remove HashableT in frame.pyi where possible

[Dr-Irv]

mypy has finally put in the feature allowing asymmetric getters/setters. During the testing, they had an issue with our stubs regarding where we use HashableT. See python/mypy#18510 (comment)

This fixes that issue. Then once mypy 1.16 i released, we can fix the tests for setting columns that have # type: ignore in it.

[twoertwein]

will this accept dict[str, npt.DTypeLike]?

[Dr-Irv]

Yes, tested in line 3090 in test_frame.py

[twoertwein]

it only works because of bidirectional type inference, if you change the test slightly:

+    dtypes = {"col1": np.int8, "col2": np.int16}
     check(
-        assert_type(
-            DF.to_records(False, {"col1": np.int8, "col2": np.int16}), np.recarray
-        ),
+        assert_type(DF.to_records(False, dtypes), np.recarray),
         np.recarray,
     )

it fails

tests/test_frame.py:3090: error: Argument 2 to "to_records" of "DataFrame" has incompatible type "dict[str, type[object]]"; expected "str | dtype[Any] | type[Any] | _SupportsDType[dtype[Any]] | str | tuple[Any, int] | tuple[Any, SupportsIndex | Sequence[SupportsIndex]] | list[Any] | _DTypeDict | tuple[Any, Any] | None | Mapping[Hashable, dtype[Any] | type[Any] | _SupportsDType[dtype[Any]] | str | tuple[Any, int] | tuple[Any, SupportsIndex | Sequence[SupportsIndex]] | list[Any] | _DTypeDict | tuple[Any, Any] | None] | None" [arg-type]

/Users/twoertwein/pandas-stubs/tests/test_frame.py:3090:42 - error: Argument of type "dict[str, int8 | int16]" cannot be assigned to parameter "column_dtypes" of type "DTypeLike | Mapping[Hashable, DTypeLike]" in function "to_records"

For some reason the other list[str] case was fine, maybe setter are handed differently?

[Dr-Irv]

@twoertwein I've addressed all your comments. I don't think any other changes are necessary. Let me know if you disagree.