sandr01d

> @sandr01d I refactored most of `_forgit_diff`. Can you check if `alt+e` on a diffed file still works for you? Works for me

Notes regarding d23ec19: I've exposed the new functions `git_branch_delete` and `git_checkout_commit` as forgit commands, so they can be used with `xargs`. I'm parsing the environment variables into arrays in `_forgit_parse_array`...

> This will not run on MacOS's bash, unfortunately, because MacOS uses ancient bash version 3.2.57 (I think due to licensing issues). Is there a different way to implement this?...

@carlfriedrich I've removed the nameref from `_forgit_parse_array` in 88d928d. The downside of this is that arrays parsed with this function now are always global variables. I would be okay with...

Note regarding ce37644 & e026a6d: I changed how we parse file arguments for passing them along to the preview functions. I've added `_forgit_escaped_files` for this. The only difference in behavior...

> Are there any functions left to refactor now? No, we're all set and from my side this would now be ready for review. > I am a little bit...

> Usually we're squash-merging the PRs in this repo, so the history would be gone anyway. On such a huge change, however, it might be useful to keep the history,...

> The longer I think and [read about it](https://ieftimov.com/posts/how-to-make-reviewers-love-your-big-pull-requests/), the more I am sure that we should squash commits semantically. Since you already spent a lot time refactoring, I can...

@carlfriedrich I've now squashed commits together that belong to the same functions. Could you take a quick look and let me know whether that works for you? If so, we...

> @sandr01d Thanks a lot for your further work on this. 22 commits is definitely a lot better than 55, and reviewing this PR commit by commit seems possible to...