verify_setup.py

"""
Quick Verification Script - Day 4-5
Tests that all components are working before building the full graph
"""

import sys
from pathlib import Path

# Add project root to path
project_root = Path(__file__).parent
sys.path.insert(0, str(project_root))


def test_imports():
    """Test that all modules can be imported"""
    print("\n[1/4] Testing imports...")
    
    try:
        from backend.core.relationship_extractor import RelationshipExtractor
        print("  [OK] RelationshipExtractor imported")
    except ImportError as e:
        print(f"  [ERROR] Failed to import RelationshipExtractor: {e}")
        return False
    
    try:
        from backend.database.neo4j_manager import Neo4jManager
        print("  [OK] Neo4jManager imported")
    except ImportError as e:
        print(f"  [ERROR] Failed to import Neo4jManager: {e}")
        return False
    
    try:
        from backend.core.nlp_processor import NLPProcessor
        print("  [OK] NLPProcessor imported")
    except ImportError as e:
        print(f"  [ERROR] Failed to import NLPProcessor: {e}")
        return False
    
    print("  [OK] All imports successful")
    return True


def test_neo4j_connection():
    """Test Neo4j connection"""
    print("\n[2/4] Testing Neo4j connection...")
    
    try:
        from backend.database.neo4j_manager import Neo4jManager
        
        manager = Neo4jManager()
        
        if manager.verify_connection():
            print("  [OK] Neo4j connection successful")
            
            # Get current stats
            stats = manager.get_graph_stats()
            print(f"  [INFO] Current graph size:")
            print(f"    - Nodes: {stats['total_nodes']}")
            print(f"    - Relationships: {stats['total_relationships']}")
            
            manager.close()
            return True
        else:
            print("  [ERROR] Neo4j connection failed")
            return False
            
    except Exception as e:
        print(f"  [ERROR] Neo4j test failed: {e}")
        print("  [HELP] Make sure Neo4j is running:")
        print("    docker ps")
        print("    docker-compose up -d")
        return False


def test_sample_documents():
    """Test that sample documents exist"""
    print("\n[3/4] Checking sample documents...")
    
    sample_dir = Path("data/sample/txt")
    
    if not sample_dir.exists():
        print(f"  [ERROR] Sample directory not found: {sample_dir}")
        return False
    
    sample_files = list(sample_dir.glob("*.txt"))
    
    if not sample_files:
        print(f"  [ERROR] No .txt files found in {sample_dir}")
        return False
    
    print(f"  [OK] Found {len(sample_files)} sample documents:")
    for f in sample_files:
        print(f"    - {f.name}")
    
    return True


def test_nlp_model():
    """Test that spaCy model is available"""
    print("\n[4/4] Testing spaCy model...")
    
    try:
        import spacy
        nlp = spacy.load("en_core_web_lg")
        print("  [OK] spaCy model loaded successfully")
        return True
    except Exception as e:
        print(f"  [ERROR] Failed to load spaCy model: {e}")
        print("  [HELP] Download the model with:")
        print("    python -m spacy download en_core_web_lg")
        return False


def main():
    """Run all verification tests"""
    print("\n" + "="*80)
    print("DAY 4-5 VERIFICATION TESTS")
    print("="*80)
    
    tests = [
        test_imports,
        test_neo4j_connection,
        test_sample_documents,
        test_nlp_model
    ]
    
    results = []
    for test in tests:
        result = test()
        results.append(result)
    
    print("\n" + "="*80)
    print("VERIFICATION SUMMARY")
    print("="*80)
    
    passed = sum(results)
    total = len(results)
    
    print(f"\nTests passed: {passed}/{total}")
    
    if passed == total:
        print("\n[OK] ALL TESTS PASSED!")
        print("\nYou're ready to build the knowledge graph:")
        print("  python build_knowledge_graph.py")
    else:
        print("\n[WARNING] Some tests failed. Fix the issues above before proceeding.")
    
    print("\n")


if __name__ == "__main__":
    main()