MapSturct 是一个生成类型安全,高性能且无依赖的 JavaBean 映射代码的注解处理器(annotation processor)。
(1)注解处理器
(2)可以生成 JavaBean 之间那的映射代码
(3)类型安全,高性能,无依赖性
<dependency>
<groupId>org.projectlombokgroupId>
<artifactId>lombokartifactId>
<version>1.16.20version>
<scope>providedscope>
dependency>
<dependency>
<groupId>org.mapstructgroupId>
<artifactId>mapstruct-jdk8artifactId>
<version>${org.mapstruct.version}version>
dependency>
<dependency>
<groupId>org.mapstructgroupId>
<artifactId>mapstruct-processorartifactId>
<version>${org.mapstruct.version}version>
dependency>
<dependency>
<groupId>cn.hutoolgroupId>
<artifactId>hutool-allartifactId>
<version>5.1.0version>
dependency>
<dependency>
<groupId>junitgroupId>
<artifactId>junitartifactId>
<version>4.12version>
<scope>testscope>
dependency>
@Data
public class User {
private Integer id;
private String name;
private String address;
private Date birth;
}
@Data
public class UserDto implements Serializable {
private Integer id;
private String name;
private String address;
private Date birth;
}
//可以使用abstract class代替接口
@Mapper
public interface UserMapper {
UserDto userToUserDto(User user);
//集合
List<UserDto> userToUserDto(List<User> users);
}
@Test
public void userPoToUserDto() {
User user =new User();
user.setId(1);
user.setName("myx");
user.setAddress("江苏苏州");
user.setBirth(new Date());
UserMapper mapper = Mappers.getMapper(UserMapper.class);
UserDto userDto = mapper.userToUserDto(user);
System.out.println(userDto);
}
底层通过自动取值赋值操作完成
这是相对反射来说的,反射需要去读取字节码的内容,花销会比较大。而通过 MapStruct 来生成的代码,其类似于人手写。速度上可以得到保证。
如果是完全映射的,使用起来肯定没有反射简单。用类似 BeanUtils 这些工具一条语句就搞定了。但是,如果需要进行特殊的匹配(特殊类型转换,多对一转换等),其相对来说也是比较简单的。
基本上,使用的时候,我们只需要声明一个接口,接口下写对应的方法,就可以使用了。当然,如果有特殊情况,是需要额外处理的。
生成的代码是对立的,没有运行时的依赖。
在我们生成的代码中,我们可以轻易的进行 debug。
在实现类的时候,如果属性名称相同,则会进行对应的转化。通过此种方式,我们可以快速的编写出转换的方法。(入门案例)
属性名不相同,在需要进行互相转化的时候,则我们可以通过@Mapping 注解来进行转化。
@Data
public class UserDto implements Serializable {
private Integer id;
private String name;
private String address;
private Date birth;
private String password;
}
@Data
public class User {
private Integer id;
private String name;
private String address;
private Date birth;
private String pwd;
}
@Mapper
public interface UserMapper {
//单个属性
//@Mapping(source = "pwd",target = "password")
//多个属性
@Mappings({
@Mapping(source = "pwd",target = "password")
})
UserDto userToUserDto(User user);
}
(1)source 需要转换的对接,通常是入参
(2)target 转换的对接,通常是出参
(3)ignore 忽略,默认false不忽略,需要忽略设置为true
(4)defaultValue 默认值
(5)expressions 可以通过表达式来构造一些简单的转化关系。虽然设计的时候想兼容很多语言,不过目前只能写Java代码。
@Mappings({
@Mapping(source = "birthdate", target = "birth"),//属性名不一致映射
@Mapping(target = "birthformat", expression = "java(org.apache.commons.lang3.time.DateFormatUtils.format(person.getBirthdate(),\"yyyy-MM-dd HH:mm:ss\"))"),//自定义属性通过java代码映射
})
public PersonVo PersonToPersonVo(Person person);
这里用到演示了如何使用TimeAndFormat对time和format操作,这里必须要指定需要使用的Java类的完整包名,不然编译的时候不知道你使用哪个Java类,会报错。
@Test
public void userPoToUserDto() {
User user =new User();
user.setId(1);
user.setName("myx");
user.setAddress("江苏苏州");
user.setBirth(new Date());
user.setPwd("123456");
UserMapper mapper = Mappers.getMapper(UserMapper.class);
UserDto userDto = mapper.userToUserDto(user);
System.out.println(userDto);
}
如果subUser与subUserDto字段名称相同直接配置即可完成(对象类型,包括list)
@Data
public class UserDto implements Serializable {
private Integer id;
private String name;
private String address;
private Date birth;
private String password;
private List subUserDto;
}
@Data
public class User {
private Integer id;
private String name;
private String address;
private Date birth;
private String pwd;
private List subUser;
}
@Mappings({
@Mapping(source = "pwd",target = "password"),
@Mapping(source = "subUser", target = "subUserDto")
})
UserDto userToUserDto(User user);
有时候,对于某些类型,无法通过代码生成器的形式来进行处理。那么, 就需要自定义的方法来进行转换。这时候,我们可以在接口(同一个接口,后续还有调用别的 Mapper 的方法)中定义默认方法(Java8及之后)。
@Data
public class UserDto implements Serializable {
private Integer id;
private String name;
private String address;
private Date birth;
private String password;
private SubUserDto subUserDto;
}
@Data
public class SubUserDto {
private Boolean result;
private String name;
}
@Data
public class User {
private Integer id;
private String name;
private String address;
private Date birth;
private String pwd;
private SubUser subUser;
}
@Data
public class SubUser {
private Integer deleted;
private String name;
}
@Mapper
public interface UserMapper {
@Mappings({
@Mapping(source = "pwd",target = "password"),
@Mapping(source = "subUser", target = "subUserDto")
})
UserDto userToUserDto(User user);
default SubUserDto subSource2subTarget(SubUser subUser) {
if (subUser == null) {
return null;
}
SubUserDto subUserDto = new SubUserDto();
subUserDto.setResult(!subUser.getDeleted().equals(0));
subUserDto.setName(subUser.getName()==null?"":subUser.getName());
return subUserDto;
}
}
只能存在一个default修饰的方法
@Test
public void userPoToUserDto() {
User user =new User();
user.setId(1);
user.setName("myx");
user.setAddress("江苏苏州");
user.setBirth(new Date());
user.setPwd("123456");
SubUser subUser =new SubUser();
subUser.setDeleted(0);
subUser.setName("rkw");
user.setSubUser(subUser);
UserMapper mapper = Mappers.getMapper(UserMapper.class);
UserDto userDto = mapper.userToUserDto(user);
System.out.println(userDto);
}
我们在实际的业务中少不了将多个对象转换成一个的场景。MapStruct 当然也支持多转一的操作。
@Data
public class SubUser {
private Integer deleted;
private String name;
}
@Data
public class User {
private Integer id;
private String name;
private String address;
private Date birth;
private String pwd;
}
@Mapper
public interface UserMapper {
@Mappings({
@Mapping(source = "user.pwd",target = "password"),
@Mapping(source = "subUser.name", target = "name")
})
NewUserDto userToUserDto(User user,SubUser subUser);
}
@Test
public void userPoToUserDto() {
User user =new User();
user.setId(1);
user.setName("myx");
user.setAddress("江苏苏州");
user.setBirth(new Date());
user.setPwd("123456");
SubUser subUser =new SubUser();
subUser.setDeleted(0);
subUser.setName("rkw");
UserMapper mapper = Mappers.getMapper(UserMapper.class);
NewUserDto userDto = mapper.userToUserDto(user,subUser);
System.out.println(userDto);
}
(1)遵循原则
1-当多个对象中, 有其中一个为 null, 则会直接返回 null
2-如一对一转换一样, 属性通过名字来自动匹配。因此, 名称和类型相同的不需要进行特殊处理
3-当多个原对象中,有相同名字的属性时,需要通过 @Mapping 注解来具体的指定, 以免出现歧义(不指定会报错)。如上面的 name
属性也可以直接从传入的参数来赋值
@Mapping(source = "person.description", target = "description")
@Mapping(source = "name", target = "name")
DeliveryAddress personAndAddressToDeliveryAddressDto(Person person, String name);
有时候,我们不是想返回一个新的 Bean 对象,而是希望更新传入对象的一些属性。这个在实际的时候也会经常使用到。
@Mapper
public interface UserMapper {
NewUserDto userToNewUserDto(User user);
/**
* 更新, 注意注解 @MappingTarget
* 注解 @MappingTarget后面跟的对象会被更新。
*/
void updateDeliveryAddressFromAddress(SubUser subUser,@MappingTarget NewUserDto newUserDto);
}
@Test
public void userPoToUserDto() {
User user =new User();
user.setId(1);
user.setName("myx");
user.setAddress("江苏苏州");
user.setBirth(new Date());
SubUser subUser =new SubUser();
subUser.setDeleted(0);
subUser.setName("rkw");
UserMapper mapper = Mappers.getMapper(UserMapper.class);
NewUserDto userDto = mapper.userToNewUserDto(user);
mapper.updateDeliveryAddressFromAddress(subUser,userDto);
System.out.println(userDto);
}
@MapMapping(valueDateFormat ="yyyy-MM-dd HH:mm:ss")
public Map<String ,String> DateMapToStringMap(Map<String,Date> sourceMap);
@Test
public void mapMappingTest(){
Map<String,Date> map=new HashMap<>();
map.put("key1",new Date());
map.put("key2",new Date(new Date().getTime()+9800000));
Map<String, String> stringObjectMap = TestMapper.MAPPER.DateMapToStringMap(map);
}
只需要在mapper接口中定义相关的类型转换方法即可,list类型也适用
@Data
public class User {
private Integer id;
private String name;
private String address;
private Date birth;
private Boolean isDisable;
private List<SubUser> user;
}
@Data
public class SubUser {
private Integer deleted;
private String name;
private List<SubSubUser> subUser;
}
@Data
public class SubSubUser {
private String aaa;
private String ccc;
}
@Data
public class UserDto implements Serializable {
private Integer id;
private String name;
private String address;
private Date birth;
private String isDisable;
private List<SubUserDto> user;
}
@Data
public class SubUserDto {
private Integer deleted;
private String name;
private List<SubSubUserDto> subUser;
}
@Data
public class SubSubUserDto {
private String aaa;
private String bbb;
}
@Mapper
public interface UserMapper {
UserDto userToNewUserDto(User user);
//子集字段相同方法不用编写会自动生成
//孙子集字段不相同(list会自动读取此方法生成list)
@Mapping(source = "ccc",target = "bbb")
SubSubUserDto bbb(SubSubUser subSubUser);
}
通过uses配置类型转换
@Mapper(uses = {TestMapper.class})
public interface UserMapper {
UserDto userToNewUserDto(User user);
}
@Mapper
public interface TestMapper {
@Mapping(source = "ccc",target = "bbb")
SubSubUserDto bbb(SubSubUser subSubUser);
}
我们都是通过 Mappers.getMapper(xxx.class) 的方式来进行对应 Mapper 的获取。此种方法为通过 Mapper 工厂获取。
如果是此种方法,约定俗成的是在接口内定义一个接口本身的实例 INSTANCE, 以方便获取对应的实例。
@Mapper
public interface SourceMapper {
SourceMapper INSTANCE = Mappers.getMapper(SourceMapper.class);
// ......
}
这样在调用的时候,我们就不需要在重复的去实例化对象了。类似下面
Target target = SourceMapper.INSTANCE.source2target(source);
对于 Web 开发,依赖注入应该很熟悉。MapSturct 也支持使用依赖注入,同时也推荐使用依赖注入。
@Mapper(componentModel = "spring")
可以选择是通过构造方法或者属性注入,默认是属性注入。
public enum InjectionStrategy {
/** Annotations are written on the field **/
FIELD,
/** Annotations are written on the constructor **/
CONSTRUCTOR
}
类似如此使用
@Mapper(componentModel = "cdi" injectionStrategy = InjectionStrategy.CONSTRUCTOR)
有时候,在对象转换的时候可能会出现这样一个问题,就是源对象中的类型是Boolean类型,而目标对象类型是String类型,这种情况可以通过@Mapper的uses属性来实现:
@Data
public class User {
private Integer id;
private String name;
private String address;
private Date birth;
private Boolean isDisable;
}
@Data
public class UserDto implements Serializable {
private Integer id;
private String name;
private String address;
private Date birth;
private String isDisable;
}
@Mapper(uses = {BooleanStrFormat.class})
public interface UserMapper {
UserDto userToNewUserDto(User user);
}
public class BooleanStrFormat {
public String toStr(Boolean isDisable) {
if (isDisable) {
return "Y";
} else {
return "N";
}
}
public Boolean toBoolean(String str) {
if (str.equals("Y")) {
return true;
} else {
return false;
}
}
}
要注意的是,如果使用了例如像spring这样的环境,Mapper引入uses类实例的方式将是自动注入,那么这个类也应该纳入Spring容器
@Test
public void userPoToUserDto() {
User user =new User();
user.setId(1);
user.setName("myx");
user.setAddress("江苏苏州");
user.setBirth(new Date());
user.setIsDisable(true);
SubUser subUser =new SubUser();
subUser.setDeleted(0);
subUser.setName("rkw");
UserMapper mapper = Mappers.getMapper(UserMapper.class);
UserDto userDto = mapper.userToNewUserDto(user);
System.out.println(userDto);
}