find_user_with_data.py

# -*- coding: utf-8 -*-
from pymongo import MongoClient
from bson import ObjectId
import sys
import io
import json

# Fix encoding for Windows console
if sys.platform == 'win32':
    sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8')

# MongoDB connection string
MONGODB_URI = "mongodb://expenseuser:Kem_6o%3F%3F@165.227.69.221:27017/expense?authSource=admin"

try:
    # Connect to MongoDB
    client = MongoClient(MONGODB_URI)
    db = client.expense
    
    # Test connection
    client.admin.command('ping')
    print("[OK] Successfully connected to MongoDB")
    print(f"Database: expense\n")
    
    # Find users with budgets that have category_id
    print("Searching for users with budgets that have category_id...")
    
    # Get user with most budgets
    user_with_most_budgets = "6741abd38d30ab5b7176397f"
    print(f"\nChecking user: {user_with_most_budgets} (has 243 budgets)")
    
    # Get budgets for this user
    user_budgets = list(db.budgets.find({
        "$or": [
            {"createdBy": ObjectId(user_with_most_budgets)},
            {"createdBy": user_with_most_budgets},
            {"user_id": ObjectId(user_with_most_budgets)},
            {"user_id": user_with_most_budgets}
        ]
    }).limit(20))
    
    print(f"Found {len(user_budgets)} budgets for this user\n")
    
    valid_combinations = []
    
    for budget in user_budgets:
        created_by = str(budget.get("createdBy") or budget.get("user_id", ""))
        category_id = None
        
        # Try direct category fields
        category_id = budget.get("category") or budget.get("categoryId") or budget.get("headCategory")
        
        # Check headCategories array
        if not category_id:
            head_categories = budget.get("headCategories", [])
            if head_categories and isinstance(head_categories, list):
                for head_cat in head_categories:
                    if isinstance(head_cat, dict):
                        # Get headCategory ID
                        head_cat_id = head_cat.get("headCategory")
                        if head_cat_id:
                            category_id = head_cat_id
                            break
                        
                        # Get nested category IDs
                        nested_cats = head_cat.get("categories", [])
                        if nested_cats and isinstance(nested_cats, list):
                            for nested_cat in nested_cats:
                                if isinstance(nested_cat, dict):
                                    nested_cat_id = nested_cat.get("category")
                                    if nested_cat_id:
                                        category_id = nested_cat_id
                                        break
                            if category_id:
                                break
        
        if category_id:
            valid_combinations.append({
                "user_id": created_by,
                "category_id": str(category_id),
                "budget_name": budget.get("name", "N/A"),
                "budget_id": str(budget.get("_id"))
            })
    
    print(f"{'='*60}")
    print(f"Found {len(valid_combinations)} budgets with category_id")
    print(f"{'='*60}\n")
    
    if valid_combinations:
        print("[OK] User ID and Category ID combinations for testing:")
        print("-" * 60)
        seen = set()
        for i, combo in enumerate(valid_combinations[:10], 1):
            key = (combo["user_id"], combo["category_id"])
            if key not in seen:
                seen.add(key)
                print(f"{i}. user_id: {combo['user_id']}")
                print(f"   category_id: {combo['category_id']}")
                print(f"   budget_name: {combo['budget_name']}")
                print()
        
        print(f"\n{'='*60}")
        print("[RECOMMENDED] Use this for testing:")
        print(f"{'='*60}")
        first = valid_combinations[0]
        print(f"user_id: {first['user_id']}")
        print(f"category_id: {first['category_id']}")
        print(f"{'='*60}\n")
    else:
        print("[WARNING] No budgets found with category_id")
        print("Checking sample budget structure...\n")
        
        # Show structure of first budget
        if user_budgets:
            sample = user_budgets[0]
            print("Sample budget structure:")
            print(f"  _id: {sample.get('_id')}")
            print(f"  createdBy: {sample.get('createdBy')}")
            print(f"  user_id: {sample.get('user_id')}")
            print(f"  name: {sample.get('name')}")
            print(f"  category: {sample.get('category')}")
            print(f"  categoryId: {sample.get('categoryId')}")
            print(f"  headCategory: {sample.get('headCategory')}")
            print(f"  has headCategories array: {bool(sample.get('headCategories'))}")
            if sample.get('headCategories'):
                print(f"  headCategories length: {len(sample.get('headCategories', []))}")
                if len(sample.get('headCategories', [])) > 0:
                    print(f"  First headCategory structure:")
                    print(json.dumps(sample.get('headCategories')[0], indent=2, default=str))
    
except Exception as e:
    print(f"[ERROR] Error: {e}")
    import traceback
    traceback.print_exc()