Fix AST verification and historical comparisons

[akaihola]

This pull request fixes three bugs:

AST verification no longer erroneously fails when using --isort.

This bug was introduced in #94 and is present in version 1.2.3.

Historical comparisons, e.g. comparing two past commits like darker --diff v1.0..v1.1 would still compare against the file on disk in the working tree, not the second commit. This has been fixed.

Darker would write historical reformats on disk, overwriting a newer version in the working tree. Now an exception is raised instead if a historical comparison like darker --revision=commit1..commit2 is used without --diff or --check.

[akaihola]

@samoylovfp, I assigned this to you since we collaborated also on #94 and you may remember the change I made there.

The description isn't actually accurate currently, I'm trying to rethink this through to figure out what's happening exactly.

Edit: Can't figure it out now, just made the description a bit more general. Note that I found another bug which is related to the same code paths, and included the fix for that one in this PR as well.

[akaihola]

@ivanov, I added you as a reviewer just in case you'd be interested – you mentioned this PR in the discussion of #140.

[akaihola]

@ivanov and/or @samoylovfp, do you think you'll be able to review this?

One option I'd like to try at some point in this project is to have a pair/group video call session where a pull request author walks through their proposed changes. Maybe this PR is a candidate for such?

[sourcery-ai]

Sourcery Code Quality Report

❌  Merging this PR will decrease code quality in the affected files by 0.95%.

Quality metrics	Before	After	Change
Complexity	5.00 ⭐	5.09 ⭐	0.09 👎
Method Length	63.22 🙂	65.04 🙂	1.82 👎
Working memory	12.23 😞	12.67 😞	0.44 👎
Quality	63.59% 🙂	62.64% 🙂	-0.95% 👎

Other metrics	Before	After	Change
Lines	1293	1400	107

Changed files	Quality Before	Quality After	Quality Change
src/darker/main.py	40.40% 😞	35.62% 😞	-4.78% 👎
src/darker/git.py	70.26% 🙂	72.29% 🙂	2.03% 👍
src/darker/utils.py	82.28% ⭐	82.28% ⭐	0.00%
src/darker/tests/test_git.py	63.79% 🙂	62.98% 🙂	-0.81% 👎
src/darker/tests/test_main.py	61.16% 🙂	62.56% 🙂	1.40% 👍

Here are some functions in these files that still need a tune-up:

File	Function	Complexity	Length	Working Memory	Quality	Recommendation
src/darker/main.py	format_edited_parts	21 😞	243 ⛔	17 ⛔	27.06% 😞	Refactor to reduce nesting. Try splitting into smaller methods. Extract out complex expressions
src/darker/main.py	main	19 😞	262 ⛔	16 ⛔	28.52% 😞	Refactor to reduce nesting. Try splitting into smaller methods. Extract out complex expressions
src/darker/tests/test_main.py	test_main	0 ⭐	257 ⛔	33 ⛔	39.08% 😞	Try splitting into smaller methods. Extract out complex expressions
src/darker/tests/test_git.py	test_git_get_modified_files	4 ⭐	177 😞	17 ⛔	46.33% 😞	Try splitting into smaller methods. Extract out complex expressions
src/darker/tests/test_git.py	test_git_check_output_lines	2 ⭐	127 😞	40 ⛔	47.30% 😞	Try splitting into smaller methods. Extract out complex expressions

Legend and Explanation

The emojis denote the absolute quality of the code:

• ⭐ excellent
• 🙂 good
• 😞 poor
• ⛔ very poor

The 👍 and 👎 indicate whether the quality has improved or gotten worse with this pull request.

---

Please see our documentation here for details on how these metrics are calculated.

We are actively working on this report - lots more documentation and extra metrics to come!

Let us know what you think of it by mentioning @sourcery-ai in a comment.

[akaihola]

I re-reviewed this myself. Merging now since other contributors are busy.