。
有没有更好的方法来重写多个嵌套的子查询,以便更容易理解和调试,而不会降低性能?
(注意:在我的情况下,我得到了一堆相当复杂的子查询,许多子查询具有多个连接,其中子句和一些子句得到了案例语句等。 将所有内容编写为一个大块代码是调试的噩梦(
例如:我想重写这个:
SELECT DoSomething(x) AS x
FROM (SELECT DoSomething(x) AS x
FROM (SELECT DoSomething(x) AS x
FROM (SELECT DoSomething(x) AS x
FROM (SELECT DoSomething(x) AS x
FROM (SELECT x
FROM tbl_1) AS Q1) AS Q2) AS Q3) AS Q4) AS Q5
变成这样的东西:
--Q1
SELECT *
INTO #Q1
FROM (SELECT x
FROM tbl_1) AS Q1
--Q2
SELECT *
INTO #Q2
FROM (SELECT DoSomething(x)
FROM Q1) AS Q2
--Q3
SELECT *
INTO #Q3
FROM (SELECT DoSomething(x)
FROM Q2) AS Q3
--Q4
SELECT *
INTO #Q4
FROM (SELECT DoSomething(x)
FROM Q3) AS Q3
--Q5
SELECT *
INTO #Q5
FROM (SELECT DoSomething(x)
FROM Q4) AS Q4
--Do Something with Q5 then cleanup
DROP TABLE #Q1
DROP TABLE #Q2
DROP TABLE #Q3
DROP TABLE #Q4
DROP TABLE #Q5
上面的例子虽然有更多的代码,但(希望(允许调试器更轻松地查看/选择单个子查询,并帮助他/她缩小潜在问题的范围
您可能想看看 CTE。
您可以像这样编写示例:
WITH
Q1
AS
(
SELECT x
FROM tbl_1
),
Q2
AS
SELECT DoSomething(x)
FROM Q1
),
Q3
AS
(
SELECT DoSomething(x)
FROM Q2
),
Q4
AS
(
SELECT DoSomething(x)
FROM Q3
),
Q5
AS
(
SELECT DoSomething(x)
FROM Q4
)
SELECT *
FROM Q5;