Hibernate处理Oracle的分页，是用rownum。

代码如下：

Java代码

public String getLimitString(String sql, boolean hasOffset);    
{   
  StringBuffer pagingSelect = new StringBuffer(sql.length(); + 100);;   
  if (hasOffset); {   
    pagingSelect.append(   
        "select * from ( select row_.*, rownum rownum_ from ( ");;   
  }   
  else {   
    pagingSelect.append("select * from ( ");;   
  }   
  pagingSelect.append(sql);;   
  if (hasOffset); {   
    pagingSelect.append(" ); row_ where rownum <= ?); where rownum_ > ?");;   
  }   
  else {   
    pagingSelect.append(" ); where rownum <= ?");;   
  }   
  return pagingSelect.toString();;   
}  
[java] view plain copy
public String getLimitString(String sql, boolean hasOffset);   
{  
  StringBuffer pagingSelect = new StringBuffer(sql.length(); + 100);;  
  if (hasOffset); {  
    pagingSelect.append(  
        "select * from ( select row_.*, rownum rownum_ from ( ");;  
  }  
  else {  
    pagingSelect.append("select * from ( ");;  
  }  
  pagingSelect.append(sql);;  
  if (hasOffset); {  
    pagingSelect.append(" ); row_ where rownum <= ?); where rownum_ > ?");;  
  }  
  else {  
    pagingSelect.append(" ); where rownum <= ?");;  
  }  
  return pagingSelect.toString();;  
}  
  public String getLimitString(String sql, boolean hasOffset); 
  {
    StringBuffer pagingSelect = new StringBuffer(sql.length(); + 100);;
    if (hasOffset); {
      pagingSelect.append(
          "select * from ( select row_.*, rownum rownum_ from ( ");;
    }
    else {
      pagingSelect.append("select * from ( ");;
    }
    pagingSelect.append(sql);;
    if (hasOffset); {
      pagingSelect.append(" ); row_ where rownum <= ?); where rownum_ > ?");;
    }
    else {
      pagingSelect.append(" ); where rownum <= ?");;
    }
    return pagingSelect.toString();;
  }

出错前提

如果这里的sql是不带order by的sql，则查询结果没有任何问题。

但是，如果sql中带有order by，则会引起混乱，即相同记录会出现在不同页中。但是，这种混乱的出现通常是在下面的情况下：

1、纪录数足够多（如果表中有lob字段更好：P）

2、插入记录数大于3页，每页最好10＋条记录

3、order by字段至少需要有2个值

4、具备相同order by字段的记录数大于3页

5、插入记录后，最好做删除、修改操作，然后再插入记录。保证记录在磁盘环境中的顺序是无序的。

6、如果满足上述条件，但还没有出现混乱现象，则适当的加大纪录数。

大家测试这种现象，不需要通过hibernate，直接通过sql就可查到。下面有由hibernate生成的sql，以供大家试验：

select * from
  ( select row_.*, rownum rownum_ from
     ( select * from T_TABLE tTable where tTable.field1 order by  tTable.field2 desc ) row_ where rownum <= ?) where rownum_ > ?;

错误原因

之所以出现这样的问题，是和oracle处理ROWNUM的原理相关的。

以下是Oracle参考手册上的一段话：

引用

ROWNUM返回第一次从表中选择时返回的行的序列号。第一行的ROWNUM为1，第二行的为2，依此类推。但要注意， 即使select语句中一条简单的order by都可能会搞乱ROWNUM（因为ROWNUM是排序前分配给各行的）。

解决办法（来自使用手册）：

9.3.3. Scrollable iteration

If your JDBC driver supports scrollable ResultSets, the Query interface may be used to obtain a ScrollableResults which allows more flexible navigation of the query results. (Oracle 8.1.6+)

Java代码

Query q = sess.createQuery("select cat.name, cat from DomesticCat cat " +   
                            "order by cat.name");;   
ScrollableResults cats = q.scroll();;   
if ( cats.first(); ); {   
  
    // find the first name on each page of an alphabetical list of cats by name   
    firstNamesOfPages = new ArrayList();;   
    do {   
        String name = cats.getString(0);;   
        firstNamesOfPages.add(name);;   
    }   
    while ( cats.scroll(PAGE_SIZE); );;   
  
    // Now get the first page of cats   
    pageOfCats = new ArrayList();;   
    cats.beforeFirst();;   
    int i=0;   
    while( ( PAGE_SIZE > i++ ); && cats.next(); ); pageOfCats.add( cats.get(1); );;   
  
}  
[java] view plain copy
Query q = sess.createQuery("select cat.name, cat from DomesticCat cat " +  
                            "order by cat.name");;  
ScrollableResults cats = q.scroll();;  
if ( cats.first(); ); {  
  
    // find the first name on each page of an alphabetical list of cats by name   
    firstNamesOfPages = new ArrayList();;  
    do {  
        String name = cats.getString(0);;  
        firstNamesOfPages.add(name);;  
    }  
    while ( cats.scroll(PAGE_SIZE); );;  
  
    // Now get the first page of cats   
    pageOfCats = new ArrayList();;  
    cats.beforeFirst();;  
    int i=0;  
    while( ( PAGE_SIZE > i++ ); && cats.next(); ); pageOfCats.add( cats.get(1); );;  
  
}  
Query q = sess.createQuery("select cat.name, cat from DomesticCat cat " +
                            "order by cat.name");;
ScrollableResults cats = q.scroll();;
if ( cats.first(); ); {
 
    // find the first name on each page of an alphabetical list of cats by name
    firstNamesOfPages = new ArrayList();;
    do {
        String name = cats.getString(0);;
        firstNamesOfPages.add(name);;
    }
    while ( cats.scroll(PAGE_SIZE); );;
 
    // Now get the first page of cats
    pageOfCats = new ArrayList();;
    cats.beforeFirst();;
    int i=0;
    while( ( PAGE_SIZE > i++ ); && cats.next(); ); pageOfCats.add( cats.get(1); );;
 
}

我觉得，如果处理数据库类是一个统一的基类，这种方法不适合用。

11.13 Tips & Tricks

1、Collection elements may be ordered or grouped using a query filter:

Java代码

Collection orderedCollection = s.filter( collection, "order by this.amount" );;   
Collection counts = s.filter( collection, "select this.type, count(this); group by this.type" );;  
[java] view plain copy
Collection orderedCollection = s.filter( collection, "order by this.amount" );;  
Collection counts = s.filter( collection, "select this.type, count(this); group by this.type" );;  
Collection orderedCollection = s.filter( collection, "order by this.amount" );;
Collection counts = s.filter( collection, "select this.type, count(this); group by this.type" );;

2、Collections are pageable by using the Query interface with a filter:

Java代码

Query q = s.createFilter( collection, "" );; // the trivial filter   
q.setMaxResults(PAGE_SIZE);;   
q.setFirstResult(PAGE_SIZE * pageNumber);;   
List page = q.list();;  
[java] view plain copy
Query q = s.createFilter( collection, "" );; // the trivial filter   
q.setMaxResults(PAGE_SIZE);;  
q.setFirstResult(PAGE_SIZE * pageNumber);;  
List page = q.list();;  
Query q = s.createFilter( collection, "" );; // the trivial filter
q.setMaxResults(PAGE_SIZE);;
q.setFirstResult(PAGE_SIZE * pageNumber);;
List page = q.list();;

这种方法的效率有待考察。

所有sql不用ROWNUM的解决办法

如果执行所有SQL都不用ROWNUM，那么最简单的办法如下：

1、派生Dialect类

Java代码

package com.xxx.data.db.hibernate;   
  
import  net.sf.hibernate.dialect.Oracle9Dialect;   
  
public class Oracle9ThunisoftDialect extends Oracle9Dialect   
{   
  public Oracle9ThunisoftDialect();   
  {   
    super();;   
  }   
  
  public boolean supportsLimit();   
  {   
    return false;   
  }   
}  
[java] view plain copy
package com.xxx.data.db.hibernate;  
  
import  net.sf.hibernate.dialect.Oracle9Dialect;  
  
public class Oracle9ThunisoftDialect extends Oracle9Dialect  
{  
  public Oracle9ThunisoftDialect();  
  {  
    super();;  
  }  
  
  public boolean supportsLimit();  
  {  
    return false;  
  }  
}  
package com.xxx.data.db.hibernate;
 
import  net.sf.hibernate.dialect.Oracle9Dialect;
 
public class Oracle9ThunisoftDialect extends Oracle9Dialect
{
  public Oracle9ThunisoftDialect();
  {
    super();;
  }
 
  public boolean supportsLimit();
  {
    return false;
  }
}

2、修改hibernate.cfg.xml配置文件

Java代码

后记

<property name="hibernate.dialect">com.xxx.data.db.hibernate.Oracle9ThunisoftDialect</property>   
  
<!-- oracle 8.1.6+ -->   
<property name="hibernate.jdbc.use_scrollable_resultset">true</property>  
[java] view plain copy
 <property name="hibernate.dialect">com.xxx.data.db.hibernate.Oracle9ThunisoftDialect</property>  
  
<!-- oracle 8.1.6+ -->  
<property name="hibernate.jdbc.use_scrollable_resultset">true</property>  
 <property name="hibernate.dialect">com.xxx.data.db.hibernate.Oracle9ThunisoftDialect</property>
 
<!-- oracle 8.1.6+ -->
<property name="hibernate.jdbc.use_scrollable_resultset">true</property>

通常情况下，这种现象很少出现，因为我们程序中缺省的order by字段的“选择性”都很大，比如缺省以日期时间排序，order by字段很少出现重复。上面的现象基本上不会出现。不过，我们公司在做压力测试的时候，同一日期时间的记录插入了N多条，从而满足了上面提到的5个前提条件，因此出现了同一记录在不同页中出现的现象。