python-testing

Maintainable pytest fixtures prevent shared state bugs by composing scoped fixtures and sharing them via conftest.py. This approach removes duplication and isolates test state, ensuring reliable test execution across your suite.

The best way to mock external dependencies in pytest is using pytest-mock-style workflows for patching, mocking exceptions, and handling context managers. This keeps tests fast, isolated, and non-brittle by avoiding real external calls.

pytest-asyncio handles async testing by enabling clear execution of async test functions and async fixtures. It supports async function mocking assertions, ensuring your async test execution remains correctly aligned and robust.

Yes, you can use @pytest.mark.parametrize alongside custom markers for marker-based filtering. This allows you to categorize and selectively run unit, integration, or slow tests, improving test selection and execution speed.

Apply TDD red-green-refactor by writing failing tests first, making them pass, then refactoring. Measure coverage with pytest --cov during this flow to tie tests to invariants and risk surfaces, ensuring robust coverage.

Flaky pytest tests often stem from shared state bugs or improper mocking. Diagnose them by reviewing fixture scope composition, ensuring correct pytest configuration, and verifying that mocking best practices are followed for external calls.