src/utils.py

# src/utils.py
import re
import datetime
from typing import Dict, List, Any
import salt.constants

def get_language_name(lang_code: str) -> str:
    """Get full language name from ISO code."""
    if lang_code is None:
        return "Unknown"
    return salt.constants.SALT_LANGUAGE_NAMES.get(lang_code, str(lang_code))

def format_model_name(model_path: str) -> str:
    """Format model name for display in leaderboard."""
    if model_path == 'google-translate':
        return 'Google Translate'
    
    # Extract model name from HuggingFace path
    if '/' in model_path:
        return model_path.split('/')[-1]
    return model_path

def validate_model_path(model_path: str) -> bool:
    """Validate if model path is supported."""
    if model_path == 'google-translate':
        return True
    
    # Check if it's a valid HuggingFace model path format
    pattern = r'^[a-zA-Z0-9._-]+/[a-zA-Z0-9._-]+$'
    return bool(re.match(pattern, model_path)) or not '/' in model_path

def get_model_type(model_path: str) -> str:
    """Determine model type from path."""
    model_path_lower = model_path.lower()
    
    if model_path == 'google-translate':
        return 'google-translate'
    elif 'gemma' in model_path_lower:
        return 'gemma'
    elif 'qwen' in model_path_lower:
        return 'qwen'
    elif 'llama' in model_path_lower:
        return 'llama'
    elif 'nllb' in model_path_lower:
        return 'nllb'
    else:
        return 'other'

def create_submission_id() -> str:
    """Create unique submission ID."""
    return datetime.datetime.now().strftime("%Y%m%d_%H%M%S")

def sanitize_input(text: str) -> str:
    """Sanitize user input."""
    if not text:
        return ""
    return re.sub(r'[^\w\-./]', '', text.strip())