PDF-TEI Editor

A comprehensive viewer/editor web application for comparing PDF sources with TEI extraction and annotation results, specifically designed for creating gold standard datasets of TEI documents from legal and humanities literature.

Key Features

• Dual-pane interface with synchronized PDF viewer and XML editor
• AI-powered extraction supporting multiple extraction engines (GROBID, etc.) via frontier and open LLMs
• Visual annotation mode — hide raw XML and annotate text with colored badges using a point-and-click interface
• Search-in-PDF navigation — automatically locates and highlights the PDF region corresponding to the cursor position in the XML editor
• Version management with branching, merging, and comparison tools
• Schema validation with automatic TEI compliance checking
• Access control with role-based permissions and collection management
• Collection organization for managing document sets
• Revision tracking with detailed change documentation
• Modular architecture - easily add new features and behavior with frontend and backend plugins

Target Use Cases

• Creating gold standard datasets for reference extraction research
• Manual validation and correction of AI-extracted bibliographic data
• Collaborative annotation of legal and humanities literature
• Training data preparation for machine learning models
• Quality assurance for large-scale digitization projects

🚀 Quick Start

Try with Docker (Fastest)

The fastest way to try PDF-TEI Editor using our deployment script:

# Clone the repository
git clone https://github.com/mpilhlt/pdf-tei-editor.git
cd pdf-tei-editor

# Deploy demo container (builds and starts automatically)
npm run deploy .env.deploy.demo.localhost

Then visit: http://localhost:8080

• Login: admin / admin or demo / demo

Or use Docker directly:

docker run -p 8000:8000 -e APP_ADMIN_PASSWORD=admin123 cboulanger/pdf-tei-editor:latest

Visit: http://localhost:8000 - Login: admin / admin123

📖 For detailed setup: See the Docker Deployment Guide

Development Setup

# Clone the repository
git clone https://github.com/mpilhlt/pdf-tei-editor.git
cd pdf-tei-editor

# Configure environment
cp .env.development .env

# Install dependencies
npm install

# Start development server
npm run start:dev

Visit: http://localhost:8000

📖 For complete setup instructions: See the Installation Guide

Documentation

📚 Complete Documentation

For End Users

• User Manual - How to use the application
• Getting Started - First-time user guide
• Interface Overview - Understanding the interface
• Workflows - Extraction, editing, and sync workflows

For Developers

• Developer Documentation - Architecture and development guides
• Architecture Overview - System design and structure
• Testing Guide - Running and writing tests

General reference

• Installation Guide - Development environment setup
• CLI - Command line interface

For Code Assistants

• Code Assistant Documentation - Concise technical guides

Technology Stack

• Backend: FastAPI (Python 3.13+), SQLite, lxml
• Frontend: ES6 modules, CodeMirror 6, PDF.js, Shoelace
• Testing: Playwright (E2E), pytest (backend), Node.js test runner (API)

License

See the project repository for license information.

Contributing

Developers interested in contributing should:

1. Read the Developer Documentation
2. Follow Coding Standards
3. Write tests following the Testing Guide
4. Submit pull requests with proper documentation

About

This repository is part of the "Legal Theory Knowledge Graph" project at the Max Planck Institute of Legal History and Legal Theory.

Related repositories:

• llamore
• bibliographic-tei

---

Questions or issues? Check the documentation or open an issue.