最佳方法：定制@NamedEntityGraph、定制查询和定制VO，可以做到按照需要最佳查询，需要注意的地方：定制VO的字段一定要等于或小于实际查询的字段，才不会复制的时候触发N+1查询。
1 问题复现
1.1 项目结构
1.2 entity
package com.xkzhangsan.jpa.entity;import lombok.Getter; import lombok.Setter;import javax.persistence.*;@Entity @Getter @Setter @Table(name = "user") public class User {@Id@GeneratedValue(strategy = GenerationType.AUTO)private Long id;private String name;@OneToOne(cascade = CascadeType.DETACH, fetch = FetchType.LAZY)@JoinColumn(name = "user_detail_id")private UserDetail userDetail; }
package com.xkzhangsan.jpa.entity;import lombok.Getter; import lombok.Setter;import javax.persistence.*;@Entity @Getter @Setter @Table(name = "user_detail") public class UserDetail {@Id@GeneratedValue(strategy = GenerationType.AUTO)private Long id;private String address; }
1.3 repository
package com.xkzhangsan.jpa.repository;import com.xkzhangsan.jpa.entity.User; import org.springframework.data.jpa.repository.JpaRepository;public interface UserRepository extends JpaRepository<User, Integer> { }
1.4 service
package com.xkzhangsan.jpa.service;import com.xkzhangsan.jpa.entity.User; import com.xkzhangsan.jpa.repository.UserRepository; import org.springframework.beans.factory.annotation.Autowired; import org.springframework.stereotype.Service;import java.util.List;@Service public class UserService {@Autowiredprivate UserRepository userRepository;public List<User> findAll(){return userRepository.findAll();}}
1.5 controller
package com.xkzhangsan.jpa.controller;import com.xkzhangsan.jpa.entity.User; import com.xkzhangsan.jpa.service.UserService; import com.xkzhangsan.jpa.vo.UserDetailVO; import com.xkzhangsan.jpa.vo.UserVO; import org.modelmapper.ModelMapper; import org.springframework.beans.factory.annotation.Autowired; import org.springframework.util.CollectionUtils; import org.springframework.web.bind.annotation.RequestMapping; import org.springframework.web.bind.annotation.RestController;import java.util.List; import java.util.stream.Collectors;@RestController public class UserController {@Autowiredprivate UserService userService;@RequestMapping(value = "/")public List<UserVO> getPersons() {ModelMapper modelMapper = new ModelMapper();List<User> userList = userService.findAll();if (!CollectionUtils.isEmpty(userList)) {return userList.stream().map(user -> {UserVO userVO = modelMapper.map(user, UserVO.class);UserDetailVO userDetailVO = modelMapper.map(user.getUserDetail(), UserDetailVO.class);userVO.setUserDetailVO(userDetailVO);return userVO;}).collect(Collectors.toList());}return null;} }
1.6 测试
查询1次，实际查询了4次，这里的N指的是集合的数量3，所以 N+1，就是3+1=4
1.7 问题原因
User关联对象懒加载导致
2 最佳解决方法
定制@NamedEntityGraph、定制查询和定制VO，可以做到按照需要最佳查询。
2.1 定制@NamedEntityGraph
package com.xkzhangsan.jpa.entity;import lombok.Getter; import lombok.Setter;import javax.persistence.*;@Entity @Getter @Setter @Table(name = "user") @NamedEntityGraph(name = "user.userDetail", attributeNodes = {@NamedAttributeNode(value = "userDetail") }) public class User {@Id@GeneratedValue(strategy = GenerationType.AUTO)private Long id;private String name;@OneToOne(cascade = CascadeType.DETACH, fetch = FetchType.LAZY)@JoinColumn(name = "user_detail_id")private UserDetail userDetail; }
从代码种可以看出增加注解，使用user.userDetail会联表查询UserDetail
@NamedEntityGraph(name = "user.userDetail", attributeNodes = {@NamedAttributeNode(value = "userDetail") })
2.1.2 定制查询
package com.xkzhangsan.jpa.repository;import com.xkzhangsan.jpa.entity.User; import org.springframework.data.jpa.repository.EntityGraph; import org.springframework.data.jpa.repository.JpaRepository;import java.util.List;public interface UserRepository extends JpaRepository<User, Integer> {@Override@EntityGraph(value = "user.userDetail", type = EntityGraph.EntityGraphType.FETCH)List<User> findAll(); }
从代码种可以看出，重写了findAll，使用了user.userDetail
2.3 定制VO（可选）
根据需要返回的字段定义一个VO，如果没有需要则不需要定制，比如当前实例种就不需要。
测试结果如下，只查询了一次，但联表查询了，left outer join user_detail
select user0_.id as id1_0_0_, userdetail1_.id as id1_1_1_, user0_.name as name2_0_0_, user0_.user_detail_id as user_det3_0_0_, userdetail1_.address as address2_1_1_ from user user0_ left outer join user_detail userdetail1_ on user0_.user_detail_id=userdetail1_.id
2.3.1 需要定制VO的实例
比如，只想返回用户id和name，可以定制一个UserSimpleVO，如下：
package com.xkzhangsan.jpa.vo;import lombok.Getter; import lombok.Setter;@Getter @Setter public class UserSimpleVO {private Long id;private String name; }
因为不包括UserDetailVO userDetailVO，所以复制属性的时候不会触发更多的查询。
需要注意的地方：定制VO的字段一定要等于或小于实际查询的字段，才不会复制的时候触发N+1查询。
还有一种方式不用定制VO，通过设置ModelMapper跳过不需要的字段，但这样有2个问题
（1）需要设置ModelMapper，比较麻烦
（2）既然不需要一些字段，定制VO是最有效的方法，这样符合迪米特法则，比如不能展示的敏感字段，如果查询或处理过程中没有处理好，可能导致误返回，最好定制VO，在VO删除这个字段。
3 使用EAGER加载方式
既然是因为LAZY导致的，改成EAGER是否可以解决问题？经过验证是不可行的，仍然会查询多次，比如：
package com.xkzhangsan.jpa.entity;import lombok.Getter; import lombok.Setter;import javax.persistence.*;@Entity @Getter @Setter @Table(name = "user") public class User {@Id@GeneratedValue(strategy = GenerationType.AUTO)private Long id;private String name;@OneToOne(cascade = CascadeType.DETACH, fetch = FetchType.EAGER)@JoinColumn(name = "user_detail_id")private UserDetail userDetail; }
发现仍然会执行多个sql
Hibernate: select user0_.id as id1_0_, user0_.name as name2_0_, user0_.user_detail_id as user_det3_0_ from user user0_ Hibernate: select userdetail0_.id as id1_1_0_, userdetail0_.address as address2_1_0_ from user_detail userdetail0_ where userdetail0_.id=? Hibernate: select userdetail0_.id as id1_1_0_, userdetail0_.address as address2_1_0_ from user_detail userdetail0_ where userdetail0_.id=? Hibernate: select userdetail0_.id as id1_1_0_, userdetail0_.address as address2_1_0_ from user_detail userdetail0_ where userdetail0_.id=?
