Our code:
from dials.test.algorithms.spot_prediction.test_scan_static_reflection_predictor import ( # noqa: F401
data as static_test,
)And flake8 --select F401 test_noqa.py gives no error. When isort is run in --profile=black mode it moves the comment somewhere that flake8 no longer recognises:
from dials.test.algorithms.spot_prediction.test_scan_static_reflection_predictor import (
data as static_test, # noqa: F401
)Now, flake8 counts this as an unused import, because flake8 only handles first-line or continuations. There was a ticket (https://gitlab.com/pycqa/flake8/-/issues/385) documenting this but this was closed as a duplicate of a ticket adding continuation handling, so this is still a problem.
This... sounds pretty similar to #1004 except there's no splitting involved. The pre-isort noqa example there doesn't seem to work in flake8 in my tests anyway, so I'm not sure if the intentions overlap.
Can be worked around with isort:skip on the noqa line.
flake8: 3.8.3 (mccabe: 0.6.1, pycodestyle: 2.6.0, pyflakes: 2.2.0) CPython 3.6.10`
isort: 5.4.2 (though running off develop because I needed the fix from 19221a3)
Our code:
And
flake8 --select F401 test_noqa.pygives no error. When isort is run in--profile=blackmode it moves the comment somewhere that flake8 no longer recognises:Now, flake8 counts this as an unused import, because flake8 only handles first-line or continuations. There was a ticket (https://gitlab.com/pycqa/flake8/-/issues/385) documenting this but this was closed as a duplicate of a ticket adding continuation handling, so this is still a problem.
This... sounds pretty similar to #1004 except there's no splitting involved. The pre-isort noqa example there doesn't seem to work in flake8 in my tests anyway, so I'm not sure if the intentions overlap.
Can be worked around with
isort:skipon the noqa line.flake8: 3.8.3 (mccabe: 0.6.1, pycodestyle: 2.6.0, pyflakes: 2.2.0) CPython 3.6.10`
isort: 5.4.2 (though running off develop because I needed the fix from 19221a3)