Jackson ContextualDeserializer不会反序列化所有字段 [英] Jackson ContextualDeserializer does not deserialize all fields

问题描述

我正在为我的一个实体实现自定义jackson反序列化器。

I'm implementing a custom jackson deserializer for one of my entities.

我的实体如下：

@Value
@JsonDeserialize
@AllArgsConstructor
public class TestModel {

    private final FieldUpdate<UUID> field1Update;
    private final FieldUpdate<UUID> field2Update;
    private final FieldUpdate<String> field3Update;

    public String toString() {
        return "TestModel. Field1="+(field1Update != null ? field1Update.toString() : null)+
                " Field2="+(field2Update != null ? field2Update.getClass().getName() : null) +
                " Field3="+(field3Update != null ? field3Update.getClass().getName() : null);
    }
}

我的问题是序列按预期工作 - 成功序列化对象如下：

My problem is that serialiation works as expected - the successfully serialized object is as follow:

{
  "field1Update" : {
    "type" : "update",
    "value" : "f59c4ef9-52c4-4f3d-99e5-a33a13ae12f3"
  },
  "field2Update" : {
    "type" : "keep"
  },
  "field3Update" : {
    "type" : "reset"
  }
}

=>这是正确的。
（有3种类型更新，保留和重置）。只有更新需要一个值。

=> which is correct. (There are the 3 Types Update, Keep and Reset). Only update needs a value.

问题是：当我反序列化时，只有第一个字段（field1Update）被反序列化。反序列化完成后，其他2个字段（field2Update和field3Update）为空。

The problem is: When i deserialize this, only the first field (field1Update) gets deserialized. The other 2 fields (field2Update and field3Update) are null after deserialization completes.

我的反序列化器如下：

public class FieldUpdateDeserializer extends StdDeserializer implements ContextualDeserializer {

    private JavaType contentType;

    public FieldUpdateDeserializer(JavaType contentType) {
        this(null,contentType);
    }

    public FieldUpdateDeserializer() {
        this(null,null);
    }

    public FieldUpdateDeserializer(Class<?> vc, JavaType contentType) {
        super(vc);
        this.contentType = contentType;
    }

    public JsonDeserializer<?> createContextual(DeserializationContext ctxt,
                                                BeanProperty property) throws JsonMappingException {
        JavaType t = property.getType();
        JavaType boundType = t.getBindings().getBoundType(0);
        return new FieldUpdateDeserializer(boundType);
    }

    @Override
    public Object deserialize(JsonParser jp, DeserializationContext ctx) throws IOException {
        if(!"type".equals(jp.nextFieldName()) )throw new JsonParseException(jp,"'type' expected");
        String typeVal = jp.nextTextValue();
        if("update".equals(typeVal)) {
            jp.nextValue(); //consume type.
            try {
                JsonDeserializer deser = ctx.findNonContextualValueDeserializer(contentType);
                return new Update<>(deser.deserialize(jp,ctx));
            } catch (Exception ex) {
                throw new IllegalStateException("Could not handle deserialization for type", ex);
            }
        } else if("keep".equals(typeVal)) {
            return new Keep<>();
        } else if("reset".equals(typeVal)) {
            return new Reset<>();
        } else {
            return ctx.handleUnexpectedToken(FieldUpdate.class, jp);
        }
    }
}

一个有趣的事实是杰克逊只调用deserialize（...）方法一次，我无法弄清楚为什么....

An interesting fact is that jackson calls the deserialize(...) method only one time and i can't figure out why....

如果有人能给我一个提示，我很高兴。

Glad if somebody can drop me a hint.

问候，
Michael

greetings, Michael

推荐答案

好的 - 经过一段时间的睡眠并分析杰克森序列化器中发生的事情，我发现我的反序列化器中没有消耗足够的令牌。

Ok - after some sleep and analyzing what happens in the jackson serializer, i discovered that i did not consume enough tokens in my deserializer.

我的解串器的工作版本是：

The working version for my deserializer is:

public Object deserialize(JsonParser jp, DeserializationContext ctx) throws IOException {
        if(!"type".equals(jp.nextFieldName()) )throw new JsonParseException(jp,"'type' expected");
        String typeVal = jp.nextTextValue();
        if("update".equals(typeVal)) {
            jp.nextValue(); //consume type.
            try {
                JsonDeserializer deser = ctx.findNonContextualValueDeserializer(contentType);
                return new Update<>(deser.deserialize(jp,ctx));
            } catch (Exception ex) {
                throw new IllegalStateException("Could not handle deserialization for type", ex);
            } finally {
                jp.nextToken();
            }
        } else if("keep".equals(typeVal)) {
            jp.nextToken();
            return new Keep<>();
        } else if("reset".equals(typeVal)) {
            jp.nextToken();
            return new Reset<>();
        } else {
            return ctx.handleUnexpectedToken(FieldUpdate.class, jp);
        }
    }

这篇关于Jackson ContextualDeserializer不会反序列化所有字段的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！