Book Keeper - PDF Quality Analyzer 📚🔍

Book Keeper v2.0 - A comprehensive AI-powered quality assurance tool that analyzes PDF documents for contradictions, content flow, redundancy, code quality, and theoretical accuracy.

✨ Features

Core Analysis Modules

1. 📚 Contradiction Detection: Identifies logical contradictions between chapters
2. 📊 Flow Analysis: Checks content progression and prerequisite violations
3. 🔁 Redundancy Detection: Finds duplicate or unnecessarily repeated content
4. 🐛 Code Validation: Validates code snippets for syntax and best practices
5. 📖 Theory Verification: Verifies against software engineering standards (SOLID, Design Patterns, etc.)
6. 📝 Terminology Consistency: Ensures consistent use of technical terms throughout the document

Technical Features

• 📄 Automatic PDF Chapter Extraction: Supports various chapter delimiter patterns
• 🧠 Vector Embeddings: Semantic text analysis using OpenAI Embeddings API
• 🗄️ Vector Database: Efficient similarity search powered by Qdrant
• 🤖 Dual LLM Support: Choose between Claude 3.5 Sonnet (default) or GPT-4o
• 📊 Comprehensive Reports: JSON and Markdown formats with quality scores
• 🌍 Multilingual Support: Works with PDFs in multiple languages

📋 System Requirements

• Python 3.10+
• Conda (Anaconda or Miniconda)
• Docker (for running Qdrant)
• API Keys:
  • ANTHROPIC_API_KEY (for Claude 3.5 Sonnet - default)
  • OPENAI_API_KEY (for GPT-4o - optional)

🚀 Quick Start

1. Clone the repository

git clone https://github.com/no-ai-labs/book-keeper.git
cd book-keeper

2. Set up the environment

# Install Conda (if not already installed)
./install_conda.sh  # macOS only

# Create and activate the environment
./setup.sh
conda activate book-keeper

3. Configure environment variables

cp .env_example .env
# Edit .env and add your API keys:
# - ANTHROPIC_API_KEY (for Claude - default)
# - OPENAI_API_KEY (for GPT-4o - optional)

4. Start Qdrant

docker-compose up -d

5. Place PDF files

Put your PDF files in the pdf/ directory.

📖 Usage

Comprehensive Analysis (Default)

Runs all quality checks:

python rag_pdf_checker.py

Specific Checks

Run only selected analyzers:

# Single check
python rag_pdf_checker.py --check contradiction

# Multiple checks
python rag_pdf_checker.py --check contradiction,flow,code

# Available checks: contradiction, flow, redundancy, code, theory, terminology

Test Mode

Limited analysis for testing (first 3 chapters only):

# Test with default model (Claude)
python rag_pdf_checker.py --test

# Test with GPT-4o
python rag_pdf_checker.py --test --openai

Model Selection

# Use Claude 3.5 Sonnet (default)
python rag_pdf_checker.py

# Use GPT-4o
python rag_pdf_checker.py --openai
# or
python rag_pdf_checker.py --gpt

Custom PDF Directory

python rag_pdf_checker.py --pdf-dir /path/to/pdfs

📊 Output

The tool generates two report files:

1. quality_report_v2.json

Detailed JSON report with all findings and scores.

2. quality_report_v2.md

Beautiful Markdown report with:

• 🎯 Overall quality score (0-100%)
• 📈 Individual module scores
• 📋 Detailed findings by category
• 🔍 Key insights summary

Quality Scoring

• 90-100%: Excellent ✅
• 80-89%: Good ⚠️
• 70-79%: Fair ⚠️
• 60-69%: Needs Improvement ❌
• Below 60%: Poor ❌

🗂️ Project Structure

book-keeper/
├── rag_pdf_checker.py      # Main v2.0 application
├── analyzers/              # Analysis modules
│   ├── base.py            # Base analyzer class
│   ├── contradiction.py   # Contradiction detector
│   ├── flow.py           # Content flow analyzer
│   ├── redundancy.py     # Redundancy detector
│   ├── code.py           # Code validator
│   └── theory.py         # Theory verifier
├── pdf/                   # PDF files directory
├── environment.yml        # Conda environment
├── requirements.txt       # Python dependencies
├── docker-compose.yml     # Qdrant setup
└── quality_report_v2.*    # Generated reports

🔧 Advanced Settings

Environment Variables (.env)

# API Keys
ANTHROPIC_API_KEY=your_key_here
OPENAI_API_KEY=your_key_here

# Qdrant Configuration
QDRANT_HOST=localhost
QDRANT_PORT=6345
QDRANT_API_KEY=optional_key

Docker Compose Configuration

The docker-compose.yml configures Qdrant with:

• REST API on port 6345
• gRPC on port 6346
• Persistent storage in ./data/qdrant
• Health checks
• Auto-restart

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Built with Claude 4 Opus and Claude 3.5 Sonnet
• Powered by OpenAI Embeddings
• Vector search by Qdrant