hibernate 的级联保存和关联关系
Customer 和 Orders 是一对多的关系，Orders和Customer是多对一关系，即这两个双向关联.
Customer 和Bid是一对一的关系，一对一的关系就不存在双向和单向了
如下是他们的类：
customer.java
package com.reflesh;
import java.util.Date;
import java.util.HashSet;
import java.util.Set;
public class Customer {
private int id;
private String name;
private Date date;
private String tele;
private Bid bidId;
private Set ordersId = new HashSet(); //省略set 和get 方法
}
orders.java
public class Orders {
private Long orderId;
private String name;
private Customer customer;
// 省略掉equals和haseCode（）
Bid.java
private Long bidId; private String name; private Customer customerBId; }
customer.hbm.xml
<id name="id" column="customerId" type="int"> <generator class="sequence"> <param name="sequence">customerId</param> </generator> </id> <property name="name" column="name" type="string" /> <property name="date" column="orderDate" type="date"/> <property name="tele" column="telephone" type="string"/> <set name="ordersId" cascade="save-update" inverse="true"> <key column="customerOId"/> <one-to-many class="com.reflesh.Orders" /> </set> <one-to-one name="bidId" class="com.reflesh.Bid" property-ref="customerBId" cascade="save-update"/>
orders.hbm.xml
<id name="orderId" column="orderId2" type="long"> <generator class="sequence"> <param name="sequence">seq_order</param> </generator> </id> <property name="name" column="name" type="string" /> <many-to-one name="customer" class="com.reflesh.Customer" column="customerOId" />
bid.hbm.xml
<id name="bidId" column="bidId" type="long"> <generator class="sequence" > <param name="sequence">seq_bid</param> </generator> </id> <property name="name" column="name" type="string" />
<many-to-one name="customerBId" class="com.reflesh.Customer" column="customerBId" unique="true" />
test.java
Session session =HibernateSessionFactory.getSession(); Transaction tr = null;
/* 关联类*/
Customer c = new Customer();
c.setName("bellwind");
Date date = RefleshActionHelper.FormatDate("2004-05-09", "yyyy-mm-dd");
c.setDate(date);
c.setTele("123456");
Serializable id = null;
Orders o = new Orders();
o.setName("订单1dfdf");
o.setCustomer(c);
Orders o2 = new Orders();
o2.setName("订单2");
o2.setCustomer(c);
c.getOrdersId().add(o);
c.getOrdersId().add(o2);
Bid bid = new Bid();
bid.setName("ijd");
bid.setCustomerBId(c);
c.setBidId(bid);
/* 关联类完*/
try {
tr = session.beginTransaction();
id = session.save(c); //三个都保存
tr.commit();
} catch (Exception e) {
e.printStackTrace();
tr.rollback();
}
finally {
session.close();
}
return id;
}
说明：a,b,c为级联的步骤
a.在set属性上加 cascade="save-update"
b. 关联类 见代码test.java的注释
c. 保存customer一个就行了，因为customer表是主表，orders表的外键customerOId和Bid表的外键customerBId 都关系到customer表的主键；
注意：
a.一般都选择保存customer来级联保存orders和bids，不选择保存orders或者bids来级联保存customer；
b. 没有.inverse="true",执行如下语句
insert into customer (name, orderDate, telephone, customerId) values (?, ?, ?, ?)
Hibernate: insert into orders (name, customerOId, orderId2) values (?, ?, ?)
Hibernate: insert into orders (name, customerOId, orderId2) values (?, ?, ?)
Hibernate: insert into bid (name, customerBId, bidId) values (?, ?, ?)
Hibernate: update orders set customerOId=? where orderId2=?
Hibernate: update orders set customerOId=? where orderId2=?
有了inverse="true",执行如下语句
insert into customer (name, orderDate, telephone, customerId) values (?, ?, ?, ?)
Hibernate: insert into orders (name, customerOId, orderId2) values (?, ?, ?)
Hibernate: insert into orders (name, customerOId, orderId2) values (?, ?, ?)
Hibernate: insert into bid (name, customerBId, bidId) values (?, ?, ?)
可以看出减少了2条语句，提高了性能
customer单向关联orders的步骤：
1.
customer.hbm.xml
<set name="ordersId" cascade="save-update" inverse="true">
<key column="customerOId"/>
<one-to-many class="com.reflesh.Orders" />
</set>
2.orders.hbm.xml 不用配关联到customer， 删除<many-to-one >元素
3.orders类的customer属性删除
4.只需单向关联
Customer c = new Customer();
Bid bid = new Bid();
c.setBidId(bid);
5.save(c) 就行了
注意：
如果是customer单向关联orders，那么inverse="true" 不可以少，否则orders表的外键为null，它需要
Hibernate: update orders set customerOId=? where orderId2=?
Hibernate: update orders set customerOId=? where orderId2=?
这两句来更新orders表的外键