用户定义的参数不适用于存储过程子查询

Question

我正在创建存储过程以根据基于用户的参数作为过滤器生成报告。我已经让几乎每个 SELECT 查询正常工作，但最后一个包含一堆 LEFT JOIN 的查询最终在整个 table.[=16= 中返回 NULL ]

我这样执行存储过程：

exec promediosDot
        @periodo1a = 201612,
        @periodo1b = 201703,
        @periodo2a = 201712,
        @periodo2b = 201803,
        @periodo3a = 201812,
        @periodo3b = 201903,
        @tipGas = 'VENTAS'

这是查询：

CREATE PROCEDURE promediosDot
    @periodo1a INT,
    @periodo1b INT,
    @periodo2a INT,
    @periodo2b INT,
    @periodo3a INT,
    @periodo3b INT,
    @tipGas NVARCHAR(6)
AS
    SELECT DISTINCT 
        T1.HOMOLOG_VP, 
        PROMEDIO1 = ((T2.CANT_PERS1 + T3.CANT_PERS2 + T4.CANT_PERS3 + T5.CANT_PERS4) / 4),
        PROMEDIO2 = ((T6.CANT_PERS1 + T7.CANT_PERS2 + T8.CANT_PERS3 + T9.CANT_PERS4) / 4),
        PROMEDIO3 = ((T10.CANT_PERS1 + T11.CANT_PERS2 + T12.CANT_PERS3 + T13.CANT_PERS4) / 4)
    FROM 
        (SELECT * 
         FROM USOS_CP 
         WHERE PER_PRO > 201611) T1
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, 
             COUNT(DISTINCT ID_HR) AS CANT_PERS1 
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo1a 
         GROUP BY 
             HOMOLOG_VP) T2 ON T1.HOMOLOG_VP = T2.HOMOLOG_VP
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS2 
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo1a + 1 
         GROUP BY 
             HOMOLOG_VP) T3 ON T1.HOMOLOG_VP = T3.HOMOLOG_VP
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS3 
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo1a + 2 
         GROUP BY 
             HOMOLOG_VP) T4 ON T1.HOMOLOG_VP = T4.HOMOLOG_VP
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS4 
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo1a + 3 
         GROUP BY 
             HOMOLOG_VP) T5 ON T1.HOMOLOG_VP = T5.HOMOLOG_VP
    --
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS1 
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo2a 
         GROUP BY 
             HOMOLOG_VP) T6 ON T1.HOMOLOG_VP = T6.HOMOLOG_VP
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS2
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo2a + 1
         GROUP BY 
             HOMOLOG_VP) T7 ON T1.HOMOLOG_VP = T7.HOMOLOG_VP
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS3
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo2a + 2
         GROUP BY 
             HOMOLOG_VP) T8 ON T1.HOMOLOG_VP = T8.HOMOLOG_VP
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS4
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo2a + 3
         GROUP BY 
             HOMOLOG_VP) T9 ON T1.HOMOLOG_VP = T9.HOMOLOG_VP
    --
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS1
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo3a
         GROUP BY 
             HOMOLOG_VP) T10 ON T1.HOMOLOG_VP = T10.HOMOLOG_VP   
     LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS2
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo3a + 1
         GROUP BY 
             HOMOLOG_VP) T11 ON T1.HOMOLOG_VP = T11.HOMOLOG_VP   
    LEFT JOIN 
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS3
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo3a + 2
         GROUP BY 
             HOMOLOG_VP) T12 ON T1.HOMOLOG_VP = T12.HOMOLOG_VP   
    LEFT JOIN
        (SELECT 
             HOMOLOG_VP, COUNT(DISTINCT ID_HR) AS CANT_PERS4
         FROM 
             DOTACION 
         WHERE 
             PERPRO = @periodo3a + 3
         GROUP BY 
             HOMOLOG_VP) T13 ON T1.HOMOLOG_VP = T13.HOMOLOG_VP   
GO

结果（摘录）：

HOMOLOG_VP  PROMEDIO1   PROMEDIO2   PROMEDIO3
---------------------------------------------
a3          NULL    NULL    NULL
a4          NULL    NULL    NULL

我不确定这只是参数的性质，直到某个时间点才起作用，或者在子查询上不起作用。通过手动编写参数，查询可以完美运行。感谢任何帮助。

Answer 1

您发布的代码中有几个参数未使用。但是您可以大大简化该查询。像这样的东西应该是同一件事。无需一遍又一遍地加入相同的 table。

SELECT cp.HOMOLOG_VP
    , PROMEDIO1 = count(case when d.PERPRO >= @periodo1a and d.PERPRO <= @periodo1a + 3 then 1 end) / 4
    , PROMEDIO2 = count(case when d.PERPRO >= @periodo2a and d.PERPRO <= @periodo2a + 3 then 1 end) / 4
    , PROMEDIO3 = count(case when d.PERPRO >= @periodo3a and d.PERPRO <= @periodo3a + 3 then 1 end) / 4
FROM USOS_CP cp
left join DOTACION d on d.HOMOLOG_VP = cp.HOMOLOG_VP
WHERE cp.PER_PRO > 201611
group by cp.HOMOLOG_VP

Answer 2

可能会出现问题，因为您将 periodo 设置为 12 月为 201612，如果您添加一个，它将变为 201613 而不是 201701。这是使用整数作为日期值的问题。正如 Sean 所提到的，这也可以转换为更简单的查询。

CREATE PROCEDURE promediosDot
        @periodo1a int,
        @periodo1b int,
        @periodo2a int,
        @periodo2b int,
        @periodo3a int,
        @periodo3b int,
        @tipGas nvarchar(6)

AS
SELECT  
    U.HOMOLOG_VP, 
    PROMEDIO1 = (   COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 0, CAST( @periodo1a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END) 
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 1, CAST( @periodo1a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 2, CAST( @periodo1a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 3, CAST( @periodo1a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)) /4,
    PROMEDIO2 = (   COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 0, CAST( @periodo2a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END) 
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 1, CAST( @periodo2a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 2, CAST( @periodo2a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 3, CAST( @periodo2a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)) /4,
    PROMEDIO3 = (   COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 0, CAST( @periodo3a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END) 
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 1, CAST( @periodo3a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 2, CAST( @periodo3a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)
                  + COUNT(DISTINCT CASE WHEN D.PERPRO = CONVERT( CHAR(6), DATEADD( MM, 3, CAST( @periodo3a AS VARCHAR(8)) + '01'), 112)  THEN D.ID_HR END)) /4
FROM USOS_CP  U
JOIN DOTACION D ON U.HOMOLOG_VP = D.HOMOLOG_VP
WHERE U.PER_PRO > 201611
AND (D.PERPRO BETWEEN @periodo1a AND @periodo1a + 3
    OR D.PERPRO BETWEEN @periodo2a AND @periodo2a + 3
    OR D.PERPRO BETWEEN @periodo3a AND @periodo3a + 3)
GROUP BY U.HOMOLOG_VP
ORDER BY HOMOLOG_VP;

用户定义的参数不适用于存储过程子查询

User defined parameters not working on stored procedure sub-query

sql

sql-server

parameters

stored-procedures

sql-server-2008