minimal_scanner.py

"""
Minimal Singtel Bill Scanner - No Heavy Downloads
This version shows you how to use the models without downloading them immediately
"""

def show_usage_examples():
    """Display usage examples for the models"""
    
    print("=" * 60)
    print("SINGTEL BILL SCANNER - USAGE GUIDE")
    print("=" * 60)
    
    print("\n1. BASIC SETUP:")
    print("-" * 20)
    print("# Install required packages:")
    print("pip install torch transformers Pillow requests")
    print()
    
    print("2. SIMPLE TEXT EXTRACTION:")
    print("-" * 30)
    print("""
from transformers import pipeline
from PIL import Image

# Initialize TrOCR model (downloads ~1.3GB first time)
pipe = pipeline("image-to-text", model="microsoft/trocr-base-handwritten")

# Process your bill image
image = Image.open("your_singtel_bill.jpg")
result = pipe(image)
extracted_text = result[0]['generated_text']

print(f"Extracted text: {extracted_text}")
""")
    
    print("3. BILL PROCESSING FUNCTION:")
    print("-" * 35)
    print("""
def process_singtel_bill(image_path):
    # Load the model
    pipe = pipeline("image-to-text", model="microsoft/trocr-base-handwritten")
    
    # Extract text
    image = Image.open(image_path)
    result = pipe(image)
    text = result[0]['generated_text']
    
    # Parse bill information
    import re
    
    # Extract total amount
    amount_match = re.search(r'Total[:]\s*\$?([0-9,]+\.?[0-9]*)', text, re.IGNORECASE)
    total_amount = float(amount_match.group(1).replace(',', '')) if amount_match else None
    
    # Extract due date
    date_match = re.search(r'Due[:]\s*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})', text, re.IGNORECASE)
    due_date = date_match.group(1) if date_match else None
    
    # Extract account number
    account_match = re.search(r'Account[:]\s*([0-9A-Z-]+)', text, re.IGNORECASE)
    account_number = account_match.group(1) if account_match else None
    
    return {
        'raw_text': text,
        'total_amount': total_amount,
        'due_date': due_date,
        'account_number': account_number
    }

# Usage
result = process_singtel_bill("your_bill.jpg")
print(f"Total: ${result['total_amount']}")
print(f"Due: {result['due_date']}")
print(f"Account: {result['account_number']}")
""")
    
    print("4. TROUBLESHOOTING:")
    print("-" * 20)
    print("If you get import errors:")
    print("- Make sure you're in the virtual environment")
    print("- Run: pip install --upgrade torch transformers Pillow")
    print("- Check Python version (needs 3.8+)")
    print()
    
    print("If models download slowly:")
    print("- Models are ~1.3GB and download once")
    print("- Cached in ~/.cache/huggingface/")
    print("- Use good internet connection")
    print()
    
    print("5. STEP-BY-STEP PROCESS:")
    print("-" * 28)
    steps = [
        "1. Take clear photo of your Singtel bill",
        "2. Save as JPG/PNG file",
        "3. Run the processing script",
        "4. Model downloads automatically (first time only)",
        "5. Text is extracted and parsed",
        "6. Get structured bill data as output"
    ]
    
    for step in steps:
        print(f"   {step}")
    
    print("\n" + "=" * 60)
    print("Ready to start? Run one of these files:")
    print("  - quick_test.py (interactive test)")
    print("  - singtel_scanner.py (full scanner)")
    print("  - test_model.py (comprehensive examples)")
    print("=" * 60)

def check_environment():
    """Check if the environment is set up correctly"""
    print("Checking environment setup...")
    
    try:
        import sys
        print(f"✅ Python version: {sys.version}")
        
        # Try importing required packages
        packages = ['torch', 'transformers', 'PIL', 'requests']
        missing_packages = []
        
        for package in packages:
            try:
                __import__(package)
                print(f"✅ {package} is installed")
            except ImportError:
                print(f"❌ {package} is NOT installed")
                missing_packages.append(package)
        
        if missing_packages:
            print(f"\nTo install missing packages, run:")
            print(f"pip install {' '.join(missing_packages)}")
            return False
        else:
            print("\n🎉 All packages are installed! You're ready to go!")
            return True
            
    except Exception as e:
        print(f"Error checking environment: {e}")
        return False

if __name__ == "__main__":
    print("SINGTEL BILL SCANNER - MINIMAL VERSION")
    print("This version helps you get started without heavy downloads\n")
    
    # Check environment first
    env_ok = check_environment()
    
    if env_ok:
        print("\n" + "🚀" * 20)
        print("ENVIRONMENT IS READY!")
        print("🚀" * 20)
        
        choice = input("\nWhat would you like to do?\n1. See usage examples\n2. Exit\nEnter choice (1-2): ")
        
        if choice == "1":
            show_usage_examples()
        
    else:
        print("\n" + "⚠️" * 20)
        print("SETUP REQUIRED")
        print("⚠️" * 20)
        print("\nRun this to install packages:")
        print("pip install torch transformers Pillow requests huggingface-hub")
        print("\nThen run this script again.")
    
    input("\nPress Enter to exit...")