resets the Bloom filter.
func (bl *Bloom) Clear() {
for i := range bl.bitset {
bl.bitset[i] = 0
}
}
// Set sets the bit[idx] of bitset.
func (bl *Bloom) Set(idx uint64) {
ptr := unsafe.Pointer(uintptr(unsafe.Pointer(&bl.bitset[idx>>6])) + uintptr((idx%64)>>3))
*(*uint8)(ptr) |= mask[idx%8]
}
// IsSet checks if bit[idx] of bitset is set, returns true/false.
func (bl *Bloom) IsSet(idx uint64) bool {
ptr := unsafe.Pointer(uintptr(unsafe.Pointer(&bl.bitset[idx>>6])) + uintptr((idx%64)>>3))
r := ((*(*uint8)(ptr)) >> (idx % 8)) & 1
return r == 1
}
// bloomJSONImExport
// Im/Export structure used by JSONMarshal / JSONUnmarshal
type bloomJSONImExport struct {
FilterSet []byte
SetLocs uint64
}
// NewWithBoolset takes a []byte slice and number of locs per entry,
// returns the bloomfilter with a bitset populated according to the input []byte.
func newWithBoolset(bs *[]byte, locs uint64) *Bloom {
bloomfilter := NewBloomFilter(float64(len(*bs)<<3), float64(locs))
for i, b := range *bs {
*(*uint8)(unsafe.Pointer(uintptr(unsafe.Pointer(&bloomfilter.bitset[0])) + uintptr(i))) = b
}
return bloomfilter
}
// JSONUnmarshal takes JSON-Object (type bloomJSONImExport) as []bytes
// returns bloom32 / bloom64 object.
func JSONUnmarshal(dbData []byte) (*Bloom, error) {
bloomImEx := bloomJSONImExport{}
if err := json.Unmarshal(dbData, &bloomImEx); err != nil {
return nil, err
}
buf := bytes.NewBuffer(bloomImEx.FilterSet)
bs := buf.Bytes()
bf := newWithBoolset(&bs, bloomImEx.SetLocs)
return bf, nil
}
// JSONMarshal returns JSON-object (type bloomJSONImExport) as []byte.
func (bl Bloom) JSONMarshal() []byte {
bloomImEx := bloomJSONImExport{}
bloomImEx.SetLocs = bl.setLocs
bloomImEx.FilterSet = make([]byte, len(bl.bitset)<<3)
for i := range bloomImEx.FilterSet {
bloomImEx.FilterSet[i] = *(*byte)(unsafe.Pointer(uintptr(unsafe.Pointer(&bl.bitset[0])) +
uintptr(i)))
}
data, err := json.Marshal(bloomImEx)
if err != nil {
log.Fatal("json.Marshal failed: ", err)
}
return data
}
ristretto-0.0.3/z/bbloom_test.go 0000664 0000000 0000000 00000004152 13700665644 0016677 0 ustar 00root root 0000000 0000000 package z
import (
"crypto/rand"
"fmt"
"testing"
"github.com/stretchr/testify/require"
)
var (
wordlist1 [][]byte
n = 1 << 16
bf *Bloom
)
func TestMain(m *testing.M) {
wordlist1 = make([][]byte, n)
for i := range wordlist1 {
b := make([]byte, 32)
rand.Read(b)
wordlist1[i] = b
}
fmt.Println("\n###############\nbbloom_test.go")
fmt.Print("Benchmarks relate to 2**16 OP. --> output/65536 op/ns\n###############\n\n")
m.Run()
}
func TestM_NumberOfWrongs(t *testing.T) {
bf = NewBloomFilter(float64(n*10), float64(7))
cnt := 0
for i := range wordlist1 {
hash := MemHash(wordlist1[i])
if !bf.AddIfNotHas(hash) {
cnt++
}
}
fmt.Printf("Bloomfilter New(7* 2**16, 7) (-> size=%v bit): \n Check for 'false positives': %v wrong positive 'Has' results on 2**16 entries => %v %%\n", len(bf.bitset)<<6, cnt, float64(cnt)/float64(n))
}
func TestM_JSON(t *testing.T) {
const shallBe = int(1 << 16)
bf = NewBloomFilter(float64(n*10), float64(7))
cnt := 0
for i := range wordlist1 {
hash := MemHash(wordlist1[i])
if !bf.AddIfNotHas(hash) {
cnt++
}
}
Json := bf.JSONMarshal()
// create new bloomfilter from bloomfilter's JSON representation
bf2, err := JSONUnmarshal(Json)
require.NoError(t, err)
cnt2 := 0
for i := range wordlist1 {
hash := MemHash(wordlist1[i])
if !bf2.AddIfNotHas(hash) {
cnt2++
}
}
require.Equal(t, shallBe, cnt2)
}
func BenchmarkM_New(b *testing.B) {
for r := 0; r < b.N; r++ {
_ = NewBloomFilter(float64(n*10), float64(7))
}
}
func BenchmarkM_Clear(b *testing.B) {
bf = NewBloomFilter(float64(n*10), float64(7))
for i := range wordlist1 {
hash := MemHash(wordlist1[i])
bf.Add(hash)
}
b.ResetTimer()
for r := 0; r < b.N; r++ {
bf.Clear()
}
}
func BenchmarkM_Add(b *testing.B) {
bf = NewBloomFilter(float64(n*10), float64(7))
b.ResetTimer()
for r := 0; r < b.N; r++ {
for i := range wordlist1 {
hash := MemHash(wordlist1[i])
bf.Add(hash)
}
}
}
func BenchmarkM_Has(b *testing.B) {
b.ResetTimer()
for r := 0; r < b.N; r++ {
for i := range wordlist1 {
hash := MemHash(wordlist1[i])
bf.Has(hash)
}
}
}
ristretto-0.0.3/z/rtutil.go 0000664 0000000 0000000 00000004656 13700665644 0015722 0 ustar 00root root 0000000 0000000 // MIT License
// Copyright (c) 2019 Ewan Chou
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
// in the Software without restriction, including without limitation the rights
// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
// copies of the Software, and to permit persons to whom the Software is
// furnished to do so, subject to the following conditions:
// The above copyright notice and this permission notice shall be included in all
// copies or substantial portions of the Software.
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
// SOFTWARE.
package z
import (
"unsafe"
)
// NanoTime returns the current time in nanoseconds from a monotonic clock.
//go:linkname NanoTime runtime.nanotime
func NanoTime() int64
// CPUTicks is a faster alternative to NanoTime to measure time duration.
//go:linkname CPUTicks runtime.cputicks
func CPUTicks() int64
type stringStruct struct {
str unsafe.Pointer
len int
}
//go:noescape
//go:linkname memhash runtime.memhash
func memhash(p unsafe.Pointer, h, s uintptr) uintptr
// MemHash is the hash function used by go map, it utilizes available hardware instructions(behaves
// as aeshash if aes instruction is available).
// NOTE: The hash seed changes for every process. So, this cannot be used as a persistent hash.
func MemHash(data []byte) uint64 {
ss := (*stringStruct)(unsafe.Pointer(&data))
return uint64(memhash(ss.str, 0, uintptr(ss.len)))
}
// MemHashString is the hash function used by go map, it utilizes available hardware instructions
// (behaves as aeshash if aes instruction is available).
// NOTE: The hash seed changes for every process. So, this cannot be used as a persistent hash.
func MemHashString(str string) uint64 {
ss := (*stringStruct)(unsafe.Pointer(&str))
return uint64(memhash(ss.str, 0, uintptr(ss.len)))
}
// FastRand is a fast thread local random function.
//go:linkname FastRand runtime.fastrand
func FastRand() uint32
ristretto-0.0.3/z/rtutil.s 0000664 0000000 0000000 00000000000 13700665644 0015532 0 ustar 00root root 0000000 0000000 ristretto-0.0.3/z/rtutil_test.go 0000664 0000000 0000000 00000012226 13700665644 0016751 0 ustar 00root root 0000000 0000000 package z
import (
"hash/fnv"
"math/rand"
"sync/atomic"
"testing"
"time"
"github.com/dgryski/go-farm"
)
func BenchmarkMemHash(b *testing.B) {
buf := make([]byte, 64)
rand.Read(buf)
b.ReportAllocs()
b.ResetTimer()
for i := 0; i < b.N; i++ {
_ = MemHash(buf)
}
b.SetBytes(int64(len(buf)))
}
func BenchmarkMemHashString(b *testing.B) {
s := "Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua."
b.ReportAllocs()
b.ResetTimer()
for i := 0; i < b.N; i++ {
_ = MemHashString(s)
}
b.SetBytes(int64(len(s)))
}
func BenchmarkSip(b *testing.B) {
buf := make([]byte, 64)
rand.Read(buf)
for i := 0; i < b.N; i++ {
SipHash(buf)
}
}
func BenchmarkFarm(b *testing.B) {
buf := make([]byte, 64)
rand.Read(buf)
for i := 0; i < b.N; i++ {
farm.Fingerprint64(buf)
}
}
func BenchmarkFnv(b *testing.B) {
buf := make([]byte, 64)
rand.Read(buf)
f := fnv.New64a()
for i := 0; i < b.N; i++ {
f.Write(buf)
f.Sum64()
f.Reset()
}
}
func SipHash(p []byte) (l, h uint64) {
// Initialization.
v0 := uint64(8317987320269560794) // k0 ^ 0x736f6d6570736575
v1 := uint64(7237128889637516672) // k1 ^ 0x646f72616e646f6d
v2 := uint64(7816392314733513934) // k0 ^ 0x6c7967656e657261
v3 := uint64(8387220255325274014) // k1 ^ 0x7465646279746573
t := uint64(len(p)) << 56
// Compression.
for len(p) >= 8 {
m := uint64(p[0]) | uint64(p[1])<<8 | uint64(p[2])<<16 | uint64(p[3])<<24 |
uint64(p[4])<<32 | uint64(p[5])<<40 | uint64(p[6])<<48 | uint64(p[7])<<56
v3 ^= m
// Round 1.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
// Round 2.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
v0 ^= m
p = p[8:]
}
// Compress last block.
switch len(p) {
case 7:
t |= uint64(p[6]) << 48
fallthrough
case 6:
t |= uint64(p[5]) << 40
fallthrough
case 5:
t |= uint64(p[4]) << 32
fallthrough
case 4:
t |= uint64(p[3]) << 24
fallthrough
case 3:
t |= uint64(p[2]) << 16
fallthrough
case 2:
t |= uint64(p[1]) << 8
fallthrough
case 1:
t |= uint64(p[0])
}
v3 ^= t
// Round 1.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
// Round 2.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
v0 ^= t
// Finalization.
v2 ^= 0xff
// Round 1.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
// Round 2.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
// Round 3.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
// Round 4.
v0 += v1
v1 = v1<<13 | v1>>51
v1 ^= v0
v0 = v0<<32 | v0>>32
v2 += v3
v3 = v3<<16 | v3>>48
v3 ^= v2
v0 += v3
v3 = v3<<21 | v3>>43
v3 ^= v0
v2 += v1
v1 = v1<<17 | v1>>47
v1 ^= v2
v2 = v2<<32 | v2>>32
// return v0 ^ v1 ^ v2 ^ v3
hash := v0 ^ v1 ^ v2 ^ v3
h = hash >> 1
l = hash << 1 >> 1
return l, h
}
func BenchmarkNanoTime(b *testing.B) {
for i := 0; i < b.N; i++ {
NanoTime()
}
}
func BenchmarkCPUTicks(b *testing.B) {
for i := 0; i < b.N; i++ {
CPUTicks()
}
}
// goos: linux
// goarch: amd64
// pkg: github.com/dgraph-io/ristretto/z
// BenchmarkFastRand-16 1000000000 0.292 ns/op
// BenchmarkRandSource-16 1000000000 0.747 ns/op
// BenchmarkRandGlobal-16 6822332 176 ns/op
// BenchmarkRandAtomic-16 77950322 15.4 ns/op
// PASS
// ok github.com/dgraph-io/ristretto/z 4.808s
func benchmarkRand(b *testing.B, fab func() func() uint32) {
b.RunParallel(func(pb *testing.PB) {
gen := fab()
for pb.Next() {
gen()
}
})
}
func BenchmarkFastRand(b *testing.B) {
benchmarkRand(b, func() func() uint32 {
return FastRand
})
}
func BenchmarkRandSource(b *testing.B) {
benchmarkRand(b, func() func() uint32 {
s := rand.New(rand.NewSource(time.Now().Unix()))
return func() uint32 { return s.Uint32() }
})
}
func BenchmarkRandGlobal(b *testing.B) {
benchmarkRand(b, func() func() uint32 {
return func() uint32 { return rand.Uint32() }
})
}
func BenchmarkRandAtomic(b *testing.B) {
var x uint32
benchmarkRand(b, func() func() uint32 {
return func() uint32 { return uint32(atomic.AddUint32(&x, 1)) }
})
}
ristretto-0.0.3/z/z.go 0000664 0000000 0000000 00000003101 13700665644 0014630 0 ustar 00root root 0000000 0000000 /*
* Copyright 2019 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
package z
import (
"github.com/cespare/xxhash"
)
// TODO: Figure out a way to re-use memhash for the second uint64 hash, we
// already know that appending bytes isn't reliable for generating a
// second hash (see Ristretto PR #88).
//
// We also know that while the Go runtime has a runtime memhash128
// function, it's not possible to use it to generate [2]uint64 or
// anything resembling a 128bit hash, even though that's exactly what
// we need in this situation.
func KeyToHash(key interface{}) (uint64, uint64) {
if key == nil {
return 0, 0
}
switch k := key.(type) {
case uint64:
return k, 0
case string:
return MemHashString(k), xxhash.Sum64String(k)
case []byte:
return MemHash(k), xxhash.Sum64(k)
case byte:
return uint64(k), 0
case int:
return uint64(k), 0
case int32:
return uint64(k), 0
case uint32:
return uint64(k), 0
case int64:
return uint64(k), 0
default:
panic("Key type not supported")
}
}
ristretto-0.0.3/z/z_test.go 0000664 0000000 0000000 00000002772 13700665644 0015704 0 ustar 00root root 0000000 0000000 /*
* Copyright 2019 Dgraph Labs, Inc. and Contributors
*
* 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.
*/
package z
import (
"math"
"testing"
"github.com/stretchr/testify/require"
)
func verifyHashProduct(t *testing.T, wantKey, wantConflict, key, conflict uint64) {
require.Equal(t, wantKey, key)
require.Equal(t, wantConflict, conflict)
}
func TestKeyToHash(t *testing.T) {
var key uint64
var conflict uint64
key, conflict = KeyToHash(uint64(1))
verifyHashProduct(t, 1, 0, key, conflict)
key, conflict = KeyToHash(1)
verifyHashProduct(t, 1, 0, key, conflict)
key, conflict = KeyToHash(int32(2))
verifyHashProduct(t, 2, 0, key, conflict)
key, conflict = KeyToHash(int32(-2))
verifyHashProduct(t, math.MaxUint64-1, 0, key, conflict)
key, conflict = KeyToHash(int64(-2))
verifyHashProduct(t, math.MaxUint64-1, 0, key, conflict)
key, conflict = KeyToHash(uint32(3))
verifyHashProduct(t, 3, 0, key, conflict)
key, conflict = KeyToHash(int64(3))
verifyHashProduct(t, 3, 0, key, conflict)
}
.svg)
.svg)
.svg)
.svg)
.svg)
.svg)