Add possibility to compress log files with Zstandard

[cristiano-prato]

I would like to be able to compress log files with zstandard instead of gzip.

[bconn98]

Hey, thanks for the contribution! Can you update the compression section of the readme to include the new compression method? Also, in the Configuration docs under the docs dir. It would also be greatly appreciated if you could create a benchmark to compare the two. If you don't have the time, we can look at adding it after the fact. For a reference to a benchmark from gzip testing a while back #117.

[cristiano-prato]

Hi, thanks for the feedback! I added some documentation, benchmark and proper error checking, let me know if it is ok for you!

[bconn98]

Awesome, thanks for the docs + benchmark updates. Couple notes on the docs. I'm still thinking about the windows question

[cristiano-prato]

I changed the documentation as suggested

[estk]

@cristiano-prato Thanks so much for the excellent contribution. One question I had was: what happens when a user does not have zstd installed?

Other than that it lgtm

[bconn98]

Agreed with @estk on that concern. I was including that in my decision process about the windows comment. Possibly a startup check with an error out?

[cristiano-prato]

@cristiano-prato Thanks so much for the excellent contribution. One question I had was: what happens when a user does not have zstd installed?

Other than that it lgtm

If the zstd executable is missing, the test will fail. If you agree, I will change the test in order to use the library to decompress the archive, instead of the executable.

[cristiano-prato]

I changed the unit test to avoid dependencies with zstd cli. This also allows the test to run on windows.

[cristiano-prato]

Hello, is there anything I can do to improve this PR?

[bconn98]

@cristiano-prato Ill take another look later today, but no I don't think so. I'm waiting to get a group of good PRs before bugging the repo owner to take a look.

[bconn98]

This looks good to me. However, we need to wait for #366 & its associated PR to close first.

[estk]

This lgtm, only thing that would be nice is a usage example.