Snappy(Google家用的快速压缩算法,以前的Zippy)

原创文章,转载请注明: 转载自Erlang非业余研究

本文链接地址: Snappy(Google家用的快速压缩算法,以前的Zippy)

刚刚雕梁同学告诉我google刚刚开源了他自己家用的快速压缩算法,AKA Zippy, 看来下貌似不错。

项目主页: http://code.google.com/p/snappy/

Snappy is a compression/decompression library. It does not aim for maximum compression, or compatibility with any other compression library; instead, it aims for very high speeds and reasonable compression. For instance, compared to the fastest mode of zlib, Snappy is an order of magnitude faster for most inputs, but the resulting compressed files are anywhere from 20% to 100% bigger. On a single core of a Core i7 processor in 64-bit mode, Snappy compresses at about 250 MB/sec or more and decompresses at about 500 MB/sec or more.

Snappy is widely used inside Google, in everything from BigTable and MapReduce to our internal RPC systems. (Snappy has previously been referred to as “Zippy” in some presentations and the likes.)

$ svn checkout http://snappy.googlecode.com/svn/trunk/ snappy && cd snappy
$ ./configure && make
$ ./snappy_unittest
Running microbenchmarks.
WARNING: Compiled with assertions enabled, will be slow.
Benchmark            Time(ns)    CPU(ns) Iterations
---------------------------------------------------
BM_UFlat/0             223919     223930        893 436.1MB/s  html
BM_UFlat/1            2161619    2162717         92 309.6MB/s  urls
BM_UFlat/2              54322      54137       3269 2.2GB/s  jpg
BM_UFlat/3             102384     102334       1915 879.1MB/s  pdf
BM_UFlat/4             904663     899863        220 434.1MB/s  html4
BM_UFlat/5              86095      86380       2315 271.6MB/s  cp
BM_UFlat/6              40602      40607       5023 261.9MB/s  c
BM_UFlat/7              13244      13198      15151 268.9MB/s  lsp
BM_UFlat/8            3956500    3939400         50 249.3MB/s  xls
BM_UFlat/9             762839     762337        261 190.3MB/s  txt1
BM_UFlat/10            653573     655639        305 182.1MB/s  txt2
BM_UFlat/11           1997280    1999700        100 203.5MB/s  txt3
BM_UFlat/12           2717739    2711917         73 169.5MB/s  txt4
BM_UFlat/13           1051349    1047460        189 467.3MB/s  bin
BM_UFlat/14            147776     147252       1358 247.7MB/s  sum
BM_UFlat/15             16914      16913      11173 238.3MB/s  man
BM_UFlat/16            239393     238573        834 399.7MB/s  pb
BM_UFlat/17            791722     793527        252 221.5MB/s  gaviota
BM_UValidate/0         108440     108560       1842 899.6MB/s  html
BM_UValidate/1        1187755    1190291        168 562.5MB/s  urls
BM_UValidate/2            657        658     289855 179.5GB/s  jpg
BM_UValidate/3          37172      37224       5372 2.4GB/s  pdf
BM_UValidate/4         438133     438530        456 890.8MB/s  html4
BM_ZFlat/0             918760     916912        217 106.5MB/s  html (23.57 %)
BM_ZFlat/1            9691700    9698550         20 69.0MB/s  urls (50.89 %)
BM_ZFlat/2             169311     168772       1173 717.4MB/s  jpg (99.88 %)
BM_ZFlat/3             417202     416255        478 216.1MB/s  pdf (82.13 %)
BM_ZFlat/4            3691074    3703129         54 105.5MB/s  html4 (23.55 %)
BM_ZFlat/5             356185     357089        560 65.7MB/s  cp (48.12 %)
BM_ZFlat/6             155500     155339       1268 68.5MB/s  c (42.40 %)
BM_ZFlat/7              52200      52129       3836 68.1MB/s  lsp (48.37 %)
BM_ZFlat/8           12006875   11998187         16 81.8MB/s  xls (41.34 %)
BM_ZFlat/9            2785472    2777361         72 52.2MB/s  txt1 (59.81 %)
BM_ZFlat/10           2381506    2385180         83 50.1MB/s  txt2 (64.07 %)
BM_ZFlat/11           7528884    7498846         26 54.3MB/s  txt3 (57.11 %)
BM_ZFlat/12           9649700    9598500         20 47.9MB/s  txt4 (68.35 %)
BM_ZFlat/13           3784346    3787884         52 129.2MB/s  bin (18.21 %)
BM_ZFlat/14            564691     566487        353 64.4MB/s  sum (51.88 %)
BM_ZFlat/15             67413      67366       2909 59.8MB/s  man (59.36 %)
BM_ZFlat/16            903805     904841        221 105.4MB/s  pb (23.49 %)
BM_ZFlat/17           2509924    2493291         79 70.5MB/s  gaviota (38.27 %)

Running correctness tests.
All tests passed.

粗粗跑了下,压缩速度确实很吓人,到时候好好测试下。有兴趣的同学可以对比下这里

玩得开心!

Post Footer automatically generated by wp-posturl plugin for wordpress.

你可能感兴趣的:(算法,linux,压缩,源码分析,snappy)