一目了然的MyBatis一对一、一对多查询

在使用SQL的时候，重点就在于多表联合查询，而MyBatis中重点也是多表联合查询，并且将表与表的关系体现了出来。而MyBatis中的关系分为一对一、一对多、多对多，这里我们先解释前两种。

先看表结构

根据图中所示，张三具有两个住址。

看对象

//Customer对象
@Data
@AllArgsConstructor
@NoArgsConstructor
public class Customer {
    
    
    private int id;
    private String userName;
    private String jobs;
    private String phone;
    private List<Address> address;
}

//Address对象
@Data
@AllArgsConstructor
@NoArgsConstructor
public class Address {
    
    
    private int id;
    private int customerId;
    private String address;
    //private Customer customer;
}

注解形式：t_customer为主表，t_address为副表
XML形式：t_address为主表 ，t_customer为副表

一、注解形式

1、注解实现一对一

public interface CustomerService {
    
    

    @Select("SELECT * FROM t_customer WHERE id = #{id}") //查询语句
    @Results(id = "customerMapper", value = {
    
     //用户表的字段映射
            @Result(column = "id",property = "id",javaType = int.class),
            @Result(column = "username",property = "userName",javaType = String.class),
            //实现用户类中地址对象的类型映射，通过@One来实现，column表示的是需要传给子查询的参数，为主查询的列名
            @Result(property = "address", column = "id", one = @One(select = "service.CustomerService.getAddressList"))
    })
    public Customer getCustomerList(Integer id);


    @Select("SELECT * FROM t_address WHERE c_id = #{c_id}") //子查询语句
    @Results(id = "addressMapper",value = {
    
     //地址表的字段映射
            @Result(column = "c_id", property = "customerId",javaType = int.class),
            @Result(column = "c_address",property = "address",javaType = String.class)
    })
    public Address getAddressList(Integer c_id);
 }

2、注解实现一对多

public interface CustomerService {
    
    

    @Select("SELECT * FROM t_customer WHERE id = #{id}") //查询语句
    @Results(id = "customerMapper", value = {
    
     //用户表的字段映射
            @Result(column = "id",property = "id",javaType = int.class),
            @Result(column = "username",property = "userName",javaType = String.class),
            //实现用户类中地址对象的类型映射，通过@One来实现，column表示的是需要传给子查询的参数，为主查询的列名
            @Result(property = "address", column = "id", many= @Many(select = "service.CustomerService.getAddressList"))
    })
    public Customer getCustomerList(Integer id);


    @Select("SELECT * FROM t_address WHERE c_id = #{c_id}") //子查询语句
    @Results(id = "addressMapper",value = {
    
     //地址表的字段映射
            @Result(column = "c_id", property = "customerId",javaType = int.class),
            @Result(column = "c_address",property = "address",javaType = String.class)
    })
    public List<Address> getAddressList(Integer c_id);
 }

3、注解实现的区别

二、XML形式

先将两个表查询出来的字段设置一个别名，方便后面的属性映射。

<!--用户表的字段设置别名-->
<sql id="customerTemplate">
    customer.id AS customerId, customer.username AS customerName,
    customer.jobs AS customerJobs, customer.phone AS customerPhone,
</sql>

<!--地址表的字段设置别名-->
<sql id="addressTemplate">
    address.id AS addressId, address.c_id AS addressCustomerId,
    address.c_address AS address
</sql>

1、实现一对一

查询标签

<select id="getAddress" parameterType="Integer" resultMap="addressMapper">
    SELECT
        <include refid="customerTemplate"/> <!--引入代码片段-->
        <include refid="addressTemplate"/>
    FROM t_address address INNER JOIN t_customer customer
    ON customer.id = address.c_id
    WHERE address.c_id = #{id};
</select>

结果映射标签

<resultMap id="addressMapper" type="address"> <!--先为address主表设置映射-->
    <id column="addressId" property="id"/>
    <result column="addressCustomerId" property="customerId"/>
    <result column="address" property="address"/>
    <!--再为t_customer副标设置映射，property为Address类中的属性名，javaType表示property的类型-->
    <association property="customer" javaType="customer">
        <id column="customerId" property="id"/>
        <result column="customerName" property="userName"/>
        <result column="customerJobs" property="jobs"/>
        <result column="customerPhone" property="phone"/>
    </association>
</resultMap>

2、实现一对多

查询标签（与一对一的查询一致）

 <select id="getAddress" parameterType="Integer" resultMap="addressMapper">
        SELECT
            <include refid="customerTemplate"/>
            <include refid="addressTemplate"/>
        FROM t_address address INNER JOIN t_customer customer
        ON customer.id = address.c_id
        WHERE address.c_id = #{id};
 </select>

结果集映射标签

<resultMap id="addressMapper" type="address">
    <id column="addressId" property="id"/>
    <result column="addressCustomerId" property="customerId"/>
    <result column="address" property="address"/>
    <!---进行映射-->
    <collection property="customer" ofType="customer">
        <id column="customerId" property="id"/>
        <result column="customerName" property="userName"/>
        <result column="customerJobs" property="jobs"/>
        <result column="customerPhone" property="phone"/>
    </collection>
</resultMap>

3、XML一对一和一对多的不同

<!--实现一对一-->
 <association property="customer" javaType="customer"></association>
 <!--
	property、javaType都是主类中副类的类型
-->

<!--实现一对多-->
 <collection property="customer" ofType="customer"></collection>
 <!--
	property、ofType都是主类中副类的类型
-->

三、注解形式和XML形式的区别

1、注解写法简单，XML形式写法复杂。
2、注解使用的是子查询的方式，XML形式使用的是多表联合查询的方式。
3、当数据量庞大的时候，注解形式执行的方式会慢与XML形式。

主要的区别就在写法和效率上，注解使用子查询，会执行两次SQL查询，而XML形式使用的是多表联合查询，一次查询会执行一次SQL，除了写法复杂一点其他的都会比使用注解的情况好很多。

我们可以使用Junit单元测试看得到情况：

1、注解的子查询情况：

2、XML实现多表联合查询情况：

所以使用XML完成多表联合查询实现一对一、一对多会比使用注解的效率高。