pytest

pytest standardizes Python unit testing by organizing tests in a tests/ directory using simple assert statements. Install pytest via pip, write test functions prefixed with test_, and run pytest to execute them. It requires no boilerplate setup compared to other frameworks.

Fixtures are reusable setup and teardown functions that prepare test data and clean up resources before and after tests run. Define them with @pytest.fixture decorator and pass them as parameters to test functions to reduce code duplication across tests.

pytest parametrization lets you run the same test with different inputs using @pytest.mark.parametrize. This covers multiple scenarios in one test function, reducing code and improving maintainability across edge cases.

pytest supports mocking and patching via the unittest.mock module and pytest plugins. Isolate units by replacing external calls with mock objects, verify they were called correctly, and test behavior in isolation without hitting real APIs or databases.

pytest supports async testing through the pytest-asyncio plugin. Mark async test functions with @pytest.mark.asyncio and pytest handles event loop setup, enabling you to test async/await code and coroutines directly.

pytest markers like @pytest.mark.slow or @pytest.mark.integration categorize tests for selective execution. Run specific groups with pytest -m marker_name, enabling faster feedback cycles by skipping expensive tests during development.