repo.blame and repo.blame_incremental now accept None as the rev parameter.

[Gaubbe]

Fixes #1835.

As this pull request only affects the typings and does not change the behavior in any way, I did not write a test for the changes I made.

While a test definitely could be written to see if the library does parse a blame output with uncommited changes, I don't perticularily think this is necessary, as the output format of the git blame -p command is the same as if there had been a commit, just with hard-coded values (such as the null SHA as the commit hash). Therefore, there is no reason the parsing would fail.

I did manually test both methods with the None rev parameter and they do give the expected behavior.

Let me know if you want me to change anything else!

[Byron]

Thanks a lot for your help!

After having had to look back to the issue it resolves to learn what None meant I realized that the docs for rev really need an overhaul. I think it would suffice to mention that None does in both instances.

Then, I think this will be a neat improvement!

[Byron]

Thanks a lot for your help with this!
And the docs are great now 🙏.