monitor.py

"""
Monitor the model download progress and provide helpful information
"""

import os
import time
from pathlib import Path

def check_huggingface_cache():
    """Check what's in the Hugging Face cache directory"""
    cache_dir = Path.home() / ".cache" / "huggingface"
    
    print(f"Checking cache directory: {cache_dir}")
    
    if cache_dir.exists():
        print("✅ Cache directory exists")
        
        # Check for transformers cache
        transformers_cache = cache_dir / "transformers"
        if transformers_cache.exists():
            print("✅ Transformers cache found")
            
            # List cached models
            for item in transformers_cache.iterdir():
                if item.is_dir():
                    size = sum(f.stat().st_size for f in item.rglob('*') if f.is_file())
                    size_mb = size / (1024 * 1024)
                    print(f"   📁 {item.name}: {size_mb:.1f} MB")
        
        # Check for hub cache
        hub_cache = cache_dir / "hub"
        if hub_cache.exists():
            print("✅ Hub cache found")
            
            total_size = 0
            model_count = 0
            
            for item in hub_cache.iterdir():
                if item.is_dir() and "microsoft" in item.name.lower():
                    try:
                        size = sum(f.stat().st_size for f in item.rglob('*') if f.is_file())
                        size_mb = size / (1024 * 1024)
                        if size_mb > 1:  # Only show models > 1MB
                            print(f"   📁 {item.name}: {size_mb:.1f} MB")
                            total_size += size_mb
                            model_count += 1
                    except:
                        pass
            
            if total_size > 0:
                print(f"\n📊 Total cached models: {model_count}")
                print(f"📊 Total cache size: {total_size:.1f} MB")
                
                if total_size > 1000:  # > 1GB
                    print("✅ Large model detected - TrOCR might be ready!")
                
    else:
        print("❌ Cache directory not found yet")
    
    return cache_dir.exists()

def monitor_download():
    """Monitor the download progress"""
    print("🔍 DOWNLOAD MONITOR")
    print("=" * 40)
    
    print("Looking for model download activity...")
    
    for i in range(30):  # Check for 30 iterations
        print(f"\n⏱️  Check #{i+1}/30:")
        
        cache_exists = check_huggingface_cache()
        
        if cache_exists:
            print("📥 Download activity detected!")
        
        print("⏳ Waiting 10 seconds before next check...")
        time.sleep(10)
    
    print("\n✅ Monitoring complete!")

def show_status():
    """Show current status"""
    print("SINGTEL BILL SCANNER - STATUS CHECK")
    print("=" * 50)
    
    # Check Python environment
    print("🐍 Python Environment:")
    print(f"   Working directory: {os.getcwd()}")
    
    # Check for test files
    test_files = ["test_bill.png", "test_singtel_bill.png"]
    print("\n📁 Test Files:")
    for file in test_files:
        if os.path.exists(file):
            size = os.path.getsize(file)
            print(f"   ✅ {file} ({size} bytes)")
        else:
            print(f"   ❌ {file} (not found)")
    
    # Check cache
    print("\n📦 Model Cache:")
    check_huggingface_cache()
    
    print("\n💡 What's happening:")
    print("   - The TrOCR model is downloading in the background")
    print("   - This is a one-time process (~1.3GB)")
    print("   - Once complete, all future runs will be instant")
    print("   - The model will be cached permanently")

if __name__ == "__main__":
    import sys
    
    if len(sys.argv) > 1 and sys.argv[1] == "monitor":
        monitor_download()
    else:
        show_status()
        
        choice = input("\nWould you like to monitor download progress? (y/n): ")
        if choice.lower() in ['y', 'yes']:
            monitor_download()
        else:
            print("\n💡 Tip: Run 'python monitor.py monitor' to watch progress")
            print("The download will continue in the background.")