关于hiberante FetchMode

以Person, Address 为例， 其关系为一对多，数据库中有3条person, 每个分别对应2条address。查询语句如下：




Java代码 

1.Criteria c = session.createCriteria(Person.class);  
2.c.setResultTransformer(Criteria.DISTINCT_ROOT_ENTITY);  
3.List<Person> list = (List<Person>)c.list();  
4.for (Person p : list) {  
5.        System.out.println(p.getName());  
6.              
7.    Set<Address> addressSet = p.getAddressSet();  
8.    for (Address a : addressSet) {  
9.        System.out.println(a);  
10.    }             
11.}  

1 默认不设置FetchMode



Java代码 

1.@OneToMany(cascade = CascadeType.ALL, fetch = FetchType.LAZY)  
2.@JoinColumn(name = "P_ID")  
3.public Set<Address> getAddressSet() {  
4.    return addressSet;  
5.}  

结果生成如下



Java代码 

1.Hibernate: select this_.PERSON_ID as PERSON1_0_0_, this_.PERSON_NAME as PERSON2_0_0_   
2.from PERSON this_  
3.P1  
4.Hibernate: select addressset0_.P_ID as P3_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_0_, addressset0_.ADDRESS_NAME as ADDRESS2_1_0_   
5.from ADDRESS addressset0_   
6.where addressset0_.P_ID=?  
7.2   P1 A2  
8.1   P1 A1  
9.P2  
10.Hibernate: select addressset0_.P_ID as P3_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_0_, addressset0_.ADDRESS_NAME as ADDRESS2_1_0_   
11.from ADDRESS addressset0_   
12.where addressset0_.P_ID=?  
13.4   P2 A2  
14.3   P2 A1  
15.P3  
16.Hibernate: select addressset0_.P_ID as P3_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_0_, addressset0_.ADDRESS_NAME as ADDRESS2_1_0_   
17.from ADDRESS addressset0_   
18.where addressset0_.P_ID=?  
19.5   P3 A1  
20.6   P3 A2  

即：先查所有Person的id， 然后根据id查对应的Address。产生N＋1问题

2. FetchMode.SELECT




Java代码 

1.@OneToMany(cascade = CascadeType.ALL, fetch = FetchType.LAZY)  
2.@Fetch(FetchMode.SELECT)  
3.@JoinColumn(name = "P_ID")  
4.public Set<Address> getAddressSet() {  
5.    return addressSet;  
6.}  

效果同默认，即hibernate默认的FetchMode是SELECT

3. FetchMode.JOIN



Java代码 

1.@OneToMany(cascade = CascadeType.ALL, fetch = FetchType.LAZY)  
2.@Fetch(FetchMode.JOIN)  
3.@JoinColumn(name = "P_ID")  
4.public Set<Address> getAddressSet() {  
5.    return addressSet;  
6.}  

效果如下



Java代码 

1.Hibernate: select this_.PERSON_ID as PERSON1_0_1_, this_.PERSON_NAME as PERSON2_0_1_, addressset2_.P_ID as P3_3_, addressset2_.ADDRESS_ID as ADDRESS1_3_, addressset2_.ADDRESS_ID as ADDRESS1_1_0_, addressset2_.ADDRESS_NAME as ADDRESS2_1_0_   
2.from PERSON this_   
3.   left outer join ADDRESS addressset2_   
4.        on this_.PERSON_ID=addressset2_.P_ID  
5.  
6.P1  
7.2   P1 A2  
8.1   P1 A1  
9.P2  
10.4   P2 A2  
11.3   P2 A1  
12.P3  
13.5   P3 A1  
14.6   P3 A2  

采用外联，用一条sql取出person及其address

4. FetchMode.SUBSELECT



Java代码 

1.@OneToMany(cascade = CascadeType.ALL, fetch = FetchType.LAZY)  
2.@Fetch(FetchMode.SUBSELECT)  
3.@JoinColumn(name = "P_ID")  
4.public Set<Address> getAddressSet() {  
5.    return addressSet;  
6.}  

效果



Java代码 

1.Hibernate: select this_.PERSON_ID as PERSON1_0_0_, this_.PERSON_NAME as PERSON2_0_0_   
2.from PERSON this_  
3.P1  
4.Hibernate: select addressset0_.P_ID as P3_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_0_, addressset0_.ADDRESS_NAME as ADDRESS2_1_0_   
5.from ADDRESS addressset0_   
6.where addressset0_.P_ID in (select this_.PERSON_ID from PERSON this_)  
7.2   P1 A2  
8.1   P1 A1  
9.P2  
10.4   P2 A2  
11.3   P2 A1  
12.P3  
13.6   P3 A2  
14.5   P3 A1 

生成2条sql， 第二句用in 查关联Address数据


5. batchSize



Java代码 

1.@OneToMany(cascade = CascadeType.ALL, fetch = FetchType.LAZY)  
2.@BatchSize(size=4)  
3.@JoinColumn(name = "P_ID")  
4.public Set<Address> getAddressSet() {  
5.    return addressSet;  
6.}  

效果



Java代码 

1.Hibernate: select this_.PERSON_ID as PERSON1_0_0_, this_.PERSON_NAME as PERSON2_0_0_   
2.from PERSON this_  
3.P1  
4.Hibernate: select addressset0_.P_ID as P3_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_, addressset0_.ADDRESS_ID as ADDRESS1_1_0_, addressset0_.ADDRESS_NAME as ADDRESS2_1_0_   
5.from ADDRESS addressset0_   
6.where addressset0_.P_ID in (?, ?, ?)  
7.1   P1 A1  
8.2   P1 A2  
9.P2  
10.4   P2 A2  
11.3   P2 A1  
12.P3  
13.5   P3 A1  
14.6   P3 A2  

生成2条sql，第二条用in。 这里数据库中共有3条person。
如果设size＝0或1， 则效果同select一样，产生N＋1问题。
如size＝2， 生成3条sql，第一条相同，第二条为



Java代码 

1.select 。。。。  
2.from ADDRESS addressset0_   
3.where addressset0_.P_ID in (?, ?)  

第三条为



Java代码 

1.select。。  
2.from ADDRESS addressset0_   
3.where addressset0_.P_ID=?  

如果设size=3则生成2条sql，第一条相同，第一条为



Java代码 

1.select 。。。。  
2.from ADDRESS addressset0_   
3.where addressset0_.P_ID in (?, ?， ？)  

由上，size即为in中数据个数。