Python Backend

🌐 Translations & Contributions: Want to translate or improve this document in your language? See our Translation Guidelines.

The Python backend provides build orchestration for Python projects.

Detection

The Python backend is detected when a pyproject.toml or setup.py file is present.

Configuration

Configure the Python backend via fish.toml in your project or workspace root:

[build]
backend = "py"
jobs = 4

[pipelines.build]
inputs = ["**/*.py", "pyproject.toml", "requirements.txt"]
outputs = ["dist/*"]

[pipelines.test]
depends_on = ["build"]
inputs = ["tests/**/*.py", "**/*.py"]

Supported Package Managers

• uv: Fast Python package installer
• poetry: Modern Python packaging and dependency management
• pip: Standard Python package installer

Tasks Generated

Build Task

uv build
# or
poetry build
# or
pip install -e .

Test Task

uv run pytest
# or
poetry run pytest
# or
pytest

Lint Task

uv run ruff check
# or
poetry run ruff check

Dependency Extraction

The Python backend extracts dependencies from:

• pyproject.toml dependencies
• requirements.txt files
• setup.py setup_requires
• Poetry lock file

Fingerprinting

The Python backend fingerprints:

• pyproject.toml content
• requirements.txt content
• Source files (excluding venv/)
• Python version

Examples

Basic Python Project

cd my-python-project
fish build

Poetry Project

cd my-poetry-project
fish build

Project with Tests

cd my-python-project
fish test

Performance Optimization

The Python backend uses:

• uv for speed: Uses uv for fast dependency management
• Virtual environment isolation: Clean build environments
• Dependency caching: Efficient dependency resolution

Troubleshooting

Python not found

Install Python from python.org or use pyenv.

uv not found

Install uv: pip install uv

Poetry not found

Install Poetry: curl -sSL https://install.python-poetry.org | python3 -

Cache not working

Clear cache: Fish cache prune and rebuild.