[flake8-future-annotations] Implement FA102

[akx]

Summary

As discussed in #4553, this adds FA102 to complement FA100 (see #3072, #3797):

FA102 Missing import when code uses simplified types (list, dict, set, etc)

This is only enabled for the target versions where the simplified annotations aren't valid:

• PEP585 list[...], dict[...] are complained about when targeting Python < 3.9
• PEP604 | unions are complained about when targeting Python < 3.10

Test Plan

So far: A couple of new tests (utilizing the same fixtures as the FA100 tests).

Let me know if more is needed...

TODO

I'm not quite sure what would be the proper way to add the autofix for this, since it'd mean two Diagnostics, if I used add_required_import..?

[github-actions]

PR Check Results

Ecosystem

✅ ecosystem check detected no changes.

Benchmark

Linux

group                                      main                                   pr
-----                                      ----                                   --
linter/all-rules/large/dataset.py          1.00     14.2±0.53ms     2.9 MB/sec    1.04     14.9±0.65ms     2.7 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.02      3.6±0.15ms     4.6 MB/sec    1.00      3.6±0.17ms     4.7 MB/sec
linter/all-rules/numpy/globals.py          1.00   443.6±22.85µs     6.7 MB/sec    1.01   448.7±22.88µs     6.6 MB/sec
linter/all-rules/pydantic/types.py         1.07      6.3±0.31ms     4.1 MB/sec    1.00      5.8±0.21ms     4.4 MB/sec
linter/default-rules/large/dataset.py      1.00      6.9±0.30ms     5.9 MB/sec    1.02      7.0±0.27ms     5.8 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.10  1696.3±52.89µs     9.8 MB/sec    1.00  1545.4±69.32µs    10.8 MB/sec
linter/default-rules/numpy/globals.py      1.02    183.7±8.40µs    16.1 MB/sec    1.00    180.9±8.14µs    16.3 MB/sec
linter/default-rules/pydantic/types.py     1.04      3.4±0.12ms     7.6 MB/sec    1.00      3.2±0.17ms     7.9 MB/sec
parser/large/dataset.py                    1.00      5.3±0.21ms     7.7 MB/sec    1.06      5.6±0.20ms     7.2 MB/sec
parser/numpy/ctypeslib.py                  1.00  1045.9±50.48µs    15.9 MB/sec    1.02  1067.1±70.81µs    15.6 MB/sec
parser/numpy/globals.py                    1.00    109.8±5.37µs    26.9 MB/sec    1.00    109.2±6.69µs    27.0 MB/sec
parser/pydantic/types.py                   1.05      2.3±0.11ms    11.0 MB/sec    1.00      2.2±0.09ms    11.6 MB/sec

Windows

group                                      main                                   pr
-----                                      ----                                   --
linter/all-rules/large/dataset.py          1.03     17.0±0.11ms     2.4 MB/sec    1.00     16.5±0.10ms     2.5 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.01      4.2±0.05ms     3.9 MB/sec    1.00      4.2±0.05ms     4.0 MB/sec
linter/all-rules/numpy/globals.py          1.01    502.5±7.22µs     5.9 MB/sec    1.00    499.7±9.19µs     5.9 MB/sec
linter/all-rules/pydantic/types.py         1.02      7.1±0.09ms     3.6 MB/sec    1.00      7.0±0.06ms     3.7 MB/sec
linter/default-rules/large/dataset.py      1.01      8.3±0.13ms     4.9 MB/sec    1.00      8.2±0.06ms     5.0 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.00  1760.5±15.49µs     9.5 MB/sec    1.00  1761.2±30.61µs     9.5 MB/sec
linter/default-rules/numpy/globals.py      1.00    205.2±3.82µs    14.4 MB/sec    1.01    206.7±5.37µs    14.3 MB/sec
linter/default-rules/pydantic/types.py     1.00      3.7±0.03ms     6.8 MB/sec    1.00      3.7±0.04ms     6.8 MB/sec
parser/large/dataset.py                    1.00      6.4±0.05ms     6.3 MB/sec    1.01      6.5±0.05ms     6.3 MB/sec
parser/numpy/ctypeslib.py                  1.00  1224.8±12.68µs    13.6 MB/sec    1.01  1231.3±13.92µs    13.5 MB/sec
parser/numpy/globals.py                    1.00    125.1±3.77µs    23.6 MB/sec    1.00    124.6±1.54µs    23.7 MB/sec
parser/pydantic/types.py                   1.00      2.8±0.03ms     9.2 MB/sec    1.00      2.7±0.03ms     9.3 MB/sec

[charliermarsh]

Thanks :)