【Python】フォルダーのファイル台帳を作る：パス・サイズ・更新日時の収集

from datetime import datetime, timezone
from pathlib import Path
from tempfile import TemporaryDirectory
from unittest.mock import patch
import os
import stat


def inventory(root):
    root = Path(root)
    pending, rows, errors = [root], [], []
    while pending:
        directory = pending.pop()
        try:
            with os.scandir(directory) as entries:
                items = sorted(entries, key=lambda entry: entry.name)
        except OSError as error:
            errors.append((directory.relative_to(root).as_posix(), type(error).__name__))
            continue
        for entry in items:
            path = Path(entry.path)
            relative = path.relative_to(root).as_posix()
            try:
                info = entry.stat(follow_symlinks=False)
                if stat.S_ISDIR(info.st_mode):
                    pending.append(path)
                elif stat.S_ISREG(info.st_mode):
                    modified = datetime.fromtimestamp(info.st_mtime, timezone.utc).isoformat()
                    rows.append((relative, info.st_size, modified))
            except OSError as error:
                errors.append((relative, type(error).__name__))
    return sorted(rows), sorted(errors)


with TemporaryDirectory() as temporary:
    root = Path(temporary)
    (root / "blocked").mkdir()
    (root / "a.txt").write_bytes(b"abc")
    os.utime(root / "a.txt", (0, 0))
    (root / "link.txt").symlink_to(root / "a.txt")
    real_scandir = os.scandir

    def controlled_scandir(path):
        if Path(path).name == "blocked":
            raise PermissionError("simulated for this example")
        return real_scandir(path)

    with patch("os.scandir", side_effect=controlled_scandir):
        rows, errors = inventory(root)
    assert rows == [("a.txt", 3, "1970-01-01T00:00:00+00:00")]
    assert errors == [("blocked", "PermissionError")]
    print("files:", rows)
    print("errors:", errors)