leocrm/scripts/seed_perf_data.py

#!/usr/bin/env python3
"""Seed performance test data — creates N contacts in the test database.

Usage:
    python scripts/seed_perf_data.py --count 200000
    python scripts/seed_perf_data.py --count 5000 --tenant-id <uuid>

Requires DATABASE_URL environment variable or .env file.
"""

from __future__ import annotations

import argparse
import asyncio
import os
import sys
import uuid
from datetime import UTC, datetime

# Add project root to path
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))

from sqlalchemy import select
from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker, create_async_engine

from app.config import get_settings
from app.models.contact import Contact
from app.models.tenant import Tenant
from app.models.user import User


async def seed_contacts(count: int, tenant_id_str: str | None = None) -> None:
    """Seed `count` contacts into the database for performance testing."""
    settings = get_settings()
    engine = create_async_engine(
        settings.database_url,
        pool_size=settings.db_pool_size,
        max_overflow=settings.db_max_overflow,
    )
    session_factory = async_sessionmaker(bind=engine, expire_on_commit=False)

    async with session_factory() as session:
        # Get or create a tenant
        if tenant_id_str:
            tenant_id = uuid.UUID(tenant_id_str)
            result = await session.execute(select(Tenant).where(Tenant.id == tenant_id))
            tenant = result.scalar_one_or_none()
            if not tenant:
                print(f"ERROR: Tenant {tenant_id_str} not found")
                return
        else:
            # Get first tenant or create one
            result = await session.execute(select(Tenant).limit(1))
            tenant = result.scalar_one_or_none()
            if not tenant:
                tenant = Tenant(name="Perf Test Tenant", slug="perf-test")
                session.add(tenant)
                await session.flush()

        # Get or create a user for this tenant
        result = await session.execute(
            select(User).where(User.tenant_id == tenant.id).limit(1)
        )
        user = result.scalar_one_or_none()
        if not user:
            from app.core.auth import hash_password

            user = User(
                tenant_id=tenant.id,
                email="perf@test.local",
                name="Perf Test User",
                password_hash=hash_password("TestPass123!"),
                role="admin",
                is_active=True,
                preferences={},
            )
            session.add(user)
            await session.flush()

        tenant_id = tenant.id
        user_id = user.id

        # Seed contacts in batches
        batch_size = 1000
        total_seeded = 0
        first_names = ["Hans", "Maria", "Peter", "Anna", "Klaus", "Ursula", "Werner", "Brigitte", "Jürgen", "Helga"]
        last_names = ["Müller", "Schmidt", "Schneider", "Fischer", "Weber", "Meyer", "Wagner", "Becker", "Hoffmann", "Schäfer"]

        while total_seeded < count:
            batch = []
            batch_count = min(batch_size, count - total_seeded)
            for i in range(batch_count):
                idx = total_seeded + i
                fn = first_names[idx % len(first_names)]
                ln = last_names[idx % len(last_names)]
                batch.append(Contact(
                    tenant_id=tenant_id,
                    first_name=f"{fn}-{idx}",
                    last_name=f"{ln}-{idx}",
                    email=f"{fn.lower()}.{ln.lower()}-{idx}@example.com",
                    phone=f"+49-555-{idx:06d}",
                    created_by=user_id,
                    updated_by=user_id,
                ))
            session.add_all(batch)
            await session.commit()
            total_seeded += batch_count
            if total_seeded % 10000 == 0 or total_seeded == count:
                print(f"Seeded {total_seeded}/{count} contacts...")

    await engine.dispose()
    print(f"Done: {total_seeded} contacts seeded.")


def main():
    """Parse arguments and run the seeding script."""
    parser = argparse.ArgumentParser(
        description="Seed performance test data into the LeoCRM database."
    )
    parser.add_argument(
        "--count",
        type=int,
        default=200000,
        help="Number of contacts to create (default: 200000)",
    )
    parser.add_argument(
        "--tenant-id",
        type=str,
        default=None,
        help="Tenant UUID to seed data into (default: first available tenant)",
    )
    args = parser.parse_args()

    print(f"Starting seed: {args.count} contacts")
    asyncio.run(seed_contacts(args.count, args.tenant_id))


if __name__ == "__main__":
    main()