Tibetan Autocomplete

A smart text editor with AI-powered autocomplete support for both Tibetan and English languages. The editor uses Claude AI (Anthropic's API) to provide context-aware sentence completions and detect gibberish text.

Features

• Real-time sentence completion for Tibetan and English text
• Smart prediction after 1.5 seconds of typing pause
• Gibberish detection for both languages
• UTF-8 support for proper Tibetan text handling
• File operations (Open/Save) with UTF-8 encoding
• Undo/Redo functionality
• Clean and intuitive user interface

Requirements

• Python 3.x
• Anthropic API key
• Noto Serif Tibetan font (for Tibetan text rendering)

Installation

1. Clone the repository:

git clone https://github.com/yourusername/tibetan_autocomplete.git
cd tibetan_autocomplete

2. Install required packages:

pip install anthropic

3. Set up your Anthropic API key:

export ANTHROPIC_API_KEY='your-api-key-here'

Usage

1. Start the editor:

python editor.py

2. Writing and Autocomplete:

   • Type any text in Tibetan or English
   • After 1.5 seconds of pause in typing, a grey-colored suggestion appears
   • Press the Tab key to accept and insert the suggested completion
   • If you don't want the suggestion, just keep typing and it will disappear

3. File Operations:

   • Save your work using File → Save or File → Save As...
   • Files are saved as .txt format with full UTF-8 support
   • You can write and save any type of text content
   • Open existing .txt files using File → Open
   • Create new documents with File → New

4. Additional Features:

   • Undo/Redo support for all text operations
   • Real-time gibberish detection
   • Clean interface with intuitive controls

Code Structure

• editor.py: Main text editor implementation with GUI
• predict.py: AI prediction logic using Claude API
  • get_tibetan_completion(): Get completion for Tibetan text
  • get_english_completion(): Get completion for English text
  • is_gibberish(): Check if input text is gibberish

Example Usage

Contributing

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

License

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

Stashed changes