Testing

Testing#

Running tests locally#

Tests require Docker. QLever containers are managed automatically by pytest fixtures in test/conftest.py. Counter handling uses temporary directories with FilesystemCounterHandler (no Redis needed for counters).

1. Install dependencies#

uv sync

2. Run tests#

uv run coverage run --rcfile=test/coverage/.coveragerc

Docker containers start automatically at the beginning of the test session and stop when tests complete.

Test infrastructure:

  • QLever data triplestore on port 8805

  • QLever provenance triplestore on port 8806

  • Temporary directories for filesystem-based counters

View coverage report:

uv run coverage report

Generate HTML report:

uv run coverage html -d htmlcov

Running specific tests#

Run a single test file:

uv run python -m pytest test/curator_test.py -v

Run tests matching a pattern:

uv run python -m pytest test/ -k "test_doi" -v

Test structure#

Tests are in the test/ directory:

File

Tests

curator_test.py

Data validation and normalization

creator_test.py

RDF generation

meta_process_test.py

End-to-end pipeline

editor_test.py

Post-processing modifications

finder_test.py

Entity lookup

group_entities_test.py

Merge grouping algorithm

Test fixtures use minimal datasets in test/ subdirectories.

CI/CD#

GitHub Actions workflow#

Tests run automatically on push and pull request via .github/workflows/run_tests.yml:

name: Run tests

on:
  push:
    branches: [master]
  pull_request:
    branches: [master]

jobs:
  test:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: ["3.10", "3.11", "3.12", "3.13"]

The workflow:

  1. Sets up Python with UV

  2. Installs dependencies

  3. Runs tests with coverage (Docker containers managed by pytest)

  4. Uploads coverage report

Test matrix#

Tests run on Python 3.10, 3.11, 3.12, and 3.13.

Writing tests#

Test naming#

Test files: *_test.py Test functions: test_*

Fixtures#

Use pytest fixtures for common setup:

@pytest.fixture
def counter_handler(tmp_path):
    return FilesystemCounterHandler(
        info_dir=str(tmp_path / "info_dir"),
        supplier_prefix="060"
    )

def test_counter_increment(counter_handler):
    initial = counter_handler.read_counter("br")
    counter_handler.increment_counter("br")
    assert counter_handler.read_counter("br") == initial + 1

Triplestore tests#

Tests that need SPARQL use the test QLever instance:

SPARQL_ENDPOINT = "http://localhost:8805"

def test_sparql_query():
    finder = ResourceFinder(ts=SPARQL_ENDPOINT, base_iri="https://w3id.org/oc/meta")
    # Test queries...

Cleanup#

Tests should clean up after themselves:

def test_with_cleanup(counter_handler):
    try:
        # Test code...
    finally:
        # Cleanup handled automatically via tmp_path fixture
        pass

Or use fixtures with cleanup:

@pytest.fixture
def temp_graph():
    g = Graph()
    yield g
    # Cleanup happens automatically after test

Coverage requirements#

Aim for high coverage on:

  • oc_meta.core.curator - Data validation logic

  • oc_meta.core.creator - RDF generation

  • oc_meta.lib.finder - Entity lookup

  • oc_meta.lib.cleaner - Identifier normalization

Lower coverage is acceptable for:

  • CLI scripts (tested via integration tests)

  • Error handling paths (hard to trigger in tests)