上一篇下一篇 排序 (非ID字段排序)

　　网上看了很多关于"上一篇下篇"的文章，可大都是按ID排序。

　　实际上，很少有按ID排序的。

　　分享下我的单独排序字段的写法，主要分为ms sql2000 和 ms 2005及以上版本。

　　SQL 2005及以上版本写法。

 declare @currentID int
 set @currentID=4

 ;with temp
 as
 (
     select row_number() over( order by short desc , createtime desc) as rowNum, *
     from ch_ItemInformation
 )
 select * from temp where rowNum in ((select rowNum from temp where id=@currentID)+1)
 union all
 select * from temp where rowNum in ((select rowNum from temp where id=@currentID)-1)

SQL 2000写法

 --前提：排序后要唯一，即排序后不能有两条
 declare @Num int
 set @Num=1
 select @Num=@Num+1 from dbo.ch_ItemInformation
 where Short>=0 and ID >=10
 order by Short desc, ID desc

 set @Num=@Num+1
 select @Num

 select top 3 * from (
 select top (select @Num) * from ch_ItemInformation order by Short desc, ID desc
 ) as a
 order by a.Short desc, a.ID desc 

有网友问: 既然使用了row_number() 函数，where 中为何还用in？

     说明下: 只是为了找到ID对应的排序后的排序号. 改成等于号也行. 如下:

declare @currentID int
set @currentID=1

;with temp
as
(
    select row_number() over( order by short desc , createtime desc) as rowNum, *
    from dbo.Product
)
select * from temp where rowNum = ((select rowNum from temp where id=@currentID)+1)
union all
select * from temp where rowNum = ((select rowNum from temp where id=@currentID)-1)