Hibernate代码分析追踪

阅读更多

Hibernate一般被我们作为数据库持久化工具使用，下面是我在使用过程中对工具使用过程中调试跟踪的一些记录。

我们一般在applicationContext.xml里如上配置，打开类AnnotationSessionFactoryBean.java，看该类的构造函数及父类的构造函数

        
            classpath:hibernate.cfg.xml
        

public AnnotationSessionFactoryBean() {
         setConfigurationClass(AnnotationConfiguration.class);
 }

public AnnotationSessionFactoryBean() {

        setConfigurationClass(AnnotationConfiguration.class);        
}        
@Override        
public void setConfigurationClass(Class configurationClass) {                
        if (configurationClass == null || !AnnotationConfiguration.class.isAssignableFrom(configurationClass)) {
                throw new IllegalArgumentException( "AnnotationSessionFactoryBean only supports AnnotationConfiguration or subclasses");                
        }                
        super.setConfigurationClass(configurationClass);        
}

org.hibernate.cfg.Configuration  doConfig(File f)会使用hibernate.cfg.xml配置好之后调用buildSessionFactory()

org.springframework.beans.factory.InitializingBean  void afterPropertiesSet()方法可见，bean的配置参数设置之后，该函数会被调用


org.springframework.orm.hibernate3.AbstractSessionFactoryBean 作为AnnotationSessionFactoryBean的父类实现了该接口，于是sessionFactory被构造出来。

public void afterPropertiesSet() throws Exception {                
        SessionFactory rawSf = buildSessionFactory();                
        this.sessionFactory = wrapSessionFactoryIfNecessary(rawSf);
        afterSessionFactoryCreation();        
}

我们在dao层一般使用getSession().load(id);getSession().save(obj);getSession().delete(obj)getSession().update(obj);等等方式实现CRUD,这些都会调用到org.session.impl.sessionImpl类下面的fireDelete(pram...),fireLoad(pram...),fireSaveOrUpdate(pram...)等方法，以delete为例子添加进去的deleteEventListener调用onDelete(DeleteEventevent)方法---》

publicvoid onDelete(DeleteEvent event, Set transientEntities)


EntityEntry entityEntry = persistenceContext.getEntry( entity );                
if ( entityEntry == null ) {}                
else {                       
 log.trace( "deleting a persistent instance" );                        
if ( entityEntry.getStatus() == Status.DELETED || entityEntry.getStatus() == Status.GONE ) {                               
 log.trace( "object was already deleted" );                               
 return;                        
}                        
persister = entityEntry.getPersister();                       
 id = entityEntry.getId();                        
version = entityEntry.getVersion();                
}
deleteEntity( source, entity, entityEntry, event.isCascadeDeleteEnabled(), persister, transientEntities );

可以看出，需要现在持久层查找该实体，找不到需要添加到持久层，找到了就可以准备删除。然后跳转deleteEntity方法，

session.getActionQueue().addAction(
				new EntityDeleteAction(
						entityEntry.getId(),
						deletedState,
						version,
						entity,
						persister,
						isCascadeDeleteEnabled,
						session
				)
		);

查看该Action的excute方法，

    final CacheKey ck;if (persister.hasCache() ) {
    ck = new CacheKey( ...);
lock = persister.getCacheAccessStrategy().lockItem( ck, version );}

并锁定缓存中的该数据，下方代码显示，需在缓存中删除，但我看不出来persistenceContext.removeEntity()和cacheAccessStrategy.remove的区别，知道的还请告知。

 if ( !isCascadeDeleteEnabled && !veto ) {                        
         persister.delete( id, version, instance, session );                
}                                
final PersistenceContext persistenceContext = session.getPersistenceContext();
EntityEntry entry = persistenceContext.removeEntry( instance );  
if ( entry == null ) {                        
         throw new AssertionFailure( "possible nonthreadsafe access to session" );               
 } 
 entry.postDelete(); 
persistenceContext.removeEntity( entry.getEntityKey() );
persistenceContext.removeProxy( entry.getEntityKey() );                                
if ( persister.hasCache() ) { 
      persister.getCacheAccessStrategy().remove( ck );                
}

而查询load不同，会涉及到sessionCache与secondLevelCache。

Object doLoad(final LoadEvent event,                        
final EntityPersister persister,                       
final EntityKey keyToLoad,                        
final LoadEventListener.LoadType options) {
    Object entity = loadFromSessionCache( event, keyToLoad, options ); 
              return entity;
                }       
         entity = loadFromSecondLevelCache(event, persister, options);        
        if ( entity != null ) {           
          return entity
             }
             return loadFromDatasource(event, persister, keyToLoad, options);
}

顺序是先一级缓存，然后二级缓存，然后数据库。在不同位置找到有不同处理，如果从数据库来，要将该实体添加至SessionCache里，如果从二级缓存来我看不出来做什么了。

暂时到这，github: http://github.com/leechedan/startSpring