SQL Server 2008 透视查询 - 日期时间答案

【问题标题】：SQL Server 2008 Pivot Query - DatetimeSQL Server 2008 透视查询 - 日期时间
【发布时间】：2016-10-21 05:25:53
【问题描述】：

我在多行中有员工进出时间的记录，如何查询具有相同 id、名称但进出时间不同的数据透视？

如下图所示：

【问题讨论】：

标签： sql-server-2008 pivot pivot-table

【解决方案1】：

您可以像这样使用旋转：

;WITH dtrTable AS (  --just a test sample of your table
SELECT *
FROM (VALUES
(1, 'emp1', '2016-10-20', '2016-10-20 10:00:00.000', '2016-10-20 15:00:00.000'),
(1, 'emp1', '2016-10-20', '2016-10-20 15:30:00.000', '2016-10-20 17:00:00.000'),
(1, 'emp1', '2016-10-20', '2016-10-20 18:30:00.000', '2016-10-20 19:00:00.000'),
(2, 'emp2', '2016-10-20', '2016-10-20 10:00:00.000', '2016-10-20 19:00:00.000'),
(2, 'emp2', '2016-10-20', '2016-10-20 21:00:00.000', '2016-10-20 22:00:00.000'),
(2, 'emp2', '2016-10-21', '2016-10-20 11:00:00.000', '2016-10-20 21:00:00.000')
) as t(empId, empName, dtrDate, dtrIn, dtrOut)
)

SELECT *
FROM (
    SELECT  empId, 
            empName, 
            dtrDate,
            [Columns]+seq as [Columns],
            [Values]
    FROM (
        SELECT *,
                CAST(ROW_NUMBER() OVER (PARTITION BY empId, dtrDate ORDER BY dtrIn) as nvarchar(100)) as seq
        FROM dtrTable
    ) as t
    UNPIVOT (
        [Values] FOR [Columns] IN ([dtrIn],[dtrOut])
    ) as unpvt
) as d
PIVOT (
    MAX([Values]) FOR [Columns] IN ([dtrIn1],[dtrOut1],[dtrIn2],[dtrOut2],[dtrIn3],[dtrOut3])
) as pvt

输出：

empId   empName dtrDate     dtrIn1                  dtrOut1                 dtrIn2                  dtrOut2                 dtrIn3                  dtrOut3
1       emp1    2016-10-20  2016-10-20 10:00:00.000 2016-10-20 15:00:00.000 2016-10-20 15:30:00.000 2016-10-20 17:00:00.000 2016-10-20 18:30:00.000 2016-10-20 19:00:00.000
2       emp2    2016-10-20  2016-10-20 10:00:00.000 2016-10-20 19:00:00.000 2016-10-20 21:00:00.000 2016-10-20 22:00:00.000 NULL                    NULL
2       emp2    2016-10-21  2016-10-20 11:00:00.000 2016-10-20 21:00:00.000 NULL                    NULL                    NULL                    NULL

Unpivot 部分将为您提供此表：

empId   empName dtrDate     Columns Values
1       emp1    2016-10-20  dtrIn1  2016-10-20 10:00:00.000
1       emp1    2016-10-20  dtrOut1 2016-10-20 15:00:00.000
1       emp1    2016-10-20  dtrIn2  2016-10-20 15:30:00.000
1       emp1    2016-10-20  dtrOut2 2016-10-20 17:00:00.000
1       emp1    2016-10-20  dtrIn3  2016-10-20 18:30:00.000
1       emp1    2016-10-20  dtrOut3 2016-10-20 19:00:00.000
2       emp2    2016-10-20  dtrIn1  2016-10-20 10:00:00.000
2       emp2    2016-10-20  dtrOut1 2016-10-20 19:00:00.000
2       emp2    2016-10-20  dtrIn2  2016-10-20 21:00:00.000
2       emp2    2016-10-20  dtrOut2 2016-10-20 22:00:00.000
2       emp2    2016-10-21  dtrIn1  2016-10-20 11:00:00.000
2       emp2    2016-10-21  dtrOut1 2016-10-20 21:00:00.000

这里我使用ROW_NUMBER() 和分区来给进出一些顺序。还有由ROW_NUMBER() 生成的数字有助于创建列，我们不能在数据透视部分使用相同的列名。

【讨论】：