Testing

Setup

cd ~/capcat && source venv/bin/activate
pip install -e ".[dev]"

Running Tests

# All unit tests
pytest tests/unit/ -v

# Single file
pytest tests/unit/test_feed_discovery.py -v

# Single test
pytest tests/unit/test_feed_discovery.py::test_validate_rss_feed -v

# With coverage
pytest tests/unit/ --cov=capcat --cov-report=term-missing

Test Structure

tests/
└── unit/                    # Unit tests (pytest)
    ├── test_cli.py
    ├── test_feed_discovery.py
    ├── test_feed_parser.py
    ├── test_session_pool.py
    ├── test_bundle_expansion.py
    ├── test_tui_completion.py
    └── ...

All tests live under tests/unit/. Integration and end-to-end tests are manual (see docs/troubleshooting.md).

TDD Order

1. Write a failing test that describes the expected behavior
2. Run it to confirm it fails (and fails for the right reason)
3. Write the minimal implementation to make it pass
4. Run tests to confirm pass
5. Commit test + implementation together

Test Design Rules

• One behavior per test function
• Test names describe the behavior: test_validate_rss_feed, test_reject_garbage
• Use pytest.raises for expected exceptions
• Mock at boundaries: HTTP calls, filesystem, sys.exit
• Do not mock the production logic under test
• Avoid time-dependent tests; use tmp_path fixture for filesystem tests

Coverage Target

80%+ line coverage on capcat/ before releasing. Check with:

pytest tests/unit/ --cov=capcat --cov-report=term-missing

Regression Tests

Each fixed bug must have a corresponding regression test committed before or with the fix. Name the test file after the component: test_<module>.py.

Current regression coverage:

• test_feed_discovery.py - feedparser-based validate_feed() (lxml removed)
• test_feed_parser.py - FeedParserFactory without lxml
• test_session_pool.py - Accept-Encoding excludes br
• test_bundle_expansion.py - bundle all / --all expands to source IDs
• test_tui_completion.py - TUI post-completion screen behavior