重新学习 Hibernate fetch lazy cascade inverse 关键字

Hibernate最让人头大的就是对集合的加载形式。
书看了N次了,还是没有真正理解Hibernate。

所以下午专门做了下测试,对配置文件的意思加深了认识。

假设有两个表,Photos(一)  ---  picture(多)Photo包含picture集合

结论1: HQL代码 > fetch(配置) > lazy (配置)
结论2: 默认 lazy="true"
结论3: fetch 和 lazy 主要是用来级联查询的,   而 cascade 和 inverse 主要是用来级联插入和修改的
结论4: 如果你是用spring来帮你管理你的session, 并且是自动提交,延迟加载就等于没加载~_~(当然
                除非你手动重新打开session然后手动Hibernate.initialize(set);然后关闭session.
结论5:     cascade主要是简化了在代码中的级联更新和删除。
j结论6:老爸可以有多个孩子,一个孩子不能有多个老爸,而且老爸说的算, 孩子围着老爸转。
               所以Photos老爸要有权力所以 cascade 这个关键子都是送给老爸的, 也就是级联更新,
               老爸改姓了,儿子也得跟着改,呵呵。“不然,就没有零花钱咯”。
                而Picture儿子整体挨骂,但是还是要维护父子之间良好的关系,对老爸百依百顺,所
               以老爸就说,儿子,“关系,由你来维护(inverse="true") ,不然就不给零花钱。呵。”。
               <set name="pictures" inverse="true" cascade="all">
                    <key>
                       <column name="photosid" not-null="true" />
                    </key>
                 <one-to-many class="girl.domain.Picture" />
             </set>
               
测试代码:

   Photos p = ps.getById(1);
  Set<Picture> set = p.getPictures();
  for(Picture pic : set){
     System.out.println(pic.getId());
  }

  配置文件的一部分:
       <set name="pictures" inverse="true" cascade="all" >
            <key>
                <column name="photosid" not-null="true" />
            </key>
            <one-to-many class="girl.domain.Picture" />
        </set>

测试过程会对配置文件不断修改:并且从来不曾手动重新打开session

测试结构:

当配置条件为 lazy=true 一句查询 测试代码中没有调用getPicture()  正常
Hibernate: select photos0_.id as id0_0_, photos0_.userid as userid0_0_, photos0_.typeid as typeid0_0_, photos0_.name as name0_0_, photos0_.createtime as createtime0_0_, photos0_.description as descript6_0_0_, photos0_.faceid as faceid0_0_, photos0_.uri as uri0_0_ from super.photos photos0_ where photos0_.id=?

lazy=true 一句查询 有getPicture()
Hibernate: select photos0_.id as id0_0_, photos0_.userid as userid0_0_, photos0_.typeid as typeid0_0_, photos0_.name as name0_0_, photos0_.createtime as createtime0_0_, photos0_.description as descript6_0_0_, photos0_.faceid as faceid0_0_, photos0_.uri as uri0_0_ from super.photos photos0_ where photos0_.id=?


lazy=true 一句查询  有getPicture() 并且访问了里面的元数Picture 且有异常抛出
Hibernate: select photos0_.id as id0_0_, photos0_.userid as userid0_0_, photos0_.typeid as typeid0_0_, photos0_.name as name0_0_, photos0_.createtime as createtime0_0_, photos0_.description as descript6_0_0_, photos0_.faceid as faceid0_0_, photos0_.uri as uri0_0_ from super.photos photos0_ where photos0_.id=?


lazy="false" 两句查询  肯定没问题,因为全部数据都个查了出来 所以怎么调用都正常
Hibernate: select photos0_.id as id0_0_, photos0_.userid as userid0_0_, photos0_.typeid as typeid0_0_, photos0_.name as name0_0_, photos0_.createtime as createtime0_0_, photos0_.description as descript6_0_0_, photos0_.faceid as faceid0_0_, photos0_.uri as uri0_0_ from super.photos photos0_ where photos0_.id=?
Hibernate: select pictures0_.photosid as photosid1_, pictures0_.id as id1_, pictures0_.id as id2_0_, pictures0_.photosid as photosid2_0_, pictures0_.name as name2_0_, pictures0_.clicked as clicked2_0_, pictures0_.uploaddate as uploaddate2_0_, pictures0_.size as size2_0_, pictures0_.description as descript7_2_0_, pictures0_.uri as uri2_0_ from super.picture pictures0_ where pictures0_.photosid=?


fetch="join"  一句查询  效果 == lazy="false" 呵呵,哪个效率高,我就不知道了。。。。。。。。。。。
Hibernate: select photos0_.id as id0_1_, photos0_.userid as userid0_1_, photos0_.typeid as typeid0_1_, photos0_.name as name0_1_, photos0_.createtime as createtime0_1_, photos0_.description as descript6_0_1_, photos0_.faceid as faceid0_1_, photos0_.uri as uri0_1_, pictures1_.photosid as photosid3_, pictures1_.id as id3_, pictures1_.id as id2_0_, pictures1_.photosid as photosid2_0_, pictures1_.name as name2_0_, pictures1_.clicked as clicked2_0_, pictures1_.uploaddate as uploaddate2_0_, pictures1_.size as size2_0_, pictures1_.description as descript7_2_0_, pictures1_.uri as uri2_0_ from super.photos photos0_ left outer join super.picture pictures1_ on photos0_.id=pictures1_.photosid where photos0_.id=?

不加fetch="join" 一句查询  没有getPicture() 正常
Hibernate: select photos0_.id as id0_0_, photos0_.userid as userid0_0_, photos0_.typeid as typeid0_0_, photos0_.name as name0_0_, photos0_.createtime as createtime0_0_, photos0_.description as descript6_0_0_, photos0_.faceid as faceid0_0_, photos0_.uri as uri0_0_ from super.photos photos0_ where photos0_.id=?

不加fetch="join" 一句查询  有getPicture() 正常
Hibernate: select photos0_.id as id0_0_, photos0_.userid as userid0_0_, photos0_.typeid as typeid0_0_, photos0_.name as name0_0_, photos0_.createtime as createtime0_0_, photos0_.description as descript6_0_0_, photos0_.faceid as faceid0_0_, photos0_.uri as uri0_0_ from super.photos photos0_ where photos0_.id=?

不加fetch="join" 一句查询 有getPicture() 并且访问里面的元素Picture的ID 有异常抛出
Hibernate: select photos0_.id as id0_0_, photos0_.userid as userid0_0_, photos0_.typeid as typeid0_0_, photos0_.name as name0_0_, photos0_.createtime as createtime0_0_, photos0_.description as descript6_0_0_, photos0_.faceid as faceid0_0_, photos0_.uri as uri0_0_ from super.photos photos0_ where photos0_.id=?

来个两兵交战 fetch="join" lazy="true"  呵呵 结果,一句查询, 结构正常 所以就当lazy不存在好了。 看来fetch 是老大。、、、、、、、、、、、、、
Hibernate: select photos0_.id as id0_1_, photos0_.userid as userid0_1_, photos0_.typeid as typeid0_1_, photos0_.name as name0_1_, photos0_.createtime as createtime0_1_, photos0_.description as descript6_0_1_, photos0_.faceid as faceid0_1_, photos0_.uri as uri0_1_, pictures1_.photosid as photosid3_, pictures1_.id as id3_, pictures1_.id as id2_0_, pictures1_.photosid as photosid2_0_, pictures1_.name as name2_0_, pictures1_.clicked as clicked2_0_, pictures1_.uploaddate as uploaddate2_0_, pictures1_.size as size2_0_, pictures1_.description as descript7_2_0_, pictures1_.uri as uri2_0_ from super.photos photos0_ left outer join super.picture pictures1_ on photos0_.id=pictures1_.photosid where photos0_.id=?

评论 1
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值