Function auto_mark_directory

scripts/update_lib/cmd_auto_mark.py:842–996

Run tests and auto-mark failures in a test directory. Runs the test once for the whole directory, then applies results to each file. Args: test_dir: Path to the test directory mark_failure: If True, add @expectedFailure to ALL failing tests verbose: Print progr

(
    test_dir: pathlib.Path,
    mark_failure: bool = False,
    verbose: bool = True,
    original_methods_per_file: dict[pathlib.Path, set[tuple[str, str]]] | None = None,
    skip_build: bool = False,
)

def auto_mark_directory(
    test_dir: pathlib.Path,
    mark_failure: bool = False,
    verbose: bool = True,
    original_methods_per_file: dict[pathlib.Path, set[tuple[str, str]]] | None = None,
    skip_build: bool = False,
) -> tuple[int, int, int]:
    """
    Run tests and auto-mark failures in a test directory.

    Runs the test once for the whole directory, then applies results to each file.

    Args:
        test_dir: Path to the test directory
        mark_failure: If True, add @expectedFailure to ALL failing tests
        verbose: Print progress messages
        original_methods_per_file: If provided, only auto-mark failures for NEW methods
                                   even without mark_failure. Dict maps file path to
                                   set of (class_name, method_name) tuples.

    Returns:
        (num_failures_added, num_successes_removed, num_regressions)
    """
    test_dir = pathlib.Path(test_dir).resolve()
    if not test_dir.exists():
        raise FileNotFoundError(f"Directory not found: {test_dir}")
    if not test_dir.is_dir():
        raise ValueError(f"Not a directory: {test_dir}")

    # Get all .py files in directory
    test_files = sorted(test_dir.glob("**/*.py"))

    # Strip reason-less markers from ALL files before running tests so those
    # tests fail normally and we capture their error messages.
    stripped_per_file: dict[pathlib.Path, set[tuple[str, str]]] = {}
    original_per_file: dict[pathlib.Path, str] = {}
    for test_file in test_files:
        contents = test_file.read_text(encoding="utf-8")
        stripped_contents, stripped = strip_reasonless_expected_failures(contents)
        if stripped:
            original_per_file[test_file] = contents
            test_file.write_text(stripped_contents, encoding="utf-8")
            stripped_per_file[test_file] = stripped

    test_name = get_test_module_name(test_dir)
    if verbose:
        print(f"Running test: {test_name}")

    results = run_test(test_name, skip_build=skip_build)

    # Check if test run failed entirely (e.g., import error, crash)
    if (
        not results.tests_result
        and not results.tests
        and not results.unexpected_successes
    ):
        # Restore original contents before raising
        for fpath, original in original_per_file.items():