如何深度理解mybatis？

2022-11-13 网易号

深度自定义MyBatis

回顾mybatis的操作的核心步骤编写核心类SqlSessionFacotryBuild进行解析配置文件深度分析解析SqlSessionFacotryBuild干的核心工作编写核心类SqlSessionFacotry 深度分析解析SqlSessionFacotry干的核心工作编写核心类SqlSession 深度分析解析SqlSession干的核心工作总结自定义mybatis用的技术点

一. 回顾mybatis的操作的核心步骤

声明一点我们本篇主要探讨的是mybatis的注解方式的操作, 完全从头开始都是小编从头开搞的, 如果与其他大神的代码思维有出入请多指教

我们首先需要准备mybatis的核心配置文件(当然导入相关的坐标这里不在啰嗦)

准备好结果的实体类以及在mApper接口上编写需要执行的sql语句

public class User { private Integer uid; private String username; private String password; private String nickname; } package cn.itcast.mapper; import cn.itcast.pojo.User; import org.Apache.ibatis.annotations.Select; import JAVA.util.List; public interface UserMapper { @Select("select * from users") List findAll(); }

使用mybatis的api来帮助我们完成sql语句的执行以及结果集的封装

//1.关联主配置文件 InputStream in = Resources.getResourceAsStream("mybatis-config.xml"); //2.解析配置文件 SqlSessionFactoryBuilder builder = new SqlSessionFactoryBuilder(); SqlSessionFactory sqlSessionFactory = builder.build(in); //3.创建会话对象 SqlSession sqlSession = sqlSessionFactory.openSession(); //4.可以采用接口代理的方式 UserMapper mapper = sqlSession.getMapper(UserMapper.class); List all = mapper.findAll(); System.out.println(all); //5.释放资源 sqlSession.close();

思考: mybatis大致是如何帮我们完成相关操作的 ?

我们通过Resources的getResourceAsStream告诉了mybatis我们编写的核心配置文件的位置, mybatis就可以找到我们数据库的连接信息, 也同时找到我们编写的sql语句的地方, 然后可以将其解析按照某种规则存放起来, 我们通过调用接口代理的方式执行方法时, 可以找到对应方法上的sql语句然后执行将结果封装返回给我们

二. 编写核心类SqlSessionFacotryBuild进行解析配置文件

那么我们废话不多说开始我们自定义mybatis的旅程,

1.首先我们需要用户编写配置文件, 然后通过我们自己的Resources来告诉我们配置文件所在位置 package com.itheima.ibatis.configuration; import java.io.InputStream; public class Resources { public static InputStream getResourceAsStream(String path) { return ClassLoader.getSystemClassLoader().getResourceAsStream(path); } } 2. 然后需要定义SqlSessionFacotryBuild来对配置文件进行解析分发 package com.itheima.ibatis.configuration; import com.itheima.ibatis.core.session.SqlSessionFactory; import com.itheima.ibatis.core.session.impl.DefaultSqlSessionFactory; import org.dom4j.Document; import org.dom4j.DocumentException; import org.dom4j.Element; import org.dom4j.Node; import org.dom4j.io.SAXReader; import javax.sql.DataSource; import java.io.File; import java.io.InputStream; import java.lang.reflect.Method; import java.util.List; import java.util.Properties; public class SqlSessionFactoryBuilder { private Configuration configuration = new Configuration(); public SqlSessionFactory build(InputStream in) { SAXReader saxReader = new SAXReader(); Document document = null; try { document = saxReader.read(in); } catch (DocumentException e) { e.printStackTrace(); } Element rootElement = document.getRootElement(); parseEnvironment(rootElement.element("environments")); parseMapper(rootElement.element("mappers")); return new DefaultSqlSessionFactory(configuration); } private void parseMapper(Element mapper) { String pack = mapper.element("package").attributeValue("name"); String directory = pack.replace(".", "/"); String path = ClassLoader.getSystemClassLoader().getResource("").getPath(); File mapperDir = new File(path, directory); if (!mapperDir.exists()) { throw new RuntimeException("找不到mapper映射"); } findMapper(mapperDir, pack); // System.out.println(configuration.getSql()); } private void findMapper(File mapperDir, String base) { File[] files = mapperDir.listFiles(); if (files != null) { for (File file : files) { if (file.isFile()) { if (file.getName().endsWith(".class")) { String name = file.getName(); name = name.substring(0, name.lastIndexOf(".")); String className = base + "." + name; initMapper(className); } } else { findMapper(file, base + "." + file.getName()); } } } } private void initMapper(String className) { try { Class clazz = Class.forName(className); Method[] methods = clazz.getMethods(); for (Method method : methods) { if(method.getAnnotations().length>0){ Mapper mapper = ParseMapper.parse(method); this.configuration.getMappers().put(className + "." + method.getName(), mapper); } } } catch (Exception e) { e.printStackTrace(); } } private void parseEnvironment(Element environments) { String defEnv = environments.attributeValue("default"); Node node = environments.selectSingleNode("//environment[@id='" + defEnv + "']"); List list = node.selectNodes("//property"); Properties properties = new Properties(); for (Element element : list) { String name = element.attributeValue("name"); String value = element.attributeValue("value"); properties.put(name, value); } DataSource dataSource = new DefaultDataSource().getDataSource(properties); configuration.setDataSource(dataSource); } } 三. 深度分析解析SqlSessionFacotryBuild干的核心工作 1. build(InputStream in) 方法做的工作

①借助Dom4j的来解析了xml文件, 将environments解析工作分发给了parseEnvironment(Element environments)

②将mappers的解析工作分发给了parseMapper(Element mapper)

2. parseEnvironment(Element environments)方法做的工作

①主要解析了连接数据库的参数们, 并且创建了数据库连接池

自定义连接池非本章节的重点,所以这里内部本质采用的Druid连接池来做了简化

package com.itheima.ibatis.configuration; import com.alibaba.druid.pool.DruidDataSourceFactory; import javax.sql.DataSource; import java.util.Properties; public class DefaultDataSource { public DataSource getDataSource(Properties properties) { try { return DruidDataSourceFactory.createDataSource(properties); } catch (Exception e) { e.printStackTrace(); } return null; } }

②将解析好的连接池放入configuration对象中,mappers成员变量先别纠结下一章节会讲解

package com.itheima.ibatis.configuration; import lombok.Data; import javax.sql.DataSource; import java.util.HashMap; import java.util.Map; @Data public class Configuration { private Map mappers = new HashMap<>(); private DataSource dataSource; }

详细图解如下图

3.parseMapper(Element mapper) 方法做的工作

①解析出用户配置的package找到sql语句所在接口的文件夹, 交给initMapper来处理

②递归找到这个包下所有的.class文件,并且获取到接口的全类名, 然后交给initMapper来处理

③initMapper通过反射获取类中的每一个方法,将方法交给一个专门解析方法上的注解的工具类ParseMapper的parse方法处理,处理完后将其放到configuration中的mappers的集合中

④ParseMapper的parse方法做的工作, 这是解析配置的核心地方

package com.itheima.ibatis.configuration; import java.lang.annotation.Annotation; import java.lang.reflect.InvocationTargetException; import java.lang.reflect.Method; import java.lang.reflect.ParameterizedType; import java.lang.reflect.Type; import java.util.ArrayList; import java.util.List; import java.util.regex.Matcher; import java.util.regex.Pattern; public class ParseMapper { public static Mapper parse(Method method) throws NoSuchMethodException, InvocationTargetException, IllegalAccessException { Annotation[] annotations = method.getAnnotations(); Object value = annotations[0].getClass().getMethod("value").invoke(annotations[0]); Mapper mapper = new Mapper(); Class resultType = method.getReturnType(); String val = (String) value; Pattern pattern = Pattern.compile("\#\{\s*\w+\s*\}"); Matcher matcher = pattern.matcher(val); List paramNames = new ArrayList<>(); while (matcher.find()) { String group = matcher.group(); String fieldName = group.substring(2, group.length() - 1).trim(); paramNames.add(fieldName); } String sql = val.replaceAll("\#\{\s*\w+\s*\}", "?"); mapper.setSql(sql); mapper.setParameterNames(paramNames); mapper.setSql(sql); if (resultType == List.class) { mapper.setSelectList(true); Type genericReturnType = method.getGenericReturnType(); ParameterizedType parameterizedType = (ParameterizedType) genericReturnType; Type actualTypeArgument = parameterizedType.getActualTypeArguments()[0]; mapper.setResultType(actualTypeArgument.getTypeName()); mapper.setType("SELECT"); } else if (resultType == Integer.class || resultType == int.class) { mapper.setType("UPDATE"); } else { mapper.setType("SELECT"); mapper.setResultType(resultType.getName()); } return mapper; } }

首先拿到方法上的注解,得到用户填入的sql语句

然后处理sql语句#{参数}的这些数据, 然后将参数的顺序保存起来, 用来后期设置参数的数据做准备, 一个

方法对应一个Mapper对象

然后再根据结果类型, 判断是什么类型相关的操作,方便后期执行对应的sql语句

四. 编写核心类SqlSessionFacotry 1.回顾那个地方创建的SqlSessionFacotry对象

经过SqlSessionFacotryBuilder的努力, 我们成功的将配置文件中核心的信息解析出来并放入了configuration对象中了, 然后我们此时将解析好的configuration传入到SqlSessionFacotry中

SqlSessionFactory的实现类如下:

public class DefaultSqlSessionFactory implements SqlSessionFactory { private final Configuration configuration; private TransactionManagement defaultTransactionManagement; public DefaultSqlSessionFactory(Configuration configuration) { this.configuration =configuration; defaultTransactionManagement = new DefaultTransactionManagement(configuration.getDataSource()); } @Override public SqlSession openSession() { return new DefaultSqlSession(configuration,defaultTransactionManagement,false); } } 2.添加事务管理器

事务管理是一个小的功能, 里面希望使用ThreadLocal集合来保证一个用户拿到的链接是同一个

事务管理的代码如下:

public class DefaultTransactionManagement implements TransactionManagement { private ThreadLocal threadLocal = new ThreadLocal<>(); private DataSource dataSource; public DefaultTransactionManagement(DataSource dataSource) { this.dataSource = dataSource; } public Connection getConnection() { Connection connection = threadLocal.get(); if (connection == null) { try { connection = dataSource.getConnection(); } catch (SQLException e) { e.printStackTrace(); } threadLocal.set(connection); } return connection; } @Override public void commit() { Connection connection = threadLocal.get(); if (connection != null ) { try { connection.commit(); } catch (Exception e) { e.printStackTrace(); } } } @Override public void rollback() { Connection connection = threadLocal.get(); if (connection != null) { try { connection.rollback(); } catch (SQLException e) { e.printStackTrace(); } } } public void close() { Connection connection = threadLocal.get(); if (connection != null) { try { connection.close(); threadLocal.remove(); } catch (SQLException e) { e.printStackTrace(); } } } @Override public void begin() { Connection connection = threadLocal.get(); if (connection != null) { try { connection.setAutoCommit(false); } catch (SQLException e) { e.printStackTrace(); } } } } 五. 深度分析解析SqlSessionFacotry干的核心工作 1. SqlSession openSession() 方法做的工作

可以看的出来我们在这个方法创建了DefaultSqlSession对象,并传入封装好的configuration,默认的事务管理器

默认通过openSession事务是开启的等等相关的参数

六.编写核心类SqlSession

其实有SqlSession的接口,我们使用的实现类是DefaultSession, 这里记录了解析的配置对象configuration

默认事务管理器对象transactionManagement, 默认事务开启的状态tx标记

package com.itheima.ibatis.core.session.impl; import com.itheima.ibatis.configuration.Configuration; import com.itheima.ibatis.configuration.Mapper; import com.itheima.ibatis.core.BaseExecutor; import com.itheima.ibatis.core.annotation.Param; import com.itheima.ibatis.core.session.SqlSession; import com.itheima.ibatis.core.transaction.TransactionManagement; import java.lang.reflect.*; import java.util.HashMap; import java.util.List; import java.util.Map; public class DefaultSqlSession implements SqlSession { private final Configuration configuration; private final boolean tx; private TransactionManagement transactionManagement; public DefaultSqlSession(Configuration configuration, TransactionManagement transactionManagement, boolean tx) { this.configuration = configuration; this.transactionManagement = transactionManagement; this.tx = tx; } public void close() { transactionManagement.close(); } @Override public void commit() { transactionManagement.commit(); } @Override public void rollback() { transactionManagement.rollback(); } @Override public List selectList(String sqlId) { return selectList(sqlId, null); } @Override public List selectList(String sqlId, Object param) { List list = new BaseExecutor(transactionManagement, tx).queryList(getMapper(sqlId), param); return (List) list; } @Override public T selectOne(String sqlId) { return selectOne(sqlId, null); } @Override public T selectOne(String sqlId, Object param) { return new BaseExecutor(transactionManagement, tx).query(getMapper(sqlId), param); } @Override public int delete(String sqlId) { return update0(sqlId, null); } @Override public int delete(String sqlId, Object param) { return update0(sqlId, param); } @Override public int update(String sqlId) { return update0(sqlId, null); } @Override public int update(String sqlId, Object param) { return update0(sqlId, param); } @Override public int insert(String sqlId) { return update0(sqlId, null); } @Override public int insert(String sqlId, Object param) { return update0(sqlId, param); } @Override public T getMapper(Class clazz) { Object o = Proxy.newProxyInstance( clazz.getClassLoader(), new Class[]{clazz}, new InvocationHandler() { @Override public Object invoke(Object proxy, Method method, Object[] args) throws Throwable { String sqlId = clazz.getName() + "." + method.getName(); Mapper mapper = configuration.getMappers().get(sqlId); String type = mapper.getType(); Object findParam = null; if (args != null) { if (args.length == 1) { Object param = args[0]; boolean isArray = param.getClass().isArray(); if (!isArray) { findParam = param; } } else { Map map = new HashMap<>(); Parameter[] parameters = method.getParameters(); for (int i = 0; i < parameters.length; i++) { Param param = parameters[i].getAnnotation(Param.class); String key = "arg"+i; if(param !=null){ key = param.value(); } map.put(key, args[i]); } findParam = map; } } if (type.equals("SELECT")) { boolean selectList = mapper.isSelectList(); if (selectList) return selectList(sqlId, findParam); else return selectOne(sqlId, findParam); } else { return update0(sqlId, findParam); } } }); return (T) o; } private int update0(String sqlId, Object param) { return new BaseExecutor(transactionManagement, tx).update(getMapper(sqlId), param); } public Mapper getMapper(String sqlId) { Mapper mapper = configuration.getMappers().get(sqlId); if (mapper == null) { throw new RuntimeException("没有找到sql映射,请检查"); } return mapper; } }七.深度分析解析SqlSession干的核心工作1.selectOne & selectList做的工作

主要是分发了下功能, 执行sql语句避免不了有参数和无参数的, 都让调用有参数的方便管理