preface
A few days ago, I had a chat with my friend. He said that their project had a requirement to realize automatic generation of primary keys, and they did not want to manually set the primary keys every time they were added. So I asked him, that your database table set the primary key automatically increase not got. His answer is that the current ID of their project is generated by snowflake algorithm, so for the stability of the project, the generation mode of ID will not be changed.
My friend asked me if I had any ideas on how to implement it. The ORM framework of their company is Mybatis, SO I suggested him to change MyBatis into MyBatis – Plus. Mybatis – Plus supports annotated id generation automatically, and Mybatis – Plus only enhances MyBatis without changing it. My friend said that in order to stabilize the project, the project team didn’t have experience in using Mybatis – Plus before, so I don’t know if there will be any pit if I switch it rashly. There is no way back, I told him otherwise you use Mybatis interceptor to achieve a bar. Then there is another blowing water creation theme.
Front knowledge
Mybatis interceptor: Mybatis interceptor: MyBatis interceptor: MyBatis interceptor: MyBatis interceptor
1. The role of mybatis interceptor
Mybatis interceptor is designed to allow users to implement their own logic at some point without having to deal with myBatis’ inherent logic
Interceptor = Interceptor
Each custom interceptor is implemented
org.apache.ibatis.plugin.Interceptor
Copy the codeThis interface, and add the @intercepts annotation to the custom interceptor class
What types can interceptors intercept
-
Executor: Method of intercepting an Executor.
-
ParameterHandler: processing interception parameters.
-
ResultHandler: Intercepts the processing of result sets.
-
StatementHandler: Intercepts the processing of Sql syntax builds.
4, the order of interception
A. Execution sequence of different types of interceptors
Executor -> ParameterHandler -> StatementHandler -> ResultSetHandler
Copy the codeB. Multiple interceptors intercept the same target method of the same type, and the order of execution is the last configured interceptor
For example, mybatis is configured as follows
<plugins>
<plugin interceptor="com.lybgeek.InterceptorA" />
<plugin interceptor="com.lybgeek.InterceptorB" />
</plugins>
Copy the codeInterceptorB executes first.
If you integrate with Spring, inject the interceptor of the Spring IOC container first and execute later. For example, there is a mybatisConfig with the following interceptor bean configuration
@Bean
public InterceptorA interceptorA(a){
return new InterceptorA();
}
@Bean
public InterceptorB interceptorB(a){
return new InterceptorB();
}
Copy the codeInterceptorB executes first. Of course, if you use the @Component annotation directly, you can use the @order annotation to control the load Order
Introduction to interceptor annotations
@intercepts: Identifies this class as an interceptor
@signature: specifies which type and method the custom interceptor needs to intercept. The type in the @signature attribute indicates that one of four types (Executor, ParameterHandler, ResultHandler, StatementHandler) can be intercepted. Method specifies the method of the corresponding type (Executor, ParameterHandler, ResultHandler, StatementHandler). Args indicates the parameter type in method
6. Introduction to interceptor methods
A. Intercept method
public Object intercept(Invocation invocation) throws Throwable
Copy the codeThe method is to implement our own business logic, such as our primary key auto-generation logic is implemented here.
The target attribute in the Invocation class is the type in the @Signature; Method is @signature method; Args is a concrete instance object of the args parameter type in @signature
B. Plugin method
public Object plugin(Object target)
Copy the codeThis is either a return proxy object or a native proxy object. If you want to return a proxy object, the return value can be set to
Plugin.wrap(target, this);
thisFor the interceptorCopy the codeIf the return is a proxy object, the interceptor’s business logic is executed, if the target is returned directly, the business logic is not interceptor. To tell mybatis whether to intercept, if to intercept, generate proxy object, do not intercept is to generate native object
C, setProperties method
public void setProperties(Properties properties)
Copy the codeUsed to specify some properties in the Mybatis profile
The primary key automatically generates ideas
Define an interceptor
Mainly to intercept
`Executor#update(MappedStatement ms, Object parameter)`}
Copy the codeThis method. Mybatis insert, update, delete are all through this method, so we intercept this method, to achieve automatic generation of primary keys. The code block is as follows
@Intercepts(value={@Signature(type = Executor.class,method = "update",args = {MappedStatement.class,Object.class})})
public class AutoIdInterceptor implements Interceptor {}
Copy the code2. Determine the type of SQL operation
An Executor update method contains the types of new, modified, and deleted operations. You can use the SqlCommandType attribute of the MappedStatement class to determine the types of new, modified, and deleted operations. This class contains all operation types
public enum SqlCommandType {
UNKNOWN, INSERT, UPDATE, DELETE, SELECT, FLUSH;
}
Copy the codeThe primary key increment is performed when SqlCommandType is INSERT
3. Fill in primary key values
3.1. Write auto-generated ID annotations
Target(ElementType.FIELD)
@Retention(RetentionPolicy.RUNTIME)
@Documented
public @interface AutoId {
/** * primary key *@return* /
String primaryKey(a);
/** * Supported primary key algorithm types *@return* /
IdType type(a) default IdType.SNOWFLAKE;
enum IdType{
SNOWFLAKE
}
}
Copy the code3.2 Implementation of snowflake algorithm
We can directly take the toolkit provided by hutool idUtil to directly implement the algorithm.
The introduction of
<dependency>
<groupId>cn.hutool</groupId>
<artifactId>hutool-all</artifactId>
</dependency>
Copy the codeSnowflake snowflake = IdUtil.createSnowflake(0.0);
long value = snowflake.nextId();
Copy the code3.3. Fill in primary key values
The core of its implementation is the use of reflection. The core code snippet is shown below
ReflectionUtils.doWithFields(entity.getClass(), field->{
ReflectionUtils.makeAccessible(field);
AutoId autoId = field.getAnnotation(AutoId.class);
if(! ObjectUtils.isEmpty(autoId) && (field.getType().isAssignableFrom(Long.class))){switch (autoId.type()){
case SNOWFLAKE:
SnowFlakeAutoIdProcess snowFlakeAutoIdProcess = new SnowFlakeAutoIdProcess(field);
snowFlakeAutoIdProcess.setPrimaryKey(autoId.primaryKey());
finalIdProcesses.add(snowFlakeAutoIdProcess);
break; }}});Copy the codepublic class SnowFlakeAutoIdProcess extends BaseAutoIdProcess {
private static Snowflake snowflake = IdUtil.createSnowflake(0.0);
public SnowFlakeAutoIdProcess(Field field) {
super(field);
}
@Override
void setFieldValue(Object entity) throws Exception{
longvalue = snowflake.nextId(); field.set(entity,value); }}Copy the codeIf mapper.xml in your project already has an ID in its insert statement, for example
insert into sys_test( `id`,`type`, `url`,`menu_type`,`gmt_create`)values( #{id},#{type}, #{url},#{menuType},#{gmtCreate})
Copy the codeYou just need to fill in the ID value. The interceptor’s job is done. If mapper. XML insert does not contain id, it looks like
insert into sys_test( `type`, `url`,`menu_type`,`gmt_create`)values( #{type}, #{url},#{menuType},#{gmtCreate})
Copy the codeYou also need to rewrite the INSERT statement and add the ID parameter
4, Rewrite insert statement and add id parameter (optional)
4.1 Rewriting insert statements
Method one: Obtain the BoundSql object from the MappedStatement object, obtain the BoundSql object from the SqlSource object, use the BoundSql#getSql method to obtain the original SQL, and append the ID to the original SQL
Method 2:
The introduction of
<dependency>
<groupId>com.alibaba</groupId>
<artifactId>druid</artifactId>
<version>${druid.version}</version>
</dependency>
Copy the codethrough
com.alibaba.druid.sql.dialect.mysql.parser.MySqlStatementParser
Copy the codeGet the corresponding table name, the name of the field to insert. Then reassemble the new INSERT statement
4.2 Reset the new SQL to the Invocation
The main idea is to create a new MappedStatement that binds the new SQL and assigns the new MappedStatement to the Invocation ARgs [0]
private void resetSql2Invocation(Invocation invocation, BoundSqlHelper boundSqlHelper,Object entity) throws SQLException {
final Object[] args = invocation.getArgs();
MappedStatement statement = (MappedStatement) args[0];
MappedStatement newStatement = newMappedStatement(statement, new BoundSqlSqlSource(boundSqlHelper));
MetaObject msObject = MetaObject.forObject(newStatement, new DefaultObjectFactory(), new DefaultObjectWrapperFactory(),new DefaultReflectorFactory());
msObject.setValue("sqlSource.boundSqlHelper.boundSql.sql", boundSqlHelper.getSql());
args[0] = newStatement;
}
Copy the code4.3 Adding the ID Parameter
Its core is utilization
org.apache.ibatis.mapping.ParameterMapping
Copy the codeThe core code snippet is shown below
private void setPrimaryKeyParaterMapping(String primaryKey) {
ParameterMapping parameterMapping = new ParameterMapping.Builder(boundSqlHelper.getConfiguration(),primaryKey,boundSqlHelper.getTypeHandler()).build();
boundSqlHelper.getBoundSql().getParameterMappings().add(parameterMapping);
}
Copy the codeInject the MyBatis interceptor into the Spring container
You can add it directly to the interceptor
@org.springframework.stereotype.Component
Copy the codeAnnotation. Also through
@Bean
public AutoIdInterceptor autoIdInterceptor(a){
return new AutoIdInterceptor();
}
Copy the code6. Add the following comments to the entity fields where you want to implement auto-increment primary keys
@AutoId(primaryKey = "id")
private Long id;
Copy the codetest
1. The corresponding test entity and unit test code are as follows
@Data
public class TestDO implements Serializable {
private static final long serialVersionUID = 1L;
@AutoId(primaryKey = "id")
private Long id;
private Integer type;
private String url;
private Date gmtCreate;
private String menuType;
}
Copy the code@Autowired
private TestService testService;
@Test
public void testAdd(a){
TestDO testDO = new TestDO();
testDO.setType(1);
testDO.setMenuType("1");
testDO.setUrl("www.test.com");
testDO.setGmtCreate(new Date());
testService.save(testDO);
testService.get(110L);
}
@Test
public void testBatch(a){
List<TestDO> testDOList = new ArrayList<>();
for (int i = 0; i < 3; i++) {
TestDO testDO = new TestDO();
testDO.setType(i);
testDO.setMenuType(i+"");
testDO.setUrl("www.test"+i+".com");
testDO.setGmtCreate(new Date());
testDOList.add(testDO);
}
testService.saveBatch(testDOList);
}
Copy the code2, when a mapper insert statement contains an ID, look like this
<insert id="save" parameterType="com.lybgeek.TestDO" useGeneratedKeys="true" keyProperty="id">
insert into sys_test(`id`,`type`, `url`,`menu_type`,`gmt_create`)
values( #{id},#{type}, #{url},#{menuType},#{gmtCreate})
</insert>
Copy the codeAnd batch insert SQL
<insert id="saveBatch" parameterType="java.util.List" useGeneratedKeys="false">
insert into sys_test( `id`,`gmt_create`,`type`,`url`,`menu_type`)
values
<foreach collection="list" item="test" index="index" separator=",">
( #{test.id},#{test.gmtCreate},#{test.type}, #{test.url},
#{test.menuType})
</foreach>
</insert>
Copy the codeView the console SQL print statements
15:52:04 [main] DEBUG com.lybgeek.dao.TestDao.save - ==> Preparing: insert into sys_test(`id`,`type`, `url`,`menu_type`,`gmt_create`) values(? ,? ,? ,? ,?)
15:52:04 [main] DEBUG com.lybgeek.dao.TestDao.save - ==> Parameters: 356829258376544258(Long), 1(Integer), www.test.com(String), 1(String), 2020- 09 -11 15:52:04.738(Timestamp)
15:52:04 [main] DEBUG com.nlybgeek.dao.TestDao.save - <== Updates: 1
Copy the code15:52:04 [main] DEBUG c.n.lybgeek.dao.TestDao.saveBatch - ==> Preparing: insert into sys_test( `id`,`gmt_create`,`type`,`url`,`menu_type`) values (? ,? ,? ,? ,?) , (? ,? ,? ,? ,?) , (? ,? ,? ,? ,?)
15:52:04 [main] DEBUG c.n.lybgeek.dao.TestDao.saveBatch - ==> Parameters: 356829258896637961(Long), 2020- 09 -11 15:52:04.847(Timestamp), 0(Integer), www.test0.com(String), 0(String), 356829258896637960(Long), 2020- 09 -11 15:52:04.847(Timestamp), 1(Integer), www.test1.com(String), 1(String), 356829258896637962(Long), 2020- 09 -11 15:52:04.847(Timestamp), 2(Integer), www.test2.com(String), 2(String)
15:52:04 [main] DEBUG c.n.lybgeek.dao.TestDao.saveBatch - <== Updates: 3
Copy the codeViewing a Database
3, if the mapper insert statement does not contain an ID, look like this
<insert id="save" parameterType="com.lybgeek.TestDO" useGeneratedKeys="true" keyProperty="id">
insert into sys_test(`type`, `url`,`menu_type`,`gmt_create`)
values(#{type}, #{url},#{menuType},#{gmtCreate})
</insert>
Copy the codeAnd batch insert SQL
<insert id="saveBatch" parameterType="java.util.List" useGeneratedKeys="false">
insert into sys_test(`gmt_create`,`type`,`url`,`menu_type`)
values
<foreach collection="list" item="test" index="index" separator=",">
(#{test.gmtCreate},#{test.type}, #{test.url},
#{test.menuType})
</foreach>
</insert>
Copy the codeView the console SQL print statements
15:59:46 [main] DEBUG com.lybgeek.dao.TestDao.save - ==> Preparing: insert into sys_test(`type`,`url`,`menu_type`,`gmt_create`,id) values (? ,? ,? ,? ,?)
15:59:46 [main] DEBUG com.lybgeek.dao.TestDao.save - ==> Parameters: 1(Integer), www.test.com(String), 1(String), 2020- 09 -11 15:59:46.741(Timestamp), 356831196144992264(Long)
15:59:46 [main] DEBUG com.lybgeek.dao.TestDao.save - <== Updates: 1
Copy the code15:59:46 [main] DEBUG c.n.lybgeek.dao.TestDao.saveBatch - ==> Preparing: insert into sys_test(`gmt_create`,`type`,`url`,`menu_type`,id) values (? ,? ,? ,? ,?).(? ,? ,? ,? ,?).(? ,? ,? ,? ,?)
15:59:46 [main] DEBUG c.n.lybgeek.dao.TestDao.saveBatch - ==> Parameters: 2020- 09 -11 15:59:46.845(Timestamp), 0(Integer), www.test0.com(String), 0(String), 356831196635725829(Long), 2020- 09 -11 15:59:46.845(Timestamp), 1(Integer), www.test1.com(String), 1(String), 356831196635725828(Long), 2020- 09 -11 15:59:46.845(Timestamp), 2(Integer), www.test2.com(String), 2(String), 356831196635725830(Long)
15:59:46 [main] DEBUG c.n.lybgeek.dao.TestDao.saveBatch - <== Updates: 3
Copy the codeFrom the console, we can see that when mapper.xml is not configured with an ID field, the interceptor will append the ID field for us automatically
Viewing a Database
conclusion
Mybatis interceptor to achieve primary key generation, but more on how to implement an interceptor and primary key generation ideas, and the intercept implementation of primary key method posted. The main reason is that the primary key is automatically generated in Mybatis – Plus, followed by ideas, we can achieve their own. Finally, if you are interested in the implementation, you can check out the demo link at the end of this article
Reference documentation
Mybatis interceptor
Mybatis plugin implements custom overwrite table names
Mybatis interceptor, dynamic modification of SQL statements
The demo link
Github.com/lyb-geek/sp…