Linux Kernel Patch Statistic among Universities

This project analyzes the Linux kernel git history to rank universities by their patch contributions. It matches commit author email domains against a university domain list, aggregates statistics (patch count, lines changed, contributors), and publishes the results as a static website via GitHub Pages.

Quick Start

# Install PDM (Linux/macOS)
curl -sSL https://pdm-project.org/install-pdm.py | python3 -

# Install dependencies
pdm install

# Clone the Linux kernel repo (full history required)
git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git

# Generate results
pdm start

The output is result.json, result.js, and localized paginated HTML detail pages in detail/<locale>/. Serve index.html with any web server to view the rankings, or open it directly as a local file after generation.

Internationalization

The website supports en, zh-CN, zh-TW, ja, and ko. Use the language selector in the page header, or open the page with ?lang=<locale> such as ?lang=zh-CN.

University names, author names, domains, emails, and commit summaries remain in their source language. The localized content covers the website UI and generated detail-page navigation.

Requirements

• Python 3.9+
• PDM package manager
• A local clone of the Linux kernel repository (with full git history)

Development

pdm install -G dev       # Install with dev dependencies (pylint, pytest)
pdm lint                 # Run pylint (threshold 9.0)
pdm test                 # Run all tests

How It Works

1. Fetches the university domain list from Hipo/university-domains-list
2. Iterates through all git commits, matching author email domains to universities (with parent domain fallback, e.g. cs.mit.edu matches mit.edu)
3. Aggregates per-domain statistics and merges aliases for the same university
4. Ranks universities by patch count (tiebreak: total lines changed)
5. Writes result.json and generates localized paginated HTML detail pages

CI/CD

• Pylint CI: Runs on every push/PR to master across Python 3.9-3.13
• Page Deployment: A daily scheduled workflow regenerates the data and deploys to GitHub Pages

Documentation

See the docs/ folder for more details:

• Architecture - Project structure and data pipeline
• Data Format - Schema of result.json

License

BSD-2-Clause