【Python】大量ファイルの名前を安全に変える：事前確認と衝突検出

from pathlib import Path
from tempfile import TemporaryDirectory, mkdtemp
import re


def rename_batch(root, mapping, *, dry_run=True):
    root = Path(root)
    plan = list(mapping.items())
    for old, new in plan:
        for name in (old, new):
            if not re.fullmatch(r"[A-Za-z0-9_.-]+", name) or name in {".", ".."}:
                raise ValueError("use a simple single-component name")
        source = root / old
        if source.is_symlink() or not source.is_file():
            raise ValueError("source must be a regular non-symlink file")
    target_keys = [new.casefold() for _, new in plan]
    if len(target_keys) != len(set(target_keys)):
        raise ValueError("duplicate destinations")
    sources = {old for old, _ in plan}
    for existing in root.iterdir():
        if existing.name.casefold() in target_keys and existing.name not in sources:
            raise FileExistsError(existing.name)
    if dry_run or not plan:
        return plan
    staging = Path(mkdtemp(prefix=".rename-", dir=root))
    try:
        for index, (old, _) in enumerate(plan):
            (root / old).rename(staging / str(index))
        for index, (_, new) in enumerate(plan):
            (staging / str(index)).rename(root / new)
    except OSError as error:
        raise RuntimeError("partial rename; inspect files and staging directory") from error
    else:
        staging.rmdir()
    return plan


with TemporaryDirectory() as temporary:
    root = Path(temporary)
    (root / "a.txt").write_text("A", encoding="utf-8")
    (root / "b.txt").write_text("B", encoding="utf-8")
    mapping = {"a.txt": "b.txt", "b.txt": "a.txt"}
    print("plan:", rename_batch(root, mapping))
    assert (root / "a.txt").read_text() == "A"
    for invalid in ({"a.txt": "b.txt"}, {"a.txt": "x.txt", "b.txt": "x.txt"}):
        try:
            rename_batch(root, invalid)
        except (ValueError, FileExistsError):
            pass
        else:
            raise AssertionError("collision was not detected")
    rename_batch(root, mapping, dry_run=False)
    assert (root / "a.txt").read_text() == "B"
    assert (root / "b.txt").read_text() == "A"
    assert sorted(path.name for path in root.iterdir()) == ["a.txt", "b.txt"]
    print("after:", (root / "a.txt").read_text(), (root / "b.txt").read_text())