I am trying to get my head around the issue where Jackson de-serialization to object results in the list field having duplicate class, when we added an extra getter of the list field.

Here are the classes

public class Wrapper {

    private List<String> values;

    public List<String> getAllValues() {
        if (values == null) {
            values = new ArrayList<>();
        }
        return values;
    }

    public void setValues(final List<String> values) {
        this.values = values;
    }

    public List<String> getValues() {
        return values;
    }

    public String toString() {
        return MoreObjects.toStringHelper(this).add("values", values).toString();
    }
}

My json that I am trying to de-serialize is this:

{
    "values": [
        "value1"
    ],
    "allValues": [
        "value1"
    ]
}

Now, as per my understanding, the setter is only present for values, so I am expecting the allValues entry in the JSON to be ignored since I don’t have any setter there. It is just a getter field introduced for transformation reason.

But apparently, Jackson is not ignoring the field, but using it to set it on the values list.

Below is the code which confirms the duplicate entries during de-serialization:

public static void main(String[] args) throws Exception {

    final ObjectMapper mapper = new ObjectMapper();

    final String wrapperJson = "{ "values": [ "value1" ], "allValues": ["value1"] }";

    System.out.println(mapper.readValue(wrapperJson.getBytes(), Wrapper.class));
}

The output printed is this:

Wrapper{values=[value1, value1]}

Can somebody explain how Jackson is creating the duplicate entry?

Additional information – We are using Jackson Databind – 2.13.5 (Yes, its on old one, so it maybe possible this is already fixed with the new versions)