TSQL：尝试 Streamline/Optimise 我的存储过程

Question

大家下午好，

我写了一个存储过程，因为它只做相当简单的计算，所以工作和执行速度相对较快。我想您可能会说我对程序本身的问题是 SELECT 和 ORDER BY 子句中重复 'CASE Statement' 的次数。我的 TSQL 知识仍然相当 N00bish，因为我充其量只是一个 'P' 板块。是否可以进一步简化我的代码，以便我只有 CASE WHEN 计算出现一次，我可以继续在多个地方使用它？我相信这对于将来的校对会更好，而且我只需要在根语句中进行更改，而不必在多个位置进行更改！

@Officer_Name是从用户界面传入的变量。如您所见，F_YEAR（财政年度）、F_Quarter（财政季度）字段计算在语句的 Order By 部分再次重复，我想知道是否可以避免这种情况 :) 许多非常感谢您拯救了这个处于困境中的 damoiseau，我希望有一位具有更高 TSQL 水平的慷慨专家可以帮我这个忙！非常感谢。

    BEGIN

    SELECT  TOP (100) PERCENT 
    COUNT(DISTINCT(dbo.TableA.[Account ID])) AS Applications,
    SUM(CASE WHEN [Client Claims] LIKE '%claim%' THEN 1 ELSE 0 END) AS Main_Client,
    COUNT([TableA_ID]) AS Clients,
    (CASE
    WHEN [Finalised date] < '07/01/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2) THEN 'PAST CASES'
    WHEN [Finalised date] BETWEEN '07/01/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2)  AND '06/30/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 1) THEN 'YEAR'
    WHEN MONTH([Finalised date]) BETWEEN 1  AND 3  THEN ' Q3'
    WHEN MONTH([Finalised date]) BETWEEN 4  AND 6  THEN ' Q4'
    WHEN MONTH([Finalised date]) BETWEEN 7  AND 9  THEN ' Q1'
    WHEN MONTH([Finalised date]) BETWEEN 10 AND 12 THEN ' Q2'
    END) AS F_Quarter, 
    (CASE 
    WHEN [Finalised date] < '07/01/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2) THEN CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2) + ' & Older'
    WHEN MONTH([Finalised date]) BETWEEN 1  AND 6 THEN convert(char(4), YEAR([Finalised date]) - 0)
    WHEN MONTH([Finalised date]) BETWEEN 7  AND 12 THEN convert(char(4), YEAR([Finalised date]) + 1) 
    ELSE convert(char(4), YEAR([Finalised date])) 
    END) AS F_YEAR

FROM    dbo.TableB INNER JOIN
        dbo.TableA ON dbo.TableB.[Account ID] = dbo.TableA.[Account ID] LEFT OUTER JOIN
        dbo.Officers ON dbo.TableA.[Account Officer] = dbo.Officers.FullName

WHERE   [Case Officer] = @Officer_Name AND [Finalisation] IS NOT NULL

GROUP BY 
    (CASE 
    WHEN [Finalised date] < '07/01/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2) THEN CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2) + ' & Older'
    WHEN MONTH([Finalised date]) BETWEEN 1  AND 6 THEN convert(char(4), YEAR([Finalised date]) - 0)
    WHEN MONTH([Finalised date]) BETWEEN 7  AND 12 THEN convert(char(4), YEAR([Finalised date]) + 1) 
    ELSE convert(char(4), YEAR([Finalised date])) 
    END),
    (CASE
    WHEN [Finalised date] < '07/01/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2) THEN 'PAST CASES'
    WHEN [Finalised date] BETWEEN '07/01/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 2)  AND '06/30/' +  CONVERT(VARCHAR(4), DATEPART(Year,GETDATE()) - 1) THEN 'YEAR'
    WHEN MONTH([Finalised date]) BETWEEN 1  AND 3  THEN ' Q3'
    WHEN MONTH([Finalised date]) BETWEEN 4  AND 6  THEN ' Q4'
    WHEN MONTH([Finalised date]) BETWEEN 7  AND 9  THEN ' Q1'
    WHEN MONTH([Finalised date]) BETWEEN 10 AND 12 THEN ' Q2'
    END)

ORDER BY F_YEAR DESC, F_Quarter
END

Answer 1

您可以将 CASE 表达式放在 CTE 中，并在后面的查询中通过别名多次引用它。但是，由于 F_YEAR 的 CASE 表达式不同于 F_Quarter 的 CASE 表达式，因此无法对整个查询只使用一个 CASE 表达式。在伪代码中，你可以这样做：

WITH cte AS (
   SELECT ...
   , {CASE Expression for year} AS F_Year
   , {CASE Expression for quarter} AS F_Quarter
   FROM...
)
SELECT ... F_Year, F_Quarter
FROM ... WHERE ...
GROUP BY F_Year, F_Quarter ...

TSQL：尝试 Streamline/Optimise 我的存储过程

TSQL: Trying to Streamline/Optimise my Stored-Procedure

tsql

stored-procedures

future-proof

streamline