Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
78 lines (56 loc) · 1.82 KB

CONTRIBUTING.md

File metadata and controls

78 lines (56 loc) · 1.82 KB

Contributing to ConnectAI Python Client

Development Setup

# Clone the repo
git clone <repo-url>
cd connect-ai-python

# Install the connector in editable mode with dev dependencies
cd connector
pip install -e ".[dev]"

Running Tests

The test suite is split into unit and integration tests.

Unit Tests (fast, no server needed)

cd connector
pytest tests/unit/ -v

Unit tests use unittest.mock to mock HTTP calls. They run in milliseconds and verify class contracts, parameter validation, type conversion, and exception hierarchy.

Integration Tests (requires mock server)

cd connector
pytest tests/integration/ -v

The mock server (connect-ai-mock/) auto-starts when tests target localhost. You can also start it manually:

cd connect-ai-mock
pip install -r requirements.txt
python run.py

All Tests

cd connector
pytest tests/ -v

Against a Live Endpoint

CDATA_BASE_URL=https://cloud.cdata.com/api \
CDATA_USERNAME=you@example.com \
CDATA_PASSWORD=<pat> \
SKIP_LIVE_TESTS=0 \
pytest tests/integration/ -v

Adding New Tests

Test type Directory When to use
Unit tests/unit/ Testing class contracts, validation, type conversion — anything that doesn't need the server
Integration tests/integration/ Testing end-to-end behavior against the mock (or live) server

Conventions

  • Group related tests in classes (e.g., TestConnectionLifecycle)
  • Use fixtures from the appropriate conftest.py
  • Unit tests should never make network calls — mock all HTTP via unittest.mock
  • Integration tests should use the con and cursor fixtures from tests/integration/conftest.py

Code Style

  • Python 3.10+
  • Type hints on public APIs
  • Follow existing patterns in the codebase