Fix false positive comparison-overlap for str/int enum members

[LeSingh1]

Fixes #19576.

Mypy represents an enum literal by its member name (a string) regardless of the member's actual value, so under --strict-equality it currently flags

from enum import Enum

class A(str, Enum):
    a = "b"

A.a == "b"  # Non-overlapping equality check

even though the comparison evaluates to True at runtime. A.a has literal type Literal[A.a] with literal value "a" (the member name), so is_overlapping_types doesn't see any overlap with Literal['b']. The same false positive applies to enum.StrEnum and to IntEnum members compared against int literals.

This change treats the comparison as potentially overlapping when one side is an enum literal whose enum class mixes in a non-enum base (e.g. str, int) that the other side is also a value of. The existing non-overlap checks for plain Enum (no mixin) and for mismatched mixin types are preserved, so class B(Enum): x = 1; B.x == 1 and MyStrEnum.member == 1 still error as before.

A regression test testStrEnumEqualityMemberNameDiffersFromValue is added to check-enum.test covering class A(str, Enum), enum.StrEnum, IntEnum, and the plain-Enum non-regression case.

[github-actions]

Diff from mypy_primer, showing the effect of this PR on open source code:

optuna (https://github.com/optuna/optuna)
+ tests/storages_tests/test_heartbeat.py:74: error: Unused "type: ignore" comment  [unused-ignore]
+ tests/storages_tests/test_heartbeat.py:245: error: Unused "type: ignore" comment  [unused-ignore]

psycopg (https://github.com/psycopg/psycopg)
+ tests/test_pipeline_async.py:57: error: Unused "type: ignore" comment  [unused-ignore]
+ tests/test_pipeline.py:60: error: Unused "type: ignore" comment  [unused-ignore]