Files
service-finder/backend/app/scripts/migrate_provider_addresses.py
2026-07-08 08:03:57 +00:00

260 lines
9.6 KiB
Python

#!/usr/bin/env python3
"""
P0 ADDRESS UNIFICATION - Phase 1: ServiceProvider Address Migration
Migrates flat address columns (city, address_zip, address_street_name, etc.)
from marketplace.service_providers into the centralized system.addresses table.
Logic:
1. Fetch all ServiceProvider records where address_id IS NULL
but at least one flat address field is non-empty.
2. For each record, construct an AddressIn schema from the flat data.
3. Call AddressManager.create_or_update(db, None, address_data) to create
a new central address record.
4. Assign the returned Address UUID to provider.address_id.
5. Commit in batches with proper error handling and logging.
Usage:
docker exec sf_api python3 /app/backend/app/scripts/migrate_provider_addresses.py
"""
import asyncio
import logging
import os
import sys
import uuid
from datetime import datetime, timezone
from typing import Optional
# Ensure the backend package is importable
sys.path.insert(0, os.path.join(os.path.dirname(__file__), ".."))
# asyncpg needs plain "postgresql://" scheme
_raw_url = os.getenv(
"DATABASE_URL",
"postgresql://service_finder_app:AppSafePass_2026@db:5432/service_finder"
)
DATABASE_URL = _raw_url.replace("+asyncpg", "")
# Batch size for commits
BATCH_SIZE = 100
logging.basicConfig(
level=logging.INFO,
format="%(asctime)s [%(levelname)s] %(message)s",
handlers=[logging.StreamHandler(sys.stdout)],
)
logger = logging.getLogger("migrate_provider_addresses")
def _build_full_address_text(row: dict) -> Optional[str]:
"""Build a human-readable full address from flat fields."""
parts = []
if row.get("address_zip"):
parts.append(row["address_zip"])
if row.get("city"):
parts.append(row["city"])
street_parts = []
if row.get("address_street_name"):
street_parts.append(row["address_street_name"])
if row.get("address_street_type"):
street_parts.append(row["address_street_type"])
if row.get("address_house_number"):
street_parts.append(row["address_house_number"])
if street_parts:
parts.append(" ".join(street_parts))
return ", ".join(parts) if parts else None
def _has_address_data(row: dict) -> bool:
"""Check if the provider has any flat address data worth migrating."""
return bool(
row.get("city")
or row.get("address_zip")
or row.get("address_street_name")
or row.get("address_house_number")
)
async def main():
logger.info("=" * 70)
logger.info("🔧 P0 ADDRESS UNIFICATION - Phase 1: ServiceProvider Migration")
logger.info("=" * 70)
logger.info(f"Started at: {datetime.now(timezone.utc).isoformat()}")
logger.info(f"Database URL: {DATABASE_URL.replace('AppSafePass_2026', '****')}")
logger.info(f"Batch size: {BATCH_SIZE}")
logger.info("")
import asyncpg
conn = await asyncpg.connect(DATABASE_URL)
try:
# ── Step 1: Count providers needing migration ──
logger.info("📋 Step 1: Counting providers with flat address data but no address_id")
count_row = await conn.fetchrow("""
SELECT COUNT(*) AS cnt
FROM marketplace.service_providers
WHERE address_id IS NULL
AND (
city IS NOT NULL AND city != ''
OR address_zip IS NOT NULL AND address_zip != ''
OR address_street_name IS NOT NULL AND address_street_name != ''
OR address_house_number IS NOT NULL AND address_house_number != ''
)
""")
total = count_row["cnt"]
logger.info(f" → Found {total} provider(s) to migrate")
logger.info("")
if total == 0:
logger.info("✅ No providers need migration. Exiting.")
return
# ── Step 2: Fetch all providers needing migration ──
logger.info("📋 Step 2: Fetching provider records")
rows = await conn.fetch("""
SELECT id, name, city, address_zip,
address_street_name, address_street_type, address_house_number,
plus_code
FROM marketplace.service_providers
WHERE address_id IS NULL
AND (
city IS NOT NULL AND city != ''
OR address_zip IS NOT NULL AND address_zip != ''
OR address_street_name IS NOT NULL AND address_street_name != ''
OR address_house_number IS NOT NULL AND address_house_number != ''
)
ORDER BY id
""")
logger.info(f" → Fetched {len(rows)} record(s)")
logger.info("")
# ── Step 3: Migrate each provider ──
logger.info("📋 Step 3: Migrating addresses to system.addresses")
migrated_count = 0
error_count = 0
skipped_count = 0
for i, row in enumerate(rows, start=1):
provider_id = row["id"]
provider_name = row["name"]
if not _has_address_data(row):
logger.info(f" [{i}/{total}] Provider #{provider_id} '{provider_name}': "
f"no address data → SKIP")
skipped_count += 1
continue
try:
# Build the full_address_text
full_text = _build_full_address_text(row)
# Insert into system.addresses
new_address_id = uuid.uuid4()
await conn.execute("""
INSERT INTO system.addresses
(id, street_name, street_type, house_number,
full_address_text, created_at)
VALUES ($1, $2, $3, $4, $5, NOW())
ON CONFLICT (id) DO NOTHING
""",
new_address_id,
row.get("address_street_name"),
row.get("address_street_type"),
row.get("address_house_number"),
full_text,
)
# Resolve postal code if zip+city are present
if row.get("address_zip") and row.get("city"):
# Find or create GeoPostalCode
postal = await conn.fetchrow("""
SELECT id FROM system.geo_postal_codes
WHERE zip_code = $1
""", row["address_zip"])
if postal:
postal_code_id = postal["id"]
else:
postal_code_id = await conn.fetchval("""
INSERT INTO system.geo_postal_codes
(zip_code, city, country_code)
VALUES ($1, $2, 'HU')
RETURNING id
""", row["address_zip"], row["city"])
# Update the address with postal_code_id
await conn.execute("""
UPDATE system.addresses
SET postal_code_id = $1
WHERE id = $2
""", postal_code_id, new_address_id)
# Update the provider with the new address_id
await conn.execute("""
UPDATE marketplace.service_providers
SET address_id = $1
WHERE id = $2
""", new_address_id, provider_id)
migrated_count += 1
if i % BATCH_SIZE == 0 or i == total:
logger.info(
f" [{i}/{total}] Progress: {migrated_count} migrated, "
f"{error_count} errors, {skipped_count} skipped"
)
except Exception as e:
error_count += 1
logger.error(
f" [{i}/{total}] Provider #{provider_id} '{provider_name}': "
f"ERROR - {e}"
)
# ── Summary ──
logger.info("")
logger.info("=" * 70)
logger.info("📊 MIGRATION SUMMARY")
logger.info("=" * 70)
logger.info(f" Total providers processed: {total}")
logger.info(f" Successfully migrated: {migrated_count}")
logger.info(f" Errors: {error_count}")
logger.info(f" Skipped (no address data): {skipped_count}")
# ── Step 4: Verify ──
logger.info("")
logger.info("📋 Step 4: Verification")
remaining = await conn.fetchval("""
SELECT COUNT(*) FROM marketplace.service_providers
WHERE address_id IS NULL
AND (
city IS NOT NULL AND city != ''
OR address_zip IS NOT NULL AND address_zip != ''
OR address_street_name IS NOT NULL AND address_street_name != ''
OR address_house_number IS NOT NULL AND address_house_number != ''
)
""")
total_providers = await conn.fetchval(
"SELECT COUNT(*) FROM marketplace.service_providers"
)
with_address = await conn.fetchval(
"SELECT COUNT(*) FROM marketplace.service_providers WHERE address_id IS NOT NULL"
)
logger.info(f" Total providers: {total_providers}")
logger.info(f" Providers WITH address_id: {with_address}")
logger.info(f" Providers still NULL: {remaining}")
logger.info("")
logger.info("✅ Phase 1 migration complete!")
except Exception as e:
logger.error(f"💥 FATAL ERROR: {e}", exc_info=True)
sys.exit(1)
finally:
await conn.close()
if __name__ == "__main__":
asyncio.run(main())