test_rag.py

#!/usr/bin/env python3
"""
Test script to simulate a recruiter interest email and test RAG reply suggestions
"""

import requests
import json
import sys
from datetime import datetime

BASE_URL = "http://localhost:3001/api"

def print_header(text):
    print("\n" + "="*80)
    print(f"  {text}")
    print("="*80)

def print_json(data):
    print(json.dumps(data, indent=2))

def test_rag_reply():
    print_header("🧪 Testing RAG Reply Suggestions")
    
    # Create a test email that simulates recruiter interest
    test_email = {
        "sender": "hr@techcorp.com",
        "subject": "Technical Interview Invitation - Senior Full Stack Developer",
        "body": """Hi Jason,

Thank you for applying to the Senior Full Stack Developer position at TechCorp.

Your resume has been shortlisted by our technical team. We're impressed with your experience in React, Node.js, and AI/ML technologies.

We would like to schedule a technical interview with you. When would be a good time for you to attend the technical round?

Looking forward to hearing from you.

Best regards,
Sarah Johnson
Tech Recruiter
TechCorp Inc.
""",
        "date": datetime.now().isoformat(),
        "category": "Interested",
        "folder": "INBOX",
        "account": "test@reachinbox.ai"
    }
    
    print("\n📧 Test Email (Simulating Recruiter Interest):")
    print(f"From: {test_email['sender']}")
    print(f"Subject: {test_email['subject']}")
    print(f"Body: {test_email['body'][:200]}...")
    
    # Note: In production, this email would be synced via IMAP
    # For testing, we'll use an existing email ID
    print("\n⚠️  Note: Using an existing email from the database for testing")
    
    # Get any email ID for testing
    response = requests.get(f"{BASE_URL}/emails?limit=1")
    if response.status_code != 200:
        print("❌ Failed to get test email")
        return
    
    emails = response.json()['emails']
    if not emails:
        print("❌ No emails found in database")
        return
    
    email_id = emails[0]['id']
    print(f"\n🔍 Testing with email ID: {email_id}")
    
    # Test RAG reply suggestion
    print_header("🤖 Generating AI Reply Suggestion (with RAG)")
    
    response = requests.post(f"{BASE_URL}/emails/{email_id}/suggest-reply")
    
    if response.status_code == 200:
        result = response.json()
        
        print("\n✅ Success!")
        print(f"\n📝 Suggested Reply:")
        print("-" * 80)
        print(result['suggestion'])
        print("-" * 80)
        
        print(f"\n📊 Metadata:")
        print(f"  • Tone: {result['tone']}")
        print(f"  • Vector DB Used: {'✅ Yes' if result.get('vectorDBUsed') else '❌ No (fallback mode)'}")
        print(f"  • Context Items: {result.get('contextUsed', 0)}")
        print(f"  • Timestamp: {result['timestamp']}")
        
        # Check if meeting link is included
        if 'cal.com' in result['suggestion'] or 'https://' in result['suggestion']:
            print("\n🎯 Meeting link detected in reply! ✅")
        else:
            print("\n⚠️  No meeting link detected (may not be recruiter interest email)")
        
    else:
        print(f"\n❌ Failed: {response.status_code}")
        print(response.text)
    
    # Test multiple variations
    print_header("🔄 Generating Reply Variations")
    
    response = requests.post(f"{BASE_URL}/emails/{email_id}/suggest-reply/variations")
    
    if response.status_code == 200:
        result = response.json()
        variations = result['variations']
        
        print(f"\n✅ Generated {len(variations)} variations:")
        for i, variation in enumerate(variations, 1):
            print(f"\n{'─' * 80}")
            print(f"Variation {i}:")
            print(f"{'─' * 80}")
            print(variation)
    else:
        print(f"\n❌ Failed: {response.status_code}")
        print(response.text)

def test_vector_db_status():
    print_header("📊 Vector Database Status")
    
    response = requests.get(f"{BASE_URL}/stats")
    if response.status_code == 200:
        stats = response.json()
        print(f"\n✅ Backend is running")
        print(f"  • Total Emails: {stats['totalEmails']}")
        print(f"  • Categorized: {stats['categorizedEmails']}")
        print(f"  • Interested: {stats['interestedEmails']}")
    else:
        print("❌ Backend not responding")

if __name__ == "__main__":
    try:
        test_vector_db_status()
        test_rag_reply()
        
        print("\n" + "="*80)
        print("  ✅ RAG Testing Complete!")
        print("="*80)
        print("\n💡 Next Steps:")
        print("  1. Add PINECONE_API_KEY to .env for full vector search")
        print("  2. Restart backend to initialize Pinecone")
        print("  3. Check logs for 'Training data stored in vector database'")
        print("  4. Test with real recruiter emails")
        print("\n📚 Documentation: See PINECONE_SETUP.md for setup instructions\n")
        
    except requests.exceptions.ConnectionError:
        print("\n❌ Error: Backend is not running!")
        print("   Start it with: cd backend && npm run dev")
        sys.exit(1)
    except Exception as e:
        print(f"\n❌ Error: {e}")
        sys.exit(1)