一对多 、多对多 的检索策略
在映射文件中, 用 元素来配置一对多关联及多对多关联关系. 元素有 lazy 和 fetch 属性
lazy: 主要决定 orders 集合被初始化的时机. 即到底是在加载 Customer 对象时就被初始化, 还是在程序访问 orders 集合时被初始化
fetch: 取值为 “select” 或 “subselect” 时, 决定初始化 orders 的查询语句的形式; 若取值为”join”, 则决定 orders 集合被初始化的时机若把 fetch 设置为 “join”, lazy 属性将被忽略
② 当取fasle 时，
运行的SQL
①取值为 select
运行的SQL
lazy: 主要决定 orders 集合被初始化的时机. 即到底是在加载 Customer 对象时就被初始化, 还是在程序访问 orders 集合时被初始化
fetch: 取值为 “select” 或 “subselect” 时, 决定初始化 orders 的查询语句的形式; 若取值为”join”, 则决定 orders 集合被初始化的时机若把 fetch 设置为 “join”, lazy 属性将被忽略
Clazz 类 的对象关系映射，之一中的set 属性中的配置
/**
* Set 的 lazy 属性
* 1. 1-n 或n-n 的集合属性默认使用懒加载检索策略
* 2. 可以通过设置set 的lazy 属性来修改默认的检索策略，默认为true
* 并不建议设置为false
* 3. lazy 还可以设置为extra，增强的延迟检索。该取值会尽可能的延迟集合初始化的时机
* 4. Hibernate.initialize() 静态方法显式初始化
*/Lazy 属性：默认值是true
① 当取true 时，
@Test
public void testOne2ManyLevelStrategy() {
Clazz clazz =(Clazz) session.get(Clazz.class, 1);
System.out.println(clazz.getClass());
System.out.println(clazz.getStus().size());
}
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_0_
from
CLAZZS clazz0_
where
clazz0_.CLAZZ_ID=?
class com.baidu.setStrategy.Clazz
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
2② 当取fasle 时，
运行的SQL
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_0_
from
CLAZZS clazz0_
where
clazz0_.CLAZZ_ID=?
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
class com.baidu.setStrategy.Clazz
2③ 当取extra 时，
运行的SQL
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_0_
from
CLAZZS clazz0_
where
clazz0_.CLAZZ_ID=?
class com.baidu.setStrategy.Clazz
Hibernate:
select
count(STU_ID)
from
STUDENTS
where
CLAZZ_ID =?
2
@Test
public void testOne2ManyLevelStrategy() {
Clazz clazz =(Clazz) session.get(Clazz.class, 1);
System.out.println(clazz.getClass());
System.out.println(clazz.getStus().size());
Student student = new Student();
student.setStuId(1);
System.out.println(clazz.getStus().contains(student));
}
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_0_
from
CLAZZS clazz0_
where
clazz0_.CLAZZ_ID=?
class com.baidu.setStrategy.Clazz
Hibernate:
select
count(STU_ID)
from
STUDENTS
where
CLAZZ_ID =?
2
Hibernate:
select
1
from
STUDENTS
where
CLAZZ_ID =?
and STU_ID =?
true
@Test
public void testOne2ManyLevelStrategy() {
Clazz clazz =(Clazz) session.get(Clazz.class, 1);
System.out.println(clazz.getClass());
System.out.println(clazz.getStus().size());
Student student = new Student();
student.setStuId(1);
System.out.println(clazz.getStus().contains(student));
Hibernate.initialize(clazz.getStus());
}
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_0_
from
CLAZZS clazz0_
where
clazz0_.CLAZZ_ID=?
class com.baidu.setStrategy.Clazz
Hibernate:
select
count(STU_ID)
from
STUDENTS
where
CLAZZ_ID =?
2
Hibernate:
select
1
from
STUDENTS
where
CLAZZ_ID =?
and STU_ID =?
true
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
/**
* Set 的 batch-size 属性: 设定一次初始化 set 集合的数量
*/
@Test
public void testSetBatchSize() {
List clazzs = session.createQuery("FROM Clazz").list();
System.out.println(clazzs.size());
for(Clazz clazz:clazzs ){
if(clazz.getStus() != null)
System.out.println(clazz.getStus().size());
}
} 运行的SQL
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_
from
CLAZZS clazz0_
3
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
2
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
2
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
2
运行的SQL
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_
from
CLAZZS clazz0_
3
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID in (
?, ?, ?
)
2
2
2fetch属性: 取值为 “select” 或 “subselect” 时, 决定初始化 students 的查询语句的形式;
若取值为”join”, 则决定orders 集合被初始化的时机
/**
* Set 的 fetch 属性:
* 1. 默认值为select：通过正常的方式来初始化set元素
* 2. 可以取值subselect：
* ① 通过子查询的方式来初始化所有的 set 集合。子查询作为where 子句的 in的条件出现，
* 子查询查询所有1 的一端的ID
* ② 此时lazy 有效，但是batch-size 属性将被忽略，不再起作用。
*/
/**
* Set 的 fetch 属性:
* 3. 可以取值为join: batch-size有效
* ① 在加载 1 的一端的对象时，使用迫切左外连接的方式检索 n 的一端的集合属性
*
* ② 会忽略lazy 属性
*
* ③ 使用HQL 查询时，会 忽略fetch=join
*
*
* 迫切左外连接：使用左外连接进行查询，且把集合属性进行初始化
*/①取值为 select
@Test
public void testSetFetchSelectOrSubselect() {
List clazzs = session.createQuery("FROM Clazz").list();
System.out.println(clazzs.size());
for(Clazz clazz:clazzs ){
if(clazz.getStus() != null)
System.out.println(clazz.getStus().size());
}
}
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_
from
CLAZZS clazz0_
3
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID in (
?, ?
)
2
2
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
2② 取值为usubselect
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_
from
CLAZZS clazz0_
3
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID in (
select
clazz0_.CLAZZ_ID
from
CLAZZS clazz0_
)
2
2
2
@Test
public void testSetFetchJoin() {
Clazz clazz = (Clazz) session.get(Clazz.class, 1);
System.out.println(clazz.getStus().size());
}使用HQL 查询时，batch-size有效， 但会 忽略fetch=join ,
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_
from
CLAZZS clazz0_
3
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID in (
?, ?
)
2
2
Hibernate:
select
stus0_.CLAZZ_ID as CLAZZ_ID3_0_1_,
stus0_.STU_ID as STU_ID1_1_1_,
stus0_.STU_ID as STU_ID1_1_0_,
stus0_.STU_NAME as STU_NAME2_1_0_
from
STUDENTS stus0_
where
stus0_.CLAZZ_ID=?
2 @Test
public void testSetFetchJoin() {
Clazz clazz = (Clazz) session.get(Clazz.class, 1);
System.out.println(clazz.getStus().size());
}运行的SQL
Hibernate:
select
clazz0_.CLAZZ_ID as CLAZZ_ID1_0_1_,
clazz0_.CLAZZ_NAME as CLAZZ_NA2_0_1_,
stus1_.CLAZZ_ID as CLAZZ_ID3_0_3_,
stus1_.STU_ID as STU_ID1_1_3_,
stus1_.STU_ID as STU_ID1_1_0_,
stus1_.STU_NAME as STU_NAME2_1_0_
from
CLAZZS clazz0_
left outer join
STUDENTS stus1_
on clazz0_.CLAZZ_ID=stus1_.CLAZZ_ID
where
clazz0_.CLAZZ_ID=?
2