Zippy

nimble install zippy

Zippy is an in-progress and experimental implementation of DEFLATE and ZLIB.

The goal of this library is to be a dependency-free Nim implementation that is as small and straightforward as possible while still focusing on performance.

To ensure Zippy is compatible with other Zip implementations, tests/validate.nim can be run. This script verifies that data compressed by Zippy can be uncompressed by other implementations (and that other implementations can uncompress data compressed by Zippy).

Zippy works well using Nim's relatively new --gc:arc and --gc:orc as well as the default garbage collector. This library also works using both nim c and nim cpp, in addition to --cc:vcc on Windows.

I have also verified that Zippy builds with --experimental:strictFuncs on Nim 1.4.0.

This library is an active project and not ready for production use. Currently only uncompress (inflating) has been implemented.

Performance

Benchmarks can be run comparing different Zip implementations. My benchmarking shows this library performs very well but it is not as fast as zlib itself (not a surprise). Check the performance yourself by running tests/benchmark.nim.

nim c --gc:arc -d:release -r .\tests\benchmark.nim (1000 uncompresses, lower time is better)

https://github.com/guzba/zippy results:

File	Time
randtest3.z	0.0519s
rfctest3.z	0.3295s
alice29.txt.z	1.4704s
urls.10K.z	7.3240s
fixed.z	6.6955s

https://github.com/treeform/miniz results:

File	Time
randtest3.z	0.5803s
rfctest3.z	0.5801s
alice29.txt.z	3.3442s
urls.10K.z	16.1209s
fixed.z	19.8003s

https://github.com/nim-lang/nimPNG results:

File	Time
randtest3.z	0.1648s
rfctest3.z	0.5760s
alice29.txt.z	2.2471s
urls.10K.z	11.0067s
fixed.z	10.0772s

https://github.com/nim-lang/zip results: (Requires zlib1.dll)

File	Time
randtest3.z	0.0061s
rfctest3.z	0.1285s
alice29.txt.z	0.4918s
urls.10K.z	2.2510s
fixed.z	2.1033s

Testing

nimble test

To prevent Zippy from causing a crash or otherwise misbehaving on bad input data, a fuzzer has been run against it. You can do run the fuzzer any time by running nim c -r tests/fuzz.nim

Credits

This implementation has been greatly assisted by zlib-inflate-simple which is by far the smallest and most readable implementation I've found.

API: zippy

import zippy

func uncompress

Uncompresses src into dst. This resizes dst as needed and starts writing at dst index 0.

func uncompress(src: seq[uint8]; dst: var seq[uint8]) {.raises: [ZippyError], tags: [].}

func uncompress

Uncompresses src and returns the uncompressed data seq.

func uncompress(src: seq[uint8]): seq[uint8] {.inline, raises: [ZippyError], tags: [].}

template uncompress

Helper for when preferring to work with strings.

template uncompress(src: string): string