python-testing-patterns

Define a function decorated with @pytest.fixture that performs setup, yields the resource to the test, then runs teardown code after the yield. Use scope parameters like session or module to control how often the fixture is created.

Use unittest.mock's patch to replace requests.get or requests.post with a Mock object whose return_value simulates the HTTP response. Then assert on the mock's call arguments to verify your client made the correct request.

Unit tests verify individual functions or classes in isolation, often with mocked dependencies, while integration tests exercise interactions between components such as APIs and databases. Pytest markers like @pytest.mark.integration let you run each group separately.

Yes, async test functions are supported by marking them with @pytest.mark.asyncio and awaiting the coroutine under test. Async fixtures can also yield resources with cleanup after the test completes.

Running the entire suite is slow and includes tests unrelated to your changes. Scope runs to the domain under modification or add -m "not slow" to skip slow-marked tests, matching how CI executes most jobs.

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