That is very cool. May I ask, how does the compressed stream seeking work? Does it keep state of the decompressor at certain points so arbitrary access can be faster than reading from the start of the stream?
similiar projects:
https://github.com/cybernoid/archivemount
https://github.com/google/fuse-archive
Congratulations on your v1.0.0! This is definitely a very nice tool, I'll try to play with it a bit and maybe try to make an ebuild (though the build system seems a bit complicated for proper no-network package managers). The extensive benchmark section is a nice plus.
A small note, archivemount has a living fork here: https://git.sr.ht/~nabijaczleweli/archivemount-ng
This is awesome :)
I find this project hugely helpful when working with Google Takeout archives. I normally pick a size that's not too large so that downloading them is easier, then it's simply a matter of: