（七）MyBatis从入门到入土——花式查询

时间 2021-02-17

标签 java spring sql 编程 mybatis app 性能测试 this 栏目 MyBatis 繁體版

原文原文链接

这是mybatis系列第7篇。没看前文的建议先去【Java冢狐】公众号中查看前文，方便理解和掌握。上一篇中咱们讲述了一些关于获取主键的方法，这篇文章中咱们将会更加深刻的介绍一些查询的方法，特别是多表查询。java

在开始以前先开始诸如建库建表等准备工做。spring

建库建表

建立一个db：mybatisdemosql

4张表:编程

user(用户表)
goods(商品表)
orders(订单表)
order_detail(订单明细表)

表之间的关系：mybatis

orders和user是一对一的关系，一条订单关联一个用户记录
orders和order_detail是一对多关系，每一个订单中可能包含多个子订单，每一个子订单对应一个商品

具体的建表语句以下所示：app

DROP DATABASE IF EXISTS `mybatisdemo`;
CREATE DATABASE `mybatisdemo`;
USE `mybatisdemo`;
DROP TABLE IF EXISTS user;
CREATE TABLE user(
  id int AUTO_INCREMENT PRIMARY KEY COMMENT '用户id',
  name VARCHAR(32) NOT NULL DEFAULT '' COMMENT '用户名'
) COMMENT '用户表';
INSERT INTO user VALUES (1,'冢狐'),(2,'Java冢狐');
DROP TABLE IF EXISTS goods;
CREATE TABLE goods(
  id int AUTO_INCREMENT PRIMARY KEY COMMENT '商品id',
  name VARCHAR(32) NOT NULL DEFAULT '' COMMENT '商品名称',
  price DECIMAL(10,2) NOT NULL DEFAULT 0 COMMENT '商品价格'
) COMMENT '商品信息表';
INSERT INTO goods VALUES (1,'Mybatis系列',8.00),(2,'spring系列',16.00);
DROP TABLE IF EXISTS orders;
CREATE TABLE orders(
  id int AUTO_INCREMENT PRIMARY KEY COMMENT '订单id',
  user_id INT NOT NULL DEFAULT 0 COMMENT '用户id，来源于user.id',
  create_time BIGINT NOT NULL DEFAULT 0 COMMENT '订单建立时间(时间戳，秒)',
  up_time BIGINT NOT NULL DEFAULT 0 COMMENT '订单最后修改时间(时间戳，秒)'
) COMMENT '订单表';
INSERT INTO orders VALUES (1,2,unix_timestamp(now()),unix_timestamp(now())),(2,1,unix_timestamp(now()),unix_timestamp(now()));
DROP TABLE IF EXISTS order_detail;
CREATE TABLE order_detail(
  id int AUTO_INCREMENT PRIMARY KEY COMMENT '订单明细id',
  order_id INT NOT NULL DEFAULT 0 COMMENT '订单id，来源于order.id',
  goods_id INT NOT NULL DEFAULT 0 COMMENT '商品id，来源于goods.id',
  num INT NOT NULL DEFAULT 0 COMMENT '商品数量',
  total_price DECIMAL(12,2) NOT NULL DEFAULT 0 COMMENT '商品总金额'
) COMMENT '订单表';
INSERT INTO order_detail VALUES (1,1,1,2,16.00),(2,1,1,1,16.00),(3,2,1,1,8.00);
select * from user;
select * from goods;
select * from orders;
select * from order_detail;

建好库和表后，下一步咱们就开始咱们的查询之旅，先从最基础的单表查询看起，依次介绍单表查询、一对一查询和一对多查询。性能

单表查询(3种方式)

先来介绍的就是单表查询。测试

需求

须要按照订单id查询订单信息。ui

方式1

建立每一个表对应的Model

db中表的字段是采用下划线分割的，model中咱们是采用骆驼命名法来命名的，如OrderModel:this

package com.zhonghu.chat07.demo1.model;
import lombok.*;
import java.util.List;
@Getter
@Setter
@Builder
@ToString
@NoArgsConstructor
@AllArgsConstructor
public class OrderModel {
    private Integer id;
    private Integer userId;
    private Long createTime;
    private Long upTime;
}

其余几个Model也相似。

Mapper xml

<select id="getById" resultType="com.zhonghu.chat07.demo1.model.OrderModel">
    <![CDATA[
    SELECT a.id,a.user_id as userId,a.create_time createTime,a.up_time upTime FROM orders a WHERE a.id = #{value}
    ]]>
</select>

注意上面的resultType，标识结果的类型。

Mapper接口方法

OrderModel getById(int id);

mybatis全局配置文件

<?xml version="1.0" encoding="UTF-8" ?>
<!DOCTYPE configuration
        PUBLIC "-//mybatis.org//DTD Config 3.0//EN"
        "http://mybatis.org/dtd/mybatis-3-config.dtd">
<configuration>
    <!-- 引入外部jdbc配置 -->
    <properties resource="jdbc.properties"/>
    <!-- 环境配置，能够配置多个环境 -->
    <environments default="demo4">
        <environment id="demo4">
            <!-- 事务管理器工厂配置 -->
            <transactionManager type="JDBC"/>
            <!-- 数据源工厂配置，使用工厂来建立数据源 -->
            <dataSource type="POOLED">
                <property name="driver" value="${jdbc.driver}"/>
                <property name="url" value="${jdbc.url}"/>
                <property name="username" value="${jdbc.username}"/>
                <property name="password" value="${jdbc.password}"/>
            </dataSource>
        </environment>
    </environments>
    <mappers>
        <mapper resource="demo1/mapper/UserMapper.xml" />
        <mapper resource="demo1/mapper/GoodsMapper.xml" />
        <mapper resource="demo1/mapper/OrderMapper.xml" />
        <mapper resource="demo1/mapper/OrderDetailMapper.xml" />
    </mappers>
</configuration>

测试用例

com.zhonghu.chat07.demo1.Demo1Test#getById
@Before
public void before() throws IOException {
    //指定mybatis全局配置文件
    String resource = "demo1/mybatis-config.xml";
    //读取全局配置文件
    InputStream inputStream = Resources.getResourceAsStream(resource);
    //构建SqlSessionFactory对象
    SqlSessionFactory sqlSessionFactory = new SqlSessionFactoryBuilder().build(inputStream);
    this.sqlSessionFactory = sqlSessionFactory;
}
@Test
public void getById() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById(1);
        log.info("{}", orderModel);
    }
}

运行输出

35:59.211 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById - ==>  Preparing: SELECT a.id,a.user_id as userId,a.create_time createTime,a.up_time upTime FROM orders a WHERE a.id = ? 
35:59.239 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById - ==> Parameters: 1(Integer)
35:59.258 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById - <==      Total: 1
35:59.258 [main] INFO  c.j.chat05.demo1.Demo1Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573)

原理

sql中咱们使用了别名，将orders中的字段转换成了和OrderModel中字段同样的名称，最后mybatis内部会经过反射，将查询结果按照名称到OrderModel中查找同名的字段，而后进行赋值。

方式2

若咱们项目中表对应的Model中的字段都是采用骆驼命名法，mybatis中能够进行一些配置，可使表中的字段和对应Model中骆驼命名法的字段进行自动映射。

须要在mybatis全局配置文件中加入下面配置：

<settings>
    <!-- 是否开启自动驼峰命名规则映射，及从xx_yy映射到xxYy -->
    <setting name="mapUnderscoreToCamelCase" value="true"/>
</settings>

Mapper xml

<select id="getById1" resultType="com.zhonghu.chat07.demo1.model.OrderModel">
    <![CDATA[
    SELECT a.id,a.user_id,a.create_time,a.up_time FROM orders a WHERE a.id = #{value}
    ]]>
</select>

注意上面的sql，咱们没有写别名了，因为咱们开启了自动骆驼命名映射，因此查询结果会按照下面的关系进行自动映射：

sql对应的字段	OrderModel中的字段
id	id
user_id	userId
create_time	createTime
up_time	upTime

Mapper接口

OrderModel getById1(int id);

测试用例

com.zhonghu.chat05.demo1.Demo1Test#getById1
@Test
public void getById1() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById1(1);
        log.info("{}", orderModel);
    }
}

运行输出

59:44.884 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - ==>  Preparing: SELECT a.id,a.user_id,a.create_time,a.up_time FROM orders a WHERE a.id = ? 
59:44.917 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - ==> Parameters: 1(Integer)
59:44.935 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - <==      Total: 1
59:44.935 [main] INFO  c.j.chat05.demo1.Demo1Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573)

输出中能够看出，sql中的字段是下划线的方式，OrderModel中的字段是骆驼命名法，结果也自动装配成功，这个就是开启mapUnderscoreToCamelCase产生的效果。

方式3

mapper xml中有个更强大的元素resultMap，经过这个元素能够定义查询结果的映射关系。

Mapper xml

<resultMap id="orderModelMap2" type="com.zhonghu.chat07.demo1.model.OrderModel">
    <id column="id" property="id" />
    <result column="user_id" property="userId" />
    <result column="create_time" property="createTime" />
    <result column="up_time" property="upTime" />
</resultMap>
<select id="getById2" resultMap="orderModelMap2">
    <![CDATA[
    SELECT a.id,a.user_id,a.create_time,a.up_time FROM orders a WHERE a.id = #{value}
    ]]>
</select>

上面resultMap有2个元素须要指定：

id：resultMap标识
type：将结果封装成什么类型，此处咱们须要将结果分装为OrderModel

注意上面的select元素，有个resultMap，标识查询结果使用哪一个resultMap进行映射，此处咱们使用的是orderModelMap2，因此查询结果会按照orderModelMap2关联的resultMap进行映射。

Mapper接口

OrderModel getById2(int id);

测试用例

com.zhonghu.chat07.demo1.Demo1Test#getById2
@Test
public void getById2() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById2(1);
        log.info("{}", orderModel);
    }
}

运行输出

14:12.518 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - ==>  Preparing: SELECT a.id,a.user_id,a.create_time,a.up_time FROM orders a WHERE a.id = ? 
14:12.546 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - ==> Parameters: 1(Integer)
14:12.564 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - <==      Total: 1
14:12.564 [main] INFO  c.j.chat05.demo1.Demo1Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573)

一对一关联查询(4种方式)

讲完了单表查询，下面开始连表查询，首先就是一对一的查询

需求

经过订单id查询订单的时候，将订单关联的用户信息也返回。

咱们修改一下OrderModel代码，内部添加一个UserModel，以下：

package com.zhonghu.chat07.demo2.model;
import lombok.*;
import java.util.List;
@Getter
@Setter
@Builder
@ToString
@NoArgsConstructor
@AllArgsConstructor
public class OrderModel {
    private Integer id;
    private Integer userId;
    private Long createTime;
    private Long upTime;
    //下单用户信息
    private UserModel userModel;
}

UserModel内容：

package com.zhonghu.chat07.demo2.model;
import lombok.*;
@Getter
@Setter
@Builder
@ToString
@NoArgsConstructor
@AllArgsConstructor
public class UserModel {
    private Integer id;
    private String name;
}

方式1

Mapper xml

<resultMap id="orderModelMap1" type="com.zhonghu.chat07.demo2.model.OrderModel">
    <id column="id" property="id" />
    <result column="user_id" property="userId"/>
    <result column="create_time" property="createTime"/>
    <result column="up_time" property="upTime"/>
    <result column="user_id" property="userModel.id"/>
    <result column="name" property="userModel.name"/>
</resultMap>
<select id="getById1" resultMap="orderModelMap1">
    <![CDATA[
    SELECT
        a.id,
        a.user_id,
        a.create_time,
        a.up_time,
        b.name
    FROM
        orders a,
        user b
    WHERE
        a.user_id = b.id
    AND a.id = #{value}
    ]]>
</select>

注意重点在于上面的这两行：

<result column="user_id" property="userModel.id"/>
<result column="name" property="userModel.name"/>

这个地方使用到了级联赋值，多级之间用.进行引用，此处咱们只有一级，能够有不少级。

Mapper 接口

OrderModel getById1(int id);

测试用例

com.zhonghu.chat07.demo2.Demo2Test#getById1
@Before
public void before() throws IOException {
    //指定mybatis全局配置文件
    String resource = "demo2/mybatis-config.xml";
    //读取全局配置文件
    InputStream inputStream = Resources.getResourceAsStream(resource);
    //构建SqlSessionFactory对象
    SqlSessionFactory sqlSessionFactory = new SqlSessionFactoryBuilder().build(inputStream);
    this.sqlSessionFactory = sqlSessionFactory;
}
@Test
public void getById1() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById1(1);
        log.info("{}", orderModel);
    }
}

运行输出

24:20.811 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - ==>  Preparing: SELECT a.id, a.user_id, a.create_time, a.up_time, b.name FROM orders a, user b WHERE a.user_id = b.id AND a.id = ? 
24:20.843 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - ==> Parameters: 1(Integer)
24:20.861 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - <==      Total: 1
24:20.861 [main] INFO  c.j.chat05.demo2.Demo2Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573, userModel=UserModel(id=2, name=Java冢狐))

方式2

此次咱们须要使用mapper xml中另一个元素association，这个元素能够配置关联对象的映射关系，看示例。

Mapper xml

<resultMap id="orderModelMap2" type="com.zhonghu.chat07.demo2.model.OrderModel">
    <id column="id" property="id" />
    <result column="user_id" property="userId"/>
    <result column="create_time" property="createTime"/>
    <result column="up_time" property="upTime"/>
    <association property="userModel">
        <id column="user_id" property="id"/>
        <result column="name" property="name" />
    </association>
</resultMap>
<select id="getById2" resultMap="orderModelMap2">
    <![CDATA[
    SELECT
        a.id,
        a.user_id,
        a.create_time,
        a.up_time,
        b.name
    FROM
        orders a,
        user b
    WHERE
        a.user_id = b.id
    AND a.id = #{value}
    ]]>
</select>

注意上面下面这部分代码：

<association property="userModel">
    <id column="user_id" property="id"/>
    <result column="name" property="name" />
</association>

注意上面的property属性，这个就是配置sql查询结果和OrderModel.userModel对象的映射关系，将user_id和userModel中的id进行映射,name和userModel中的name进行映射。

Mapper接口

OrderModel getById2(int id);

测试用例

@Test
public void getById2() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById2(1);
        log.info("{}", orderModel);
    }
}

运行结果

51:44.896 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - ==>  Preparing: SELECT a.id, a.user_id, a.create_time, a.up_time, b.name FROM orders a, user b WHERE a.user_id = b.id AND a.id = ? 
51:44.925 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - ==> Parameters: 1(Integer)
51:44.941 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - <==      Total: 1
51:44.942 [main] INFO  c.j.chat05.demo2.Demo2Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573, userModel=UserModel(id=2, name=Java冢狐))

从结果的最后一行能够看出，全部字段的值映射都是ok的。

方式3

先按照订单id查询订单数据，而后在经过订单中user_id去用户表查询用户数据，经过两次查询，组合成目标结果，mybatis已经内置了这种操做，以下。

UserMapper.xml

咱们先定义一个经过用户id查询用户信息的select元素，以下

<select id="getById" resultType="com.zhonghu.chat07.demo2.model.UserModel">
    <![CDATA[
    SELECT id,name FROM user where id = #{value}
    ]]>
</select>

OrderModel.xml

<resultMap id="orderModelMap3" type="com.zhonghu.chat07.demo2.model.OrderModel">
    <id column="id" property="id" />
    <result column="user_id" property="userId"/>
    <result column="create_time" property="createTime"/>
    <result column="up_time" property="upTime"/>
    <association property="userModel" select="com.zhonghu.chat07.demo2.mapper.UserMapper.getById" column="user_id" />
</resultMap>
<select id="getById3" resultMap="orderModelMap3">
    <![CDATA[
    SELECT
        a.id,
        a.user_id,
        a.create_time,
        a.up_time
    FROM
        orders a
    WHERE
        a.id = #{value}
    ]]>
</select>

OrderModel.userModel属性的值来在于另一个查询，这个查询是经过association元素的select属性指定的，此处使用的是

com.zhonghu.chat07.demo2.mapper.UserMapper.getById

这个查询是有条件的，条件经过association的column进行传递的，此处传递的是getById3查询结果中的user_id字段。

Mapper接口

OrderModel getById3(int id);

测试用例

com.zhonghu.chat07.demo2.Demo2Test#getById3
@Test
public void getById3() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById3(1);
        log.info("{}", orderModel);
    }
}

运行输出

07:12.569 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById3 - ==>  Preparing: SELECT a.id, a.user_id, a.create_time, a.up_time FROM orders a WHERE a.id = ? 
07:12.600 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById3 - ==> Parameters: 1(Integer)
07:12.619 [main] DEBUG c.j.c.d.mapper.UserMapper.getById - ====>  Preparing: SELECT id,name FROM user where id = ? 
07:12.620 [main] DEBUG c.j.c.d.mapper.UserMapper.getById - ====> Parameters: 2(Integer)
07:12.625 [main] DEBUG c.j.c.d.mapper.UserMapper.getById - <====      Total: 1
07:12.625 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById3 - <==      Total: 1
07:12.625 [main] INFO  c.j.chat05.demo2.Demo2Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573, userModel=UserModel(id=2, name=Java冢狐))

从输出中能够看出有2次查询，先按照订单id查询订单，而后经过订单记录中用户id去用户表查询用户信息，最终执行了2次查询。

方式4

方式3中给第二个查询传递了一个参数，若是须要给第二个查询传递多个参数怎么办呢？能够这么写

<association property="属性" select="查询对应的select的id" column="{key1=父查询字段1,key2=父查询字段2,key3=父查询字段3}" />

这种至关于给子查询传递了一个map，子查询中须要用过map的key获取对应的条件，看案例：

OrderMapper.xml

<resultMap id="orderModelMap4" type="com.zhonghu.chat07.demo2.model.OrderModel">
    <id column="id" property="id" />
    <result column="user_id" property="userId"/>
    <result column="create_time" property="createTime"/>
    <result column="up_time" property="upTime"/>
    <association property="userModel" select="com.zhonghu.chat07.demo2.mapper.UserMapper.getById1" column="{uid1=user_id,uid2=create_time}" />
</resultMap>
<select id="getById4" resultMap="orderModelMap4">
    <![CDATA[
    SELECT
        a.id,
        a.user_id,
        a.create_time,
        a.up_time
    FROM
        orders a
    WHERE
        a.id = #{value}
    ]]>
</select>

UserMapper.xml

<select id="getById1" resultType="com.zhonghu.chat07.demo2.model.UserModel">
    <![CDATA[
    SELECT id,name FROM user where id = #{uid1} and id = #{uid2}
    ]]>
</select>

Mapper接口

OrderModel getById4(int id);

测试用例

com.zhonghu.chat07.demo2.Demo2Test#getById4
@Test
public void getById4() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById4(1);
        log.info("{}", orderModel);
    }
}

运行输出

19:59.881 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById4 - ==>  Preparing: SELECT a.id, a.user_id, a.create_time, a.up_time FROM orders a WHERE a.id = ? 
19:59.914 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById4 - ==> Parameters: 1(Integer)
19:59.934 [main] DEBUG c.j.c.d.mapper.UserMapper.getById1 - ====>  Preparing: SELECT id,name FROM user where id = ? and id = ? 
19:59.934 [main] DEBUG c.j.c.d.mapper.UserMapper.getById1 - ====> Parameters: 2(Integer), 1610803573(Long)
19:59.939 [main] DEBUG c.j.c.d.mapper.UserMapper.getById1 - <====      Total: 0
19:59.939 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById4 - <==      Total: 1
19:59.939 [main] INFO  c.j.chat05.demo2.Demo2Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573, userModel=null)

输出中看一下第二个查询的条件，传过来的是第一个查询的user_id和create_time。

一对多查询(2种方式)

最后就是进行一队多查询的解析

需求

根据订单id查询出订单信息，而且查询出订单明细列表。

先修改一下OrderModel代码，以下：

package com.zhonghu.chat07.demo3.model;
import lombok.*;
import java.util.List;
@Getter
@Setter
@Builder
@ToString
@NoArgsConstructor
@AllArgsConstructor
public class OrderModel {
    private Integer id;
    private Integer userId;
    private Long createTime;
    private Long upTime;
    //订单详情列表
    private List<OrderDetailModel> orderDetailModelList;
}

OrderModel中添加了一个集合orderDetailModelList用来存放订单详情列表。

方式1

OrderMapper.xml

<resultMap id="orderModelMap1" type="com.zhonghu.chat07.demo3.model.OrderModel">
    <id column="id" property="id"/>
    <result column="user_id" property="userId"/>
    <result column="create_time" property="createTime"/>
    <result column="up_time" property="upTime"/>
    <collection property="orderDetailModelList" ofType="com.zhonghu.chat07.demo3.model.OrderDetailModel">
        <id column="orderDetailId" property="id"/>
        <result column="order_id" property="orderId"/>
        <result column="goods_id" property="goodsId"/>
        <result column="num" property="num"/>
        <result column="total_price" property="totalPrice"/>
    </collection>
</resultMap>
<select id="getById1" resultMap="orderModelMap1">
    <![CDATA[
    SELECT
        a.id ,
        a.user_id,
        a.create_time,
        a.up_time,
        b.id orderDetailId,
        b.order_id,
        b.goods_id,
        b.num,
        b.total_price
    FROM
        orders a,
        order_detail b
    WHERE
        a.id = b.order_id
        AND a.id = #{value}
    ]]>
</select>

注意上面的getById1中的sql，这个sql中使用到了t_order和t_order_detail链接查询，这个查询会返回多条结果，可是最后结果按照orderModelMap1进行映射，最后只会返回一个OrderModel对象，关键在于collection元素，这个元素用来定义集合中元素的映射关系，有2个属性须要注意：

property：对应的属性名称
ofType：集合中元素的类型，此处是OrderDetailModel

原理是这样的，注意orderModelMap1中有个

<id column="id" property="id"/>

查询出来的结果会按照这个配置中指定的column进行分组，即按照订单id进行分组，每一个订单对应多个订单明细，订单明细会按照collection的配置映射为ofType元素指定的对象。

实际resultMap元素中的id元素可使用result元素代替，只是用id能够提高性能，mybatis能够经过id元素配置的列的值判断惟一一条记录，若是咱们使用result元素，那么判断是不是同一条记录的时候，须要经过全部列去判断了，因此经过id能够提高性能，使用id元素在一对多中能够提高性能，在单表查询中使用id元素仍是result元素，性能都是同样的。

Mapper接口

OrderModel getById1(Integer id);

测试用例

com.zhonghu.chat07.demo3.Demo3Test#getById1
@Before
public void before() throws IOException {
    //指定mybatis全局配置文件
    String resource = "demo3/mybatis-config.xml";
    //读取全局配置文件
    InputStream inputStream = Resources.getResourceAsStream(resource);
    //构建SqlSessionFactory对象
    SqlSessionFactory sqlSessionFactory = new SqlSessionFactoryBuilder().build(inputStream);
    this.sqlSessionFactory = sqlSessionFactory;
}
@Test
public void getById1() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        Integer id = 1;
        OrderModel orderModel = mapper.getById1(id);
        log.info("{}", orderModel);
    }
}

运行输出

03:52.092 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - ==>  Preparing: SELECT a.id , a.user_id, a.create_time, a.up_time, b.id orderDetailId, b.order_id, b.goods_id, b.num, b.total_price FROM orders a, order_detail b WHERE a.id = b.order_id AND a.id = ? 
03:52.124 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - ==> Parameters: 1(Integer)
03:52.148 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById1 - <==      Total: 2
03:52.148 [main] INFO  c.j.chat05.demo3.Demo3Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573, orderDetailModelList=[OrderDetailModel(id=1, orderId=1, goodsId=1, num=2, totalPrice=16.00), OrderDetailModel(id=2, orderId=1, goodsId=1, num=1, totalPrice=16.00)])

注意最后一条输出，和指望的结果一致。

方式2

经过2次查询，而后对结果进行分装，先经过订单id查询订单信息，而后经过订单id查询订单明细列表，而后封装结果。mybatis中默认支持这么玩，仍是经过collection元素来实现的。

OrderDetailMapper.xml

<select id="getListByOrderId1" resultType="com.zhonghu.chat07.demo3.model.OrderDetailModel" parameterType="int">
    <![CDATA[
    SELECT
        a.id,
        a.order_id AS orderId,
        a.goods_id AS goodsId,
        a.num,
        a.total_price AS totalPrice
    FROM
        order_detail a
    WHERE
        a.order_id = #{value}
    ]]>
</select>

OrderMapper.xml

<resultMap id="orderModelMap2" type="com.zhonghu.chat07.demo3.model.OrderModel">
    <id column="id" property="id"/>
    <result column="user_id" property="userId"/>
    <result column="create_time" property="createTime"/>
    <result column="up_time" property="upTime"/>
    <collection property="orderDetailModelList" select="com.zhonghu.chat07.demo3.mapper.OrderDetailMapper.getListByOrderId1" column="id"/>
</resultMap>
<select id="getById2" resultMap="orderModelMap2">
    <![CDATA[
    SELECT
        a.id ,
        a.user_id,
        a.create_time,
        a.up_time
    FROM
        orders a
    WHERE
        a.id = #{value}
    ]]>
</select>

重点在于下面这句配置：

<collection property="orderDetailModelList" select="com.zhonghu.chat07.demo3.mapper.OrderDetailMapper.getListByOrderId1" column="id"/>

表示orderDetailModelList属性的值经过select属性指定的查询获取，即：

com.zhonghu.chat07.demo3.mapper.OrderDetailMapper.getListByOrderId1

查询参数是经过column属性指定的，此处使用getById2 sql中的id做为条件，即订单id。

Mapper接口

OrderModel getById2(int id);

测试用例

com.zhonghu.chat07.demo3.Demo3Test#getById2
@Test
public void getById2() {
    try (SqlSession sqlSession = this.sqlSessionFactory.openSession(true);) {
        OrderMapper mapper = sqlSession.getMapper(OrderMapper.class);
        OrderModel orderModel = mapper.getById2(1);
        log.info("{}", orderModel);
    }
}

运行输出

10:07.087 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - ==>  Preparing: SELECT a.id , a.user_id, a.create_time, a.up_time FROM ordera a WHERE a.id = ? 
10:07.117 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - ==> Parameters: 1(Integer)
10:07.135 [main] DEBUG c.j.c.d.m.O.getListByOrderId1 - ====>  Preparing: SELECT a.id, a.order_id AS orderId, a.goods_id AS goodsId, a.num, a.total_price AS totalPrice FROM order_detail a WHERE a.order_id = ? 
10:07.136 [main] DEBUG c.j.c.d.m.O.getListByOrderId1 - ====> Parameters: 1(Integer)
10:07.141 [main] DEBUG c.j.c.d.m.O.getListByOrderId1 - <====      Total: 2
10:07.142 [main] DEBUG c.j.c.d.mapper.OrderMapper.getById2 - <==      Total: 1
10:07.142 [main] INFO  c.j.chat05.demo3.Demo3Test - OrderModel(id=1, userId=2, createTime=1610803573, upTime=1610803573, orderDetailModelList=[OrderDetailModel(id=1, orderId=1, goodsId=1, num=2, totalPrice=16.00), OrderDetailModel(id=2, orderId=1, goodsId=1, num=1, totalPrice=16.00)])

输出中有2次查询，先经过订单id查询订单信息，而后经过订单id查询订单明细，mybatis内部对结果进行了组装。

总结

mybatis全局配置文件中经过mapUnderscoreToCamelCase能够开启sql中的字段和javabean中的骆驼命名法的字段进行自动映射
掌握resultMap元素常见的用法
一对一关联查询使用resultMap->association元素（2种方式）
一对多查询使用resultMap->collection元素（2种方式）
resultMap中使用id元素主要在复杂的关联查询中能够提高效率，能够经过这个来判断记录的惟一性，若是没有这个，须要经过全部的result相关的列才能判断记录的惟一性

建议

mybatis为咱们提供了强大的关联查询，不过我的建议尽可能少用，最好是采用单表的方式查询，在程序中经过屡次查询，而后本身对结果进行组装。

Model中最好只定义一些和单表字段关联的属性，不要掺杂着其余对象的引用。

最后

若是以为看完有收获，但愿能关注一下，顺便给我点个赞，这将会是我更新的最大动力，感谢各位的支持
欢迎各位关注个人公众号【java冢狐】，专一于java和计算机基础知识，保证让你看完有所收获，不信你打我
求一键三连：点赞、转发、在看。
若是看完有不一样的意见或者建议，欢迎多多评论一块儿交流。感谢各位的支持以及厚爱。

——我是冢狐，和你同样热爱编程。

欢迎关注公众号“Java冢狐”获取最新消息