tar.sz support

[spiderman666]

Hi all. I have started to come across snappy compressed tarballs (tar.sz). Keka doesn't seem to know the file type and spits an error when I try to decompress it. Will there be support for this compression type soon and can someone please point me in the direction of how I can decompress this file on a mac? Many thanks.

[gingerbeardman]

Do you have a link to an example file or two that exist in the wild online?

It seems you can install snappy on command line with brew install snzip

More info

• http://fileformats.archiveteam.org/wiki/Snappy
• https://commons.apache.org/proper/commons-compress/examples.html#Snappy

Also, side note, .sz is also the extension for 1997's SunZip http://justsolve.archiveteam.org/wiki/Sunzip

[spiderman666]

Hi @gingerbeardman , thanks for the help!

The tar.sz and .tsz was definitely made by Snzip which you have found/suggested.

I had no luck with the homebrew install of Snzip on my mac as my machine crashes when compiling cmake.
I also lucked out with the snzip windows binaries and ended up having to install a linux VM just to be able to crack open the snappy compressed tar.sz

Thanks again!

[gingerbeardman]

Are you using an old macOS?

I'm using Sonoma on an M1 Pro and brew simply downloaded binaries for snzip and dependencies, no compilation happened.

Also, to ask again, are these types of archives common in the wild? Who is using them?

[spiderman666]

Yes, old macOS and old hardware.

The tar.sz files I have came from an external drive that was a backup for web design, and website cloud backup. If i had to guess, I would pick that tar.sz has been used when piping tar over ssh due to the thousand and thousand of small files inside each archive.

The only other references I could find when trying to research it online had to do with seedboxes and their file browser.

Google themselves make it sound like the snappy compression/format is extremely common and frequently used (with Apache), but this might be an internal thing as I have not encountered it before?

[aonez]

@spiderman666 Went ahead and added Snappy extraction support in this build: Keka-v1.4.0-dev.r5389

@gingerbeardman it also recognizes and extracts by default deltaskin #84 (comment)

[spiderman666]

@aonez Thank you very much that worked a treat!