C#操作数据库，ROW_NUMBER分页。

从这里http://www.cnblogs.com/bluedoctor/archive/2011/11/01/2232297.html听说row_number分页在高版本的MSSQL中最快的。

专门看了下row_number用法

ROW_NUMBER生成以一个以表里的某一个列为排序[这里的排序效果与select * from table order by id 这种效果一样]的虚拟列（看过也可以用getdate()），并自动生成编号。

比如select ROW_NUMBER() over (order by id) as RowId,* from adonet查出的结果是以ID排序，在前面多加一列RowId。

RowId  id            name       addr                datet

1    000000    onepc0    cnblogs - 0    2011-10-25 18:25:04.093
2    000001    onepc1    cnblogs - 1    2011-10-25 18:55:04.093
3    000002    onepc2    cnblogs - 2    2011-10-25 19:25:04.093
4    000003    onepc3    cnblogs - 3    2011-10-25 19:55:04.093
5    000004    onepc4    cnblogs - 4    2011-10-25 20:25:04.093
6    000005    onepc5    cnblogs - 5    2011-10-25 20:55:04.093
7    000006    onepc6    cnblogs - 6    2011-10-25 21:25:04.093
8    000007    onepc7    cnblogs - 7    2011-10-25 21:55:04.093
9    000008    onepc8    cnblogs - 8    2011-10-25 22:25:04.093
10    000009    onepc9    cnblogs - 9    2011-10-25 22:55:04.093

 

select ROW_NUMBER() over (order by id desc) as RowId,* from adonet ---ID为desc序排列

RowId  id            name       addr                datet

1    046606    00000    cnblogs - 46606    2014-06-22 17:25:04.093
2    046605    11    cnblogs - 46605    2014-06-22 16:55:04.093
3    046604    onepc46604    cnblogs - 46604    2014-06-22 16:25:04.093
4    046603    onepc46603    cnblogs - 46603    2014-06-22 15:55:04.093
5    046602    onepc46602    cnblogs - 46602    2014-06-22 15:25:04.093
6    046601    onepc46601    cnblogs - 46601    2014-06-22 14:55:04.093
7    046600    onepc46600    cnblogs - 46600    2014-06-22 14:25:04.093
8    046599    onepc46599    cnblogs - 46599    2014-06-22 13:55:04.093
9    046598    onepc46598    cnblogs - 46598    2014-06-22 13:25:04.093
10    046597    onepc46597    cnblogs - 46597    2014-06-22 12:55:04.093

 

select ROW_NUMBER() over (order by id desc) as RowId,* from adonet order by id或者rowid 这里排的话，查下表就清楚了。

 

 

MSSQL存储过程

--select id,name,addr,datet from adonet
--select id,name,addr,datet,ROW_NUMBER() over (order by datet) aaa from adonet order by aaa desc
--select GETDATE()
--select * from (select ROW_NUMBER() over (order by id) as RowId,* from adonet) temptable 
--where RowId>=1 and RowId<=100
use wentest
if exists(select * from sys.sysobjects where type='p' and name='row_fy')
    drop procedure row_fy
go
create procedure row_fy
    @pagesize int, --分页大小
    @page int --当前第几页
    --@outfynum int output
with encryption --加密
as
    --declare @temp int
    select * from 
        (
            select ROW_NUMBER() over (order by id) as RowId,* from 
            adonet
        ) as temptable
    where RowId>=(@page-1)*@pagesize+1 and RowId<=@page*@pagesize

执行

row_fy 100,1