F402 is triggered by __future__.annotations

[mattmess1221]

Description

F402 is being triggered by __future__ imports. These imports can probably be safely ignored.

Searched terms: F402, annotations, future

Reproduction

from __future__ import annotations

for annotations in [["anno1", "anno2"]]:
    print(annotations)

Output:

% ruff --version
ruff 0.9.3
% ruff check --select F
f.py:3:5: F402 Import `annotations` from line 1 shadowed by loop variable
  |
1 | from __future__ import annotations
2 |
3 | for annotations in [["anno1", "anno2"]]:
  |     ^^^^^^^^^^^ F402
4 |     print(annotations)
  |

Found 1 error.

[InSyncWithFoo]

This seems to be by design, considering what Pyflakes does:

# https://github.com/PyCQA/pyflakes/blob/d9e32c4c/pyflakes/checker.py#L971
if isinstance(existing, Importation) and isinstance(parent_stmt, FOR_TYPES):
    self.report(messages.ImportShadowedByLoopVar,
                node, value.name, existing.source)

# https://github.com/PyCQA/pyflakes/blob/d9e32c4c/pyflakes/checker.py#L415
class FutureImportation(ImportationFrom):
    """
    A binding created by a from `__future__` import statement.

    `__future__` imports are implicitly used.
    """
    ...

[AlexWaygood]

Yes, putting from __future__ import annotations in a module does actually insert an annotations symbol into the module's namespace:

>>> from __future__ import annotations
>>> annotations
_Feature((3, 7, 0, 'beta', 1), None, 16777216)

This is documented at https://docs.python.org/3/library/__future__.html.

I appreciate that most people don't ever use the symbol that's inserted as a result of these import statements -- the main use case is the "side effect" the import statement has that causes the module to be parsed in a different way. However, your code still might be more readable if you used a different variable name to differentiate your loop variable from the symbol inserted by the import statement.