Mybatis插件的开发经验

来源：cnblogs　　作者：冬眠的山谷　　时间：2019/6/11 9:17:37　　对本文有异议

前言

MyBatis开放用户实现自己的插件，从而对整个调用过程进行个性化扩展。

这是MyBatis整个调用流程的主要参与者。

我们可以对其中的一些过程进行拦截，添加自己的功能，比如重写Sql添加分页参数。

拦截的接口

MyBatis允许拦截的接口如下

Executor

public interface Executor {
    ResultHandler NO_RESULT_HANDLER = null;
    int update(MappedStatement var1, Object var2) throws SQLException;
    <E> List<E> query(MappedStatement var1, Object var2, RowBounds var3, ResultHandler var4, CacheKey var5, BoundSql var6) throws SQLException;
    <E> List<E> query(MappedStatement var1, Object var2, RowBounds var3, ResultHandler var4) throws SQLException;
    <E> Cursor<E> queryCursor(MappedStatement var1, Object var2, RowBounds var3) throws SQLException;
    List<BatchResult> flushStatements() throws SQLException;
    void commit(boolean var1) throws SQLException;
    void rollback(boolean var1) throws SQLException;
    CacheKey createCacheKey(MappedStatement var1, Object var2, RowBounds var3, BoundSql var4);
    boolean isCached(MappedStatement var1, CacheKey var2);
    void clearLocalCache();
    void deferLoad(MappedStatement var1, MetaObject var2, String var3, CacheKey var4, Class<?> var5);
    Transaction getTransaction();
    void close(boolean var1);
    boolean isClosed();
    void setExecutorWrapper(Executor var1);
}

ParameterHandler

public interface ParameterHandler {
    Object getParameterObject();
    void setParameters(PreparedStatement var1) throws SQLException;
}

ResultSetHandler

public interface ResultSetHandler {
    <E> List<E> handleResultSets(Statement var1) throws SQLException;
    <E> Cursor<E> handleCursorResultSets(Statement var1) throws SQLException;
    void handleOutputParameters(CallableStatement var1) throws SQLException;
}

StatementHandler

public interface StatementHandler {
    Statement prepare(Connection var1, Integer var2) throws SQLException;
    void parameterize(Statement var1) throws SQLException;
    void batch(Statement var1) throws SQLException;
    int update(Statement var1) throws SQLException;
    <E> List<E> query(Statement var1, ResultHandler var2) throws SQLException;
    <E> Cursor<E> queryCursor(Statement var1) throws SQLException;
    BoundSql getBoundSql();
    ParameterHandler getParameterHandler();
}

只要拦截器定义了拦截的接口和方法，后续调用该方法时，将会被拦截。

拦截器实现

如果要实现自己的拦截器，需要实现接口Interceptor


@Slf4j
@Intercepts(@Signature(type = Executor.class,
        method ="update",
        args ={MappedStatement.class,Object.class} ))
public class MyIntercetor implements Interceptor {
    @Override
    public Object intercept(Invocation invocation) throws Throwable {
        log.info("MyIntercetor ...");
        Object result = invocation.proceed();
        log.info("result = " + result);
        return result;
    }
    @Override
    public Object plugin(Object o) {
        return Plugin.wrap(o,this);
    }
    @Override
    public void setProperties(Properties properties) {
    }
}

1. 拦截方法配置

Intercepts,Signature

public @interface Intercepts {
    Signature[] value();
}

public @interface Signature {
    Class<?> type();

    String method();

    Class<?>[] args();
}

配置

@Intercepts(@Signature(type = Executor.class,
        method ="update",
        args ={MappedStatement.class,Object.class} ))

我们知道Java中方法的签名包括所在的类，方法名称，入参。

@Signature定义方法签名

type：拦截的接口，为上节定义的四个接口

method:拦截的接口方法

args:参数类型列表，需要和方法中定义的顺序一致。

也可以配置多个

@Intercepts({@Signature(
    type = Executor.class,
    method = "query",
    args = {MappedStatement.class, Object.class, RowBounds.class, ResultHandler.class}
), @Signature(
    type = Executor.class,
    method = "query",
    args = {MappedStatement.class, Object.class, RowBounds.class, ResultHandler.class, CacheKey.class, BoundSql.class}
)})

2. intercept(Invocation invocation)

public class Invocation {
    private final Object target;
    private final Method method;
    private final Object[] args;

    public Invocation(Object target, Method method, Object[] args) {
        this.target = target;
        this.method = method;
        this.args = args;
    }

    public Object getTarget() {
        return this.target;
    }

    public Method getMethod() {
        return this.method;
    }

    public Object[] getArgs() {
        return this.args;
    }

    public Object proceed() throws InvocationTargetException, IllegalAccessException {
        return this.method.invoke(this.target, this.args);
    }
}

通过Invocation可以获取到被拦截的方法的调用对象，方法，参数。

proceed()用于继续执行并获得最终的结果。

这里使用了设计模式中的责任链模式。

3.这里不能返回null。

用于给被拦截的对象生成一个代理对象，并返回它。

@Override
    public Object plugin(Object o) {
        return Plugin.wrap(o,this);
    }

可以看下wrap方法，其实现了JDK的接口InvocationHandler，也就是为传入的target创建了一个代理对象。这里使用了JDK动态代理方式。也可以自己实现其他代理方式，比如cglib.

    public class Plugin implements InvocationHandler {
    private final Object target;
    private final Interceptor interceptor;
    private final Map<Class<?>, Set<Method>> signatureMap;
　　 public static Object wrap(Object target, Interceptor interceptor) {
        Map<Class<?>, Set<Method>> signatureMap = getSignatureMap(interceptor);
        Class<?> type = target.getClass();
        Class<?>[] interfaces = getAllInterfaces(type, signatureMap);
        return interfaces.length > 0 ? Proxy.newProxyInstance(type.getClassLoader(), interfaces, new Plugin(target, interceptor, signatureMap)) : target;
    }

　　public Object invoke(Object proxy, Method method, Object[] args) throws Throwable {
    try {
        Set<Method> methods = (Set)this.signatureMap.get(method.getDeclaringClass());
        return methods != null && methods.contains(method) ? this.interceptor.intercept(new Invocation(this.target, method, args)) : method.invoke(this.target, args);
    } catch (Exception var5) {
        throw ExceptionUtil.unwrapThrowable(var5);
    }
 }

由于使用了动态代理，方法执行时，将会被调用invoke方法，会先判断是否设置了拦截器：methods != null && methods.contains(method)，

如果设置了拦截器，则调用拦截器this.interceptor.intercept(new Invocation(this.target, method, args))

否则直接调用method.invoke(this.target, args);

4.拦截器在执行前输出"MyIntercetor ..."，在数据库操作返回后输出"result =xxx"

       log.info("MyIntercetor ...");
        Object result = invocation.proceed();
        log.info("result = " + result);

插件实现完成！

测试

在Spring中引入很简单。

第一种方式:

创建拦截器的bean

@Slf4j
@Configuration
public class IntercetorConfiguration {
    @Bean
    public MyIntercetor myIntercetor(){
        return new MyIntercetor();
    }
}

注意第一种方式和第二种方式仅适用于SpringBoot应用，并且引入以下依赖

<dependency>
      <groupId>org.mybatis.spring.boot</groupId>
      <artifactId>mybatis-spring-boot-starter</artifactId>
      <version>1.3.2</version>
</dependency>

第二种方式

手动往Configuration中添加拦截器。

@Slf4j
@Configuration
public class IntercetorConfiguration {
  @Autowired
    private List<SqlSessionFactory> sqlSessionFactoryList;
    @PostConstruct
    public void addPageInterceptor() {
        MyIntercetor interceptor = new MyIntercetor();
        Iterator var3 = this.sqlSessionFactoryList.iterator();
        while(var3.hasNext()) {
            SqlSessionFactory sqlSessionFactory = (SqlSessionFactory)var3.next();
            sqlSessionFactory.getConfiguration().addInterceptor(interceptor);
        }
    }
}

第三种方式

如果是纯Spring应用，可在mybatis配置文件中配置

<plugins>
    <plugin   intercetor="xxx.xxx.MyIntercetor">
            <property   name="xxx"  value="xxx">
    </plugin>
</plugins>

由于上面定义的拦截器是拦截Executor的update方法，所以在执行insert,update,delete的操作时，将会被拦截。

本例子使用insert来测试。具体代码查看：GitHub

2019-06-10 16:08:03.109  INFO 20410 --- [nio-8110-exec-1] c.m.user.dao.intercetor.MyIntercetor     : MyIntercetor ...
2019-06-10 16:08:03.166  INFO 20410 --- [nio-8110-exec-1] com.alibaba.druid.pool.DruidDataSource   : {dataSource-1} inited
2019-06-10 16:08:03.267 DEBUG 20410 --- [nio-8110-exec-1] o.m.s.t.SpringManagedTransaction         : JDBC Connection [com.mysql.cj.jdbc.ConnectionImpl@5cb1c36e] will not be managed by Spring
2019-06-10 16:08:03.274 DEBUG 20410 --- [nio-8110-exec-1] c.m.u.dao.mapper.UserMapper.insertList   : ==>  Preparing: insert into user (name) values (?) , (?) , (?) 
2019-06-10 16:08:03.307 DEBUG 20410 --- [nio-8110-exec-1] c.m.u.dao.mapper.UserMapper.insertList   : ==> Parameters: name:58(String), name:64(String), name:69(String)
2019-06-10 16:08:03.355 DEBUG 20410 --- [nio-8110-exec-1] c.m.u.dao.mapper.UserMapper.insertList   : <==    Updates: 3
2019-06-10 16:08:03.358 DEBUG 20410 --- [nio-8110-exec-1] c.m.u.d.m.U.insertList!selectKey         : ==>  Preparing: SELECT LAST_INSERT_ID() 
2019-06-10 16:08:03.358 DEBUG 20410 --- [nio-8110-exec-1] c.m.u.d.m.U.insertList!selectKey         : ==> Parameters: 
2019-06-10 16:08:03.380 DEBUG 20410 --- [nio-8110-exec-1] c.m.u.d.m.U.insertList!selectKey         : <==      Total: 1

2019-06-10 16:08:03.381  INFO 20410 --- [nio-8110-exec-1] c.m.user.dao.intercetor.MyIntercetor     : result = 3

可以看到拦截器被调用了。

简单的分页插件实现

这里拦截StatementHandler的prepare方法，也就是SQL语句预编译之前进行SQL改写。

@Slf4j
@Intercepts({@Signature(type = StatementHandler.class, method = "prepare", args = {Connection.class, Integer.class})})
public class PageIntercetor implements Interceptor {
    @Override
    public Object intercept(Invocation invocation) throws Throwable {
        log.info("StatementHandler  prepare ...");
        StatementHandler statementHandler = (StatementHandler) invocation.getTarget();
        ParameterHandler parameterHandler = statementHandler.getParameterHandler();
        BoundSql boundSql = statementHandler.getBoundSql();
        //获取到原始sql语句
        String sql = boundSql.getSql();
        String mSql = sql + " limit 0,1";
        //通过反射修改sql语句
        Field field = boundSql.getClass().getDeclaredField("sql");
        field.setAccessible(true);
        field.set(boundSql, mSql);
        return invocation.proceed();
    }
    @Override
    public Object plugin(Object target) {
        return Plugin.wrap(target, this);
    }
    @Override
    public void setProperties(Properties properties) {
        //此处可以接收到配置文件的property参数
        System.out.println(properties.getProperty("name"));
    }
}

分页插件实现的难点在于当使用不同的Statement时，执行流程是不一样的。

Statement需要定义statementType="STATEMENT"，这个时候SQL语句不需要进行预编译处理，参数是与xml中配饰的SQL语句拼接在一起的。

<select id="select" resultMap="BaseResultMap" statementType="STATEMENT">
  select id, name
  from user
  where
    name = '${name}'
</select>

而当使用PreparedStatement时需要定义statementType="PREPARED"，这个时候SQL语句需要进行预编译处理。CallableStatement(用于调用存储过程)同理。

  <select id="select" resultMap="BaseResultMap" statementType="PREPARED">
    select id, name
    from user
    where
      name = #{name}
  </select>