Consider not applying PEP 604 by presence of __future__

[hannseman]

Applying PEP 604 on non 3.10 code will crash on libraries depending on introspection of type hints using typing.get_type_hints. This should apply for libraries like https://github.com/samuelcolvin/pydantic and https://github.com/konradhalas/dacite.

Example:

>>> from __future__ import annotations
>>> import typing
>>> class Snake(typing.NamedTuple):
...     name: str | None
...
>>> typing.get_type_hints(Snake)

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/Cellar/[email protected]/3.9.1_1/Frameworks/Python.framework/Versions/3.9/lib/python3.9/typing.py", line 1410, in get_type_hints
    value = _eval_type(value, base_globals, localns)
  File "/usr/local/Cellar/[email protected]/3.9.1_1/Frameworks/Python.framework/Versions/3.9/lib/python3.9/typing.py", line 277, in _eval_type
    return t._evaluate(globalns, localns, recursive_guard)
  File "/usr/local/Cellar/[email protected]/3.9.1_1/Frameworks/Python.framework/Versions/3.9/lib/python3.9/typing.py", line 533, in _evaluate
    eval(self.__forward_code__, globalns, localns),
  File "<string>", line 1, in <module>
TypeError: unsupported operand type(s) for |: 'type' and 'NoneType'
class Snake(typing.NamedTuple):
    name: str | None

My proposal is to only rewrite to PEP 604 when the minimum version is 3.10.

[asottile]

I'd rather use this in python3.7 so I'm gong to reject this proposal

[antonagestam]

@asottile Would you consider enabling/disabling of this with a feature flag in non-3.10 Python? This will introduce breaking code changes for all essentially any runtime usage of type hints.

... and if not, would you consider a "documentation fix" saying e.g. "this project introduces changes that break runtime introspection of type hints"?

[asottile]

runtime types should use Annotated anyway, I don't consider their current implementation valid

[antonagestam]

@asottile So would you consider a patch that documents the breakage?

Also, this isn't in regard to runtime types that differ from static types. How do you propose using Annotated with Optional[int]? Are you saying one should use something like Annotated[Optional[int], "What here?"]?

Update:

To clarify, this code breaks by the recent changes:

from __future__ import annotations
from typing import Optional, get_type_hints


def foo() -> Optional[int]:
    ...


get_type_hints(foo)

I don't understand how Annotated could alleviate this, but I noticed using string syntax seems to guard the annotation from being broken, e.g def foo() -> "Optional[int]": .... I guess that would be a workaround.

[asottile]

#389 probably does what you want

[antonagestam]

@asottile Hey, thanks for reconsidering and fixing! 🙏