chore/backend/utils/account_deletion_scheduler.py

import logging
import os
import shutil
from datetime import datetime, timedelta
from logging.handlers import RotatingFileHandler
from apscheduler.schedulers.background import BackgroundScheduler
from tinydb import Query

from config.deletion_config import ACCOUNT_DELETION_THRESHOLD_HOURS
from config.paths import get_user_image_dir
from db.db import users_db, child_db, task_db, reward_db, image_db, pending_reward_db
from models.user import User
from events.types.event import Event
from events.types.event_types import EventType
from events.types.user_deleted import UserDeleted
from events.sse import send_to_user

# Setup dedicated logger for account deletion
logger = logging.getLogger('account_deletion_scheduler')
logger.setLevel(logging.INFO)

# Create logs directory if it doesn't exist
os.makedirs('logs', exist_ok=True)

# Add rotating file handler
file_handler = RotatingFileHandler(
    'logs/account_deletion.log',
    maxBytes=10*1024*1024,  # 10MB
    backupCount=5
)
file_handler.setFormatter(
    logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
)
logger.addHandler(file_handler)

# Also log to stdout
console_handler = logging.StreamHandler()
console_handler.setFormatter(
    logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
)
logger.addHandler(console_handler)

MAX_DELETION_ATTEMPTS = 3

def send_user_deleted_event_to_admins(user_id: str, email: str, deleted_at: str):
    """
    Send USER_DELETED event to all admin users.
    
    TODO: Currently sends to all authenticated users with active SSE connections.
    In production, this should filter to only users with admin role.
    """
    event = Event(
        EventType.USER_DELETED.value,
        UserDeleted(user_id, email, deleted_at)
    )
    
    # TODO: Get list of admin users and send only to them
    # For now, we'll skip broadcasting since we don't have a way to get all active admin connections
    # This will need to be implemented when admin role system is in place
    logger.info(f"USER_DELETED event created for {user_id} ({email}) at {deleted_at}")
    # Future implementation:
    # admin_users = get_admin_users()
    # for admin in admin_users:
    #     send_to_user(admin.id, event.to_dict())

def is_user_due_for_deletion(user: User) -> bool:
    """
    Check if a user is due for deletion based on marked_for_deletion_at timestamp
    and the configured threshold.
    """
    if not user.marked_for_deletion or not user.marked_for_deletion_at:
        return False
    
    try:
        marked_at = datetime.fromisoformat(user.marked_for_deletion_at)
        threshold_delta = timedelta(hours=ACCOUNT_DELETION_THRESHOLD_HOURS)
        due_at = marked_at + threshold_delta
        
        # Get current time - make it timezone-aware if marked_at is timezone-aware
        now = datetime.now()
        if marked_at.tzinfo is not None:
            # Convert marked_at to naive UTC for comparison
            marked_at = marked_at.replace(tzinfo=None)
            due_at = marked_at + threshold_delta
        
        return now >= due_at
    except (ValueError, TypeError) as e:
        logger.error(f"Error parsing marked_for_deletion_at for user {user.id}: {e}")
        return False

def get_deletion_attempt_count(user: User) -> int:
    """
    Calculate the number of deletion attempts based on deletion_attempted_at.
    This is a simplified version - in practice, you might track attempts differently.
    """
    # For now, we'll consider any user with deletion_attempted_at as having 1 attempt
    # In a more robust system, you'd track this in a separate field or table
    if user.deletion_attempted_at:
        return 1
    return 0

def delete_user_data(user: User) -> bool:
    """
    Delete all data associated with a user in the correct order.
    Returns True if successful, False otherwise.
    """
    user_id = user.id
    success = True
    
    try:
        # Step 1: Set deletion_in_progress flag
        logger.info(f"Starting deletion for user {user_id} ({user.email})")
        Query_ = Query()
        users_db.update({'deletion_in_progress': True}, Query_.id == user_id)
        
        # Step 2: Remove pending rewards for user's children
        try:
            children = child_db.search(Query_.user_id == user_id)
            child_ids = [child['id'] for child in children]
            
            if child_ids:
                for child_id in child_ids:
                    removed = pending_reward_db.remove(Query_.child_id == child_id)
                    if removed:
                        logger.info(f"Deleted {len(removed)} pending rewards for child {child_id}")
        except Exception as e:
            logger.error(f"Failed to delete pending rewards for user {user_id}: {e}")
            success = False
        
        # Step 3: Remove children
        try:
            removed = child_db.remove(Query_.user_id == user_id)
            if removed:
                logger.info(f"Deleted {len(removed)} children for user {user_id}")
        except Exception as e:
            logger.error(f"Failed to delete children for user {user_id}: {e}")
            success = False
        
        # Step 4: Remove user-created tasks
        try:
            removed = task_db.remove(Query_.user_id == user_id)
            if removed:
                logger.info(f"Deleted {len(removed)} tasks for user {user_id}")
        except Exception as e:
            logger.error(f"Failed to delete tasks for user {user_id}: {e}")
            success = False
        
        # Step 5: Remove user-created rewards
        try:
            removed = reward_db.remove(Query_.user_id == user_id)
            if removed:
                logger.info(f"Deleted {len(removed)} rewards for user {user_id}")
        except Exception as e:
            logger.error(f"Failed to delete rewards for user {user_id}: {e}")
            success = False
        
        # Step 6: Remove user's images from database
        try:
            removed = image_db.remove(Query_.user_id == user_id)
            if removed:
                logger.info(f"Deleted {len(removed)} images from database for user {user_id}")
        except Exception as e:
            logger.error(f"Failed to delete images from database for user {user_id}: {e}")
            success = False
        
        # Step 7: Delete user's image directory from filesystem
        try:
            user_image_dir = get_user_image_dir(user_id)
            if os.path.exists(user_image_dir):
                shutil.rmtree(user_image_dir)
                logger.info(f"Deleted image directory for user {user_id}")
            else:
                logger.info(f"Image directory for user {user_id} does not exist (already deleted or never created)")
        except Exception as e:
            logger.error(f"Failed to delete image directory for user {user_id}: {e}")
            success = False
        
        # Step 8: Remove user record
        if success:
            try:
                users_db.remove(Query_.id == user_id)
                deleted_at = datetime.now().isoformat()
                logger.info(f"Successfully deleted user {user_id} ({user.email})")
                
                # Send USER_DELETED event to admin users
                send_user_deleted_event_to_admins(user_id, user.email, deleted_at)
                
                return True
            except Exception as e:
                logger.error(f"Failed to delete user record for {user_id}: {e}")
                return False
        else:
            # Deletion failed, update flags
            logger.error(f"Deletion failed for user {user_id}, marking for retry")
            users_db.update({
                'deletion_in_progress': False,
                'deletion_attempted_at': datetime.now().isoformat()
            }, Query_.id == user_id)
            return False
            
    except Exception as e:
        logger.error(f"Unexpected error during deletion for user {user_id}: {e}")
        # Try to clear the in_progress flag
        try:
            users_db.update({
                'deletion_in_progress': False,
                'deletion_attempted_at': datetime.now().isoformat()
            }, Query_.id == user_id)
        except:
            pass
        return False

def process_deletion_queue():
    """
    Process the deletion queue: find users due for deletion and delete them.
    """
    logger.info("Starting deletion scheduler run")
    
    processed = 0
    deleted = 0
    failed = 0
    
    try:
        # Get all marked users
        Query_ = Query()
        marked_users = users_db.search(Query_.marked_for_deletion == True)
        
        if not marked_users:
            logger.info("No users marked for deletion")
            return
        
        logger.info(f"Found {len(marked_users)} users marked for deletion")
        
        for user_dict in marked_users:
            user = User.from_dict(user_dict)
            processed += 1
            
            # Check if user is due for deletion
            if not is_user_due_for_deletion(user):
                continue
            
            # Check retry limit
            attempt_count = get_deletion_attempt_count(user)
            if attempt_count >= MAX_DELETION_ATTEMPTS:
                logger.critical(
                    f"User {user.id} ({user.email}) has failed deletion {attempt_count} times. "
                    "Manual intervention required."
                )
                continue
            
            # Skip if deletion is already in progress (from a previous run)
            if user.deletion_in_progress:
                logger.warning(
                    f"User {user.id} ({user.email}) has deletion_in_progress=True. "
                    "This may indicate a previous run was interrupted. Retrying..."
                )
            
            # Attempt deletion
            if delete_user_data(user):
                deleted += 1
            else:
                failed += 1
        
        logger.info(
            f"Deletion scheduler run complete: "
            f"{processed} users processed, {deleted} deleted, {failed} failed"
        )
        
    except Exception as e:
        logger.error(f"Error in deletion scheduler: {e}")

def check_interrupted_deletions():
    """
    On startup, check for users with deletion_in_progress=True
    and retry their deletion.
    """
    logger.info("Checking for interrupted deletions from previous runs")
    
    try:
        Query_ = Query()
        interrupted_users = users_db.search(
            (Query_.marked_for_deletion == True) & 
            (Query_.deletion_in_progress == True)
        )
        
        if interrupted_users:
            logger.warning(
                f"Found {len(interrupted_users)} users with interrupted deletions. "
                "Will retry on next scheduler run."
            )
            # Reset the flag so they can be retried
            for user_dict in interrupted_users:
                users_db.update(
                    {'deletion_in_progress': False},
                    Query_.id == user_dict['id']
                )
    except Exception as e:
        logger.error(f"Error checking for interrupted deletions: {e}")

# Global scheduler instance
_scheduler = None

def start_deletion_scheduler():
    """
    Start the background deletion scheduler.
    Should be called once during application startup.
    """
    global _scheduler
    
    if _scheduler is not None:
        logger.warning("Deletion scheduler is already running")
        return
    
    logger.info("Starting account deletion scheduler")
    
    # Check for interrupted deletions from previous runs
    check_interrupted_deletions()
    
    # Create and start scheduler
    _scheduler = BackgroundScheduler()
    
    # Run every hour
    _scheduler.add_job(
        process_deletion_queue,
        'interval',
        hours=1,
        id='account_deletion',
        name='Account Deletion Scheduler',
        replace_existing=True
    )
    
    _scheduler.start()
    logger.info("Account deletion scheduler started (runs every 1 hour)")

def stop_deletion_scheduler():
    """
    Stop the deletion scheduler (for testing or shutdown).
    """
    global _scheduler
    
    if _scheduler is not None:
        _scheduler.shutdown()
        _scheduler = None
        logger.info("Account deletion scheduler stopped")

def trigger_deletion_manually():
    """
    Manually trigger the deletion process (for admin use).
    Returns stats about the run.
    """
    logger.info("Manual deletion trigger requested")
    process_deletion_queue()
    
    # Return stats (simplified version)
    Query_ = Query()
    marked_users = users_db.search(Query_.marked_for_deletion == True)
    return {
        'triggered': True,
        'queued_users': len(marked_users)
    }