Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Is there a good reason why Mapping.__contains__ expects object, but Mapping.get expects a strict type? #8219

Open
pradeep90 opened this issue Jun 30, 2022 · 2 comments
Labels
stubs: false negative Type checkers do not report an error, but should

Comments

@pradeep90
Copy link

I ran into a case where a user intended to check whether an enum value x was present in a dictionary Dict[int, int]: x.value in d.

But instead they wrote x in d, which always fails, since the Enum itself is never a key in the dictionary. The type checker didn't complain since the typeshed stub for Mapping.__contains__ accepts any type compatible with object:

    def __contains__(self, __o: object) -> bool: ...

However, if the user had used d.get(x), the type checker would have complained, because the typeshed stub for Mapping.get is stricter:

    def get(self, __key: _KT) -> _VT_co | None: ...

Simple repro:

d: dict[int, int] = {1: 2}

# no type error
'hello' in d

# type error
d.get('hello')

It looks like the stub for __contains__ has expected object ever since typing.pyi was added to typeshed (in 2015). If the idea is that users may want to check for existence of keys of arbitrary types, a similar argument would hold for d.get(x).

Is there a good reason why the stub complains about d.get(x) but not x in d? Otherwise, it'd be good to make __contains__ stricter.

@srittau
Copy link
Collaborator

srittau commented Jun 30, 2022

I definitely see the logic in that, but I also expect that a change like this would break quite a lot. An exploratory PR to judge the impact would be interesting, though.

@srittau srittau added the stubs: false negative Type checkers do not report an error, but should label Jun 30, 2022
@AlexWaygood
Copy link
Member

This is a similar issue to #6597, though it comes at it from the opposite direction

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
stubs: false negative Type checkers do not report an error, but should
3 participants