Question

如果问题重复，我道歉但我有一些要求让我感到紧张。经过多次尝试，我无法解决它。要求是我有一个表格，其数据如下：

CHANGE_NUMBER   ITEM_NUMBER REV BOM_ITEM    FIND_NUMBER QUANTITY
MD04086         710851-1    B   230248-0           0    9
MD04086         710851-1    B   230289-0           0    4
MD04086         710851-1    B   230292-0           0    1
MD04086         710851-1    B   230298-0           0    2

我想对这些数据做些什么是按照以下格式制作：

Column_name     Column_Value

CHANGE_NUMBER   MD04086
ITEM_NUMBER     710851-1
REV             B
BOM_ITEM        230248-0
FIND_NUMBER     0
QUANTITY        9
CHANGE_NUMBER   MD04086
ITEM_NUMBER     710851-1
REV             B
BOM_ITEM        230289-0
FIND_NUMBER     0
QUANTITY        4

到目前为止，我使用了以下查询，但它没有提供我想要实现的内容

with q1 as
(select  'itemRev' as column_name,[CHANGE_NUMBER] as column_value  from  generalassemblies_data_final
union all

select  'itemNumber' as column_name,[ITEM_NUMBER] as column_value  from  generalassemblies_data_final
union all

select  'REV' as column_name,[REV] as column_value  from  generalassemblies_data_final
union all

select  'BOM_ITEM' as column_name,[BOM_ITEM] as column_value  from  generalassemblies_data_final
union all

select  'findNum' as column_name,[FIND_NUMBER] as column_value  from  generalassemblies_data_final
union all

select  'qty' as column_name,[QUANTITY] as column_value  from  generalassemblies_data_final

)

select *
from q1;

有人可以帮帮我吗。

Answer 1

Select   ColumnName
       , ColumnValue
FROM 
(
  --<-- Make sure all the columns are converted to the same data type here
     -- This is the only purpose of this sub-query 
    Select CHANGE_NUMBER      
          ,ITEM_NUMBER          
          ,REV
          ,BOM_ITEM
          ,CAST(FIND_NUMBER AS VARCHAR(100)) AS FIND_NUMBER
          ,CAST(QUANTITY    AS VARCHAR(100)) AS QUANTITY
    from TableName --<-- Your Table name here 
)t  
   UNPIVOT (ColumnValue for ColumnName 
           IN (CHANGE_NUMBER , ITEM_NUMBER , REV
               ,BOM_ITEM , FIND_NUMBER , QUANTITY)
            )UP

Answer 2

如果column_values全部转换为相同的类型（nvarchar？），您的UNION ALL列表将会起作用。除此之外，我建议您查看表示级问题，该问题由可以使用查询结果的程序处理。

我认为任何报表设计师都可以工作。如果您希望在3GL等3GL中执行此操作，则只需保留基于行的查询并写出格式化数据：

for each row in someresult
  Console.WriteLine(string.format("Change Number" & vbtab & "{0}", row.ChangeNumber)
  Console.WriteLine(string.format("Item Number" & vbtab & "{0}", row.ItemNumber)
next

我无法在这里为您提供精确的语法，但结果，输出流和格式的混合可以起作用。

Answer 3

您也可以使用APPLY运算符执行此类操作：

select Transposed.*
from <TableName>
cross apply (values
    (CHANGE_NUMBER)
  , (ITEM_NUMBER)
  , (REV)
  , (BOM_ITEM)
  , (FIND_NUMBER)
  , (QUANTITY)
  , (CHANGE_NUMBER)
  , (ITEM_NUMBER)
  , (REV)
  , (BOM_ITEM)
  , (Cast(FIND_NUMBER as varchar(10)))
  , (Cast(QUANTITY as varchar(10)))
) as Transposed (ChangeNumber, ItemNumber, Revision, BomItem, FindNumber, Quantity);

我更喜欢这个版本，因为我已经在常规版本上使用APPLY，发现它比UNPIVOT更容易记住。另外，我只使用PascalCase对列进行别名以突出显示语法;随时将它们改为你想要的任何东西。

将列转换为每行的行

3 个答案: