Question

请帮助解决此SQL Server问题。我有一个表员工以及下面列出的字段（请参阅粗体员工表）。我想得到如下所示的结果（请参阅粗体结果）。

员工表：

Name
Age
Salary
Employee Type
Employed Date

员工数据：

Name        Age   Salary      Employee_Type   Date_Hired
Employee 1  24    24,000.00   Staff           January 15, 2009
Employee 2  33    32,000.00   Staff           February 1, 2005
Employee 3  23    75,000.00   Consultant      March 20, 2002

结果：

          Employee 1           Employee 2           Employee 3
Age           24                    33                  23
Salary        24,000.00             32,000.00           75,000.00
Employee_Type Staff                 Staff               Consultant
Date_Hired    January 15, 2009      February 1, 2005    March 20, 2002

Answer 1

Create Table #Employee (Name Varchar(20), Age TinyInt, Salary SmallMoney, EmployeeType Varchar(20), EmployedDate Date)

Insert Into #Employee
Select 'Employee 1', 24, 24000.00, 'Staff', '01/15/2009' Union all 
Select 'Employee 2', 33, 32000.00, 'Staff', '02/01/2005' Union All 
Select 'Employee 3', 23, 75000.00, 'Consultant', '03/20/2002'

Declare @Names As Varchar(Max), @strSQL Varchar(Max)
Select @Names = Stuff((Select ',' + QuoteName(Name) From #Employee Order by Name For XML Path('')), 1, 1, '')
Set @strSQL = ';With CTE As (Select Name, FieldName, FieldValue From (Select Name, Convert(Varchar(20), Age) As Age, Convert(Varchar(20), Salary) As Salary, Convert(Varchar(20), EmployeeType) As EmployeeType, Convert(Varchar(20), EmployedDate, 101) As EmployedDate From #Employee )As p UnPivot (FieldValue For FieldName In (Age, Salary, EmployeeType, EmployedDate) )AS unpvt) Select * From CTE Pivot (Max(FieldValue) For Name in (' + @Names + ')) As Pvt'
print @strSQL
Execute (@strSQL)

Drop Table #Employee

感谢Arbi Baghdanian，:)）

Answer 2

这是一种做法。 UNPIVOT用于将列名称传输到行，然后使用该ColumnName（现在是一行）的标准分组来为每个员工转动名称/值组合。请注意，使用unpivot时，列应具有相同的数据类型，这就是存在转换和转换的原因。这种方法的唯一缺点是员工姓名是硬编码的，但除非使用动态旋转，否则这总是一个问题。但是，既然你提到你只有大约10名员工参加corsstab，那应该不是问题。这是脚本：

select
    unp.ColumnName,
    max(case when Name = 'Employee 1' then Value else null end) [Employee 1],
    max(case when Name = 'Employee 2' then Value else null end) [Employee 2],
    max(case when Name = 'Employee 3' then Value else null end) [Employee 3]
    from 
    (
        select 
            Name, cast(Age as varchar(20)) Age, 
            cast(Salary as varchar(20)) Salary, 
            cast(Employee_Type as varchar(20)) Employee_Type,
            convert(varchar(20), Employed_Date, 106) Employed_Date  
            from dbo.Employee
    ) src
    unpivot 
    (
        Value for ColumnName in (
            [Age], [Salary], [Employee_Type], [Employed_Date])
    ) unp
    group by unp.ColumnName;

-- the above produces something like this for 
-- the data consistent with the sample in question:

ColumnName           Employee 1           Employee 2           Employee 3
-------------------- -------------------- -------------------- -----------
Age                  24                   33                   23
Employed_Date        15 Jan 2009          01 Feb 2005          20 Mar 2002
Employee_Type        Staff                Staff                Consultant
Salary               24000.00             32000.00             75000.00

感谢Oleg。：）

Answer 3

Declare  @strSQL1 Varchar(Max)

Declare @Names1 As Varchar(Max)
Select @Names1 = Stuff((Select ',' + 'max(case when Name = '''+Name+'''then Value else null end)'+ QuoteName(Name) From Employee Order by Name For XML Path('')), 1, 1, '')
print @Names1

Set @strSQL1 = 'select
    unp.ColumnName,'+@Names1+ ' from 
    (
        select 
            Name, cast(Age as varchar(20)) Age, 
            cast(Salary as varchar(20)) Salary, 
            cast(Employee_Type as varchar(20)) Employee_Type,
            convert(varchar(20), Date_Hired, 106) Date_Hired
            from dbo.Employee
    ) src
    unpivot 
    (
        Value for ColumnName in (
            [Age], [Salary], [Employee_Type], Date_Hired)
    ) unp
    group by unp.ColumnName; ' 

print @strSQL1
Execute (@strSQL1)

另一个动态解决方案 - 升级了Oleg的解决方案

请帮助解决这个SQL Server问题

3 个答案: