I just announced the new Spring 5 modules in REST With Spring:

>> CHECK OUT THE COURSE

1. Overview

In this quick article, we’re going to look at a more advanced use-case of using Jackson – working with Maps that contain null values or null keys.

2. Ignore Nulls Values in a Map

Jackson has a simple but useful way of globally controlling what happens to null values when a Map gets serialized:

ObjectMapper mapper = new ObjectMapper();
mapper.setSerializationInclusion(Include.NON_NULL);

Now any null value in Map object serialized through this mapper is going to be ignored:

@Test
public void givenIgnoringNullValuesInMap_whenWritingMapObjectWithNullValue_thenIgnored() 
  throws JsonProcessingException {
    ObjectMapper mapper = new ObjectMapper();
    mapper.setSerializationInclusion(Include.NON_NULL);

    MyDto dtoObject1 = new MyDto();

    Map<String, MyDto> dtoMap = new HashMap<String, MyDto>();
    dtoMap.put("dtoObject1", dtoObject1);
    dtoMap.put("dtoObject2", null);

    String dtoMapAsString = mapper.writeValueAsString(dtoMap);

    assertThat(dtoMapAsString, containsString("dtoObject1"));
    assertThat(dtoMapAsString, not(containsString("dtoObject2")));
}

3. Serializing A Map With A Null Key

By default, Jackson doesn’t allow the serialization of a Map with a null key. If you do try to write out such a map, you’ll get the following exception:

c.f.j.c.JsonGenerationException: 
  Null key for a Map not allowed in JSON (use a converting NullKeySerializer?)
    at c.f.j.d.s.i.FailingSerializer.serialize(FailingSerializer.java:36)

The library is however flexible enough that you can define a custom, null key serializer and override the default behavior:

class MyDtoNullKeySerializer extends StdSerializer<Object> {
    public MyDtoNullKeySerializer() {
        this(null);
    }

    public MyDtoNullKeySerializer(Class<Object> t) {
        super(t);
    }
    
    @Override
    public void serialize(Object nullKey, JsonGenerator jsonGenerator, SerializerProvider unused) 
      throws IOException, JsonProcessingException {
        jsonGenerator.writeFieldName("");
    }
}

Now the Map with the null key will work just fine – and the null key will be written as an empty String:

@Test
public void givenAllowingMapObjectWithNullKey_whenWriting_thenCorrect() 
throws JsonProcessingException {
    ObjectMapper mapper = new ObjectMapper();
    mapper.getSerializerProvider().setNullKeySerializer(new MyDtoNullKeySerializer());

    MyDto dtoObject = new MyDto();
    dtoObject.setStringValue("dtoObjectString");
 
    Map<String, MyDto> dtoMap = new HashMap<String, MyDto>();
    dtoMap.put(null, dtoObject);

    String dtoMapAsString = mapper.writeValueAsString(dtoMap);

    assertThat(dtoMapAsString, containsString("\"\""));
    assertThat(dtoMapAsString, containsString("dtoObjectString"));
}

4. Ignore Null Fields

Besides Maps, Jackson provides a lot of configuration and flexibility for ignoring/working with null fields in general. You can check out this tutorial to see exactly how that works.

5. Conclusion

Serializing a Map object is common enough that we need a library that’s able to handle the nuances of the serialization process well. Jackson provides a few handy customization options to help you shape the output of this serialization process quite well.

It also provides a lot of solid ways to work with collections in a more general sense.

The implementation of all these examples and code snippets can be found in over on GitHub – this is a Maven-based project, so it should be easy to import and run as it is.

I just announced the new Spring 5 modules in REST With Spring:

>> CHECK OUT THE LESSONS