app_startup.py

#!/usr/bin/env python3
"""
Startup script for Hugging Face Spaces deployment
Ensures spaCy model is available before starting the main app
"""

import subprocess
import sys
import os
import logging

# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

def install_spacy_model():
    """Install spaCy English model if not already available"""
    try:
        import spacy
        
        # Try to load the model
        try:
            nlp = spacy.load("en_core_web_sm")
            logger.info("✅ spaCy model 'en_core_web_sm' is already available")
            return True
        except OSError:
            logger.info("📦 spaCy model 'en_core_web_sm' not found, downloading...")
            
        # Download the model
        subprocess.check_call([
            sys.executable, "-m", "spacy", "download", "en_core_web_sm"
        ])
        
        # Verify installation
        nlp = spacy.load("en_core_web_sm")
        logger.info("✅ spaCy model 'en_core_web_sm' downloaded and loaded successfully")
        return True
        
    except subprocess.CalledProcessError as e:
        logger.error(f"❌ Failed to download spaCy model: {e}")
        return False
    except Exception as e:
        logger.error(f"❌ Error setting up spaCy: {e}")
        return False

def main():
    """Main startup function"""
    logger.info("🚀 Starting TalentLens.AI...")
    
    # Install spaCy model
    model_success = install_spacy_model()
    
    if not model_success:
        logger.warning("⚠️ spaCy model not available, application will use fallback methods")
    
    # Import and run the main app
    try:
        from app import main as app_main
        logger.info("✅ Starting Streamlit app...")
        app_main()
    except Exception as e:
        logger.error(f"❌ Failed to start main app: {e}")
        sys.exit(1)

if __name__ == "__main__":
    main()