Hibernate继承映射总结

Hibernate继承映射三种策略中:

每个类分层结构是一张表:

    表现在数据库中父类与子类在同一张表中,这样查询起来方便.

每个子类是一张表:

   表现在数据库中父类与子类分别在不同的表中,但子类只有子类定义的属性所对应的字段,这样做查询的时候会用到inner join或者是left join等.

最后一种是每个具体类是一张表:

   表现在数据库中也是父类与子类分别在不同的表中,不过子类也包括了父类的所有字段.这样如果直接从子类load就比较快,而如果从父类load的话sql就会比较麻烦.

比如第三种策略,如果我只load pig对象:

Pig p = (Pig)session.load(Pig.class, 1);

则发出sql如下:

Hibernate: 
    /* load com.lwf.hibernate.extend1.Pig */ select
        pig0_.id as id0_0_,
        pig0_.name as name0_0_,
        pig0_.sex as sex0_0_,
        pig0_.weight as weight1_0_ 
    from
        t_pig pig0_ 
    where
        pig0_.id=?

而如果我load animal对象:

Animal a = (Animal)session.load(Animal.class, 1);

则发出sql如下:

Hibernate: 
    /* load com.lwf.hibernate.extend1.Animal */ select
        animal0_.id as id0_0_,
        animal0_.name as name0_0_,
        animal0_.sex as sex0_0_,
        animal0_.weight as weight1_0_,
        animal0_.height as height2_0_,
        animal0_.clazz_ as clazz_0_ 
    from
        ( select
            id,
            name,
            sex,
            weight,
            null as height,
            1 as clazz_ 
        from
            t_pig 
        union
        select
            id,
            name,
            sex,
            null as weight,
            height,
            2 as clazz_ 
        from
            t_bird 
    ) animal0_ 
where
    animal0_.id=?

看看就知道谁要方便...