hibernate系列（三）多对多的关联关系

以Teacher和Student为例，他们之间是多对多的关系。
手动创建的数据库的三张表为，teacher、student、teacher_student。分别如下：

CREATE TABLE `teacher` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `name` varchar(45) DEFAULT '',
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

CREATE TABLE `student` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `name` varchar(45) DEFAULT '',
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

CREATE TABLE `teacher_student` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `teacher_id` int(11) DEFAULT NULL,
  `student_id` int(11) DEFAULT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

Teacher类如下：

public class Teacher {

	private Long id;
	private String name;
	private Set<Student> students;
//省略get、set方法
}

Teacher类对应的映射文件Teacher.hbm.xml：

<hibernate-mapping>
	<class name="com.ligang.domain.Teacher" table="teacher">
		<id name="id" column="id" type="long">
			<generator class="identity"/>
		</id>
		<property name="name" column="name" type="string"/>
		<set name="students" table="teacher_student">
			<key column="teacher_id"></key>
			<many-to-many class="com.ligang.domain.Student" column="student_id"></many-to-many>
		</set>
	</class>
</hibernate-mapping>

其中set标签的name指的是Teacher类的students属性,table指的是该属性要关联到哪张表。内部的key标签的column指的是该Teacher类所属的表的主键id作为teacher_student的外键teacher_id值。<many-tomany>标签指的是Student类所属的student表的主键作为teacher_student表的student_id值。
Stusent类如下：

public class Student {

	private Long id;
	private String name;
	private Set<Teacher> teachers;
//省略get、set方法
}

Student类对应的映射文件Student.hbm.xml如下：

<hibernate-mapping>
	<class name="com.ligang.domain.Student" table="student">
		<id name="id" column="id" type="long">
			<generator class="identity"/>
		</id>
		<property name="name" column="name" type="string"/>
		<set name="teachers" table="teacher_student"  inverse="true">
			<key column="student_id"></key>
			<many-to-many class="com.ligang.domain.Teacher" column="teacher_id"></many-to-many>
		</set>
	</class>
</hibernate-mapping>

其中的set映射关系同上。
然后就来看下增添方法：

@Test
	public void testAddTeacher1(){
		Session session=hibernateDao.getSession();
		Transaction tx=session.beginTransaction();
		
		Teacher t=new Teacher();
		t.setName("teacher4");
		
		Student s1=new Student();
		s1.setName("assa");
		
		Student s2=new Student();
		s2.setName("sdfvdv");
		
		Set<Student> students=new HashSet<Student>();
		students.add(s1);
		students.add(s2);
		
		t.setStudents(students);
		
		session.save(s1);
		session.save(s2);
		session.save(t);
		
		tx.commit();
		session.close();
	}

三个save，三个insert语句，同时，由于t.setStudents(students)，所以teacher会去维护teacher_student关系表，所以又会增添两条insert语句，sql如下：

Hibernate: insert into hibernate.student (name) values (?)
Hibernate: insert into hibernate.student (name) values (?)
Hibernate: insert into hibernate.teacher (name) values (?)
Hibernate: insert into hibernate.teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into hibernate.teacher_student (teacher_id, student_id) values (?, ?)

如果增添改为如下状态：

@Test
	public void testAddTeacher(){
		Session session=hibernateDao.getSession();
		Transaction tx=session.beginTransaction();
		
		Teacher t=new Teacher();
		t.setName("teacher4");
		
		Student s1=new Student();
		s1.setName("assa");
		
		Student s2=new Student();
		s2.setName("sdfvdv");
		
		Set<Student> students=new HashSet<Student>();
		Set<Teacher> teachers=new HashSet<Teacher>();
		students.add(s1);
		students.add(s2);
		teachers.add(t);
		
		t.setStudents(students);
		s1.setTeachers(teachers);
		s2.setTeachers(teachers);
		
		session.save(s1);
		session.save(s2);
		session.save(t);
		
		tx.commit();
		session.close();
	}

此时，不仅建立起了t.setStudents(students)关系，同时s1.setTeachers(teachers);s2.setTeachers(teachers);也建立了关系，所以最终事务提交的时候，teacher、student都会去维护teacher_student表，造成了4个insert语句，造成了重复。如下：

Hibernate: insert into hibernate.student (name) values (?)
Hibernate: insert into hibernate.student (name) values (?)
Hibernate: insert into hibernate.teacher (name) values (?)
Hibernate: insert into hibernate.teacher_student (student_id, teacher_id) values (?, ?)
Hibernate: insert into hibernate.teacher_student (student_id, teacher_id) values (?, ?)
Hibernate: insert into hibernate.teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into hibernate.teacher_student (teacher_id, student_id) values (?, ?)

为了避免这一现象，就需要某一方放弃维护teacher_student表的权限。如student放弃维护这一关系，使用inverse="true",如下:

<hibernate-mapping>
	<class name="com.ligang.domain.Student" table="student">
		<id name="id" column="id" type="long">
			<generator class="identity"/>
		</id>
		<property name="name" column="name" type="string"/>
		<set name="teachers" table="teacher_student" inverse="true">
			<key column="student_id"></key>
			<many-to-many class="com.ligang.domain.Teacher" column="teacher_id"></many-to-many>
		</set>
	</class>
</hibernate-mapping>

此时再按照上述方法增添就只有teacher去维护teacher_student表了，不会造成重复。

再看下多对多关联关系的查询：

@Test
	public void getTeacher(){
		Session session=hibernateDao.getSession();
		Transaction tx=session.beginTransaction();
		
		Teacher t=(Teacher) session.get(Teacher.class,3L);
		System.out.println(t.getName());
		System.out.println(t.getStudents().size());
		
		tx.commit();
		session.close();
	}

这里同样存在这查询策略问题，此时Teacher类的映射文件对应的set标签的lazy属性有三个值，一个true、false、extra。默认为true，即实行延迟加载的策略，用到Student时才回去加载它，如下sql：

Hibernate: select teacher0_.id as id1_3_0_, teacher0_.name as name2_3_0_ from hibernate.teacher teacher0_ where teacher0_.id=?
teacher2
Hibernate: select students0_.teacher_id as teacher_1_3_0_, students0_.student_id as student_2_4_0_, student1_.id as id1_2_1_, student1_.name as name2_2_1_ from hibernate.teacher_student students0_ inner join hibernate.student student1_ on students0_.student_id=student1_.id where students0_.teacher_id=?
2

若设置为false,则表示在加载Teacher时立即去加载相关的Student，如下sql：

Hibernate: select teacher0_.id as id1_3_0_, teacher0_.name as name2_3_0_ from hibernate.teacher teacher0_ where teacher0_.id=?
Hibernate: select students0_.teacher_id as teacher_1_3_0_, students0_.student_id as student_2_4_0_, student1_.id as id1_2_1_, student1_.name as name2_2_1_ from hibernate.teacher_student students0_ inner join hibernate.student student1_ on students0_.student_id=student1_.id where students0_.teacher_id=?
teacher2
2

若设置为extra,则会更加智能化一些，即上述t.getStudents().size()并没有去访问Student的实际内容，仅仅是想获取数量，所以sql语句如下：

Hibernate: select teacher0_.id as id1_3_0_, teacher0_.name as name2_3_0_ from hibernate.teacher teacher0_ where teacher0_.id=?
teacher2
Hibernate: select count(student_id) from hibernate.teacher_student where teacher_id =?
2

再看下更新Teacher:

@Test
	public void updateTeacher(){
		Session session=hibernateDao.getSession();
		Transaction tx=session.beginTransaction();
		
		Teacher t=(Teacher) session.get(Teacher.class,6L);
		
		Student s1=new Student();
		s1.setName("assa");
		
		Student s2=new Student();
		s2.setName("sdfvdv");
		
		Set<Student> students=new HashSet<Student>();
		students.add(s1);
		students.add(s2);
		
		t.setStudents(students);
		
		session.save(s1);
		session.save(s2);
		
		tx.commit();
		session.close();
	}

更新的时候，会先删除之前的teacher_student表中的相关记录，然后再新增新的记录，如下sql:

Hibernate: select teacher0_.id as id1_3_0_, teacher0_.name as name2_3_0_ from hibernate.teacher teacher0_ where teacher0_.id=?
Hibernate: insert into hibernate.student (name) values (?)
Hibernate: insert into hibernate.student (name) values (?)
Hibernate: delete from hibernate.teacher_student where teacher_id=?
Hibernate: insert into hibernate.teacher_student (teacher_id, student_id) values (?, ?)
Hibernate: insert into hibernate.teacher_student (teacher_id, student_id) values (?, ?)

若想转载请注明出处： http://lgbolgger.iteye.com/blog/2125014
作者：iteye的乒乓狂魔