hibernate之级联cascade和关系维持inverse

hibernate的关联关系，重点在理解级联cascade和inverse

1、cascade一般用在级联保存，级联更新，级联删除上

1.1cascade注解有两种，一种是基于hibernate注解

org.hibernate.annotations.Cascade

org.hibernate.annotations.CascadeType

支持一下级联

ALL, PERSIST//级联持久化，调用session.persist()时会触发级联事件 MERGE//级联保存或者更新，jpa规范 hibernate为了支持jsr220 后面添加的，调用session.merge()时触发级联 REMOVE,//级联删除，jpa规范同上，调用session.delete()时触发 REFRESH, DELETE,//级联删除，session.delete()触发 SAVE_UPDATE,//级联保存或者更新session.save(),session.update(),session.saveOrUpdate(); REPLICATE,

配置示例

@Cascade(value={org.hibernate.annotations.CascadeType.ALL})
privateStudentInfostudentInfo;

@Cascade(value={org.hibernate.annotations.CascadeType.ALL})
	private StudentInfo studentInfo ;

1.2第二种注解是基于jpa规范，也就是apache jsr220规范，也是ejb3的持久层规范

javax.persistence.CascadeType

ALL,

PERSIST,调用session.persist()时触发 MERGE，调用session.merge()触发 REMOVE,调用session.delete()触发 REFRESH, DETACH

配置示例

@ManyToOne(cascade={CascadeType.MERGE})
@JoinColumn(name="teacher_id")
privateTeacherteacher;

@ManyToOne(cascade={CascadeType.MERGE})
	@JoinColumn(name = "teacher_id")
	private Teacher teacher;

1.3级联一般用在OneToOne和OneToMany上，这也是hibernate官方的推荐，有时候我们在开发中也用在 ManyToOne,ManyToMany上，只是想在测试上少写点代码而已，在持久化一个实体的时候级联持久其他关联实体，

如下：teacher和student是ManyToMany,加上了cascade注解，便于测试

@Test
publicvoidaddStudentToTeacher(){
Studentstudent1=newStudent("张三",20,20072733L);
Studentstudent2=newStudent("李四",20,20072734L);
Studentstudent3=newStudent("王五",20,20072735L);

Teacherteacher=newTeacher("张老师");
teacher.getStudents().add(student3);
teacher.getStudents().add(student2);
teacher.getStudents().add(student1);
this.teacherDao.save(teacher);
}

@Test
	public void addStudentToTeacher(){
		Student student1 = new Student("张三",20,20072733L);
		Student student2 = new Student("李四",20,20072734L);
		Student student3 = new Student("王五",20,20072735L);
		
		Teacher teacher = new Teacher("张老师");
		teacher.getStudents().add(student3);
		teacher.getStudents().add(student2);
		teacher.getStudents().add(student1);
		this.teacherDao.save(teacher);
	}

2、inverse，英文意思是反向，反转。在这里可以理解为控制反转，也就是说实体间的关系由谁控制，所以inverse用在实体关联上。如OneToOne，OneToMany，ManyToMany

在OneToMany中，如果不指定inverse，那么hibernate会去找默认的表来维持关系。

例如用老师和课程两个实体来说明，假设teacher和course是OneToMany的关系，

配置如下：

//通过外键teacher_id关联teacher，inverse通过mappedBy来设置

@ManyToOne(cascade={CascadeType.MERGE})
@JoinColumn(name="teacher_id")
privateTeacherteacher;

@OneToMany(mappedBy="teacher",fetch=FetchType.LAZY,cascade={CascadeType.MERGE,CascadeType.REMOVE})
privateSet<Course>courses=newHashSet<Course>();

@ManyToOne(cascade={CascadeType.MERGE})
	@JoinColumn(name = "teacher_id")
	private Teacher teacher;

@OneToMany(mappedBy="teacher",fetch=FetchType.LAZY,cascade={CascadeType.MERGE,CascadeType.REMOVE})
	private Set<Course> courses = new HashSet<Course>();

在这里指定teacher的属性courses的关系由关联实体Course的teacher属性维持，如果不配置，则会去找关联表teacher_course来维持关系。OneToOne和ManyToMany大同小异，这里不再累述！