Summary

Conclusions

Modules and packages define responsibility boundaries. A file format is an external contract: you must check not only syntax but also domain meaning. Separate calculations and explicit paths let you test a program quickly and safely in temporary directories.

Self-check questions

1. What code runs during the first import?
2. How does sys.modules differ from sys.path?
3. Why is the condition involving __name__ needed?
4. How does a relative import depend on the launch method?
5. How does the working directory differ from the module directory?
6. Which open modes can change a file?
7. Why should encoding be specified explicitly?
8. What does JSON lose when converting a tuple or dataclass?
9. Why validate structure after json.loads?
10. Why can CSV not be split reliably with split?
11. What does pytest.raises check?
12. How does parametrize affect the number of tests?
13. What isolation guarantees does tmp_path provide?
14. When are capsys and monkeypatch useful?
15. Why does 100 percent coverage not prove correctness?

Useful links

• https://docs.python.org/3.14/tutorial/modules.html.
• https://docs.python.org/3.14/library/pathlib.html.
• https://docs.python.org/3.14/library/json.html.
• https://docs.python.org/3.14/library/csv.html.
• https://docs.pytest.org/en/stable/.