89 lines
3.5 KiB
Python
89 lines
3.5 KiB
Python
|
"""为育种核心表补「部分唯一索引（未软删）」并发兜底，幂等，可 --check 预览。
|
|||
|
|
|
|||
|
|
背景：combination_code / tree_no / cultivar_name / personnel.name / (site_id, plot_code)
|
|||
|
|
无 DB 层唯一约束，service 层 select-then-insert 存在并发竞态。此处补部分唯一索引
|
|||
|
|
（WHERE is_deleted = false），与各 service 的查重语义（只查未软删）一致：
|
|||
|
|
- 未软删记录之间不允许重复 → 并发插入同一编号仅一条成功，另一条 409。
|
|||
|
|
- 已软删记录不参与约束 → 删除后仍可重新使用该编号。
|
|||
|
|
|
|||
|
|
用法（backend/ 目录）：
|
|||
|
|
python scripts/add_unique_constraints.py --check
|
|||
|
|
python scripts/add_unique_constraints.py
|
|||
|
|
"""
|
|||
|
|
import argparse
|
|||
|
|
import os
|
|||
|
|
import sys
|
|||
|
|
|
|||
|
|
BACKEND_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
|
|||
|
|
if BACKEND_DIR not in sys.path:
|
|||
|
|
sys.path.insert(0, BACKEND_DIR)
|
|||
|
|
|
|||
|
|
from sqlalchemy import create_engine, text
|
|||
|
|
|
|||
|
|
from app.config.setting import settings
|
|||
|
|
|
|||
|
|
# (表, 索引名, 唯一列, 说明)
|
|||
|
|
UNIQUE_INDEXES = [
|
|||
|
|
("bre_cross_combination", "uniq_bre_cross_combination_code", "combination_code", "组合编号"),
|
|||
|
|
("bre_tree", "uniq_bre_tree_no", "tree_no", "单株编号"),
|
|||
|
|
("bre_germplasm", "uniq_bre_germplasm_cultivar_name", "cultivar_name", "品种/种质名称"),
|
|||
|
|
("bre_personnel", "uniq_bre_personnel_name", "name", "人员姓名"),
|
|||
|
|
("bre_plot", "uniq_bre_plot_site_code", "site_id, plot_code", "基地地块编号"),
|
|||
|
|
]
|
|||
|
|
|
|||
|
|
|
|||
|
|
def main() -> None:
|
|||
|
|
parser = argparse.ArgumentParser(description="为育种表补部分唯一索引（未软删）")
|
|||
|
|
parser.add_argument("--check", action="store_true", help="仅打印将执行的 DDL，不落库")
|
|||
|
|
args = parser.parse_args()
|
|||
|
|
|
|||
|
|
engine = create_engine(settings.DB_URI)
|
|||
|
|
print(f"# 连接: {settings.DB_URI}")
|
|||
|
|
|
|||
|
|
with engine.connect() as conn:
|
|||
|
|
for table, idx_name, cols, label in UNIQUE_INDEXES:
|
|||
|
|
# 1) 未软删重复检查（部分索引只约束未软删行，故只查未软删）
|
|||
|
|
dup_sql = (
|
|||
|
|
f"SELECT {cols} FROM {table} "
|
|||
|
|
"WHERE is_deleted = false "
|
|||
|
|
f"GROUP BY {cols} HAVING COUNT(*) > 1"
|
|||
|
|
)
|
|||
|
|
try:
|
|||
|
|
dups = conn.execute(text(dup_sql)).fetchall()
|
|||
|
|
except Exception as e:
|
|||
|
|
print(f"# [跳过] {table}: 查重失败 {e!s}")
|
|||
|
|
continue
|
|||
|
|
if dups:
|
|||
|
|
print(f"# [跳过] {label}（{table}.{cols}）存在 {len(dups)} 组未软删重复，请先清洗数据:")
|
|||
|
|
for d in dups[:10]:
|
|||
|
|
print(" " + " / ".join(str(x) for x in d))
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
# 2) 索引是否已存在
|
|||
|
|
exists = conn.execute(
|
|||
|
|
text("SELECT 1 FROM pg_indexes WHERE indexname = :name"),
|
|||
|
|
{"name": idx_name},
|
|||
|
|
).fetchone()
|
|||
|
|
if exists:
|
|||
|
|
print(f"# [已存在] {idx_name}")
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
ddl = (
|
|||
|
|
f"CREATE UNIQUE INDEX {idx_name} ON {table} ({cols}) "
|
|||
|
|
"WHERE is_deleted = false"
|
|||
|
|
)
|
|||
|
|
print(" " + ddl)
|
|||
|
|
if not args.check:
|
|||
|
|
try:
|
|||
|
|
with engine.begin() as c:
|
|||
|
|
c.execute(text(ddl))
|
|||
|
|
print("OK " + ddl)
|
|||
|
|
except Exception as e:
|
|||
|
|
print("ERR " + ddl + " -> " + str(e)[:160])
|
|||
|
|
|
|||
|
|
print("# --check 模式：未执行任何变更。" if args.check else "# 完成。")
|
|||
|
|
|
|||
|
|
|
|||
|
|
if __name__ == "__main__":
|
|||
|
|
main()
|