testing

Unit tests in pytest are functions prefixed with `test_` that verify individual components. Write assertions to check expected behavior, use fixtures for setup and teardown, and run `pytest` to execute your test suite. This isolates code units and catches regressions early.

Unit tests verify individual functions or classes in isolation using mocks to stub dependencies. Integration tests verify multiple components working together with real or near-real interactions. Both are essential: unit tests catch logic bugs fast, integration tests validate system behavior end-to-end.

Fixtures provide reusable test setup via `@pytest.fixture` decorated functions; pass them as test arguments. Mocking isolates code by replacing real dependencies with controlled substitutes using libraries like `unittest.mock`. This enables testing specific logic without external services or databases.

Test coverage reports show what percentage of your code is executed by tests using tools like `coverage.py`. Run `coverage run -m pytest` then `coverage report` to identify untested branches and gaps. High coverage reduces the risk of bugs reaching production.

Property-based testing with Hypothesis generates random inputs to verify system invariants and edge cases automatically. Instead of writing specific test cases, you define properties that must always hold true. Use it for mathematical functions, data transformations, and algorithms to catch unexpected failures.

Configure your CI/CD pipeline to run `pytest` and coverage checks on every commit or pull request. Fail the build if tests don't pass or coverage drops below a threshold. This prevents untested code from merging and ensures code quality gates are enforced automatically.