feat: Set up comprehensive Python testing infrastructure with Poetry

[llbbl]

Set up Python Testing Infrastructure

Summary

This PR establishes a comprehensive testing infrastructure for the SeFa (Closed-Form Factorization of Latent Semantics in GANs) Python project. The setup provides a solid foundation for writing and maintaining tests, with Poetry for dependency management, pytest for testing, and coverage reporting.

Changes Made

Package Management

• Poetry Setup: Created pyproject.toml with Poetry configuration as the package manager
• Dependencies: Migrated core dependencies (torch, numpy, streamlit, tqdm) to Poetry
• Testing Dependencies: Added pytest, pytest-cov, and pytest-mock as development dependencies

Testing Infrastructure

• Directory Structure:

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

Configuration

• pytest Configuration in pyproject.toml:

  • Test discovery patterns for test_*.py and *_test.py files
  • Coverage reporting with HTML and XML outputs
  • Custom markers: @pytest.mark.unit, @pytest.mark.integration, @pytest.mark.slow
  • Strict mode with verbose output and short traceback format

• Coverage Settings:

  • Source directories: sefa, utils, interface
  • Excluded: test files, init.py, SessionState.py
  • Report formats: terminal, HTML, XML
  • Currently set to 0% threshold (to be increased as tests are added)

Test Fixtures (conftest.py)

• temp_dir: Temporary directory for test files
• mock_model_config: Mock configuration for model testing
• sample_latent_codes: Sample PyTorch tensors for latent code testing
• sample_numpy_array: Sample numpy arrays for data testing
• mock_generator_output: Mock generator model outputs
• device: PyTorch device fixture (CPU by default)
• test_data_dir: Temporary data directory with test files
• reset_torch_seed: Auto-fixture to ensure reproducible tests
• mock_streamlit_session: Mock Streamlit session state
• capture_stdout: Stdout capture for testing print statements
• Custom markers for GPU tests and model-dependent tests

Other Updates

• Updated .gitignore with:
  • Testing artifacts: .pytest_cache/, .coverage, htmlcov/, coverage.xml
  • Claude settings: .claude/*
  • Virtual environments and IDE files
  • Note: poetry.lock is intentionally NOT ignored

Running Tests

After pulling this branch, install dependencies and run tests:

# Install Poetry if not already installed
curl -sSL https://install.python-poetry.org | python3 -

# Install project dependencies
poetry install

# Run tests (both commands work)
poetry run test
poetry run tests

# Run specific test categories
poetry run pytest -m unit        # Run only unit tests
poetry run pytest -m integration # Run only integration tests
poetry run pytest -m "not slow"  # Skip slow tests

Validation

The setup includes validation tests (test_setup_validation.py) that verify:

• All dependencies are properly installed
• The testing infrastructure is correctly configured
• All fixtures work as expected
• Directory structure is properly created
• Coverage reporting is functional

All 20 validation tests pass successfully.

Next Steps

With this infrastructure in place, developers can now:

1. Write unit tests for individual model components in tests/unit/
2. Create integration tests for end-to-end workflows in tests/integration/
3. Add more specialized fixtures as needed
4. Gradually increase the coverage threshold as more tests are added

Notes

• The project uses Poetry for modern Python dependency management
• Coverage is configured but currently excludes model files from the report (can be adjusted later)
• The testing setup follows pytest best practices with clear separation of unit and integration tests
• All standard pytest features and plugins are available for use