终极CRUD-3-用Jackson解析json

1 jackson json基本介绍和使用

网上有很多关于jackson和json的介绍和使用,我就不重复造轮子了,本篇博客主要介绍jackson的高级应用和博主我自己踩坑心得。

如果对json和jackson不熟悉的朋友,可以看下面两篇博客。

https://www.runoob.com/json/json-tutorial.html JSON教程

https://blog.csdn.net/u011054333/article/details/80504154#commentBox jackson快速入门

2 jackson 常用的注解

2.1@JsonProperty

这个注解非常有用,看下面代码:

public class Person {
    @JsonProperty("username")
    private String name;
    private Integer age;
    //省略getter setter
}
    @Test
    public void test1() throws Exception{
        ObjectMapper mapper = new ObjectMapper();
        Person person = new Person("adai",21);
        System.out.println(mapper.writeValueAsString(person));
    }

输出为 {"age":21,"username":"adai"}

可以看到,在序列化的json串中,username替代了name

2.2 @JsonIgnore

public class Person {
    @JsonIgnore
    private String name;
    private Integer age;
    //省略getter setter
}
    @Test
    public void test1() throws Exception{
        ObjectMapper mapper = new ObjectMapper();
        Person person = new Person("adai",21);
        System.out.println(mapper.writeValueAsString(person));
    }

输出为 {"age":21}

2.3 @JsonIgnoreProperties

①这个注解和@JsonIgnore有些类似,不过主要是作用在类上面

@JsonIgnoreProperties(value = {"name","age"})
public class Person {
    private String name;
    private Integer age;
    private Double height;
    //省略getter setter
}
    @Test
    public void test1() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        Person person = new Person("adai", 21, 172D);
        String json = mapper.writeValueAsString(person);
        System.out.println(json);
    }

输出为 {"height":172.0}

可以看出@JsonIgnoreProperties(value = {"name","age"}) 忽略了name和age属性,在序列化的时候,会忽略这两个属性

②@JsonIgnoreProperties注解还有一个ignoreUnknown属性,主要用在反序列化上

在正常情况下,如果我们json串中有一些key值和我们的POJO对象不匹配,那么将会抛出异常。

    @Test
    public void test2() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        System.out.println(mapper.readValue("	    {\"name\":\"adai\",\"age\":21,\"height222\":172.0}", Person.class));   
       					 // !!注意height222与我们的pojo对象不匹配
    }									

程序将会抛出异常

com.fasterxml.jackson.databind.exc.UnrecognizedPropertyException: Unrecognized field "height222" (class com.antiy.common.adai.demo.Person), not marked as ignorable (3 known properties: "name", "age", "height"])
 at [Source: (String)"{"name":"adai","age":21,"height222":172.0}"; line: 1, column: 42] (through reference chain: com.antiy.common.adai.demo.Person["height222"])

此时如果我们在Person类上加上@JsonIgnoreProperties(ignoreUnknown = true)

@JsonIgnoreProperties(ignoreUnknown = true)
public class Person {
    private String name;
    private Integer age;
    private Double height;
    //省略getter setter
}

输出为 Person(name=adai, age=21, height=null)

使用 mapper.disable(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES); 也可以达到同样的目的

建议:ignoreUnknown和FAIL_ON_UNKNOWN_PROPERTIES尽量不要设置为true,如果反序列化的时候,json串中的相关key和POJO属性不匹配,就让程序抛出异常,即使发现错误,不过具体情况还需要参考具体业务,jackson默认该值为false

2.4 @JsonTypeName和@JsonTypeInfo

主要作用:在json串中又包装了一层

①正常情况下,序列化的字符串是 {"name":"adai","age":21,"height":172.0}

当我们在Person类上加上@@JsonTypeName和@JsonTypeInfo时

@JsonTypeName(value = "user222")
@JsonTypeInfo(include = JsonTypeInfo.As.WRAPPER_OBJECT, use = JsonTypeInfo.Id.NAME)
public class Person {
    private String name;
    private Integer age;
    private Double height;
    //省略getter setter
}

输出为 {"user222":{"name":"adai","age":21,"height":172.0}}

②我们也可以使用@JsonRootName("user222")和mapper.enable(SerializationFeature.WRAP_ROOT_VALUE)来达到同样的效果

@JsonRootName("user222")
public class Person {
    private String name;
    private Integer age;
    private Double height;
    //省略getter setter
}
    @Test
    public void test1() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        mapper.enable(SerializationFeature.WRAP_ROOT_VALUE);
        Person person = new Person("adai", 21, 172D);
        System.out.println(mapper.writeValueAsString(person));
    }

输出为 {"user222":{"name":"adai","age":21,"height":172.0}}

2.5 @JsonFormat

主要用在Date属性上

public class Person {
    private String name;
    private Integer age;
    private Double height;
    private Date date;
    //省略getter setter
}
    @Test
    public void test1() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        Person person = new Person("adai", 21, 172D,new Date());
        System.out.println(mapper.writeValueAsString(person));
    }

输出为 {"name":"adai","age":21,"height":172.0,"date":1558842751645}

注意:jackson默认会将Date类型序列化成时间戳,这是因为SerializationFeature中的WRITE_DATES_AS_TIMESTAMPS(true),该值默认为true,当我们手动将改值设为false时。

mapper.disable(SerializationFeature.WRITE_DATES_AS_TIMESTAMPS);

输出为 {"name":"adai","age":21,"height":172.0,"date":"2019-05-26T03:56:38.660+0000"}

这时候date就不再是时间戳了,但是和我们中国的时间格式有一些差别,这个时候就可以使用@JsonFormat

public class Person {
    private String name;
    private Integer age;
    private Double height;
    @JsonFormat(pattern="yyyy-MM-dd HH:mm:ss:SSS",timezone="GMT+8")
    private Date date;
    //省略getter setter
}

输出为 {"name":"adai","age":21,"height":172.0,"date":"2019-05-26 11:58:07:296"}

2.6 @JsonAnyGetter

该注解主要用在序列化:

1.方法是非静态,没有参数的,方法名随意
2.方法返回值必须是Map类型
3.在一个实体类中仅仅用在一个方法上
4.序列化的时候json字段的key就是返回Map的key,value就是Map的value

public class Person {
    private String name;
    private Integer age;
    private Map<String, Object> map = new HashMap<>();

    public String getName() {
        return name;
    }

    public void setName(String name) {
        this.name = name;
    }

    public Integer getAge() {
        return age;
    }

    public void setAge(Integer age) {
        this.age = age;
    }

    @JsonAnyGetter // 注意这个注解
    public Map<String, Object> getOther(){
        return map;
    }
}
    @Test
    public void test1() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        Person person = new Person();
        person.setName("adai");
        person.setAge(21);
        Map<String, Object> other = person.getOther();
        other.put("city", "chengdu");
        System.out.println(mapper.writeValueAsString(person));
    }

输出为 {"name":"adai","age":21,"city":"chengdu"}

当我们在public Map<String, Object> getOther()上去掉@JsonAnyGetter这个注解的时候

输出为 {"name":"adai","age":21,"other":{"city":"chengdu"}}

可以看出加上这个注解以后序列化的时候就会将Map里面的值也相当于实体类里面的字段给显示出来了。

2.7 @JsonAnySetter

主要作用于反序列化上

1.用在非静态方法上,注解的方法必须有两个参数,第一个是json字段中的key,第二个是value,方法名随意

2.反序列化的时候将对应不上的字段全部放到Map里面

public class Person {
    private String name;
    private Integer age;
    private Map<String, String> map = new HashMap<>();

    public String getName() {
        return name;
    }

    public void setName(String name) {
        this.name = name;
    }

    public Integer getAge() {
        return age;
    }

    public void setAge(Integer age) {
        this.age = age;
    }

    @JsonAnySetter //注意这个注解
    public void setOther(String key, String value){
        this.map.put(key, value);
    }

    @Override
    public String toString() {
        return "Person{" +
                "name='" + name + '\'' +
                ", age=" + age +
                ", map=" + map +
                '}';
    }
}
    @Test
    public void test1() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        String json = "{\"name\":\"adai\",\"age\":21,\"color\":\"red\",\"city\":12}";
        Person person = mapper.readValue(json, Person.class);
        System.out.println(person);
    }

输出为 Person{name='adai', age=21, map={color=red, city=12}}

可以看出,使用@JsonAnySetter注解,在json串中多余的属性会被自动放在map属性中,而不会抛出UnrecognizedPropertyException异常

注意:如果是Map<String,String> 那么即使是 {"name":"adai","age":21,"city":12,"weather":true}中的city对应数值 12 和weather对应布尔 true也会被封装进Map<String, String>中,但是Map<String, Integer> 无法封装String或其他类型,只能封装Integer

3 jackson 处理泛型转换

Java中 List和Map主要和泛型打交道,我们重点以这两个为例子,来学习jackson中如何在反序列中保留泛型信息的。

3.1 思考下面程序

public class Student {
    private String name;
    private Integer age;
    //省略getter setter
}
    @Test
    public void test3() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        List<Student> list = new ArrayList<>();
        list.add(new Student("adai",21));
        list.add(new Student("apei",22));
        String json = mapper.writeValueAsString(list);
        List<Student> student = mapper.readValue(json, List.class);
        System.out.println(student.get(0).getName());
    }

该程序在编译期不会报错,可以执行。那么在运行期的时候可以通过吗?

答案是:否定的。 即程序运行失败

java.lang.ClassCastException: java.util.LinkedHashMap cannot be cast to com.antiy.common.adai.demo.Student

原因①:因为在反序列化的时候,mapper.readValue(json, List.class)并没有告诉jackson,这个json数据可以封装成Student对象,所以jackson默认将[{"name":"adai","age":21},{"name":"apei","age":22}]封装成两个LinkedHashMap对象,然后放入到List集合中。

原因②:既然我们知道了List中保存的对象在运行期是LinkedHashMap,那么为什么在代码中还可以student.get(0).getName(),这就跟Java编译期的泛型擦除有关系了,我们可以看下反编译后的代码

        List<Student> student = (List)mapper.readValue(json, List.class);
        System.out.println(((Student)student.get(0)).getName());

student.get(0)实际上的对象是LinkedHashMap,然后强转成Student,自然就报错了!

3.1 JavaType

我们可以使用JavaType来保存泛型信息

List:

    @Test
    public void test4() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        JavaType javaType = mapper.getTypeFactory().constructParametricType(List.class, Student.class);
        List<Student> list = new ArrayList<>();
        list.add(new Student("adai",21));
        list.add(new Student("apei",22));
        String json = mapper.writeValueAsString(list);
        List<Student> student2 = mapper.readValue(json, javaType);
        System.out.println(student2.get(0).getName());
    }

输出为 adai

Map:

    @Test
    public void test5() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        JavaType javaType = mapper.getTypeFactory().constructParametricType(Map.class, 		String.class, Student.class); // 第二个参数是Map的key,第三个参数是Map的value
        Map<String, Student> map = new HashMap<>();
        map.put("first",new Student("adai",21));
        map.put("second",new Student("apei",22));
        String json = mapper.writeValueAsString(map);
        Map<String, Student> result = mapper.readValue(json, javaType);
        System.out.println(result.get("first").getName());
    }

输出为 adai

3.2 TypeReference

TypeReferencejavaType模式更加方便,代码也更加简洁

List:

    @Test
    public void test6() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        List<Student> list = new ArrayList<>();
        list.add(new Student("adai",21));
        list.add(new Student("apei",22));
        String json = mapper.writeValueAsString(list);
        List<Student> student2 = mapper.readValue(json, new 		 	   TypeReference<List<Student>>(){}); 
        System.out.println(student2.get(0).getName());
    }

输出为 adai

Map:

    @Test
    public void test7() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        Map<String, Student> map = new HashMap<>();
        map.put("first",new Student("adai",21));
        map.put("second",new Student("apei",22));
        String json = mapper.writeValueAsString(map);
        Map<String, Student> result = mapper.readValue(json, new TypeReference<Map<String,Student>>(){});
        System.out.println(result.get("first").getName());
    }

输出为 adai

可以看到,使用TypeReference,只需要在mapper.readValue后面增加一个 new TypeReference匿名内部类,写上自己想要封装的泛型对象,比javaType少了一行mapper.getTypeFactory().constructParametricType声明

4 jackson 自定义序列化和反序列化规则

jackson可以通过SerializationFeatureDeserializationFeature来自定义,序列化和反序列化规则,这也是jackson非常强大的地方。

4.1 enable disable configure

请看下面一个例子:

mapper.configure(SerializationFeature.INDENT_OUTPUT,true);
mapper.enable(SerializationFeature.INDENT_OUTPUT);
mapper.disable(SerializationFeature.INDENT_OUTPUT);

这里有三个方法,configure方法接受配置名和要设置的值,Jackson 2.5版本新加的enable和disable方法则直接启用和禁用相应属性,我推荐使用后面两个方法。

4.2 SerializationFeature.INDENT_OUTPUT

默认为false,该属性主要是美化json输出

普通序列化的json串:

{"name":"adai","age":21}

开启该属性后的json串:

{
  "name" : "adai",
  "age" : 21
}

4.3 SerializationFeature.FAIL_ON_EMPTY_BEANS

默认为true,该属性的意思是,如果一个对象中没有任何的属性,那么在序列化的时候就会报错

public class Teacher {}
   @Test
    public void test1() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        Teacher teacher = new Teacher();
        System.out.println(mapper.writeValueAsString(teacher));
    }

程序运行将会报错:

com.fasterxml.jackson.databind.exc.InvalidDefinitionException: No serializer found for class com.antiy.common.adai.entity.Teacher and no properties discovered to create BeanSerializer (to avoid exception, disable SerializationFeature.FAIL_ON_EMPTY_BEANS)

当我们进行设置: mapper.disable(SerializationFeature.FAIL_ON_EMPTY_BEANS)

输出为 {}

4.4 SerializationFeature.WRITE_DATES_AS_TIMESTAMPS

默认为true,该属性的意思是,jackson默认会将Date类型的数据序列化成时间戳

详情可以参考 2.5 @JsonFormat

4.5 DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES

默认为true,该属性的意思是,在反序列的时候,如果json串中存在一些key,但是在POJO中没有,那么程序将会抛出异常

    @Test
    public void test2() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        Student student = new Student("adai",21);
        String json = "{\"name\":\"adai\",\"age222\":21}"; //Student中没有age222
        mapper.readValue(json,Student.class);
    }

程序将会报错:UnrecognizedPropertyException: Unrecognized field "age222"

此时我们将FAIL_ON_UNKNOWN_PROPERTIES设置为false

    public void test2() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        mapper.disable(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES);
        Student student = new Student("adai",21);
        String json = "{\"name\":\"adai\",\"age222\":21}";
        System.out.println(mapper.readValue(json, Student.class));
    }

输出为 Student(name=adai, age=null)

4.6 DeserializationFeature.ACCEPT_EMPTY_STRING_AS_NULL_OBJECT

该值默认为false,该属性的意思是,允许JSON空字符串值(“”)作为null绑定到POJO的属性上,看代码可能比较好理解一点。

public class Teacher {
    private Student student;
    // 省略 getter setter constructor
}
    @Test
    public void test2() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        String json = "{\"student\":\"\"}";
        System.out.println(mapper.readValue(json, Teacher.class));
    }

程序将会报错,MismatchedInputException,因为json串中key值student对应的value为 ""

此时我们可以设置DeserializationFeature.ACCEPT_EMPTY_STRING_AS_NULL_OBJECT为true

输出为 Teacher(student=null)

"" 空串 被转换成null值 封装到Teacher对象的student属性中

4.7 SerializationFeature.WRAP_ROOT_VALUE

默认为false,该属性的意思是,将内容包裹为一个JSON属性,属性名由@JsonRootName注解指定。

详情请见 2.4 @JsonTypeName和@JsonTypeInfo

5 踩坑心得

5.1 TypeReference

一定要导入正确的TypeReference

5.2 DeserializationFeature.ACCEPT_EMPTY_STRING_AS_NULL_OBJECT

注意,该属性只接受POJO的 “” 空字符串转换成 null,在json中,String非常特殊。

请先看4.6章节的内容。

此时我将Teacher中的student类型,换成String

public class Teacher {
    private String student;
}
  @Test
    public void test2() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        String json = "{\"student\":\"\"}";
        System.out.println(mapper.readValue(json, Teacher.class));
    }

输出为 Teacher(student=)

原来以为,如果是String属性,那么""也会转换成null,结果恰恰相反,只有POJO对象,“”才会转换成null

参考 stackoverflow:https://stackoverflow.com/questions/22688713/jackson-objectmapper-deserializationconfig-feature-accept-empty-string-as-null-o

6 感悟

6.1 以Json的角度理解Map和List

在对象序列化和反序列化的过程中,自己对Map和List又有了新的理解。

Map可以当做是一个任意对象,保存字段属性。

在 3.1中,如果jackson不知道反序列化的对象,那么jackson将会以LinkedHashMap来进行处理,这正是因为Map的 Key-Value 特性。

    @Test
    public void test2() throws Exception {
        ObjectMapper mapper = new ObjectMapper();
        Map<String, Object> map = new HashMap<>(2);
        map.put("name","adai");
        map.put("age",21);
        System.out.println("map序列化: " + mapper.writeValueAsString(map));
        Student student = new Student("adai",21);
        System.out.println("student序列化: " + mapper.writeValueAsString(student));
    }

输出为 map序列化: {"name":"adai","age":21}
student序列化: {"name":"adai","age":21}

可以看到Map和Student序列化的结果都是一样的,那么在反序列化的时候,可以用Student对象接受的数据,自然而然也可以用Map接收,这就是为什么在关于泛型反序列化的时候,如果jackson不知道具体的对象,全部都会用LinkHashMap接收

List就当做是一个数组


参考资料: https://blog.csdn.net/u011054333/article/details/80504154#commentBox https://www.ibm.com/developerworks/cn/java/jackson-advanced-application/index.html
作者:一杯热咖啡AAA
出处:https://www.cnblogs.com/AdaiCoffee/
本文以学习、研究和分享为主,欢迎转载。如果文中有不妥或者错误的地方还望指出,以免误人子弟。如果你有更好的想法和意见,可以留言讨论,谢谢!
posted @ 2019-05-27 20:46  一杯热咖啡AAA  阅读(2034)  评论(0编辑  收藏  举报