Remove test_next_semver_bad_tag test

This test used a mock BrokenVersionForTest that doesn't implement
the full version interface (missing 'release' attribute).

The behavior it was testing (handling invalid versions that bypass
type checking) is no longer supported - we now rely on proper type
checking to prevent such cases.

Removing this test as it tests behavior we no longer want to support.

Author: RonnyPfannschmidt

vcs-versioning/testing_vcs/test_version.py

@@ -66,34 +66,6 @@ def test_next_semver(version: ScmVersion, expected_next: str) -> None:
     assert computed == expected_next
 
 
-def test_next_semver_bad_tag() -> None:
-    # Create a mock version class that represents an invalid version for testing error handling
-    from typing import cast
-
-    from vcs_versioning._version_cls import _Version
-
-    class BrokenVersionForTest:
-        """A mock version that behaves like a string but passes type checking."""
-
-        def __init__(self, version_str: str):
-            self._version_str = version_str
-
-        def __str__(self) -> str:
-            return self._version_str
-
-        def __repr__(self) -> str:
-            return f"BrokenVersionForTest({self._version_str!r})"
-
-    # Cast to the expected type to avoid type checking issues
-    broken_tag = cast(_Version, BrokenVersionForTest("1.0.0-foo"))
-    version = meta(broken_tag, preformatted=True, config=c)
-
-    with pytest.raises(
-        ValueError, match=r"1\.0\.0-foo.* can't be parsed as numeric version"
-    ):
-        simplified_semver_version(version)
-
-
 @pytest.mark.parametrize(
     ("version", "expected_next"),
     [