119 lines
3.6 KiB
Python
119 lines
3.6 KiB
Python
#!/usr/bin/env python3
|
|
"""回填 crm_product / crm_customer 的 name_pinyin_abbr（仅空字段）。
|
|
|
|
用法：
|
|
cd backend
|
|
venv/bin/python scripts/backfill_name_pinyin_abbr.py --dry-run
|
|
venv/bin/python scripts/backfill_name_pinyin_abbr.py
|
|
venv/bin/python scripts/backfill_name_pinyin_abbr.py --env-file .env.production
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import argparse
|
|
import os
|
|
import sys
|
|
from pathlib import Path
|
|
from typing import Optional
|
|
|
|
BACKEND_ROOT = Path(__file__).resolve().parents[1]
|
|
if str(BACKEND_ROOT) not in sys.path:
|
|
sys.path.insert(0, str(BACKEND_ROOT))
|
|
|
|
|
|
def load_settings_env(env_file: Optional[str]) -> None:
|
|
if env_file:
|
|
os.environ["ENV_FILE"] = env_file
|
|
from dotenv import load_dotenv
|
|
|
|
candidates = []
|
|
if env_file:
|
|
p = Path(env_file)
|
|
candidates.append(p if p.is_absolute() else BACKEND_ROOT / p)
|
|
candidates.extend(
|
|
[
|
|
BACKEND_ROOT / ".env.local",
|
|
BACKEND_ROOT / ".env",
|
|
BACKEND_ROOT / ".env.production",
|
|
]
|
|
)
|
|
for path in candidates:
|
|
if path.is_file():
|
|
load_dotenv(path, override=True)
|
|
break
|
|
|
|
|
|
def _empty_pinyin_filter(model):
|
|
from sqlalchemy import or_
|
|
|
|
return or_(model.name_pinyin_abbr.is_(None), model.name_pinyin_abbr == "")
|
|
|
|
|
|
def backfill_model(db, model, name_attr: str, *, dry_run: bool, batch_size: int) -> int:
|
|
from app.utils.pinyin_abbr import build_name_pinyin_abbr
|
|
|
|
q = db.query(model).filter(model.del_flag == 0).filter(_empty_pinyin_filter(model))
|
|
total = q.count()
|
|
print(f"{model.__tablename__}: empty={total}")
|
|
if total == 0:
|
|
return 0
|
|
|
|
updated = 0
|
|
dry_printed = 0
|
|
while True:
|
|
rows = q.limit(batch_size).all()
|
|
if not rows:
|
|
break
|
|
for row in rows:
|
|
try:
|
|
name = getattr(row, name_attr) or ""
|
|
abbr = build_name_pinyin_abbr(name)
|
|
if not abbr:
|
|
continue
|
|
if dry_run:
|
|
if dry_printed < 20:
|
|
print(f" would set {row.id}: {abbr}")
|
|
dry_printed += 1
|
|
updated += 1
|
|
else:
|
|
row.name_pinyin_abbr = abbr
|
|
updated += 1
|
|
except Exception as exc: # noqa: BLE001 — batch continue
|
|
print(f" skip {getattr(row, 'id', '?')}: {exc}")
|
|
if dry_run:
|
|
if updated > dry_printed:
|
|
print(f" ... and {updated - dry_printed} more")
|
|
break
|
|
db.commit()
|
|
# filled rows leave the empty filter; no offset needed
|
|
print(f"{model.__tablename__}: updated={updated}")
|
|
return updated
|
|
|
|
|
|
def main() -> int:
|
|
parser = argparse.ArgumentParser(description="Backfill CRM name_pinyin_abbr (empty only)")
|
|
parser.add_argument("--dry-run", action="store_true", help="只统计/预览，不写库")
|
|
parser.add_argument("--env-file", default=None, help="指定 dotenv 文件")
|
|
parser.add_argument("--batch-size", type=int, default=200)
|
|
args = parser.parse_args()
|
|
|
|
load_settings_env(args.env_file)
|
|
|
|
from app.core.database import SessionLocal
|
|
from app.models.crm_models import CrmCustomer, CrmProduct
|
|
|
|
db = SessionLocal()
|
|
try:
|
|
backfill_model(
|
|
db, CrmProduct, "product_name", dry_run=args.dry_run, batch_size=args.batch_size
|
|
)
|
|
backfill_model(
|
|
db, CrmCustomer, "customer_name", dry_run=args.dry_run, batch_size=args.batch_size
|
|
)
|
|
finally:
|
|
db.close()
|
|
return 0
|
|
|
|
|
|
if __name__ == "__main__":
|
|
raise SystemExit(main())
|