What it does
pytest is a Python testing framework for small unit tests and larger functional suites. It discovers tests by naming conventions, reports assertion failures, and provides fixtures for shared setup and teardown. The project's how-to guides cover temporary directories, monkeypatching, warnings, skipped tests, and plugins, so a team can extend tests without building a custom runner.
A first test can live in test_sample.py as a function whose name begins with test_. The getting-started guide demonstrates a simple function and an assert; invoking pytest from the project directory discovers the file and prints the result. A real example might assert that a parser returns the expected fields for a short input string. The result is a report of passed and failed tests, not an application behavior change.
Fixtures help reuse setup such as a temporary file or database connection. Parametrization lets one test cover several inputs, while monkeypatch changes a dependency for a controlled test. Install pytest into the project's Python environment so the test command uses the same dependency set as CI. pytest is MIT-licensed and has no paid hosted service requirement in this entry.
Strengths
- Test functions can use ordinary Python `assert` statements with readable failure output.
- Fixtures and parametrization support reusable setup and multiple input cases.
- The docs cover temporary paths, monkeypatching, warnings, and test selection.
Watch out for
- The developer must supply the tests and the Python environment they depend on.
Best for
- Testing Python functions and packages with fast local feedback
- Adding fixture-based unit or functional tests to a Python project
Sources
- Get Started - pytest documentation (Oct 25, 2026)
- pytest documentation (Oct 25, 2026)
- The MIT License (Oct 25, 2026)