lumi-dataset-hub 0.2.0

High-performance dataset caching and streaming client bindings for Python

lumi-dataset-hub

High-performance dataset caching and streaming client bindings for Python.

---

Release Automation & Publishing

This package uses automated GitHub Workflows to compile and release multi-platform binary wheels to PyPI. The pipeline is secured using OIDC Trusted Publishing, meaning no passwords or static API tokens are stored on GitHub.

How to Release a New Version

1. Increment Version: Update the version number in pyproject.toml (e.g., version = "0.2.1") and Cargo.toml.
2. Commit and Tag:

   git add pyproject.toml Cargo.toml
   git commit -m "bump: version 0.2.1"
   git tag v0.2.1
   

3. Push to GitHub:

   git push origin main
   git push origin v0.2.1
   

4. Pipeline Execution:
   • Pushing the tag automatically triggers the Release Python Bindings workflow.
   • The workflow compiles wheels for Linux (manylinux x86_64, aarch64), macOS (x86_64, aarch64), and Windows (x86_64), and builds the source distribution (sdist).
   • After a successful build, the job requests a federated OIDC token from PyPI's trusted publisher interface and deploys all distributions seamlessly.

PyPI Trusted Publisher Configuration Reference

For administrators setting up publishing credentials, the following mappings are configured under PyPI Account Settings:

• Provider: GitHub
• Owner: Organization or Username (e.g., rd-lumi)
• Repository: dataset-caching
• Workflow Name: release.yml
• Environment Name: pypi
• PyPI Project: lumi-dataset-hub