mybatis批量更新
1.先看DAO层
?
|
1
2
3
4
5
6
|
publicvoidupdateList(List<AddService> oldAddServiceList) {
SqlSession session = sqlSessionFactory.openSession(ExecutorType.BATCH,true);
session.update("AddServiceMapper.batchUpdate",oldAddServiceList);
session.commit();
session.close();
}
|
当然DAO层最上方类中注入上sqlSessionFactory
?
|
1
2
|
@Resource(name ="sqlSessionFactory")
privateSqlSessionFactory sqlSessionFactory;
|
?
|
1
|
sqlSessionFactory.openSession(ExecutorType.BATCH,true);
|
详细了解请点击 http://www.cnblogs.com/xcch/articles/2042298.html
2.再看XML配置文件
?
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
<updateid="batchUpdate"parameterType="java.util.List">
<foreachcollection="list"item="item"index="index"separator=";">
update sp_add_service set
SERVICE_NAME = #{item.serviceName},
SERVICE_TYPE = #{item.serviceType},
PRICE = #{item.price},
DESCRIPTION = #{item.description},
PRICE_RATE = #{item.priceRate},
IS_REQUIRED = #{item.isRequired},
MODIFY_TIME = now()
where ID = #{item.id}
</foreach>
</update>
|
3.再看Service层
?
|
1
2
3
4
|
@Transactional
publicvoidupdateOldAddServces(List<AddService> oldAddServiceList) {
addServiceDao.updateList(oldAddServiceList);
}
|
注意！！！以上代码全照例子写后发现批量更新还是不行，后来百度了一下，数据库连接代码后面加上&allowMultiQueries=true
jdbc:mysql://111.111.111.111:3306/ibada?characterEncoding=UTF-8&allowMultiQueries=true
然后再执行，发现OK了。批量更新大工告成。
批量插入和批量删除大体一样，就给上XML供大家参考，不做详细描述。
?
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
|
<!-- 批量插入 -->
<insertid="batchInsert"parameterType="java.util.List">
insert into sp_add_service( <includerefid="AddService_Column_List"/> )
values
<foreachcollection="list"item="item"index="index"separator=",">
(#{item.id}, #{item.logisticsLtdId}, #{item.serviceName}, #{item.serviceType},
#{item.price},#{item.description},#{item.priceRate},
#{item.parentId}, #{item.isRequired}, now(), now())
</foreach>
</insert>
<deleteid="batchDelete">
delete from sp_add_service
where LOGISTICS_LTD_ID = #{logisticsLtdId}
and ID not in
<foreachitem="item"index="index"collection="oldServiceIds"
open="("separator=","close=")">
#{item}
</foreach>
</delete>
|
关于上面的博客里面的内容一直没仔细看，可能原博主那个true写的不对，应该是fasle，
?
|
1
|
(ExecutorType.BATCH,false);
|
测试环境，mysql数据库。
测试内容，5000条数据的插入。
测试Dao层代码：
?
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
@Repository
publicclassTruckSourceDaoextendsBaseDao{
@Resource(name ="sqlSessionFactory")
privateSqlSessionFactory sqlSessionFactory;
publicInteger batchInsertList(List<TruckSource> truckSources){
SqlSession session = sqlSessionFactory.openSession(ExecutorType.BATCH,false);
for(inti =0;i < truckSources.size(); i++) {
session.insert("TruckSource.insertOne",truckSources.get(i));
}
session.commit();
session.clearCache();
returnnull;
}
publicInteger insertList(List<TruckSource> truckSources){
returngetSqlSession().insert("TruckSource.insertBatch",truckSources);
}
publicInteger batchInsertLists(List<TruckSource> truckSources){
SqlSession session = sqlSessionFactory.openSession(ExecutorType.BATCH,false);
session.insert("TruckSource.insertBatch",truckSources);
session.commit();
session.clearCache();
returnnull;
}
}
|
Mapper.xml里面代码：
?
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
|
<sqlid="TruckSource_Column_List">
TRUCK_OWNER_ID,TRUCK_SOURCE_TYPE,INITIATION_PROVINCE_CODE,INITIATION_CITY_CODE,INITIATION_DISTRICT_CODE,DESTINATION_PROVINCE_CODE,DESTINATION_CITY_CODE,DESTINATION_DISTRICT_CODE,
SURPLUS_LOAD,SURPLUS_LENGTH,SURPLUS_VOLUME,START_TIME,VALID_DATE,STATE,NOTICE,CREATE_TIME,MODIFY_TIME
</sql>
<insertid="insertBatch"parameterType="java.util.List">
insert into sp_truck_source(<includerefid="TruckSource_Column_List"/>)
values
<foreachcollection="list"item="item"index="index"separator=",">
(#{item.truckOwnerId}, #{item.truckSourceType}, #{item.initiationProvinceCode}, #{item.initiationCityCode},
#{item.initiationDistrictCode}, #{item.destinationProvinceCode}, #{item.destinationCityCode}, #{item.destinationDistrictCode},
#{item.surplusLoad}, #{item.surplusLength}, #{item.surplusVolume}, #{item.startTime},
#{item.validDate},#{item.state},#{item.notice},
now(), now())
</foreach>
</insert>
<insertid="insertOne"parameterType="TruckSource">
insert into sp_truck_source(<includerefid="TruckSource_Column_List"/>)
values (#{truckOwnerId},#{truckSourceType},#{initiationProvinceCode},#{initiationCityCode},#{initiationDistrictCode},#{destinationProvinceCode},#{destinationCityCode},#{destinationDistrictCode},
#{surplusLoad},#{surplusLength},#{surplusVolume},#{startTime},#{validDate},#{state},#{notice},now(),now())
</insert>
|
单元测试代码：
?
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
|
@RunWith(SpringJUnit4ClassRunner.class)
@ContextConfiguration(locations ="classpath:applicationContext.xml")
publicclassTTTT {
@Resource
privateTruckSourceDao truckSourceDao;
@Test
publicvoidinsert() {
List<TruckSource> truckSources =newArrayList<TruckSource>();
for(inti =0;i <5000; i++) {
TruckSource truckSource =newTruckSource();
truckSource.setDestinationDistrictCode(i+"");
truckSource.setDestinationCityCode(i+"");
truckSource.setDestinationProvinceCode(i+"");
truckSource.setInitiationCityCode(i+"");
truckSource.setInitiationDistrictCode(i+"");
truckSource.setInitiationProvinceCode(i+"");
truckSources.add(truckSource);
}
Long beginTime = System.currentTimeMillis();
truckSourceDao.batchInsertList(truckSources);
System.out.println("batchInsert:"+(System.currentTimeMillis() - beginTime));
beginTime = System.currentTimeMillis();
truckSourceDao.insertList(truckSources);
System.out.println("foreach insert:"+(System.currentTimeMillis() - beginTime));
beginTime = System.currentTimeMillis();
truckSourceDao.batchInsertLists(truckSources);
System.out.println("batch foreach insert:"+(System.currentTimeMillis() - beginTime));
}
}
|
测试方法种类，3种：
1.batchInsertList 插入时单独开一个session用
?
|
1
|
SqlSession session = sqlSessionFactory.openSession(ExecutorType.BATCH,false);
|
然后下面for循环依次单个插入
2.在mapper.xml里面写foreach sql插入
3.加上ExecutoerType.Batch,再在mapper.xml里面写foreach sql插入
最后实验结果是第三种执行时间最短：
{insert3=5128, insert1=146085, insert2=5625}
上面是数据量为5000情况下的实验结果。
在数据量为1000的情况下，第2种是2000多毫秒，第三种是800多毫秒，所以差距还是比较大的。