本文转载自微信公众号「你呀不牛」,作者不牛。转载本文请联系你呀不牛公众号。
1需求
最近做一了个需求,调用其他服务的REST接口,感觉很简单,于是迅速就搞起来了
构造Request类
- public class User {
- private String name;
- private Integer age;
- public User(String name, Integer age) {
- this.name = name;
- this.age = age;
- }
- }
啪,我上来就一new
- service.sendRequest(new User("niu", 18));
打完,收工,又是努力工作(摸鱼)的一天。
2定位
但是,某天晚上8点,测试人员突然给我打电话,说调用失败,同时本身又缺少打印,没有办法具体哪出问题了。
我是不会认为这么简单的代码自己会出错的,不可能!!
经过网络抓包后发现,收到的参数都是null,但是我这边明明调用构造器传入参数了
难道出现灵异事件了?
经过分析,整体数据流为:
能出现问题的地方只能是序列化JSON地方,于是本地测试验证了这一结论:
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- }
虽然是出问题了,但是序列化并没有转为属性为null的对象,而是直接抛出异常
- Exception in thread "main" com.fasterxml.jackson.databind.exc.InvalidDefinitionException: No serializer found for class online.jvm.bean.User and no properties discovered to create BeanSerializer (to avoid exception, disable SerializationFeature.FAIL_ON_EMPTY_BEANS)
- at com.fasterxml.jackson.databind.exc.InvalidDefinitionException.from(InvalidDefinitionException.java:77)
通过查询异常资料,解决掉这种异常需要在增加Jackson的序列化配置FAIL_ON_EMPTY_BEANS,FAIL_ON_EMPTY_BEANS这个配置表示如果某个bean序列化为空时不会异常失败
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- objectMapper.configure(FAIL_ON_EMPTY_BEANS, false);
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- }
这种就不会报错,而是返回序列化成空串,也就导致接受方为属性都为null
通过看自研RPC框架看到是有该FAIL_ON_EMPTY_BEANS的配置
3解决
再来分析一下原因,Jackson序列化时需要调用bean的getter方法
1、写上getter后再看下结果:
- public class User {
- private String name;
- private Integer age;
- public User(String name, Integer age) {
- this.name = name;
- this.age = age;
- }
- public String getName() {
- return name;
- }
- public Integer getAge() {
- return age;
- }
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- // 输出正常 : {"name":"niu","age":18}
- }
- }
2、或者把属性访问权限改为public
- public class User {
- public String name;
- public Integer age;
- public User(String name, Integer age) {
- this.name = name;
- this.age = age;
- }
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- // 输出正常 : {"name":"niu","age":18}
- }
- }
但是如果要求不能暴露bean的属性即使是getter也不行呢?
3、注解 @JsonProperty
这是就需要使用Jackson提供的注解 @JsonProperty
- public class User {
- @JsonProperty("userName")
- private String name;
- @JsonProperty
- private Integer age;
- public User(String name, Integer age) {
- this.name = name;
- this.age = age;
- }
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- // {"userName":"niu","age":18}
- }
- }
来看下注解@JsonProperty的源码注释
- Marker annotation that can be used to define a non-static method as a "setter" or "getter" for a logical property (depending on its signature), or non-static object field to be used (serialized, deserialized) as a logical property.
大体意思是注解如果用在属性上相当于为该属性定义getter和setter。
那如果既有getter又有@JsonProperty注解,以哪个为准呢?
- public class User {
- @JsonProperty("userName")
- private String name;
- @JsonProperty
- private Integer age;
- public User(String name, Integer age) {
- this.name = name;
- this.age = age;
- }
- public String getName() {
- return name;
- }
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- // {"age":18,"userName":"niu"}
- }
- }
如果getter一个没有的属性,效果如何呢?
- public class User {
- @JsonProperty("userName")
- private String name;
- @JsonProperty
- private Integer age;
- public User(String name, Integer age) {
- this.name = name;
- this.age = age;
- }
- public String getName2() {
- return name;
- }
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- // {"age":18,"name2":"niu","userName":"niu"}
- }
- }
这说明如果有@JsonProperty注解,先以注解为准
然后利用反射找到对象类的所有get方法,接下来去get,然后小写化,作为json的每个key值,而get方法的返回值作为value。接下来再反射field,添加到json中。
4、特殊情况
还有一种比较特殊的情况, getter方法由lombok生成,且属性的次首字母是大写:
- @Getter
- public class User {
- @JsonProperty
- private String nAme;
- @JsonProperty
- private Integer age;
- public User(String name, Integer age) {
- this.nAme = name;
- this.age = age;
- }
- public static void main(String[] args) throws IOException {
- ObjectMapper objectMapper = new ObjectMapper();
- String request = objectMapper.writeValueAsString(new User("niu", 18));
- System.out.println(request);
- // {"nAme":"niu","age":18,"name":"niu"}
- }
- }
这是因为lombok生成的getter会把属性的第一个字母变成大写,
序列化时会把get后与小写字母中间的大写变成小写,也就是会把NA变成小写
所以序列化结果会有name(getter获取)和nAme(注解获取)两个属性
- public String getNAme() {
- return this.nAme;
- }
如果我们自己用idea快捷键生成getter,
此时之后序列化nAme
- public String getnAme() {
- return nAme;
- }
4小结
许多bug都是在自以为没有问题的地方产生,看似简单,更需要小心,同时也需要多注意序列化原理,整体感觉序列化还是用Gson更省心,完全不用关心Getter和Setter方法,会完全按照属性名来序列化。
本文的涉及的bug过程和解决方式希望对你也有所帮助,再见。