bakery-ia/scripts/clean_old_dashboard_data.py

#!/usr/bin/env python3
"""
Utility script to clean old purchase orders and production batches with malformed reasoning_data.

This script deletes pending purchase orders and production batches that were created before
the fix for template variable interpolation. After running this script, trigger a new
orchestration run to create fresh data with properly interpolated variables.

Usage:
    python scripts/clean_old_dashboard_data.py --tenant-id <tenant_id>
"""
import asyncio
import argparse
import sys
import os
from datetime import datetime, timedelta

# Add services to path
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../services/procurement/app'))
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../services/production/app'))
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../shared'))


async def clean_old_purchase_orders(tenant_id: str, dry_run: bool = True):
    """Clean old pending purchase orders"""
    try:
        from app.core.database import AsyncSessionLocal
        from sqlalchemy import text

        async with AsyncSessionLocal() as session:
            # Get count of pending POs
            count_result = await session.execute(
                text("""
                    SELECT COUNT(*)
                    FROM purchase_orders
                    WHERE tenant_id = :tenant_id
                    AND status = 'pending_approval'
                """),
                {"tenant_id": tenant_id}
            )
            count = count_result.scalar()

            print(f"Found {count} pending purchase orders for tenant {tenant_id}")

            if count == 0:
                print("No purchase orders to clean.")
                return 0

            if dry_run:
                print(f"DRY RUN: Would delete {count} pending purchase orders")
                return count

            # Delete pending POs
            result = await session.execute(
                text("""
                    DELETE FROM purchase_orders
                    WHERE tenant_id = :tenant_id
                    AND status = 'pending_approval'
                """),
                {"tenant_id": tenant_id}
            )
            await session.commit()

            deleted = result.rowcount
            print(f"✓ Deleted {deleted} pending purchase orders")
            return deleted

    except Exception as e:
        print(f"Error cleaning purchase orders: {e}")
        return 0


async def clean_old_production_batches(tenant_id: str, dry_run: bool = True):
    """Clean old pending production batches"""
    try:
        # Import production service dependencies
        sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../services/production/app'))
        from app.core.database import AsyncSessionLocal as ProductionSession
        from sqlalchemy import text

        async with ProductionSession() as session:
            # Get count of pending/scheduled batches
            count_result = await session.execute(
                text("""
                    SELECT COUNT(*)
                    FROM production_batches
                    WHERE tenant_id = :tenant_id
                    AND status IN ('pending', 'scheduled')
                """),
                {"tenant_id": tenant_id}
            )
            count = count_result.scalar()

            print(f"Found {count} pending/scheduled production batches for tenant {tenant_id}")

            if count == 0:
                print("No production batches to clean.")
                return 0

            if dry_run:
                print(f"DRY RUN: Would delete {count} pending/scheduled batches")
                return count

            # Delete pending/scheduled batches
            result = await session.execute(
                text("""
                    DELETE FROM production_batches
                    WHERE tenant_id = :tenant_id
                    AND status IN ('pending', 'scheduled')
                """),
                {"tenant_id": tenant_id}
            )
            await session.commit()

            deleted = result.rowcount
            print(f"✓ Deleted {deleted} pending/scheduled production batches")
            return deleted

    except Exception as e:
        print(f"Error cleaning production batches: {e}")
        return 0


async def main():
    parser = argparse.ArgumentParser(
        description='Clean old dashboard data (POs and batches) with malformed reasoning_data'
    )
    parser.add_argument(
        '--tenant-id',
        required=True,
        help='Tenant ID to clean data for'
    )
    parser.add_argument(
        '--execute',
        action='store_true',
        help='Actually delete data (default is dry run)'
    )

    args = parser.parse_args()

    dry_run = not args.execute

    print("=" * 60)
    print("Dashboard Data Cleanup Utility")
    print("=" * 60)
    print(f"Tenant ID: {args.tenant_id}")
    print(f"Mode: {'DRY RUN (no changes will be made)' if dry_run else 'EXECUTE (will delete data)'}")
    print("=" * 60)
    print()

    if not dry_run:
        print("⚠️  WARNING: This will permanently delete pending purchase orders and production batches!")
        print("   After deletion, you should trigger a new orchestration run to create fresh data.")
        response = input("   Are you sure you want to continue? (yes/no): ")
        if response.lower() != 'yes':
            print("Aborted.")
            return
        print()

    # Clean purchase orders
    print("1. Cleaning Purchase Orders...")
    po_count = await clean_old_purchase_orders(args.tenant_id, dry_run)
    print()

    # Clean production batches
    print("2. Cleaning Production Batches...")
    batch_count = await clean_old_production_batches(args.tenant_id, dry_run)
    print()

    print("=" * 60)
    print("Summary:")
    print(f"  Purchase Orders: {po_count} {'would be' if dry_run else ''} deleted")
    print(f"  Production Batches: {batch_count} {'would be' if dry_run else ''} deleted")
    print("=" * 60)

    if dry_run:
        print("\nTo actually delete the data, run with --execute flag:")
        print(f"  python scripts/clean_old_dashboard_data.py --tenant-id {args.tenant_id} --execute")
    else:
        print("\n✓ Data cleaned successfully!")
        print("\nNext steps:")
        print("  1. Restart the orchestrator service")
        print("  2. Trigger a new orchestration run from the dashboard")
        print("  3. The new POs and batches will have properly interpolated variables")


if __name__ == '__main__':
    asyncio.run(main())