A digital archive that documents and organizes historical aggression events that occurred in Lebanon between 1949 and 1985. This project transforms the contents of the original book:
“لبنان 1949–1985: الاعتداءات الإسرائيلية”
into a searchable, filterable, and user-friendly web platform.
🌐 Live site: https://lebanon-aggressions-archive.netlify.app/
- 🔎 Full-text search across all events
- 🗂️ Filter by year, month, exact date, or date range
- 📅 Beautiful date picker for easier filtering
- 📋 Copy-to-clipboard for quick sharing
- 📨 Correction request tool (RequestFix)
- 📄 Export database (JSON)
- 📱 Fully responsive design with a mobile-first search UI
- 🚀 Modern React + Tailwind architecture
- Digitize an important historical reference
- Make research and fact-checking accessible
- Preserve Lebanese historical documentation
- Enable fast navigation across decades of recorded events
The dataset is based on the published book:
“لبنان 1949–1985: الاعتداءات الإسرائيلية”
Copyright and intellectual rights belong to the original authors and publishers.
The initial digitization of the book relied on a companion project:
GitHub: https://github.com/HassanMSh/arabic-chrono-dots-ocr
This system performs OCR on Arabic chronological text using:
-
dots.ocr-4bit (Quantized Version) https://huggingface.co/helizac/dots.ocr-4bit
-
Original dots.ocr Model https://huggingface.co/rednote-hilab/dots.ocr
This OCR pipeline generated the base text used to build the structured dataset powering this archive.
- React (Vite)
- Tailwind CSS
- Netlify (Deployment)
- Modern hooks, responsive layout, and clean component structure.
npm install
npm run dev # start the local dev server
npm run build # build for production into dist/
npm run lint # run ESLint
npm run check:data # validate public/events.jsonChanges go through pull requests. GitHub Actions runs lint, the data check and the build on every pull request and on every push to main.
Notes for AI coding agents and contributors are in AGENTS.md.
For improvements, corrections, or contributions: hassan.m.shamseddine@gmail.com
This project is open-source under the MIT License. The dataset remains the property of its original creators.