Handle non Unicode env keys/values

[Jongy]

(disregard the code / commit message - I'll clean up & add tests once we agree on the implementation 😄 )

I hit the same issue as described in the linked ticket - with key PWD having a non-UTF-8 value ("invalid" from now on).

Technically, keys can be invalid too in Linux. The naive implementation I provided here filters out all invalid keys & values. Keys in config-rs are Strings so it always makes sense to filter invalid ones, as they won't be used by this library AFAIU.
As for values - I'm unsure if plainly ignoring invalid values is okay. If the value is set in a key that would have been used by the current config consumer, then ignoring it (as if the user did not provide a value through env) is not okay.
Ideally - the logic would be:

1. Valid key & value - continue
2. Invalid key - drop
3. Invalid value - drop if key is not consumed, otherwise panic.

Let me know what you think :) I can try looking into point 3.

Resolves: #579

[coveralls]

Pull Request Test Coverage Report for Build 13248291345

Warning: This coverage report may be inaccurate.

This pull request's base commit is no longer the HEAD commit of its target branch. This means it includes changes from outside the original pull request, including, potentially, unrelated coverage changes.

• For more information on this, see Tracking coverage changes with pull request builds.
• To avoid this issue with future PRs, see these Recommended CI Configurations.
• For a quick fix, rebase this PR at GitHub. Your next report should be accurate.

Details

• 10 of 13 (76.92%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.1%) to 65.124%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/env.rs	10	13	76.92%

Totals
Change from base Build 13059823868:	0.1%
Covered Lines:	943
Relevant Lines:	1448

---

💛 - Coveralls

[epage]

I'm sorry, I overlooked that I was reviewing two different PRs (#641) for the same problem.

The tests in #641 were able to also cover Windows which could be nice.

[Jongy]

Thanks for the review @epage , I addressed all comments.

The tests in #641 were able to also cover Windows which could be nice.

Fixed the tests here to support Windows. I haven't set up Wine to run them so I'll wait to see how it goes in the CI.

[epage]

Thanks!

[Jongy]

Thanks again for the review, and thanks for releasing, I'll proceed to update atuin to use the new version.