The quicklz package

[Tags: gpl, library]

This package provides a high level binding to the QuickLZ ( library for ByteStrings, under the GPLv2 license. QuickLZ is fast and compresses very well.

This package includes the QuickLZ 1.5.0 source code, with compression level 1 and streaming currently disabled.

The versioning scheme for this package is unusual. QuickLZ trys to be as fast as possible, and will break backwards compatibility to achieve it in newer versions. This versioning scheme reflects the QuickLZ versioning scheme: the first three digits of version w.x.y.z are the quicklz version, with the z component (and any further needed ones) being updates to this package, not quicklz.

[Skip to ReadMe]


Change logNone available
Dependenciesbase (>=3 && <5), bytestring (==0.9.*) [details]
AuthorAustin Seipp
Home page
Bug tracker
Source repositoryhead: git clone
UploadedMon Mar 14 07:39:19 UTC 2011 by AustinSeipp
Downloads1681 total (57 in last 30 days)
0 []
StatusDocs uploaded by user
Build status unknown [no reports yet]




memory-safeenable quickLZ memory safety checksDisabledAutomatic

Use -f <flag> to enable a flag, or -f -<flag> to disable that flag. More info


Maintainers' corner

For package maintainers and hackage trustees

Readme for quicklz-

hs-quicklz: haskell bindings to quicklz

This library implements haskell bindings to QuickLZ, a fast compression library. The interface is very simple: compress :: ByteString -> ByteString decompress :: ByteString -> ByteString decompress' :: ByteString -> ByteString -- overlapping decompression As the name states, QuickLZ is fast - very fast at both compression and decompression. Per its own benchmarks, at compression level 1, QuickLZ 1.5.0 has a compression speed of 308Mbyte/s, and a decompression speed of 358Mbyte/s (their benchmarks, Core i7 920.) This package has some benchmarks and tests associated - I would like to expand the benchmarks to include bigger data sets in the future. There is currently only an interface for strict bytestrings. I would like to implement the streaming mode functionality sometime in the future, perhaps tying it to John Millikin's enumerator package.