namedquerycollectioninitializer.java

一个Java持久层类库

//$Id: NamedQueryCollectionInitializer.java 10019 2006-06-15 07:50:12Z [e-mail] $
package org.hibernate.persister.collection;

import java.io.Serializable;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.hibernate.FlushMode;
import org.hibernate.HibernateException;
import org.hibernate.engine.SessionImplementor;
import org.hibernate.impl.AbstractQueryImpl;
import org.hibernate.loader.collection.CollectionInitializer;

/**
 * A wrapper around a named query.
 * @author Gavin King
 */
public final class NamedQueryCollectionInitializer implements CollectionInitializer {
	private final String queryName;
	private final CollectionPersister persister;
	
	private static final Logger log = LoggerFactory.getLogger(NamedQueryCollectionInitializer.class);

	public NamedQueryCollectionInitializer(String queryName, CollectionPersister persister) {
		super();
		this.queryName = queryName;
		this.persister = persister;
	}

	public void initialize(Serializable key, SessionImplementor session) 
	throws HibernateException {
		
		if ( log.isDebugEnabled() ) {
			log.debug(
					"initializing collection: " + 
					persister.getRole() + 
					" using named query: " + 
					queryName 
				);
		}
		
		//TODO: is there a more elegant way than downcasting?
		AbstractQueryImpl query = (AbstractQueryImpl) session.getNamedSQLQuery(queryName); 
		if ( query.getNamedParameters().length>0 ) {
			query.setParameter( 
					query.getNamedParameters()[0], 
					key, 
					persister.getKeyType() 
				);
		}
		else {
			query.setParameter( 0, key, persister.getKeyType() );
		}
		query.setCollectionKey( key )
				.setFlushMode( FlushMode.MANUAL )
				.list();

	}
}