python-testing-patterns

Write test functions prefixed with test_ that follow the Arrange-Act-Assert pattern, then run them with the pytest command. Use assert statements directly and pytest.raises to verify exceptions are raised with expected messages.

Use unittest.mock's patch to replace requests.get or requests.post with Mock objects that return controlled responses. Configure mock_response.json.return_value and verify calls with assert_called_once_with to test API clients without network access.

Fixtures provide setup and teardown for test resources like database connections using yield, while mocking replaces external dependencies with controlled fakes. Fixtures manage real test infrastructure; mocks simulate behavior of external services.

Yes, pytest supports async tests through the pytest-asyncio plugin using the @pytest.mark.asyncio decorator on async test functions. Async fixtures can also be defined with yield for setup and teardown of asynchronous resources.

Install pytest-cov and run pytest with --cov=yourpackage to measure coverage. Add --cov-report=term-missing to see uncovered lines and --cov-fail-under=80 to enforce a minimum coverage threshold in CI pipelines.

Use property-based testing with hypothesis when you want to verify invariants across many generated inputs, such as reversing a string twice returns the original. It complements example-based tests by discovering edge cases you did not anticipate.