Telegram Chat PDF Processor

Convert your Telegram chat exports to optimized PDF files ready for AI processing and vector databases.

🚀 Quick Start (Windows)

1. Clone the repository:

   git clone [email protected]:rldyourmnd/telegram_to_pdfVectorDB.git
   cd telegram_to_pdfVectorDB

2. Export your Telegram chats:

   • Open Telegram Desktop
   • Go to Settings → Advanced → Export Telegram data
   • Select "Personal chats" and "Machine-readable JSON"
   • Save the export as result.json in the project folder

3. Run the processor:

   • Windows: Double-click launch_windows.bat
   • Linux: Run ./launch_linux.sh
   • First-time setup: Use install_and_run_windows.bat (installs Python automatically)
   • The script will automatically install dependencies and process your chats

📁 Output Structure

project/
├── chats_clean_pdf/          # Generated PDF files
├── metadata/                 # Processing metadata
│   └── metadata_summary.json
├── result.json              # Your Telegram export
└── launch_windows.bat       # Easy launcher

⚙️ Configuration

The tool works out-of-the-box, but you can customize settings by creating a .env file:

# Copy the example configuration
cp .env.example .env

# Edit with your settings
notepad .env  # Windows
nano .env     # Linux/macOS

Key settings to configure:

# IMPORTANT: Replace with your actual Telegram data
USER_NAME=Your Actual Telegram Name
USER_ID=user123456789

# PDF optimization (default values work well)
MAX_FILE_SIZE_KB=200
PDF_FONT_SIZE=10
PDF_LINE_SPACING=12

# Chunking algorithm (recommended defaults)
CHUNK_SIZE_SHORT=25
CHUNK_SIZE_MEDIUM=18
CHUNK_SIZE_LONG=12

# Processing options
MIN_MESSAGE_LENGTH=2
SHORT_MESSAGE_THRESHOLD=50
LONG_MESSAGE_THRESHOLD=150

# Debug options
VERBOSE_LOGGING=true
SHOW_PROGRESS=true

🔍 Finding Your User Information

To correctly identify your messages vs received messages:

1. Open result.json in any text editor
2. Search for a message you wrote (recognize by your writing style)
3. Find these fields in your message:

   "from": "Your Actual Name",
   "from_id": "user123456789"

4. Copy exact values to your .env file
5. Test: If messages still show as "From [Name]:" instead of "Me:", check your settings

🔧 Manual Installation

If you prefer manual setup:

# Install Python dependencies
pip install -r requirements.txt

# Run the processor
python process_telegram_chats.py

📊 Features

• Optimized for AI: PDFs sized for vector databases (max 200KB by default)
• Smart chunking: Dynamic chunk sizing based on message length
• Memory efficient: Processes large chats in parts
• Clean formatting: Optimized text format for AI processing
• Metadata tracking: Complete processing information
• Cross-platform: Works on Windows, macOS, and Linux

🤖 n8n Integration

Perfect for n8n workflows:

1. Text Splitter settings:

   • Chunk size: 800
   • Overlap: 200

2. Batch processing: 5-8 files at a time for optimal memory usage

3. Search patterns:

   • Me: for your messages
   • From [NAME]: for contact messages

4. Embedding models:

   • OpenAI: text-embedding-ada-002 (1536 dimensions)
   • Local: Any 768-dimension model

📋 Processing Statistics

The processor provides detailed statistics:

• Total chats processed
• Messages per chat
• PDF files created
• Chunk distribution
• Large chats split into multiple parts

🛠️ Requirements

• Python 3.7+
• Windows/macOS/Linux
• Telegram Desktop (for export)

🚨 Troubleshooting

Problem: Messages show as "From [Your Name]:" instead of "Me:"

Solution: Configure your user identification in .env file:

1. Copy .env.example to .env
2. Open result.json and find a message you sent
3. Look for "from": "Your Name" and "from_id": "user123456789"
4. Update .env with these exact values:

   USER_NAME=Your Exact Telegram Name
   USER_ID=user123456789

Problem: "Python not found" error

Solutions:

• Windows: Use install_and_run_windows.bat (auto-installs Python)
• Manual: Download Python from python.org and check "Add to PATH"
• Linux: sudo apt install python3 python3-pip

Problem: "No such file 'result.json'"

Solution: Export your Telegram data correctly:

1. Telegram Desktop → Settings → Advanced → Export Telegram data
2. Select "Personal chats" and "Machine-readable JSON"
3. Save as result.json in the project folder

Problem: Large files / GitHub limits

The .gitignore protects against committing:

• result.json (your personal chat data)
• Generated PDFs (chats_clean_pdf/)
• Configuration files (.env)

Problem: Processing fails or crashes

1. Check if result.json is valid JSON (not corrupted)
2. Try with VERBOSE_LOGGING=true in .env
3. Ensure enough disk space (chat exports can be large)
4. For very large exports, process in smaller batches

🔐 Privacy & Security

• Your data stays local - no data is sent anywhere
• Git protection - .gitignore prevents accidental data commits
• Configuration files - Never commit .env files with personal data
• Generated PDFs - Review before sharing (contain your chat history)

📝 License

MIT License - feel free to use and modify!