PostgreSQL的upsert实例操作(insert on conflict do)

建表语句：

DROP TABLE IF EXISTS "goods";
CREATE TABLE "goods" (
 "store_cd" int4 NOT NULL,
 "good_cd" varchar(50) COLLATE "pg_catalog"."default" NOT NULL,
 "name" varchar(255) COLLATE "pg_catalog"."default"
);
 
INSERT INTO "goods" VALUES (101, '1', '张三');
INSERT INTO "goods" VALUES (102, '2', '李四');
INSERT INTO "goods" VALUES (103, '3', '王五');
 
ALTER TABLE "goods" ADD CONSTRAINT "pr_cd_key" PRIMARY KEY ("store_cd", "good_cd");

表数据：

数据存在则更新数据，不存在则插入数据

INSERT INTO GOODS VALUES ( 104, '4', '赵六' ) 
ON CONFLICT ON CONSTRAINT pr_key_cd DO
UPDATE 
 SET NAME = '更新' 
WHERE
 GOODS.STORE_CD = '104' 
 AND GOODS.GOOD_CD = '4'

pr_key_cd为必须为唯一主键，也可以用下面写法（注意：必须保证筛选出数据唯一）

INSERT INTO GOODS VALUES ( 104, '4', '赵六' ) 
ON CONFLICT ( STORE_CD, GOOD_CD ) DO
UPDATE 
 SET NAME = '更新' 
WHERE
 GOODS.STORE_CD = '104' 
 AND GOODS.GOOD_CD = '4'

上面的两种的写法，是先执行insert如果主键冲突则执行update，没有冲突就执行insert了。要是想先执行update语句呢？

update更新失败执行insert，更新成功则执行update。

WITH TABLE1 AS ( UPDATE GOODS SET NAME = '更新' WHERE STORE_CD = '104' AND GOOD_CD = '4' RETURNING * ) 
INSERT INTO GOODS SELECT 104, '4', '赵六' 
WHERE NOT EXISTS ( SELECT 1 FROM TABLE1 WHERE STORE_CD = '104' AND GOOD_CD = '4' )

补充：Postgresql插入或更新操作upsert

幂等性的一个要求是多次操作的结果一致。对于update操作，多次直接的结果都是最后update的值，是满足需求的。但对于insert，如果已经插入，第二次会报错，duplicate error， 主键重复或者unique key duplicate。所以需要做一下处理。

最简单的就是，try-catch，当报错的时候，调用update去更新，或者策略更简单点，直接返回就行，不需要更新，以第一条为准。

PostgreSQL从9.5之后就提供了原子的upsert语法： 不存在则插入，发生冲突可以update。

Inert语法

官方文档： https://www.postgresql.org/docs/devel/sql-insert.html

[ WITH [ RECURSIVE ] with_query [, ...] ]
INSERT INTO table_name [ AS alias ] [ ( column_name [, ...] ) ]
  [ OVERRIDING { SYSTEM | USER} VALUE ]
  { DEFAULT VALUES | VALUES ( { expression | DEFAULT } [, ...] ) [, ...] | query }
  [ ON CONFLICT [ conflict_target ] conflict_action ]
  [ RETURNING * | output_expression [ [ AS ] output_name ] [, ...] ]
where conflict_target can be one of:
  ( { index_column_name | ( index_expression ) } [ COLLATE collation ] [ opclass ] [, ...] ) [ WHERE index_predicate ]
  ON CONSTRAINT constraint_name
and conflict_action is one of:
  DO NOTHING
  DO UPDATE SET { column_name = { expression | DEFAULT } |
          ( column_name [, ...] ) = [ ROW ] ( { expression | DEFAULT } [, ...] ) |
          ( column_name [, ...] ) = ( sub-SELECT )
         } [, ...]
       [ WHERE condition ]

index_column_name

The name of a table_name column. Used to infer arbiter indexes. Follows CREATE INDEX format. SELECT privilege on index_column_name is required.

index_expression

Similar to index_column_name, but used to infer expressions on table_name columns appearing within index definitions (not simple columns). Follows CREATE INDEX format. SELECT privilege on any column appearing within index_expression is required.

使用示例

创建表

CREATE TABLE "test"."upsert_test" (
 "id" int4 NOT NULL,
 "name" varchar(255) COLLATE "pg_catalog"."default"
)
;

当主键id冲突时，更新其他字段

INSERT INTO test.upsert_test(id, "name")
  VALUES(1, 'm'),(2, 'n'),(4, 'c')
  ON conflict(id) DO UPDATE
  SET "name" = excluded.name;

did 冲突的主键

EXCLUDED 代指要插入的记录

当主键或者unique key发生冲突时，什么都不做

INSERT INTO test.upsert_test(id, "name")
VALUES(1, 'm'),(2, 'n'),(4, 'c')
ON conflict(id) DO NOTHING;

以上为个人经验，希望能给大家一个参考，也希望大家多多支持潘少俊衡。如有错误或未考虑完全的地方，望不吝赐教。