Postgresql，MySQL, SQL Server的多表链接（join）update操做

数据库更新时常常会join其余表作判断更新，PostgreSQL的写法与其余关系型数据库更有不一样，下面以SQL Server, MySQL，PostgreSQL的数据库作对比和展现。

先造数据源。

create table A(id int, city varchar(20))；
create table B(id int, name varchar(20))；
insert into A values(1, 'beijing');
insert into A values(2, 'shanghai');
insert into A values(3, 'guangzhou');
insert into A values(4, 'hangzhou');

insert into B values(1, 'xiaoming');
insert into B values(2, 'xiaohong');
insert into B values(3, 'xiaolv');

如今我要将xiaohong的城市改成shenzhen，看看不一样的数据库产品是怎么作的：

SQL Server:

update A
set A.city = 'shenzhen'
from A join B
on A.id = B.id
where B.name = 'xiaohong'

MySQL:

update A
join B ON A.id= B. id
set A.city='shenzhen'
where B.name = 'xiaohong'

PostgreSQL:

update A
set city = 'shenzhen'
from B
where A.id = B.id
and B.name = 'xiaohong'

需求更新：

若是要将 a表多余的id的city更新为 ‘abcd’, 即 4 -> ‘abcd’, 实现update left join 

PostgreSQL

update a
set city = 'abcd'
from a a1
left join b 
on a1.id = b.id
where a.id = a1.id
and b.id is null

若是要将a表中的city用b表中的那么更新， 即 1- >xiaoming, 2 -> xiaohong, 3 ->xiaolv

update a
set city = b.name
from a a1
join b
on a.id = b.id
where a1.id = a.id