test_dependencies.py

#!/usr/bin/env python3
"""
Test script to verify all dependencies and PDF handling capabilities.
"""
import os
import sys

print("=== CogniChat Dependencies & PDF Handling Test ===")

# Test imports
try:
    print("\nTesting core imports...")
    from langchain_community.embeddings import HuggingFaceEmbeddings
    from langchain_community.retrievers import BM25Retriever
    from langchain.text_splitter import RecursiveCharacterTextSplitter
    from langchain_core.documents import Document
    print("✓ Core LangChain imports successful!")
    
except ImportError as e:
    print(f"✗ Import error: {e}")
    if "rank_bm25" in str(e):
        print("  → Missing dependency: pip install rank-bm25==0.2.2")
    sys.exit(1)

# Test PDF loading capabilities
try:
    print("\nTesting PDF loading capabilities...")
    
    # Test PyPDF availability
    try:
        from langchain_community.document_loaders import PyPDFLoader
        print("✓ PyPDFLoader available")
    except ImportError:
        print("✗ PyPDFLoader not available")
    
    # Test PyMuPDF availability
    try:
        import fitz
        print("✓ PyMuPDF (fitz) available - can handle corrupted PDFs")
    except ImportError:
        print("✗ PyMuPDF (fitz) not available")
    
    # Test pdfplumber availability
    try:
        import pdfplumber
        print("✓ pdfplumber available - additional PDF parsing method")
    except ImportError:
        print("✗ pdfplumber not available")
        
except Exception as e:
    print(f"✗ Error testing PDF capabilities: {e}")

# Test BM25 Retriever
try:
    print("\nTesting BM25 Retriever...")
    
    # Create some test documents
    test_docs = [
        Document(page_content="This is the first test document about machine learning."),
        Document(page_content="This is the second document discussing natural language processing."),
        Document(page_content="The third document covers artificial intelligence topics."),
    ]
    
    # Create BM25 retriever
    bm25_retriever = BM25Retriever.from_documents(test_docs)
    bm25_retriever.k = 2
    
    # Test retrieval
    query = "machine learning"
    results = bm25_retriever.get_relevant_documents(query)
    print(f"✓ BM25 retriever created and tested successfully!")
    print(f"✓ Retrieved {len(results)} documents for query: '{query}'")
    
except Exception as e:
    print(f"✗ Error testing BM25 retriever: {e}")
    import traceback
    traceback.print_exc()
    sys.exit(1)

print("\n=== All tests completed successfully! ===")
print("\nThe application should now handle:")
print("  • Regular file uploads and processing")
print("  • Corrupted PDF files with multiple fallback methods")
print("  • BM25 and FAISS hybrid retrieval")
print("  • Proper error messages for failed file processing")
print("\nMake sure to install all dependencies with:")
print("  pip install -r requirements.txt")

# Display dependency summary
print("\n=== Key Dependencies Added/Updated ===")
print("  • rank-bm25==0.2.2 (for BM25 retrieval)")
print("  • pymupdf==1.23.26 (PDF fallback method)")
print("  • pdfplumber==0.10.3 (additional PDF parsing)")