isort: fix bad interaction between force-sort-within-sections and force-to-top

[bluetech]

Fixes #3630.

To ease review I split to 2 commits - simple refactor + the logic change.

I've tested this on our code base, works well now.

[github-actions]

PR Check Results

Ecosystem

✅ ecosystem check detected no changes.

Benchmark

Linux

group                                      main                                   pr
-----                                      ----                                   --
linter/all-rules/large/dataset.py          1.00     15.5±0.16ms     2.6 MB/sec    1.01     15.6±0.23ms     2.6 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.01      4.1±0.04ms     4.0 MB/sec    1.00      4.1±0.05ms     4.0 MB/sec
linter/all-rules/numpy/globals.py          1.00    568.8±8.12µs     5.2 MB/sec    1.02    577.6±6.66µs     5.1 MB/sec
linter/all-rules/pydantic/types.py         1.00      6.8±0.08ms     3.7 MB/sec    1.01      6.9±0.07ms     3.7 MB/sec
linter/default-rules/large/dataset.py      1.01      8.4±0.06ms     4.9 MB/sec    1.00      8.3±0.08ms     4.9 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.02  1924.4±13.96µs     8.7 MB/sec    1.00  1894.1±21.54µs     8.8 MB/sec
linter/default-rules/numpy/globals.py      1.00    216.6±2.47µs    13.6 MB/sec    1.01    218.7±1.66µs    13.5 MB/sec
linter/default-rules/pydantic/types.py     1.00      4.0±0.03ms     6.4 MB/sec    1.00      4.0±0.04ms     6.4 MB/sec

Windows

group                                      main                                    pr
-----                                      ----                                    --
linter/all-rules/large/dataset.py          1.06     15.4±0.79ms     2.6 MB/sec     1.00     14.5±0.42ms     2.8 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.04      4.2±0.21ms     4.0 MB/sec     1.00      4.0±0.22ms     4.1 MB/sec
linter/all-rules/numpy/globals.py          1.13   587.4±26.36µs     5.0 MB/sec     1.00   520.6±29.79µs     5.7 MB/sec
linter/all-rules/pydantic/types.py         1.12      7.2±0.32ms     3.5 MB/sec     1.00      6.4±0.23ms     4.0 MB/sec
linter/default-rules/large/dataset.py      1.12      9.5±0.56ms     4.3 MB/sec     1.00      8.4±0.29ms     4.8 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.06  1894.8±155.65µs     8.8 MB/sec    1.00  1787.1±94.85µs     9.3 MB/sec
linter/default-rules/numpy/globals.py      1.00   193.4±10.95µs    15.3 MB/sec     1.13   218.1±11.40µs    13.5 MB/sec
linter/default-rules/pydantic/types.py     1.00      3.8±0.12ms     6.8 MB/sec     1.10      4.1±0.24ms     6.2 MB/sec

[JonathanPlasse]

I think you can simply this to

Suggested change

	match (force_to_top.contains(name1), force_to_top.contains(name2)) {
	(true, true) => Ordering::Equal,
	(false, false) => Ordering::Equal,
	(true, false) => Ordering::Less,
	(false, true) => Ordering::Greater,
	}
	force_to_top.contains(name2).cmp(&force_to_top.contains(name1))

[bluetech]

Changed to something similar, thanks.