Course contentsShow
← Python
Lesson 542 of 1,0608. Professional Tooling and PackagingFree lesson

Automating Uploads with Continuous Integration

Configure GitHub Actions or similar CI pipelines to automatically build and publish new package versions upon tagging a release.

In this lesson, you will learn how to configure a GitHub Actions workflow to automatically build and publish your Python package to PyPI whenever you create a formal release tag in your repository.

The Role of Automation

Manually building wheels with build and uploading them via twine is prone to human error—you might forget to increment your version number or upload an outdated artifact. By using Continuous Integration (CI), you turn your distribution process into a repeatable, automated pipeline that triggers only when you are ready to ship.

Configuring the Workflow

GitHub Actions uses YAML files stored in .github/workflows/ to define these tasks. Since we are using modern standards like pyproject.toml and PEP 517 build backends, we can use the official pypa/gh-action-pypi-publish action. This action integrates seamlessly with PyPI’s "Trusted Publishers" mechanism, removing the need to store long-lived passwords in your repository settings.

Here is a simplified example of what that workflow configuration looks like:

name: Publish Package
on:
  release:
    types: [published]

jobs:
  pypi-publish:
    runs-on: ubuntu-latest
    permissions:
      id-token: write  # Required for Trusted Publishers
    steps:
      - uses: actions/checkout@v4
      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'
      - name: Install build tool
        run: pip install build
      - name: Build package
        run: python -m build
      - name: Publish to PyPI
        uses: pypa/gh-action-pypi-publish@release/v1

How It Works

  1. Trigger: The workflow listens for the release event. When you create a new release on GitHub, the pipeline wakes up.
  2. Environment: It spins up a clean, isolated environment to ensure the build is reproducible.
  3. Build: It runs python -m build, which utilizes your pyproject.toml to create your source distribution and wheel.
  4. Publish: The gh-action-pypi-publish action handles the authentication using an OIDC token and sends your artifacts to PyPI automatically.

This ensures that the exact code you "tagged" as a release is the exact code that gets distributed to users, maintaining a clean and professional release history.

Key Takeaway: Automating your upload process with CI ensures consistent, verifiable releases by removing manual build steps and leveraging secure authentication methods.