https://github.com/shevek/parallelgzip

ParallelGZIPOutputStream and ParallelGZIPInputStream
https://github.com/shevek/parallelgzip

Last synced: 10 days ago
JSON representation

ParallelGZIPOutputStream and ParallelGZIPInputStream

• Host: GitHub
• URL: https://github.com/shevek/parallelgzip
• Owner: shevek
• License: apache-2.0
• Created: 2014-09-20T00:05:04.000Z (about 10 years ago)
• Default Branch: master
• Last Pushed: 2023-02-24T06:55:47.000Z (over 1 year ago)
• Last Synced: 2024-08-01T12:18:47.475Z (3 months ago)
• Language: Java
• Homepage:
• Size: 246 KB
• Stars: 57
• Watchers: 8
• Forks: 8
• Open Issues: 5
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
Parallel GZIPOutputStream and GZIPInputStream

=============================================

This library contains a parallelized GZIP implementation which is a

high performance drop-in replacement for the standard java.util.zip

classes. It is a pure Java equivalent of the pigz parallel compresssor.

The performance of ParallelGZIPOutputStream is excellent: it scales

linearly with the number of cores, and spends 95% of all thread time

in the native compression routines on 24 and c32-core Xeon systems.

ParallelGZIPOutputStream has exactly the same memory contract as

GZIPOutputStream: it extends FilterOutputStream. If a single thread

writes data into the compressor, it will write compressed data to the

underlying output stream on the same thread without any externally

visible synchronization or flush calls. The user never needs to know

or change anything as a result of the parallelism.

Currently, ParallelGZIPInputStream is a subclass of the standard

GZIPInputStream, but parallelism may be added in future.

API Documentation

=================

The [JavaDoc API](http://shevek.github.io/parallelgzip/docs/javadoc/)

is available.

References

==========

* http://zlib.net/pigz/pigz.pdf

* http://www.gzip.org/zlib/rfc-gzip.html

Credits

=======

I needed this at work, but I was inspired to

publish it by Paul Eggert's CS131 coursework at

http://www.cs.ucla.edu/classes/fall11/cs131/hw/hw3.html - I wonder

what grade I will get.