pytest-coverage

Pytest Coverage Improvement

Run pytest with coverage analysis and iteratively add tests until all lines are covered.

Instructions

Step 1: Generate Coverage Report

Run pytest with annotated coverage output:

pytest --cov --cov-report=annotate:cov_annotate

To scope to a specific module:

pytest --cov=<module_name> --cov-report=annotate:cov_annotate

To run specific test files:

pytest tests/test_<module>.py --cov=<module_name> --cov-report=annotate:cov_annotate

Step 2: Identify Uncovered Lines

Open the cov_annotate/ directory. Each source file has a corresponding annotated file:

• Lines starting with ! (exclamation mark) are not covered by tests.
• Files with 100% coverage can be skipped.

Step 3: Add Missing Tests

For each file with less than 100% coverage:

1. Open the annotated file and identify all !-marked lines.
2. Determine what test cases would exercise those code paths:
   • Conditional branches (if/elif/else)
   • Exception handlers (try/except)
   • Edge cases (empty inputs, None values, boundary conditions)
   • Error paths (invalid arguments, missing data)
3. Write focused tests that specifically target the uncovered lines.
4. Follow existing test conventions in the project (fixtures, naming, structure).

Step 4: Verify and Iterate

Re-run the coverage command after adding tests. Repeat steps 2-3 until all lines are covered.

Step 5: Clean Up

After reaching target coverage:

• Remove the cov_annotate/ directory (or add it to .gitignore)
• Ensure all new tests pass independently: pytest tests/ -v

Tips

• Prefer testing behavior over implementation details.
• Use pytest.raises() to cover exception paths.
• Use @pytest.mark.parametrize to cover multiple branches concisely.
• Mock external dependencies (database, APIs) to test error-handling paths.
• If a line is genuinely unreachable, mark it with # pragma: no cover.