python-testing

Pytest fixtures manage setup and teardown across function, class, module, and session scopes defined in conftest.py. Define fixtures with @pytest.fixture decorator, specify scope parameter, and use them as test function arguments to automate initialization and cleanup workflows.

Running tests via `uv run pytest` ensures consistent isolated environments and dependency resolution. UV manages Python version and package isolation, preventing environment drift and making test results reproducible across machines.

Async testing uses pytest plugins like pytest-asyncio to handle coroutines. Mark async test functions with @pytest.mark.asyncio, define async fixtures with scope parameters, and pytest executes them in the event loop without manual setup.

Pytest parametrization uses @pytest.mark.parametrize decorator to run a test function across multiple input sets defined as tuples or lists. This reduces code duplication and provides clear test reports showing each input combination's result.

Configure coverage in pyproject.toml or .coveragerc with pytest-cov plugin. Specify source directories, omit patterns, and branch coverage settings; run `uv run pytest --cov` to generate coverage reports and monitor test quality.

Use unittest.mock.patch or pytest fixtures with monkeypatch to isolate code under test. Mock external services, APIs, and dependencies as fixtures scoped appropriately, injecting them into test functions to verify behavior without side effects.