run_template_test.py

#!/usr/bin/env python3
"""
Simple test runner for template.docx conversion
Tests only the core functionality without LibreOffice
"""

import os
import sys
from pathlib import Path

# Add current directory to path
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))

from app import (
    validate_docx_structure, 
    preprocess_docx_for_perfect_conversion,
    analyze_template_font_sizes,
    setup_local_arial_font
)

def main():
    """Test the template conversion system"""
    print("🎯 Template.docx Conversion System Test")
    print("=" * 50)

    # Get script directory
    script_dir = Path(__file__).parent.absolute()
    print(f"📁 Script directory: {script_dir}")

    # Check files
    print("📁 Checking required files...")

    arial_path = script_dir / "arial.ttf"
    template_path = script_dir / "template.docx"
    
    if not arial_path.exists():
        print(f"❌ Arial font not found: {arial_path}")
        return False
    print(f"✅ Arial font found: {arial_path}")
    
    if not template_path.exists():
        print(f"❌ Template not found: {template_path}")
        return False
    print(f"✅ Template found: {template_path}")
    
    # Test Arial font setup
    print("\n🔤 Setting up Arial font...")
    if setup_local_arial_font():
        print("✅ Arial font setup successful")
    else:
        print("⚠️ Arial font setup had issues (may still work)")
    
    # Test template analysis
    print("\n📏 Analyzing template font sizes...")
    font_mapping = analyze_template_font_sizes(str(template_path))
    
    if font_mapping:
        print(f"✅ Found {len(font_mapping)} text patterns with font sizes")
        
        # Show specific patterns we care about
        important_patterns = {
            'size_12': ['{{serial_number}}', '{{date}}', 'الرقم التسلسلي', 'التاريخ'],
            'size_13': ['{{name_1}}', '{{location_1}}', 'اسم المالك', 'يسكن'],
            'size_14': ['الطرف البائع', 'الطرف المشتري']
        }
        
        for size_name, patterns in important_patterns.items():
            found_patterns = []
            for pattern in patterns:
                for text, size in font_mapping.items():
                    if pattern in text:
                        found_patterns.append(f"{pattern}→{size}pt")
                        break
            
            if found_patterns:
                print(f"  • {size_name}: {', '.join(found_patterns[:3])}")
    else:
        print("❌ Font size analysis failed")
        return False
    
    # Test DOCX validation
    print("\n🔍 Validating DOCX structure...")
    validation_info = validate_docx_structure(str(template_path))
    
    print(f"✅ Validation completed:")
    print(f"  • Tables: {validation_info.get('has_tables', False)}")
    print(f"  • Images: {validation_info.get('has_images', False)}")
    print(f"  • RTL content: {validation_info.get('rtl_content_detected', False)}")
    print(f"  • Placeholders: {validation_info.get('placeholder_count', 0)}")
    print(f"  • Font families: {len(validation_info.get('font_families', set()))}")
    
    # Test preprocessing
    print("\n🔧 Testing preprocessing...")
    try:
        processed_path = preprocess_docx_for_perfect_conversion(str(template_path), validation_info)
        
        if processed_path != str(template_path):
            print("✅ Preprocessing applied successfully")
            print(f"  • Font settings applied")
            print(f"  • Arial font set as default")
            print(f"  • Specific font sizes applied")
            
            # Clean up
            try:
                os.unlink(processed_path)
                print("  • Temporary file cleaned up")
            except:
                pass
        else:
            print("ℹ️ No preprocessing needed")
        
    except Exception as e:
        print(f"❌ Preprocessing failed: {e}")
        return False
    
    # Summary
    print("\n" + "=" * 50)
    print("🎉 Template Conversion System Ready!")
    print("\n📋 Summary:")
    print("✅ Arial font from fonts/ directory will be used")
    print("✅ Font sizes will be preserved:")
    print("   • Size 12: Serial numbers, dates, times")
    print("   • Size 13: Names, IDs, locations, phones")
    print("   • Size 14: 'الطرف البائع', 'الطرف المشتري'")
    print("   • Size 12: All other text (default)")
    print("✅ RTL Arabic text will be handled correctly")
    print("✅ Tables and images will be preserved")
    print(f"✅ {validation_info.get('placeholder_count', 0)} placeholders will be maintained")
    
    print("\n🚀 To use the system:")
    print("1. Run: python app.py")
    print("2. Open the Gradio interface")
    print("3. Upload template.docx")
    print("4. Download the converted PDF")
    
    return True

if __name__ == "__main__":
    success = main()
    if success:
        print("\n✅ All tests passed! System is ready to use.")
    else:
        print("\n❌ Some tests failed. Please check the setup.")
    
    input("\nPress Enter to exit...")
    sys.exit(0 if success else 1)