Hatchling PyO3 Plugin¶

A Hatchling build hook plugin for building PyO3 Rust extensions

Documentation: https://frankie567.github.io/hatchling-pyo3-plugin/

Source Code: https://github.com/frankie567/hatchling-pyo3-plugin

Overview¶

PyO3 is a popular framework for creating Python bindings to Rust code. While Maturin is the recommended build backend for PyO3 projects, this plugin enables PyO3 extensions to be built using Hatchling as an alternative build backend.

This is inspired by setuptools-rust, which provides similar functionality for setuptools.

Features¶

✅ Build PyO3 Rust extensions as part of the Hatchling build process
✅ Automatic detection of Cargo.toml
✅ Cross-platform support (Linux, macOS, Windows)
✅ Configurable build profiles (release/debug)
✅ Custom Cargo arguments support
✅ Integration with standard Python packaging workflow

Usage¶

Installation¶

Add the plugin to your pyproject.toml:

[build-system]
requires = ["hatchling", "hatchling-pyo3-plugin"]
build-backend = "hatchling.build"

[tool.hatch.build.hooks.pyo3]
# Plugin will automatically detect and build Rust extensions

Project Structure¶

my-project/
├── pyproject.toml
├── Cargo.toml          # Rust project configuration
├── src/
│   └── lib.rs          # Rust source code with PyO3 bindings
└── my_package/
    └── __init__.py     # Python package

Example pyproject.toml¶

[build-system]
requires = ["hatchling", "hatchling-pyo3-plugin"]
build-backend = "hatchling.build"

[project]
name = "my-pyo3-project"
version = "0.1.0"
description = "A project with PyO3 extensions"
requires-python = ">=3.8"

[tool.hatch.build.hooks.pyo3]
# Optional: specify Rust extensions explicitly
# If not specified, plugin will look for Cargo.toml in project root

# Optional configuration:
# cargo-manifest = "Cargo.toml"  # Path to Cargo.toml (default: "Cargo.toml")
# profile = "release"            # Build profile (default: "release", can be "debug")
# target-dir = "target"          # Cargo target directory (default: "target")
# cargo-args = ["--features", "special"]  # Additional cargo arguments

Development¶

Setup environment¶

We use uv to manage the development environment and production build, and just to manage command shortcuts. Ensure they are installed on your system.

Run unit tests¶

You can run all the tests with:

just test

Format the code¶

Execute the following command to apply linting and check typing:

just lint

Publish a new version¶

You can bump the version, create a commit and associated tag with one command:

just version patch

just version minor

just version major

Your default Git text editor will open so you can add information about the release.

When you push the tag on GitHub, the workflow will automatically publish it on PyPi and a GitHub release will be created as draft.

Serve the documentation¶

You can serve the Mkdocs documentation with:

just docs-serve

It'll automatically watch for changes in your code.

License¶

This project is licensed under the terms of the MIT license.