Question

我有项目表＆amp;用项目映射的多个任务。如果所有任务都在Project中完成，那么我尝试显示YES。如果未映射任务，则显示NA。这两个条件非常直接。

但是，如果我有多个任务映射到项目A＆amp;如果任务中的任何一个没有完成，并且在该项目A中完成了所有其他任务，我想显示NO。

我试过这个查询以获得低于o / p。

            SELECT DISTINCT PROJECT.[PROJ_ID],  
            CASE WHEN TASK_STA.TASK_STA_DS = 'Completed' THEN 'YES'  WHEN TASK_STA.TASK_STA_DS is null THEN 'NA' ELSE 'NO' END AS [Are all Tasks Completed]
            FROM PROJECT LEFT JOIN [PROJ_TASK] ON PROJECT.[PROJ_ID] = [PROJ_TASK].[PROJ_ID]
            LEFT JOIN dbo.TASK_STA
            ON [PROJ_TASK].TASK_STA_ID = TASK_STA.TASK_STA_ID

        PROJ_ID Completed
        20      NO
        20      YES
        21      NA
        22      YES
        24      NA

有没有办法让o / p如下所示。

        PROJ_ID Completed
        20      NO
        21      NA
        22      YES
        24      NA

Answer 1

您可以将group_concat函数用于已完成的列。

Answer 2

试试这个..........

    ;with cte as
  (
    SELECT  X.PROJECT.[PROJ_ID],X.[Are all Tasks Completed]
    FROM
    (
     SELECT DISTINCT PROJECT.[PROJ_ID],  
            CASE WHEN TASK_STA.TASK_STA_DS = 'Completed' THEN 'YES'  WHEN TASK_STA.TASK_STA_DS is null THEN 'NA' ELSE 'NO' END AS [Are all Tasks Completed]
            FROM PROJECT LEFT JOIN [PROJ_TASK] ON PROJECT.[PROJ_ID] = [PROJ_TASK].[PROJ_ID]
            LEFT JOIN dbo.TASK_STA
            ON [PROJ_TASK].TASK_STA_ID = TASK_STA.TASK_STA_ID
     ) X 
     GROUP BY X.PROJECT.[PROJ_ID],X.[Are all Tasks Completed]

    ) 

    select * from cte  a where a.[Are all Tasks Completed] = 'NO'  GROUP BY a.PROJECT.[PROJ_ID],a.[Are all Tasks Completed]
    union all
    select * from cte  b where  b.PROJECT.[PROJ_ID] not in ( select a.PROJECT.[PROJ_ID] from cte  a where a.[Are all Tasks Completed] = 'NO' )
            GROUP BY b.PROJECT.[PROJ_ID],b.[Are all Tasks Completed]

Answer 3

做一些额外的左连接来过滤掉不需要的条件。工作示例：

DECLARE @PROJECT TABLE (PROJ_ID int)
DECLARE @PROJTASK TABLE (PROJ_ID int, TASK_STA_ID int)
DECLARE @TASK_STA TABLE (TASK_STA_ID int, TASK_STA_DS varchar(10))

INSERT INTO @PROJECT
(PROJ_ID)
VALUES(20),
(21),
(22),
(24)

INSERT INTO @PROJTASK
(PROJ_ID, TASK_STA_ID)
VALUES(20, 201),
(20, 202),
(22, 221)

INSERT INTO @TASK_STA
(TASK_STA_ID, TASK_STA_DS)
VALUES(201, 'NO'),
(202, 'Completed'),
(221, 'Completed')

SELECT 
DISTINCT 
p.[PROJ_ID],  
    CASE 
    WHEN ts.TASK_STA_DS = 'Completed' THEN 'YES'  
    WHEN ts.TASK_STA_DS is null THEN 'NA' 
    ELSE 'NO' END AS [Are all Tasks Completed]
FROM @PROJECT p 
    LEFT JOIN @PROJTASK pt ON p.[PROJ_ID] = pt.[PROJ_ID]
    LEFT JOIN @TASK_STA ts ON pt.TASK_STA_ID = ts.TASK_STA_ID 

SELECT 
DISTINCT 
    p.[PROJ_ID],  
    CASE 
    WHEN ts.TASK_STA_DS = 'Completed' THEN 'YES'  
    WHEN ts.TASK_STA_DS is null THEN 'NA' 
    ELSE 'NO' END AS [Are all Tasks Completed]
FROM @PROJECT p 
    LEFT JOIN @PROJTASK pt ON p.[PROJ_ID] = pt.[PROJ_ID]
    LEFT JOIN @TASK_STA ts ON pt.TASK_STA_ID = ts.TASK_STA_ID 
    LEFT JOIN @PROJTASK pt2 ON p.[PROJ_ID] = pt.[PROJ_ID] AND pt.TASK_STA_ID > pt2.TASK_STA_ID 
    LEFT JOIN @TASK_STA ts_NOs ON pt2.TASK_STA_ID = ts_NOs.TASK_STA_ID AND ts.TASK_STA_DS = 'Completed' AND ts_NOs.TASK_STA_DS != 'Completed'
WHERE ts_NOs.TASK_STA_ID IS NULL

这是输出

PROJ_ID Are all Tasks Completed
20  NO
21  NA
22  YES
24  NA

SQL Query以避免重复ID

3 个答案: