Special-case .pop() on TypedDict

Mypy doesn't respect TypedDict object changes. As an example, the following snippet

from typing import TypedDict

class A(TypedDict, total=False):
    a: str

def func(a: str) -> None:
    pass

data: A = {"a": "a"}
a = data.pop("a", "")
func(a, **data)

Raises the error:

error: "func" gets multiple values for keyword argument "a" [misc]

Formally, it is correctly, but it will be nice if mypy respects object keys changes. Especially in Unpack case:

from typing import Unpack, TypedDict

class A(TypedDict, total=False):
    a: str
    
def func(a: str) -> None:
    pass
    
def func2(**data: Unpack[A]) -> None:
    a = data.pop("a", "")
    func(a, **data) 

Start by reproducing the two TypedDict examples with mypy, focusing on .pop() followed by keyword unpacking and Unpack handling. Trace the existing TypedDict key-tracking and duplicate-key diagnostic paths; done means valid code using the removed key no longer reports a duplicate keyword argument while invalid cases remain diagnosed.