ibatis 实现 物理级别的 分页 兼容多种数据库(转载)

最近在看iBatis时，想做用动态Sql做个分布。因为在做项目时用iBator工具生成没有分页的功能，只有一些我们常用的功能。所以要对生成后的代码做修改。我在Java高手真经的一书中看到有做了MySql SqlServer Oracle的分页。实现如下： 
MySql--> 
首先说说MySql的分页语句

select  *  from   user   where ... order   by ... limit 10,25     


根据以上的语句我们可以对ibator生成后的代码进行修改。 第一修改User_sqlMap.xml 修改前

< select   id = "ibatorgenerated_selectByExample"           resultMap = "ibatorgenerated_BaseResultMap"
     parameterClass = "com.demo.ibatis.beans.UserExample" >
    select ID, LOGIN_NAME, PASSWORD
    from user
     < isParameterPresent >
         < include   refid = "user.ibatorgenerated_Example_Where_Clause"   />
                < isNotNull   property = "orderByClause" >
                order by $orderByClause$
                        </ isNotNull >
         </ isParameterPresent >
</ select >

修改后

< select   id = "ibatorgenerated_selectByExample"   resultMap = "ibatorgenerated_BaseResultMap"
         parameterClass = "com.demo.ibatis.beans.UserExample" >
        select ID, LOGIN_NAME, PASSWORD
        from user
     < isParameterPresent >
         < include   refid = "user.ibatorgenerated_Example_Where_Clause"   />
             < isNotNull   property = "orderByClause" >
                order by $orderByClause$
                         </ isNotNull >
            [b] < isNotNull   property = "limitClauseStart" >
                limit
                $limitClauseStart$,$limitClauseCount$
             </ isNotNull > [/b]
     </ isParameterPresent >
</ select >

重要的是黑色的字体是添加到原来的xml中的。接下来修改UserExample.java 
在该类中添加两个字段和相应的getter/setter方法

protected  String limitClauseStart; //起始参数
protected  String limitClauseCount; //数量参数

接下来就可以通过Dao调用selectByExample()方法测试,好的以上就是MySql的分页。SqlServer的分页比较难了，因为 SqlServer中没有limit的关键字。 
看看SqlServer的分页语句，那当然 SqlServer人分布方法有很多种。这里我采用其中的一种。 
语句如下：

   SELECT   TOP  页大小 *
   FROM  TestTable
   WHERE  (ID  NOT   IN
          ( SELECT   TOP  页大小*页数 id
          FROM  表
          ORDER   BY  id))
ORDER   BY  ID

接下来和上面的MySql一样也进行修改 
修改xml

< select   id = "ibatorgenerated_selectByExample"   resultMap = "ibatorgenerated_BaseResultMap"
         parameterClass = "com.demo.ibatis.beans.UserExample" >
        [b]  < isNotNull   property = "limitClauseStart" >
            SELETE TOP $limitClauseCount$ * FROM user where (ID NOT IN
                       (SELECT TOP $limitClauseStart$ ID
                   </ isNotNull > [/b]
                    < isNull   property = "limitClauseStart" >
            SELETE *
           </ isNull >
                    FROM user
        from user
     < isParameterPresent >
         < include   refid = "user.ibatorgenerated_Example_Where_Clause"   />
             < isNotNull   property = "orderByClause" >
                order by $orderByClause$
                               [b]   < isNotNull   property = "limitClauseStart" >
                     ))order by $orderByClause$
                     </ isNotNull > [/b]
                         </ isNotNull >
     </ isParameterPresent >
</ select >   


接下来是ORacle的，和上面都是差不多的这里不多说。 语法： select* from(select rownum tid,user.* FROM (select * from user where id>1 order by ID desc)user where where rownum<35) where tid>10; 修改Xml,

< select   id = "ibatorgenerated_selectByExample"       resultMap = "ibatorgenerated_BaseResultMap"
         parameterClass = "com.demo.ibatis.beans.UserExample" >
        [b]  < isNotNull   property = "limitClauseStart" >
            select* from(select rownum tid,user.* FROM (
                   </ isNotNull > [/b]
                   select * from user
     < isParameterPresent >
         < include   refid = "user.ibatorgenerated_Example_Where_Clause"   />
             < isNotNull   property = "orderByClause" >
                order by $orderByClause$
                         </ isNotNull >
                        [b]   < isNotNull   property = "limitClauseStart" >
                    <![[CDATA )user where where rownum<$limitClauseCount$+$limitClauseStart$+1) where tid>$limitClauseStart$;]]>
                  </ isNotNull > [/b]
     </ isParameterPresent >
</ select >