feat: Add comprehensive Python testing infrastructure with Poetry

[llbbl]

Add Python Testing Infrastructure

Summary

This PR sets up a comprehensive testing infrastructure for the Graph Transformer project. While the project uses Conda for its main dependencies, I've added Poetry alongside it specifically for managing testing and development tools, providing a modern Python testing environment.

Changes Made

Package Management

• Added pyproject.toml with Poetry configuration
• Configured Poetry to work alongside the existing Conda environment setup
• Note: DGL and other conda-specific packages remain managed by Conda

Testing Dependencies

• pytest: Main testing framework
• pytest-cov: Coverage reporting with 80% threshold
• pytest-mock: Mocking utilities for unit tests

Testing Configuration

• Configured pytest settings in pyproject.toml:
  • Test discovery patterns for test_*.py and *_test.py
  • Coverage reporting (terminal, HTML, XML)
  • Custom markers: unit, integration, slow
  • Strict mode with detailed output
  • Coverage fail threshold set to 80%

Directory Structure

tests/ ├── __init__.py ├── conftest.py # Shared fixtures ├── test_setup_validation.py # Validation tests ├── unit/ │ └── __init__.py └── integration/ └── __init__.py 

Shared Fixtures (conftest.py)

• temp_dir: Temporary directory management
• mock_config: Configuration dictionary for tests
• sample_graph_data: Graph data structures for testing
• sample_molecular_data: Molecular data for testing
• mock_model: PyTorch model mock
• mock_dataloader: DataLoader mock
• sample_checkpoint: Model checkpoint fixture
• device: CPU device for consistent testing
• random_seed: Reproducible random seeds
• cleanup_cuda: Automatic CUDA cache cleanup
• mock_tensorboard_writer: TensorBoard writer mock

Development Tools

• Added linting (flake8), formatting (black), type checking (mypy)
• Updated .gitignore with testing artifacts and build files
• Added Poetry lock file to gitignore (as requested)

How to Use

Initial Setup

# Install Poetry if not already installed curl -sSL https://install.python-poetry.org | python3 - # Install testing dependencies poetry install # For the main project dependencies, continue using Conda: conda env create -f environment_cpu.yml # or environment_gpu.yml

Running Tests

# Run all tests poetry run test # or poetry run tests # Run with specific markers poetry run pytest -m unit poetry run pytest -m integration poetry run pytest -m "not slow" # Run with coverage report poetry run pytest --cov-report=html

Writing Tests

1. Place unit tests in tests/unit/
2. Place integration tests in tests/integration/
3. Use the provided fixtures from conftest.py
4. Mark tests appropriately with @pytest.mark.unit, @pytest.mark.integration, or @pytest.mark.slow

Notes

• The project maintains its original Conda-based dependency management for scientific packages
• Poetry is used specifically for testing and development tools
• DGL (Deep Graph Library) must still be installed via Conda as it's not available on PyPI
• The validation test file (test_setup_validation.py) verifies the testing setup works correctly

Next Steps

Developers can now immediately start writing tests for the codebase using this infrastructure. The 80% coverage threshold will help maintain code quality as the project evolves.