Question

我打算使用pivot

将以下数据转换为透视数据

SRNO        DATE        TYPE        USERNAME
--------------------------------------------
1           16/07/2016  OPENING     JOHN
2           13/07/2016  SALES       JOHN
3           12/07/2016  OPENING     JOHN
4           11/07/2016  PURCHASE    ADWARD
5           16/07/2016  OPENING     JOHN
6           13/07/2016  SALES       JOHN
7           12/07/2016  SALES       JOHN
8           12/07/2016  OPENING     JOHN
9           19/07/2016  OPENING     JOHN
10          12/07/2016  PURCHASE    ADWARD

我想要以下格式的上表数据

OPENING_SRNO        OPENING_DATE        OPENING_USERNAME        SALES_SRNO        SALES_DATE        SALES_USERNAME        PURCHASE_SRNO        PURCHASE_DATE        PURCHASE_USERNAME
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1                   16/07/2016          JOHN                    2                 13/07/2016        JOHN                  4                    11/07/2016           ADWARD     
3                   12/07/2016          JOHN                    6                 13/07/2016        JOHN                 10                    12/07/2016           ADWARD    
5                   16/07/2016          JOHN                    7                 12/07/2016        JOHN
8                   12/07/2016          JOHN
9                   19/07/2016          JOHN

TYPE列具有动态值，我发现在pivot in集合中应用动态列时遇到了困难。

更新

我尝试过关注，但NULL列中的一些TYPE值给了我错误。

DECLARE @cols AS NVARCHAR(MAX),
    @query  AS NVARCHAR(MAX),
    @COLS1 AS NVARCHAR(MAX)

SELECT @COLS = STUFF((SELECT DISTINCT ','+QUOTENAME(case when TYPE IS NULL then 'empty' else TYPE end) 
                        FROM TEMP
          FOR XML PATH(''), TYPE
        ).value('.', 'NVARCHAR(MAX)') 
    ,1,1,'')
PRINT(@COLS) 

set @query = 'SELECT ' + @cols + ' from 
             (
                select SRNO,DATE, USERNAME
                from TEMP
            ) x
            pivot 
            (
                MAX(SRNO)
                for TYPE in (' + @cols + ')
            ) p '

            PRINT @QUERY
execute(@query);

Answer 1

不确定，此问题适用于PIVOT ...
不过，这应该可以让你开始：

SELECT
  opening_srno, opening_date, opening_username, sales_srno, sales_date,
  sales_username, purchase_srno, purchase_date, purchase_username
FROM
  (SELECT
    srno OPENING_SRNO
    , d_date OPENING_DATE
    , username OPENING_USERNAME
    , RANK() OVER (ORDER BY srno) rnk
  FROM Data
  WHERE d_type = 'OPENING'
  ) opening
LEFT OUTER JOIN
    (SELECT
      srno SALES_SRNO
      , d_date SALES_DATE
      , username SALES_USERNAME
      , RANK() OVER (ORDER BY srno) rnk
    FROM Data
    WHERE d_type = 'SALES'
    ) sales
  ON opening.rnk = sales.rnk
LEFT OUTER JOIN
    (SELECT
      srno PURCHASE_SRNO
      , d_date PURCHASE_DATE
      , username PURCHASE_USERNAME
      , RANK() OVER (ORDER BY srno) rnk
    FROM Data
    WHERE d_type = 'PURCHASE'
    ) purchase
  ON opening.rnk = purchase.rnk
ORDER BY opening.rnk
;

抱歉 - 无法访问要验证的SQL Server实例。（经过Oracle测试。）

^{未使用 TYPE 和 DATE 作为列名，因为它们是SQL保留字。}

请评论是否需要调整/进一步详细说明。

SQL Server：将行数据作为列

1 个答案: