Question

我有一个返回多个结果集的存储过程，它看起来像这样

BEGIN
    SET NOCOUNT ON;

    SELECT c1, c2, c3
    FROM t1
    WHERE id = @id

    IF (@@ROWCOUNT = 0)
    BEGIN
        SELECT c1, c2, c3
        FROM t2
        WHERE id = @id
    END
END

我在前端ASP.NET网站中使用此存储过程。

如果第一个SELECT语句没有返回任何行，我在SqlDataReader中得到2个结果集（第一个显然为空）。 有没有办法只返回上一个SELECT语句的结果集？

Answer 1

您可以在这里选择几个选项，您必须在您的环境中进行测试，看看哪种方法效果最好。

第一个选项是将第一个语句包装在if块中，类似于您对第二个块所做的操作：

BEGIN
    SET NOCOUNT ON;

    if exists
    (
        SELECT c1, c2, c3
        FROM t1
        WHERE id = @id  
    )
    begin
        SELECT c1, c2, c3
        FROM t1
        WHERE id = @id
    end
    else
    begin
        SELECT c1, c2, c3
        FROM t2
        WHERE id = @id
    END
END

第二个选项是使用临时表/变量：

BEGIN
    SET NOCOUNT ON;

    declare @t1 table (c1 int, c2 int, c3 int)

    insert @t1 (c1,c2,c3)
    SELECT c1, c2, c3
    FROM t1
    WHERE id = @id  

    IF (@@ROWCOUNT = 0)
    BEGIN
        SELECT c1, c2, c3
        FROM t2
        WHERE id = @id
    END
    ELSE
    BEGIN
        select c1,c2,c3
        from @t1
    end

END

Answer 2

与其他好的答案略有不同，您可以使用联合：

SELECT c1, c2, c3 FROM t1 WHERE id = @id
UNION ALL
SELECT c1, c2, c3 FROM t2 WHERE id = @id
AND NOT EXISTS (
    SELECT * FROM t1 WHERE id = @id
)

Answer 3

DECLARE @Count int;

SELECT @Count = Count(*)
FROM t1
WHERE id = @id


IF(@Count = 0)
BEGIN

SELECT c1, c2, c3
FROM t2
WHERE id = @id

END
ELSE
BEGIN

SELECT c1, c2, c3
FROM t1
WHERE id = @id

END

Answer 4

保持简单：

BEGIN        

SET NOCOUNT ON;        

if exists (SELECT 1 FROM t1 WHERE id = @id)

  SELECT c1, c2, c3                
  FROM t1
  WHERE id = @id

else        

  SELECT c1, c2, c3
  FROM t2
  WHERE id = @id

END

Answer 5

;
WITH
rws_1 AS(
    SELECT c1, c2, c3 FROM t1 WHERE id = @id
),
rws_2 AS(
    SELECT c1, c2, c3 FROM t2 WHERE id = @id
)
SELECT * FROM rws_1
UNION ALL
SELECT * FROM rws_2 WHERE NOT EXISTS(SELECT * FROM rws_1)

如何阻止从存储过程返回结果集？

5 个答案: