File input and output is simple until paths, encodings, interruptions, permissions, and large files appear. Use path objects to make file operations clearer and context managers to ensure files close even when an error occurs. Always decide whether the file contains text or binary data.
Encoding should be explicit when files cross systems. For structured data, use CSV or JSON tools rather than manually splitting strings. Large files should be streamed instead of loaded entirely into memory. Writes that replace important data should use a temporary file and an atomic replacement strategy where appropriate.
Tests should operate in temporary directories and include missing files, invalid content, permission failures, and empty input. Avoid building paths from untrusted input without validation. Good file I/O is not only about getting a successful result; it is about preserving data when the environment behaves unexpectedly.