Python Testing Patterns

Write tests that make behavior safer to change. Prefer focused, readable tests over clever fixtures or coverage theater.

For advanced patterns such as async tests, monkeypatching, property-based tests, database testing, CI, and coverage tooling, read references/advanced-patterns.md.

Core Principles

• One behavior per test. A failure should point to one broken behavior.
• Arrange / Act / Assert. Keep setup, execution, and verification distinct.
• Test boundaries and errors. Happy-path tests alone are weak.
• Isolate external effects. Mock network, time, randomness, and services when they are not the behavior under test.
• Use meaningful coverage. Cover important behavior and risks, not every line.

Basic Pytest Pattern