Nepali Font Mapper

A Python package for mapping legacy Nepali fonts to Unicode, with built-in language detection capabilities.

Features

• Language Detection:

  • Uses Levenshtein distance algorithm for word similarity matching
  • Employs cosine similarity with pre-trained sentence embeddings for accurate language detection
  • Configurable similarity threshold for fine-tuning detection accuracy

• Font Mapping:

  • Maps characters from legacy Nepali fonts (like Preeti) to Unicode Nepali
  • Handles complex character combinations and special cases
  • Preserves non-Nepali text in mixed-language content
  • Supports multiple legacy font formats

Installation

git clone https://github.com/amritregmi26/np-font-mapper.git
cd np-font-mapper

pip install -e .

This command will install the package with the name np-font-mapper.

Quick Start

from np_font_mapper import FontMapper

# Initialize the mapper
mapper = FontMapper()

# Optional: Adjust similarity threshold for language detection (default is 0.70)
mapper.initialize_detector()
mapper.similarity_threshold = 0.8

# Basic usage - maps Preeti font to Unicode
text = ";Ë|fxno" 
result = mapper.map(text)
print(result)  # Output: सङ्ग्राहलय

Working with Different Fonts

# Check available fonts
print(mapper.available_fonts)

# Specify a different font
result = mapper.map("your_text", font="kantipur")

How It Works

1. Language Detection:

   • Words are analyzed using Levenshtein distance and cosine similarity
   • Pre-trained embeddings help identify English language content
   • English text is preserved in its original form

2. Font Mapping:

   • Characters are mapped using a comprehensive character map
   • Special character combinations are handled appropriately
   • Post-processing rules ensure correct rendering

FontMapper Class

class FontMapper:
    def __init__(self, similarity_threshold=0.70):
        """
        Initialize FontMapper with optional similarity threshold.
        
        Args:
            similarity_threshold (float): Threshold for language detection (0.0 to 1.0)
        """

    def map(self, text: str, font: str = None) -> str:
        """
        Map input text to Unicode Nepali.
        
        Args:
            text (str): Input text in legacy font
            font (str, optional): Font to use for mapping. Defaults to 'preeti'
            
        Returns:
            str: Mapped text in Unicode Nepali
        """

Supported Fonts

• Preeti (Default)
• Kantipur
• Himali-TT
• PCS-Nepali
• Sagarmatha

Contributing

Contributions are welcome! Here's how you can help:

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'feat: add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Areas for Contribution

• Additional font support
• Improved language detection
• Bug fixes and edge cases
• Documentation improvements

License

This project is licensed under the MIT License.

Acknowledgments

Font mapping rules and character sets are referenced from nep-ttf2utf and ttf-to-unicode.

Support

Give a Star ⭐ to support this repository!

For issues, questions, or suggestions open an issue on GitHub.