查询需求:
如果第一个SQL能查出结果,则返回结果,否则,执行第二条SQL,返回第二条SQL的结果。

SQL Server中使用表变量的方式:
如果查出有“wangwang”用户则返回,否则查询“zhangzhang”用户的id返回。有优先级

declare @temp table(id int)
insert into @temp
select id
from user
where name = 'wangwang'
if not exists(select id from @temp)
begin
  select id
  from user
  where name='zhangzhang'
end else
begin
  select id from @temp
end

MySQL中没有表变量,只能使用临时表代替:

CREATE TEMPORARY TABLE temp1(id int ); -- 创建临时表1
insert into temp1 
select id
from user  
where name='wangwang'; 

CREATE TEMPORARY TABLE temp2(id int ); -- 创建临时表2
insert into temp2 
select id
from user 
where name='zhangzhang'; 

CREATE TEMPORARY TABLE temp1_re select id from temp1; -- 复制临时表1

select (case when (SELECT count(*) FROM temp1_re)>0 then -- 使用case when判断
 (select id from temp1 order by id desc limit 1) else (
select id from temp2 order by id desc limit 1) end ) id ; 

注意:
(1)需要复制临时表1,是因为“在同一个query语句中,你只能查找一次临时表”,否则报错“can’t reopen ”。针对表中数据不多的情况下,可以复制表进行操作。
(2)因为要在mybatis的xml中执行多条SQL,数据库连接需要设置allowMultiQueries=true。

2019-10-29更新:
发现MySQL有个类似SQLserver中的表变量的语句,那就是with as,也叫做子查询部分(subquery factoring),可以定义一个SQL片断,该SQL片断会被整个SQL语句用到,例如:

with temp_a as(
select id,name from user where name='wangwang'
),
temp_b as(
select id,name from user where name='zhangzhang'
)
select (case when (SELECT count(*) FROM temp_a)>0 then -- 使用case when判断
 			(select id from temp_a order by id desc limit 1) 
 		else 
			(select id from temp_b order by id desc limit 1) 
		end 
	) id 

这种写法是MySQL 8.0的特性
参考:https://www.cnblogs.com/Niko12230/p/5945133.html
https://stackoverflow.com/questions/324935/mysql-with-clause

Logo

更多推荐