lz1-compression-stream

lZ77-nodejs-streams

Compression and decompression streams for the Lempel-Ziv 77 compression algorithm. The compression stream convert streams of bytes into a stream of JSON objects that represent the data in their compressed form. The decompression stream produces the input byte stream from the JSON objects. This is designed as a dependency to another project (that doesn't exist yet) which will be responsible for serialization/de-serialization in a space efficient way. The prototype can be found here https://github.com/spacekitcat/libz77

Progress

I've implemented the compression stream and it uses Tiny Toatie Cache as it's storage engine. In comparison with the original prototype, this version seems to be several times faster.

sample.js

sample.js runs the compressor against any file path you give as the first argument. The Saved number is a tally of the number of bytes above 4 that the packet expands to, you would need about 4 bytes in overhead to serialize the packet to a file.

lz77-nodejs-streams ‹master*› % node sample.js resources/512m-test.txt
PROGRESS   ▓▓▓▓▓▓▓▓▓▓▓▓▓░░░░░░░░░░░░░░░░░░░  41%   ETA 64s   SAVING 209.98 MB   SPEED 5.61 MB/s   PROGRESS 210.8/512 MB
 
    ... 'N' seconds pass ...
 
PROGRESS   ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓  100%   ETA 1s   SAVING 510 MB   SPEED 3.5 MB/s   PROGRESS 512/512 MB
 
I compressed the devil outta resources/512m-test.txt
 
             Input size: 536870912
     Theoretical output: 2099778

Unit tests

The unit tests use Jest and the Yarn command below runs them.

/lz77-nodejs-streams ‹master*› % yarn test