Question

我的sql表数据设置如下

 SheetNo  VersioId ColumnId Value
  Sheet1    1      column1     3
  Sheet1    3      column1     7
  Sheet1    1      column2     A
  Sheet1    1      column3     Z
  Sheet1    4      column3     T
  Sheet1    1      column5     R
  Sheet1    2      column5     Q

我想编写一个SQL查询来转换此数据集，如下所示

 SheetNo  VersioId ColumnId Value
 Sheet1    3      column1     7
 Sheet1    1      column2     A
 Sheet1    4      column3     T
 Sheet1    2      column5     Q

我真正需要的是获取其最大版本ID的每个列值。有人可以帮我这个

Answer 1

这可能会对您有所帮助：

select y1.* from YourTable y1
join
(select Columnid,max(versionid) versionid from YourTable 
group by Columnid) as y2
on y1.Columnid = y2.Columnid and y1.versionid = y2.versionid

Answer 2

基本上，从基表中获取SheetNo和ColumnId，然后交叉应用以通过VersioId获取VersioId和Value的顶部。分组结果以折叠行，并按您的喜好排序。可以使用相同的结构通过最大的某个不同列来获取其他数据。请参阅SqLFiddle example。

select A.SheetNo, B.VersioId, A.ColumnId, B.Value
from Sheets A
  cross apply (
    select top 1 VersioId, Value
    from Sheets C
    where C.SheetNo = A.SheetNo
      and C.ColumnId = A.ColumnId
    order by C.VersioId desc
    ) B
group by A.SheetNo, B.VersioId, A.ColumnId, B.Value
order by A.SheetNo, A.ColumnId

转换SQL数据集

2 个答案: