Share Archiver’s vision is to achieve faster downloads for all by improving online file sharing through small archive size with fast decompression.
Since files shared online are only once compressed and uploaded by the uploader, but downloaded and decompressed numerous times. The focus was therefore on achieving smallest archive size, which leads to faster downloads because less data transmitted overall, while still having fast decompression speeds to still be practical for the downloaders of the uploaded share archive. It's approach is to only use strong asymmetric compressors with fast decompression speeds, and benchmarking each file for smallest archive size for compressor selection.
The current version 0.1 can make use of these great compressors: Zlib, Balz, LZMA (XZ), GZip, Zstandard, PPmd, LZ4, Brotli, LZHAM, ZOPFLI, Lzturbo, GLZA, Balz, BSC, CSArc, NNCP, ZPAQ Shoutout to the developers of these great compressors, they did such just an amazing job and invested a lot of time for the community to profit thereof and make use of it.
There are two folders, v0.1 which includes the first version with reasonings in the notes folder, and executable files created with pyinstaller on Ubuntu 18 and Ubuntu 20. And the v0.2 folder which is work in progress and contains a tryout for a gui with pysimplegui but also a version that would make use of precomp, srep and also compressor-grouping of files with a recompression mechanism as to enable the specific compressor to take advantage of inter-file redundancies.
This was a small side hobby project of mine which is not actively in development anymore so i archived this repo
This application will compress an input folder and decompress a share archive to an output folder. For the argument inputs, absolute paths as well as relatives paths are accepted. There is a help menu integrated, which will show by running share -h
As a simple example, to compress, run
share compress /path/to/input/folder /path/to/output/filename
So the first argument is a folder path, the second argument is a path with the filename. It will create a .share file with the provided filename, therefore ./silesia will generate a silesia.share file
To decompress, run
share decompress /path/to/input/archive.share /path/to/output/folder
The first argument is the share archive (dont forget the .share) and the second is the extraction path, it will generate a folder with the archive filename. So "share decompress ../input/silesia.share ../output" will generate a ../output/silesia folder with the extracted files in it.
It this folder to be generated already exists (a subfolder in the output path provided with the same name as the archive), share archiver will quit and show a prompting to resolve this conflict manually. This is simply because I did not want share to automatically overwrite any already extracted files - share assumes that this specific archive has already been extracted to this output folder path previously if this directory is already present. To resolve it, create a intermediate output folder and include it in the output path, or remove said folder with its content.
There are also profiles integrated. Using those limits or extends the number of compressors used for one compression.
Fast (0)
Default (1)
Strong (2)
Max (3)
which can be activated by using the levels (some examples):
share -lvl 0 compress /home/phhofm/Dokumente/silesia /home/phhofm/Dokumente/output/silesia_fast
share -v -lvl 2 -ms 1 compress ../silesia ../output/silesia_strong
share decompress ../output/silesia_strong.share ../extract
share -v -lvl 3 -ms 1 compress ../silesia ../output/silesia_max
These profiles will mainly affect compression speed and compression ratio, with the exception of the Max profile:
The Max profile shall not be used for online file sharing, since it will make use of compressors with long decompression times, its goal is only small archive size, and can be used for personal interests.
Info for the arguments should also be availabe by running share -h
Thanks for trying out Share Archiver :D
Philip Hofmann
The default profile has been used for all of these
11735040 maximumcompressioncorpus.share
12177048 maximumcompressioncorpus.7z
12403504 maximumcompressioncorpus.tar.xz
12412975 maximumcompressioncorpus.tar.lzma
12412988 maximumcompressioncorpus.tar.lz
12523696 maximumcompressioncorpus.tar.7z
12654720 maximumcompressioncorpus.exe
13610427 maximumcompressioncorpus.tar.bz2
14405261 maximumcompressioncorpus.zip
14405261 maximumcompressioncorpus.cbz
15095465 maximumcompressioncorpus.tar.gz
15102263 maximumcompressioncorpus.war
15102263 maximumcompressioncorpus.jar
15102263 maximumcompressioncorpus.ear
15102263 maximumcompressioncorpus.crx
15102263 maximumcompressioncorpus.apk
15311263 maximumcompressioncorpus.tar.zst
20907094 maximumcompressioncorpus.tar.lz4
21334955 maximumcompressioncorpus.tar.lzo
21427739 maximumcompressioncorpus.tar.Z
53135336 maximumcompressioncorpus.ar
53136148 maximumcompressioncorpus.cpio
53144064 maximumcompressioncorpus.tar
53522432 maximumcompressioncorpus.iso
Times Share Archiver:
Compression, Fast:
Original Size: 53134726
Archive Size: 11755520
Compression Ratio: 4.5199809110953835
Program Execution Time (PET): 11.13 seconds
Decompression:
Program Execution Time (PET): 0.36 seconds
Compression, Default:
Original Size: 53134726
Archive Size: 11735040
Compression Ratio: 4.527869184936736
Program Execution Time (PET): 156.91 seconds
Decompression:
Program Execution Time (PET): 0.34 seconds
Compression, Strong:
Original Size: 53134726
Archive Size: 11376640
Compression Ratio: 4.670511328476597
Program Execution Time (PET): 254.39 seconds
Decompression:
Program Execution Time (PET): 0.31 seconds
Compression, Max:
Original Size: 53134726
Archive Size: 11100160
Compression Ratio: 4.786843252709871
Program Execution Time (PET): 16464.23 seconds
Decompression:
Program Execution Time (PET): 6479.34 seconds
You can find a generated share archive example in v0.1/test
The contents of the maximumcompressioncorpus_strong.share looks like this (we can see the different compressors that were selected for each file:
There is also an info file included which gets generated, it includes infos on the system and the mode used. The idea behind this was to be able to handle compatibility in the future. Meaning version missmatched when decompressing, or operating system missmatches when decompressing, like a windows user getting the message that decompressions needs to happen on a linux system since compression was done there, or wise versa, or when the future wine version was used for compression which would feature more compressors, or to simply display the note that maximum setting had been used by the compressing user and that decompression therefore will take a very long time.
This is what the generated info file in the archive looks like:
