Question

我有一个动态PIVOT查询，其中列是动态生成的。

我的桌子：ATTENDANCE_MASTER 包含：ID，Stud_id，ATT_DATE，PRESENT

存储数据，如：

ID  Stud_id ATT_DATE   PRESENT
1     1     2015-08-1    1
2     2     2015-08-1    0
3     3     2015-08-1    1
4     1     2015-08-2    0
5     2     2015-08-2    1
6     3     2015-08-2    1

我创建了PIVOT查询

DECLARE @columns NVARCHAR(MAX), @sql NVARCHAR(MAX);

SET @columns = N'';
SELECT @columns += N', p.' + QUOTENAME(ATT_DATE)
  FROM (SELECT p.ATT_DATE FROM dbo.ATTENDANCE_MASTER AS p
  GROUP BY p.ATT_DATE) AS x;

SET @sql = N'SELECT Stud_id, ' + STUFF(@columns, 1, 2, '') + '
FROM
(
  SELECT p.ATT_DATE, p.Stud_id, p.PRESENT FROM dbo.ATTENDANCE_MASTER AS p
) AS j
PIVOT
(
  SUM(PRESENT) FOR ATT_DATE IN ('+ STUFF(REPLACE(@columns, ', p.[', ',['), 1, 1, '') + ')
) AS p;';
PRINT @sql;
EXEC sp_executesql @sql;

我需要像

这样的列之和

Stud_ID  2015-08-01   2015-08-2 2015-08-3 Total
1            1            0         1      2
2            1            1         1      3
3            1            1         0      2
4            0            0         1      1

请建议我解决。

先谢谢。

Answer 1

您可以添加GROUP BY WITH CUBE以获得以下内容：

DECLARE @columns NVARCHAR(MAX), @sql NVARCHAR(MAX);

SET @columns = N'';
SELECT @columns += N', p.' + QUOTENAME(ATT_DATE)
  FROM (SELECT p.ATT_DATE FROM dbo.ATTENDANCE_MASTER AS p
  GROUP BY p.ATT_DATE) AS x;

SET @sql = N'SELECT Stud_id, ' + STUFF(@columns, 1, 2, '') + '
FROM
(
  SELECT p.ATT_DATE, p.Stud_id, p.PRESENT 
  FROM dbo.ATTENDANCE_MASTER AS p
  GROUP BY p.ATT_DATE, p.Stud_id, p.PRESENT 
  WITH CUBE
) AS j
PIVOT
(
  SUM(PRESENT) FOR ATT_DATE IN ('+ STUFF(REPLACE(@columns, ', p.[', ',['), 1, 1, '') + ')
) AS p;';
PRINT @sql;
EXEC sp_executesql @sql;

Answer 2

我首先建议不要使用变量连接来创建列列表。 It's behaviour is undefined and can be unexpected。而是使用SQL Server的XML扩展：

SET @Columns = (SELECT  N', p.' + QUOTENAME(p.Att_Date)
                FROM    dbo.ATTENDANCE_MASTER AS p
                GROUP BY p.ATT_DATE
                ORDER BY p.ATT_DATE
                FOR XML PATH(''), TYPE
                ).value('.', 'NVARCHAR(MAX)');

然后你可以简单地使用@Columns为总数创建一个表达式，所以使用：

', Total = ' + STUFF(REPLACE(@columns, ', p.[', ' + p.['), 1, 3, '')

你会得到类似的东西：

, Total = p.[2015-08-01] + p.[2015-08-02]

您可以将其添加到动态SQL中，因此对于完整的工作示例：

IF OBJECT_ID(N'tempdb..#T', 'U') IS NOT NULL DROP TABLE #T;
CREATE TABLE #T ([ID] int, [Stud_id] int, [ATT_DATE] datetime, [PRESENT] int);

INSERT INTO #T ([ID], [Stud_id], [ATT_DATE], [PRESENT])
VALUES
    (1, 1, '2015-08-01 00:00:00', 1),
    (2, 2, '2015-08-01 00:00:00', 0),
    (3, 3, '2015-08-01 00:00:00', 1),
    (4, 1, '2015-08-02 00:00:00', 0),
    (5, 2, '2015-08-02 00:00:00', 1),
    (6, 3, '2015-08-02 00:00:00', 1);

DECLARE @columns NVARCHAR(MAX), @sql NVARCHAR(MAX);             

SET @Columns = (SELECT  N', p.' + QUOTENAME(REPLACE(CONVERT(VARCHAR(10), p.Att_Date, 111), '/', '-'))
                FROM    #T AS p
                GROUP BY p.ATT_DATE
                ORDER BY p.ATT_DATE
                FOR XML PATH(''), TYPE
                ).value('.', 'NVARCHAR(MAX)');

SET @sql = N'SELECT Stud_id, ' + STUFF(@columns, 1, 2, '') + ', Total = ' + STUFF(REPLACE(@columns, ', p.[', ' + p.['), 1, 3, '') + '
FROM
(
  SELECT p.ATT_DATE, p.Stud_id, p.PRESENT FROM #T AS p
) AS j
PIVOT
(
  SUM(PRESENT) FOR ATT_DATE IN ('+ STUFF(REPLACE(@columns, ', p.[', ',['), 1, 1, '') + ')
) AS p;';
PRINT @sql;
EXEC sp_executesql @sql;

SQL Server中PIVOT表中动态列的SUM

2 个答案: