Use multi-fix semantics for inplace removal

[charliermarsh]

Summary

Long ago, we added the ability to "apply a single fix to a string in-place", which we used as a hack to enable multiple autofixes. (In short: we'd apply one fix, look at the fixed code, apply another fix, then tell Ruff that the fix is the replace the entire expression with that fixed output.) Ruff now supports multiple edits per fix, so we can instead model this behavior as two distinct fixes.

While I was here, I also made the inplace rule a little more robust.

[github-actions]

PR Check Results

Ecosystem

✅ ecosystem check detected no changes.

Benchmark

Linux

group                                      main                                   pr
-----                                      ----                                   --
linter/all-rules/large/dataset.py          1.03     18.6±1.36ms     2.2 MB/sec    1.00     18.0±0.84ms     2.3 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.00      4.6±0.25ms     3.7 MB/sec    1.00      4.5±0.20ms     3.7 MB/sec
linter/all-rules/numpy/globals.py          1.10   646.1±38.11µs     4.6 MB/sec    1.00   586.1±28.39µs     5.0 MB/sec
linter/all-rules/pydantic/types.py         1.12      8.3±0.35ms     3.1 MB/sec    1.00      7.4±0.34ms     3.5 MB/sec
linter/default-rules/large/dataset.py      1.09     10.0±0.61ms     4.0 MB/sec    1.00      9.2±0.35ms     4.4 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.05      2.3±0.13ms     7.3 MB/sec    1.00      2.2±0.11ms     7.7 MB/sec
linter/default-rules/numpy/globals.py      1.17   287.5±16.50µs    10.3 MB/sec    1.00    244.9±9.91µs    12.1 MB/sec
linter/default-rules/pydantic/types.py     1.06      4.5±0.29ms     5.6 MB/sec    1.00      4.3±0.27ms     6.0 MB/sec

Windows

group                                      main                                   pr
-----                                      ----                                   --
linter/all-rules/large/dataset.py          1.00     18.2±0.62ms     2.2 MB/sec    1.12     20.3±0.65ms     2.0 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.00      4.8±0.23ms     3.5 MB/sec    1.07      5.2±0.18ms     3.2 MB/sec
linter/all-rules/numpy/globals.py          1.00   583.6±27.90µs     5.1 MB/sec    1.10   643.5±29.58µs     4.6 MB/sec
linter/all-rules/pydantic/types.py         1.00      8.0±0.33ms     3.2 MB/sec    1.08      8.6±0.32ms     3.0 MB/sec
linter/default-rules/large/dataset.py      1.00      9.7±0.51ms     4.2 MB/sec    1.10     10.7±0.61ms     3.8 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.00      2.2±0.11ms     7.5 MB/sec    1.06      2.4±0.11ms     7.1 MB/sec
linter/default-rules/numpy/globals.py      1.00   276.2±15.90µs    10.7 MB/sec    1.06   293.7±23.46µs    10.0 MB/sec
linter/default-rules/pydantic/types.py     1.00      4.8±0.25ms     5.3 MB/sec    1.10      5.2±0.31ms     4.9 MB/sec