rewrite tests

[figsoda]

rewrite most tests with proptest, tests single file {,de}compression and directory {,de}compression

Also found an issue with chaining lz4 compressions: tar.lz4.lz4 files are not being compressed correctly, so it is ignored for now

I don't think file permissions are working on windows and for zip files, so I disabled checking them in these situations

Closes #30
Closes #31
Closes #50

[marcospb19]

Will give a proper review in a couple days. Got a bit busy again.

[figsoda]

resolved merge conflicts

[GabrielSimonetto]

I really like this addition, but the scary part for me, as a non-initiated proptest user, is if we will be able to interpret the tests and be sure that they are doing their job along time.

Also, we probably should write something about it on CONTRIBUTING.md as to how to work with it... there should be some stuff to write about on it, but the first thing that comes to mind is just a reference to the proptest book, but also, how we should treat the proptest_regressions dir, (which, the book seems to say that it should be added to git registry? idk).

I will keep looking into it so as to help understand how we should work with proptest, but you can help guide us on how to make this workflow more understandable that would be neat (:

p.s.: I will add that I believe this crate and philosophy of testing is amazing and I would find it really fun and useful to get some experience with it, even if it requires some work

[GabrielSimonetto]

Btw, have you understood how these changes fixed #50 ? It would be nice to know what we were doing wrong, for future reference

[figsoda]

Added some more information to CONTRIBUTING.md

I have no idea how this fixes #50. I actually gave up investigating #50 and just bundled with a couple other issues and rewrote the tests.

[vrmiguel]

Very cool :D

I think that the assert_cmd approach is very interesting but it could bite us in the future if we ever decide to add code coverage metrics, I guess

[marcospb19]

I have no idea how this fixes #50. I actually gave up investigating #50 and just bundled with a couple other issues and rewrote the tests.

That's a mystery, I wish anyone knew whats causing it.