Detect inconsistent overload with TypeVar

[cdce8p]

Environment data

• Language Server version: 2021.9.0
• OS and version: darwin x64
• Python version (and distribution if applicable, e.g. Anaconda): 3.10.0rc1
• python.analysis.indexing: undefined
• python.analysis.typeCheckingMode: basic

Code Snippet

from typing import Literal, TypeVar, overload

T = TypeVar("T", int, str)

@overload
def func(option: Literal["a"], var: str) -> str:
    ...

@overload
def func(option: Literal["b"], var: int) -> str:  # Return type must be: int
    ...

def func(option: Literal["a", "b"], var: T) -> T:
    pass

Expected behaviour

Overload 2 is inconsistent with function signature.

Mypy

error: Overloaded function implementation cannot satisfy signature 2 due to inconsistencies in how they use type variables

Actual behaviour

no error

Additional information

After changing the return type of oveload 2 to list, pylance will emit an error:

Overloaded function implementation is not consistent with signature of overload 2
  Function return type "list[Unknown]" is incompatible with type "T@func"
    Type "list[Unknown]" cannot be assigned to type "int* | str*"
      "list[Unknown]" is incompatible with "int*"
      "list[Unknown]" is incompatible with "str*"

[erictraut]

Thanks for the bug report. This will be fixed in the next release.

[jakebailey]

This issue has been fixed in version 2021.9.1, which we've just released. You can find the changelog here: https://github.com/microsoft/pylance-release/blob/main/CHANGELOG.md#202191-8-september-2021