Question

有没有办法在IN（）语句中嵌套CASE-WHEN语句，以便其中一个WHEN或ELSE返回一个子查询。对我来说，这不应该是一个问题，但不知何故我得到错误：

“子查询返回的值超过1。”

IN()应该处理多个值！

以下是重现错误的一个小例子：

-- tblA will be searched for values
Declare @tblA  table (i int)
insert @tblA
    select 1
    union select 2
    union select 3

--tblB: its values will be searched in tblA
Declare @tblB  table (i int)
insert @tblB
    select 2
    union select 3
    union select 1

--@c used by the CASE statement to match
declare @c varchar(50)
set @c = 'Match'

select *
from @tblA
where i IN ( -- IN statement should accept subquery in it
    case @c
        when 'Dont Match' then 2 --If it had matched, then the single value 2 would have been returned and TSQL would be happy
        else (
            select i from @tblB --TSQL not happy and causing error when more than one values returned from the subquery
        )
    end
)

Answer 1

尝试

select *
from @tblA A
WHERE (@c = 'Dont Match' AND i = 2) OR 
(@c <> 'Dont Match' AND EXISTS (SELECT * FROM @tblB WHERE i = A.i)

Answer 2

您的问题不在于IN子句，而在于CASE。
CASE无法处理多个值。
的有效

DECLARE @TBLA TABLE (I INT) INSERT INTO @TBLA ( [I] ) SELECT 0 SELECT [Result]=CASE WHEN 1=2 THEN (SELECT I FROM @TBLA T) ELSE (SELECT I FROM @TBLA T) END

无效，因为表变量包含两个记录。

DECLARE @TBLA TABLE (I INT) INSERT INTO @TBLA ( [I] ) SELECT 0 UNION SELECT 1 SELECT [Result]=CASE WHEN 1=2 THEN (SELECT I FROM @TBLA T) ELSE (SELECT I FROM @TBLA T) END

Answer 3

IN() is supposed to handle more than one values!

是的，实际上是。如果以这种方式修改查询，就可以看到它

select *
from @tblA
where i IN ( select i from @tblB )

此查询将无错误地执行。该错误是由CASE语句生成的，因为它不能在THEN或ELSE子句中获得超过1的值。

CASE-WHEN返回子查询时出现TSQL错误

3 个答案: