7. Testing your code.

Lecture notes

• Read Python Crash Course Chapters 11 (skip the section on testing classes).
• We will use ipytest to make it work in google colab: please read the documentation
• We will also use doctest
  Just call it in a cell in colab, and it will test all functions defined up till now.
  import doctest doctest.testmod() ### or doctest.testmod(verbose=True)
• Cheat Sheet: list, Cheat Sheet: list BW
• Pytest
  • ✓ Flexible and powerful test framework with plugins and fixtures for complex tests.
  • ✓ Better error reporting with detailed tracebacks and debugging information.
  • ✓ Allows parameterized testing, making it easier to run tests with multiple inputs.
  • ✗ Requires learning additional syntax and setup.
  • ✗ Heavier and slower for small, isolated tests.
• Doctest
  • ✓ Lightweight and easy to use.
  • ✓ Promotes documentation as tests, helping ensure examples in docs are accurate.
  • ✗ Limited in handling complex tests.
  • ✗ Less informative error reporting, minimal feedback on failed tests.
  • ✗ Sensitive to formatting in test cases.
• For code you will use again, it is important to have tests
• For them to be useful, you need to actually run them regularly

Before class (code, output, test, test output)

• Write tests with both pytest and doctest for the practical code from weeks 5 and 6.
  • For IsReduplicated
  • For find_redup
  • For strip_gutenberg
  • For process_line
• Hint
  • for the doctest, write the test, and look at the failure message to see the expected output, use this in the text
  • for a multi-line string in the test, you need to escape the newlines
    so "line one\\nline two"
    instead of "line one\nline two

Practical work (code, output, test, test output)

• Write code to analyze a text in terms of characters
  Return a dictionary with the count of
  • all characters (length)
  • ascii (unaccented characters - hint use isascii() and isalpha())
  • accented characters
  • vowels
  • consonants
• Document this using docstring or pytest
  • test with some Czech
  • test with some Pinyin
  • test with some Vietnamese
• Write a swear_filter function that censors offensive words by replacing them with asterisks (e.g., "darn" -> "****").
  You should cover at least Carlin's seven dirty words: shit, piss, fuck, cunt, cocksucker, motherfucker, tits
• Write pytest tests for swear_filter
  • Verify correct censoring of offensive words in various contexts.
  • Add a test for words within other words --- we will learn how to pass this next week
    test driven development write tests, and then makes them pass
• Add doctest examples showing censored versions of sentences with sample swear words.
• ❂ Extend the function and tests to optionally
  • replace all letters except 1st and last: shit → s**t
  • replace short words with bleep and long words with bleepbleep

❂ If you haved finished everything else, try this

Summary

• Learned how to write tests for functions and classes using the pytest and doctest modules.
• Practiced writing test functions to verify specific behaviors of functions.
• Explored using fixtures to efficiently create reusable resources for multiple tests.

• Testing helps ensure that critical behaviors of functions and classes actually work.
• You should add tests on critical code behaviors, without aiming for full coverage in early projects.
• Tests provide confidence in maintaining and improving code without breaking existing functionality.

• Initial tests make projects more respected and encourage collaboration with other programmers.
• Responding to failed tests is easier and faster than handling bug reports from users.
• Contributions to collaborative projects often require passing existing tests and adding new ones for any new functionality.

• You should at least have doctests for your assignment

LAC: Language and the Computer Francis Bond.