apps/backend/python/tdc_outlook_quick.py

#!/usr/bin/env python3
"""
🔓 QUICK TDC OUTLOOK HARVEST
Direkte fra lokal Outlook - ingen admin!
"""
import win32com.client
import pythoncom
import json
import hashlib
from datetime import datetime, timedelta
from pathlib import Path
from neo4j import GraphDatabase

NEO4J_URI = "neo4j+s://054eff27.databases.neo4j.io"
NEO4J_USER = "neo4j"  
NEO4J_PASSWORD = "Qrt37mkb0xBZ7_ts5tG1J70K2mVDGPMF2L7Njlm7cg8"

KEYWORDS = ["strategi", "cyber", "NIS2", "SOC", "MDR", "cloud", "Azure", "AI", 
            "Copilot", "Columbus", "ERP", "budget", "kunde", "kontrakt", "SKI", 
            "produkt", "arkitektur", "roadmap", "sikkerhed", "incident"]

def main():
    print("=" * 60)
    print("🔓 TDC OUTLOOK QUICK HARVEST")
    print("=" * 60)
    
    pythoncom.CoInitialize()
    outlook = win32com.client.Dispatch('Outlook.Application')
    ns = outlook.GetNamespace('MAPI')
    
    print(f"✅ Forbundet til: clak@tdc.dk")
    
    # Neo4j
    neo4j = GraphDatabase.driver(NEO4J_URI, auth=(NEO4J_USER, NEO4J_PASSWORD))
    
    # Hent Inbox
    inbox = ns.GetDefaultFolder(6)  # 6 = olFolderInbox
    items = inbox.Items
    items.Sort("[ReceivedTime]", True)
    
    total = items.Count
    print(f"📧 Emails i Inbox: {total}")
    print(f"🔍 Søger efter: {len(KEYWORDS)} keywords")
    print()
    
    results = []
    stats = {"scanned": 0, "matched": 0}
    cutoff = datetime.now() - timedelta(days=180)
    
    # Scan emails
    for i, item in enumerate(items):
        try:
            if item.Class != 43:  # MailItem
                continue
            
            stats["scanned"] += 1
            
            # Check date
            received = item.ReceivedTime
            if datetime(received.year, received.month, received.day) < cutoff:
                continue
            
            subject = str(item.Subject or "").lower()
            body = str(item.Body or "")[:1500].lower()
            sender = str(item.SenderEmailAddress or "")
            
            # Match keywords
            matched = [kw for kw in KEYWORDS if kw.lower() in subject or kw.lower() in body]
            
            if matched:
                stats["matched"] += 1
                email_data = {
                    "id": item.EntryID[:50],
                    "subject": item.Subject[:200] if item.Subject else "",
                    "sender": sender,
                    "sender_name": str(item.SenderName or ""),
                    "received": received.strftime("%Y-%m-%d %H:%M"),
                    "preview": body[:400],
                    "keywords": matched,
                    "has_attachments": item.Attachments.Count > 0
                }
                results.append(email_data)
                
                # Save to Neo4j
                content_hash = hashlib.md5(f"{email_data['subject']}:{email_data['id']}".encode()).hexdigest()
                with neo4j.session() as session:
                    session.run("""
                        MERGE (e:OutlookEmail {contentHash: $hash})
                        ON CREATE SET
                            e.subject = $subject,
                            e.sender = $sender,
                            e.senderName = $senderName,
                            e.received = $received,
                            e.preview = $preview,
                            e.keywords = $keywords,
                            e.hasAttachments = $hasAtt,
                            e.harvestedAt = datetime()
                        MERGE (ds:DataSource {name: 'TDC_Outlook_Local'})
                        MERGE (e)-[:HARVESTED_FROM]->(ds)
                    """,
                        hash=content_hash,
                        subject=email_data["subject"],
                        sender=email_data["sender"],
                        senderName=email_data["sender_name"],
                        received=email_data["received"],
                        preview=email_data["preview"],
                        keywords=email_data["keywords"],
                        hasAtt=email_data["has_attachments"]
                    )
                
                print(f"  ✅ [{', '.join(matched[:2])}] {email_data['subject'][:60]}")
            
            # Progress
            if stats["scanned"] % 500 == 0:
                print(f"  ... scannet {stats['scanned']}/{total}")
            
            # Limit
            if stats["scanned"] >= 5000:
                print("  (stop ved 5000)")
                break
                
        except Exception as e:
            continue
    
    # Save JSON
    output = Path("data/outlook_local_harvest")
    output.mkdir(parents=True, exist_ok=True)
    output_file = output / f"tdc_emails_{datetime.now().strftime('%Y%m%d_%H%M')}.json"
    
    with open(output_file, 'w', encoding='utf-8') as f:
        json.dump({"stats": stats, "emails": results}, f, indent=2, ensure_ascii=False)
    
    neo4j.close()
    pythoncom.CoUninitialize()
    
    print()
    print("=" * 60)
    print("📊 RESULTAT")
    print("=" * 60)
    print(f"   Scannet:  {stats['scanned']}")
    print(f"   Matched:  {stats['matched']}")
    print(f"   Gemt:     {output_file}")
    print("=" * 60)

if __name__ == "__main__":
    main()