杰克逊将额外的字段反序列化为地图

时间:2016-08-25 23:35:57

标签: java json jackson deserialization

我希望将JSON对象中的任何未知字段反序列化为地图中的条目,该地图是pojo的成员。

例如json

{
  "knownField" : 5,
  "unknownField1" : "926f7c2f-1ae2-426b-9f36-4ba042334b68",
  "unknownField2" : "ed51e59d-a551-4cdc-be69-7d337162b691"
}

和pojo

class myObject{
  int knownField;
  Map<String, UUID> unknownFields;
  // getters/setters whatever
}

有没有办法用杰克逊配置?如果没有,是否有一种有效的方法来编写StdDeserializer来执行此操作(假设unknownFields中的值可能是更复杂但众所周知的一致类型)?

2 个答案:

答案 0 :(得分:21)

有一个功能和注释完全符合此目的。

我测试过,它可以像你的例子一样使用UUID:

class MyUUIDClass {
    public int knownField;

    Map<String, UUID> unknownFields = new HashMap<>();

    // Capture all other fields that Jackson do not match other members
    @JsonAnyGetter
    public Map<String, UUID> otherFields() {
        return unknownFields;
    }

    @JsonAnySetter
    public void setOtherField(String name, UUID value) {
        unknownFields.put(name, value);
    }
}

它会像这样工作:

    MyUUIDClass deserialized = objectMapper.readValue("{" +
            "\"knownField\": 1," +
            "\"foo\": \"9cfc64e0-9fed-492e-a7a1-ed2350debd95\"" +
            "}", MyUUIDClass.class);

更常见的类型如字符串工作:

class MyClass {
    public int knownField;

    Map<String, String> unknownFields = new HashMap<>();

    // Capture all other fields that Jackson do not match other members
    @JsonAnyGetter
    public Map<String, String> otherFields() {
        return unknownFields;
    }

    @JsonAnySetter
    public void setOtherField(String name, String value) {
        unknownFields.put(name, value);
    }
}

I found this feature in this blog post first)。

答案 1 :(得分:0)

你需要编写自定义反序列化器这样的东西

class CustomDeserializer extends StdDeserializer<MyObject>{

    public CustomDeserializer(Class<MyObject> t) {
        super(t);
    }

    @Override
    public MyObject deserialize(JsonParser parser, DeserializationContext context)
            throws IOException, JsonProcessingException {
        int knownField=0;
        Map<String, UUID> unknownFields = new HashMap<String, UUID>();

        JsonToken currentToken = null;
      while ((currentToken = parser.nextValue()) != null) {
          switch (currentToken) {
              case VALUE_NUMBER_INT:
                  if (parser.getCurrentName().equals("knownField")) {
                      knownField = parser.getIntValue();
                  }
                  break;
              case VALUE_STRING:
                  if (parser.getCurrentName().startsWith("unknownField")){
                      unknownFields.put(parser.getCurrentName(),UUID.fromString(parser.getValueAsString()));
                  }
                  break;
              default:
                  break;
          }
      }
      return new MyObject(knownField,unknownFields);
    }
}

然后按如下方式设置自定义反序列化器

ObjectMapper mapper = new ObjectMapper();
        SimpleModule mod = new SimpleModule("TestModule");
        mod.addDeserializer(MyObject.class, new CustomDeserializer(MyObject.class));
        mapper.registerModule(mod);
        MyObject myObj = mapper.readValue(contents, MyObject.class);

编辑:您可以在点击START_OBJECT令牌时反序列化任何复杂对象。并使用DeserializationContext.readValue(JsonParser, Class<MyObject>)反序列化MyObject