Effective testing with Python

Samples of pytest tests and a slide presentation explaining how to do effecting testing with Pytest and Python. Slides are in Spanish and English.

This project was created for PyConEs 2021 and used later for other venues. It has been continued as a central repository of test samples in Python with pytest and other tools

It has a Markdown based presentation in Spanish and English plus a set of test samples using a bare minimum source code.

Check out the test samples at tests/ Check out the slides at slides (markdown or PDF) Check out the dummy code at source/

Apart from tests samples, it also has some advanced tests configuration, see tests/confest.py for more detail.

Feel free to refer or reuse this content. Attribution is appreciated

Glossary

• fixture: generated artifact to use in a test, could be data, an object, ...
• factory: function to generate a fixture
• sut: system under testing

Usage

pytest
pytest -m unit

External documentation

Pytest docs Markdown slides for RevealJS

Development

Project Structure

• source/: dummy code
• tests/
  • Usual folders by domain and type:
    • types: smoke for sanity tests, unit, integration, api ..
    • domain: depends on your project
    • folder naming is used to run tests by tag later (python -m yourtag)

Test samples

• Smoke tests
• Unit tests
  • delayed assert: sometimes is useful to run several tests over the same execution, but we still want to see them as separated ones

Recommended plugins and libraries

Highly recommended:

• Faker: fake data generator
• factoryboy and polifactory: Mother or Object factory libraries
• freezegun: stops and moves time at will, use with caution
• lambdapowertools: AWS lambda utils
• any time library like timeago, arrow, pendulum ...

Some are already installed, see Pipfile

• deep diff: advanced diff library that allow you to customize your diff, especially for complex structures like dicts. Among other features you can ignore keys, take order into account or not, ect.
• time diff and ranges: arrow, pendulum, deloream, maya, moment, timeago, when.
• Some are really useful to complex time ranges or concepts like: a week from today, next monday, last Sunday of the previous month, etc.