Python Path.copy: a snapshot needs an explicit metadata policy

Operation contract

A temporary receipt file is copied into an archive path, then the source is revised. The destination retains its earlier bytes. The default call checks file content only; it does not ask to preserve all available metadata. Both paths are owned by the fixture and removed when the temporary directory closes.

Failure boundary

Path.copy may replace an existing file at the destination, and symlinks are followed by default. Choose a destination policy before copying user-selected paths. The copy is not a transactional backup: source mutation during copying, failed writes, cross-filesystem behavior, and crash recovery need separate handling. Request metadata preservation only when the application requires it.

Working program

from pathlib import Path
from tempfile import TemporaryDirectory

with TemporaryDirectory() as directory:
    source = Path(directory) / "current-receipt.txt"
    archived = Path(directory) / "archived-receipt.txt"
    source.write_text("batch-47", encoding="utf-8")
    copied = source.copy(archived)
    source.write_text("batch-48", encoding="utf-8")
    print("copied", copied.read_text(encoding="utf-8"))
    print("current", source.read_text(encoding="utf-8"))

Output

copied batch-47
current batch-48

Costs and limits

Copy cost generally grows with data size, although a filesystem may support copy-on-write. Metadata preservation and cross-device handling add filesystem work; measure on the deployment volume.

Common Mistakes

• The default copy need not preserve every file attribute.
• The destination may be replaced if it already exists.
• A copy made during concurrent writes is not automatically a consistent snapshot.

Connected lessons

• Python pathlib files: specify encoding and close the resource owner
• Python testing: assert temporary cleanup after a failure
• Python file replacement: write a sibling file before changing the name

Test this contract.