coverage

Coverage.py is the de facto standard for Python code coverage measurement, and for good reason. The CLI interface is intuitive with `coverage run`, `coverage report`, and `coverage html` commands that do exactly what you'd expect. Configuration via `.coveragerc` or `pyproject.toml` is straightforward with well-documented options for source paths, omit patterns, and branch coverage.

The HTML reports are particularly excellent - they provide clear line-by-line visualization of what's covered and what's not, with color-coding that makes it immediately obvious where gaps exist. The tool integrates seamlessly with pytest via pytest-cov, and the data file format allows combining coverage from multiple test runs, which is invaluable for complex test suites.

Error messages are generally clear, though path resolution issues can occasionally be confusing when dealing with complex project structures. The documentation is comprehensive with good examples for common scenarios like parallel execution, subprocess coverage, and CI integration.

Coverage.py is the de facto standard for Python code coverage, and for good reason. The CLI is intuitive with sensible defaults - just run `coverage run -m pytest` followed by `coverage report` and you're immediately productive. The HTML reports are particularly well-done, with clear visual indicators and line-by-line highlighting that makes identifying untested code trivial.

Configuration via `.coveragerc` or `pyproject.toml` is straightforward and well-documented. I especially appreciate the ability to exclude specific lines with `# pragma: no cover` comments and the pattern-based path exclusions for ignoring test files or migrations. The `coverage combine` command handles parallel test execution elegantly, which is crucial for CI/CD pipelines.

Error messages are clear and actionable. When coverage drops below thresholds set with `--fail-under`, you get immediate feedback. The integration with pytest via pytest-cov is seamless, though using coverage.py directly gives you more control. Documentation is comprehensive with practical examples for common scenarios like branch coverage and partial line coverage.

Coverage.py has been my go-to tool for measuring test coverage across dozens of Python projects. The initial setup is dead simple - just `pip install coverage` and `coverage run -m pytest` gets you started. The HTML reports are exceptionally clear, showing exactly which lines weren't covered with color-coded highlighting that makes gaps obvious at a glance.

The configuration through `.coveragerc` or `pyproject.toml` is well-documented and intuitive. Common tasks like excluding virtual environments, ignoring specific lines with `# pragma: no cover`, or measuring branch coverage are straightforward. Error messages are helpful - when I misconfigured source paths, it told me exactly what directories it was looking for.

Integration with pytest through pytest-cov is seamless, and the tool handles edge cases well (multiprocessing, subprocess coverage). When debugging why coverage seems wrong, the `coverage debug` command provides useful diagnostics. The documentation includes real-world examples and the maintainer is incredibly responsive on GitHub issues - I've had questions answered within hours with clear explanations.