hibernate 多对多(many-to-many)

多对多（many-to-many）：在操作和性能方面都不太理想，所以多对多的映射使用较少，实际使用中最好转换成一对多的对象模型；hibernate会为我们创建中间关联表，转换成两个一对多。

1. E-R图

 

2. 实体类：

Teacher实体类如下：

package com.reiyen.hibernate.domain;

import java.util.Set;

public class Teacher {

	private int id;
	private String name;
	private Set<Student> students;
       //setter和getter方法

}

 Student实体类如下：

package com.reiyen.hibernate.domain;

import java.util.Set;

public class Student {

	private int id;
	private String name;
	private Set<Teacher> teachers;
        //setter和getter方法

}

 3.映射文件如下：

Teacher.hbm.xml如下：

<?xml version="1.0"?>
<!DOCTYPE hibernate-mapping PUBLIC 
	"-//Hibernate/Hibernate Mapping DTD 3.0//EN"
	"http://hibernate.sourceforge.net/hibernate-mapping-3.0.dtd">
<hibernate-mapping package="com.reiyen.hibernate.domain">
	<class name="Teacher">
		<id name="id">
			<generator class="native" />
		</id>
		<property name="name" />
		<!-- 通过table项告诉hibernate中间表的名称 -->
		<set name="students" table="teacher_student">
			<!-- 通过key属性告诉hibernate在中间表里面查询teacher_id值相应的teacher记录 -->
			<key column="teacher_id" />
			<!-- 通过column项告诉hibernate对student表中查找student_id值相就的studnet记录 -->
			<many-to-many class="Student" column="student_id" />
		</set>
	</class>
</hibernate-mapping>

 Student.hbm.xml如下：

<?xml version="1.0"?>
<!DOCTYPE hibernate-mapping PUBLIC 
	"-//Hibernate/Hibernate Mapping DTD 3.0//EN"
	"http://hibernate.sourceforge.net/hibernate-mapping-3.0.dtd">
<hibernate-mapping package="com.reiyen.hibernate.domain">
	<class name="Student" >
		<id name="id" >
			<generator class="native" />
		</id>
		<property name="name" />
		<set name="teachers" table="teacher_student">
		<key column="student_id" />
		<many-to-many class="Teacher" column="teacher_id"/>
		</set>
	</class>
</hibernate-mapping>

 一定要注意映射文件中<many-to-many class="Teacher" column="teacher_id"/>中class的值，它必须与你另一个关联映射文件中的class属性的name值一致，其实就是与你的实体类的类名一致,如：<many-to-many class="Teacher" column="teacher_id"/>中class的值就不能写成"teacher"。如果写成这样的话，就会抛出如下异常：An association from the table teacher_student refers to an unmapped class: com.reiyen .hibernate.domain.teacher

4. 测试程序如下：

public class Many2Many {
	public static void main(String[] args) {
		add();
		//query(1);
	}

	static void query(int id) {
		Session s = null;
		Transaction tx = null;
		try {
			s = HibernateUtil.getSession();
			tx = s.beginTransaction();
			Teacher t = (Teacher) s.get(Teacher.class, id);
			System.out.println("students:" + t.getStudents().size());
			tx.commit();
		} finally {
			if (s != null)
				s.close();
		}
	}

	static void add() {
		Session s = null;
		Transaction tx = null;
		try {
			Set<Teacher> ts = new HashSet<Teacher>();

			Teacher t1 = new Teacher();
			t1.setName("t1 name");
			ts.add(t1);

			Teacher t2 = new Teacher();
			t2.setName("t2 name");
			ts.add(t2);

			Set<Student> ss = new HashSet<Student>();
			Student s1 = new Student();
			s1.setName("s1");
			ss.add(s1);

			Student s2 = new Student();
			s2.setName("s2");
			ss.add(s2);

			t1.setStudents(ss);  //1
			t2.setStudents(ss);  //1
//			
//			s1.setTeachers(ts);  //2
//			s2.setTeachers(ts);  //2

			s = HibernateUtil.getSession();
			tx = s.beginTransaction();
			s.save(t1);
			s.save(t2);
			s.save(s1);
			s.save(s2);
			tx.commit();
		} finally {
			if (s != null)
				s.close();
		}
	}
}

 运行此程序后：控制台打印的sql语句如下所示：

Hibernate: insert into Teacher (name) values (?)
Hibernate: insert into Teacher (name) values (?)
Hibernate: insert into Student (name) values (?)
Hibernate: insert into Student (name) values (?)
Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)

一共在中间表里面插入了4条记录。

中间表结构如下所示：

DROP TABLE IF EXISTS `test`.`teacher_student`;
CREATE TABLE  `test`.`teacher_student` (
  `teacher_id` int(11) NOT NULL,
  `student_id` int(11) NOT NULL,
  PRIMARY KEY (`student_id`,`teacher_id`),
  KEY `FK2E2EF2DE6C8A2663` (`teacher_id`),
  KEY `FK2E2EF2DE5BEEDBC3` (`student_id`),
  CONSTRAINT `FK2E2EF2DE5BEEDBC3` FOREIGN KEY (`student_id`) REFERENCES `student` (`id`),
  CONSTRAINT `FK2E2EF2DE6C8A2663` FOREIGN KEY (`teacher_id`) REFERENCES `teacher` (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

表中插入的记录如下所示：

mysql> select * from teacher_student;
+------------+------------+
| teacher_id | student_id |
+------------+------------+
|          1 |          1 |
|          1 |          2 |
|          2 |          1 |
|          2 |          2 |
+------------+------------+
4 rows in set (0.00 sec)

 

程序中注释为1的语句非常重要，它是建立Teacher与Student关联的语句，如果没有这两条语句，虽然程序照样会执行，但是在中间表teacher_student没有任何记录，也就是Teacher与Student之间未关联。

当然你也可以通过程序中注释为2的语句来建立Teacher与Student之间的关联关系，同样会产生与注释为1的语句的效果。但是你不能在程序中同时出现以上四句程序，否则会抛出异常（ PRIMARY KEY (`student_id`,`teacher_id`),所以会出现主键冲突的异常 ）,：

Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into teacher_student (student_id, teacher_id) values (?, ?)
Hibernate: insert into teacher_student (student_id, teacher_id) values (?, ?)
Hibernate: insert into teacher_student (student_id, teacher_id) values (?, ?)
Hibernate: insert into teacher_student (student_id, teacher_id) values (?, ?)
Exception in thread "main" org.hibernate.exception.ConstraintViolationException: Could not execute JDBC batch update

解决上面产生异常的办法是设置inverse属性。即在Tearcher一端或Student一端设置inverse="true"，即让他们之中的某一方放弃维护关联关系。此时，虽然上面四句程序在测试程序中同时出现了（其实就是在对象模型上相互设置了他们的关联关系），但程序照样能运行正常，因为了在数据库模型上，只会有两句程序生效，也就是没有设置inverse="true"的那一端会去维护关联关系。有关inverse的说细信息，可以参看我的文章hibernate级联（cascade和inverse）.

 

执行测试程序中的查询测试，控制台打印的信息如下所示：

Hibernate: select teacher0_.id as id5_0_, teacher0_.name as name5_0_ from Teacher teacher0_ where teacher0_.id=?
Hibernate: select students0_.teacher_id as teacher1_1_, students0_.student_id as student2_1_, student1_.id as id7_0_, student1_.name as name7_0_ from teacher_student students0_ left outer join Student student1_ on students0_.student_id=student1_.id where students0_.teacher_id=?
students:2

从打印出的sql语句可以看出，多对多关系进行查询时，效率是比较低的。

评论

1 楼 yanqlv 2015-12-11  

你说的“在数据库模型上，只会有两句程序生效”有点玄乎了。

            t1.setStudents(ss);  //1 
            t2.setStudents(ss);  //1 
//           
//          s1.setTeachers(ts);  //2 
//          s2.setTeachers(ts);  //2 

你这后两句不能执行是因为中间表teacher_student具有联合主键。

teacher, student谁能维护关系，就表示对谁调用session.save()时会让Hibernate向中间表插入数据。

但在不违反联合主键的约束前提下，双方都能维护关系的。也就是说在维护的这些关系不冲突时，Hibernate是不管这些的，即你自己写程序时保证也是可以的。