We Keep Coding

Use Jackson to deserialize JSON string or object into a String field

I am using Jackson 2 library and I am trying to read a JSON response, which looks like:
{ "value":"Hello" }
When value is empty, JSON response looks like:
{ "value":{} }
My model POJO class looks like this
public class Hello {
private String value;
public String getValue() {
return value;
}
public void setValue(String value) {
this.value = value;
}
}
The problem is that when response looks like {value:{}}, Jackson is trying to read an Object, but my model class field is a string, so it throws an Exception:
JsonMappingException: Can not deserialize instance of java.lang.String out of START_OBJECT token.
My question is how Jackson can successfully read JSONs who look like:
{"value":"something"}
and at the same time if response looks like this {"value":{}} (empty response for me), pass null to value field of my Hello model class.
I am using the code below in order to read JSON string:
String myJsonAsString = "{...}";
ObjectMapper mapper = new ObjectMapper();
mapper.readValue(myJsonAsString , Hello.class);

You can use a custom deserializer for this feld. Here is one that returns the string if it's there, or null in any other case:
public class Hello {
#JsonDeserialize(using = StupidValueDeserializer.class)
private String value;
public String getValue() {
return value;
}
public void setValue(String value) {
this.value = value;
}
}
public class StupidValueDeserializer extends JsonDeserializer<String> {
#Override
public String deserialize(JsonParser p, DeserializationContext ctxt) throws IOException, JsonProcessingException {
JsonToken jsonToken = p.getCurrentToken();
if (jsonToken == JsonToken.VALUE_STRING) {
return p.getValueAsString();
}
return null;
}
}

Txh for JB Nizet, but if you get type another than String (e.g Object), Jackson deserialiser tried to deserialize inner Object and can throw a latent exception. After that, other fields in json filled as null in Java.
To avoid this you'll ignore children
#Override
public String deserialize(JsonParser p, DeserializationContext ctxt) throws IOException, JsonProcessingException {
JsonToken jsonToken = p.currentToken();
if (jsonToken == JsonToken.VALUE_STRING) {
return p.getValueAsString();
}
p.skipChildren();
return "other_string";
}

Jackson - Transform field value on serialization

In a Spring Boot applicaion with AngularJS frontend, a "Pin" field value has to be blackened on serialization, i.e., if the Pin field value is null in the POJO, the according JSON field has to remain blank; if the field value contains data, it has to be replaced with a "***" string.
Does Jackson provide a feature to get this done?

You can do it easily like following without any Custom Serializer
public class Pojo {
#JsonIgnore
private String pin;
#JsonProperty("pin")
public String getPin() {
if(pin == null) {
return "";
} else {
return "***";
}
}
#JsonProperty("pin")
public void setPin(String pin) {
this.pin = pin;
}
#JsonIgnore
public String getPinValue() {
return pin;
}
}
You can use Pojo.getPinValue() to get the exact value.

Try the following example.
public class Card {
public int id;
public String pin;
}
public class CardSerializer extends StdSerializer<Card> {
public CardSerializer() {
this(null);
}
public CardSerializer(Class<Card> t) {
super(t);
}
#Override
public void serialize(Card value, JsonGenerator jgen, SerializerProvider provider)
throws IOException, JsonProcessingException {
jgen.writeStartObject();
jgen.writeNumberField("id", value.id);
jgen.writeStringField("pin", "****");
jgen.writeEndObject();
}
}
Then you need to register your customer serializer with the ObjectMapper
Card card = new Card(1, "12345");
ObjectMapper mapper = new ObjectMapper();
 
SimpleModule module = new SimpleModule();
module.addSerializer(Card.class, new CardSerializer());
mapper.registerModule(module);
String serialized = mapper.writeValueAsString(card);
There are some improvements you can do here like registering the serializer directly on the class, but you can read more about it here Section 4 - http://www.baeldung.com/jackson-custom-serialization

How do I make Jackson ObjectMapper use my custom deserializer (applied with contentUsing)?

I am having trouble getting jackson to respect my custom JsonDeserializer. The situation is, I have a class MyClass that contains a list of another class, OtherClass, that is outside of my control (so I can't annotate it). This OtherClass class is an interface with multiple implementations. I don't care what the original OtherClass was, I want them to always deserialize as BasicOtherClass.
Here is what I have:
#Getter
public class MyClass {
#JsonProperty("otherclasses")
#JsonSerialize(contentUsing=OtherClassSerializer.class)
#JsonDeserialize(contentUsing=OtherClassDeserializer.class)
private List<OtherClass> otherClasses;
public MyClass(
#JsonProperty("otherclasses")
#JsonDeserialize(contentUsing=OtherClassDeserializer.class)
List<OtherClass> otherClasses) {
this.otherClass = otherClass;
}
}
public static class OtherClassSerializer extends JsonSerializer<OtherClass> {
#Override
public void serialize(OtherClass otherClass, JsonGenerator gen, SerializerProvider serializers)
throws IOException, JsonProcessingException {
gen.writeStartObject();
gen.writeStringField("name", otherClass.getName());
gen.writeStringField("value", otherClass.getValue());
gen.writeEndObject();
}
/** This method is required when default typing is enabled */
#Override
public void serializeWithType(
OtherClass otherClass, JsonGenerator gen, SerializerProvider serializers, TypeSerializer typeSer)
throws JsonProcessingException, IOException {
typeSer.writeTypePrefixForScalar(value, gen, OtherClass.class);
serialize(value, gen, serializers);
typeSer.writeTypeSuffixForScalar(value, gen);
}
}
public static class OtherClassDeserializer extends JsonDeserializer<Header> {
#Override
public Header deserialize(JsonParser p, DeserializationContext ctxt) throws IOException,
JsonProcessingException {
if (p.getCurrentToken() != JsonToken.START_OBJECT) {
throw new IllegalStateException("Failed to parse OtherClass from json");
}
String name = null;
String value = null;
while (p.nextToken() != JsonToken.END_OBJECT) {
String key = p.getText();
p.nextToken();
String val = p.getText();
if (key.equals("name")) {
name = val;
} else if (key.equals("value")) {
value = val;
}
}
return new BasicOtherClass(name, value);
}
}
This is what I am trying to get to work:
ObjectMapper mapper = new ObjectMapper().enableDefaultTyping(ObjectMapper.DefaultTyping.NON_FINAL);
OtherClass otherClass = new BufferedOtherClass("name value");
MyClass myClass = new MyClass(Lists.newArrayList(otherClass));
String json = mapper.writeValueAsString(myClass);
// json == ["com.bschlenk.MyClass", {"otherclass": ["java.util.ArrayList", [["com.other.OtherClass", {"name": "name", "value", "value"}]]]}]
But when I try to read that json back into MyClass, it fails:
MyClass parsed = mapper.readValue(json, MyClass.class);
// com.fasterxml.jackson.databind.JsonMappingException:
// Can not construct instance of org.apache.http.Header, problem:
// abstract types either need to be mapped to concrete types,
// have custom deserializer,
// or be instantiated with additional type information
This works when I don't have type information enabled. However, it is other code that is serializing MyClass that I don't have control of, and it has type info on.
Is what I am trying to do even possible? Why doesn't mapper.readValue use my custom JsonDeserializer class? Is this by design?

Create a type aware Jackson deserializer

I want to configure a Jackson deserializer that act differently depending on the target type of the annotated field.
public class Car {
#JsonSerialize(using=IdSerializer.class)
#JsonDeserialize(using=IdDeserializer.class)
String id
}
public class Bus {
#JsonSerialize(using=IdSerializer.class)
#JsonDeserialize(using=IdDeserializer.class)
Id id
}
Jackson serializers know the type from which it is converting data, so this is working:
public class IdSerializer extends JsonSerializer<Object> {
#Override
public void serialize(Object value, JsonGenerator jsonGen, SerializerProvider provider) throws IOException {
// value is the annotated field class
if(value instanceof String)
jsonGen.writeObject(...);
else if (value instanceof Id)
jsonGen.writeObject(...);
else
throw new IllegalArgumentException();
}
}
Jackson deserializers seem to do not know the target type into which it will convert data:
public class IdDeserializer extends JsonDeserializer<Object> {
#Override
public Object deserialize(JsonParser jp, DeserializationContext context) throws IOException {
// what is the annotated field class?
}
}

In the serializer, you could add extra information about the type that will help you during deserialization.
Building from your posted IdSerializer...
public class IdSerializer extends JsonSerializer<Object> {
#Override
public void serialize(Object value, JsonGenerator jsonGen, SerializerProvider provider) throws IOException {
// value is the annotated field class
if(value instanceof String){
jsonGen.writeStartObject();
jsonGen.writeFieldName("id");
jsonGen.writeObject(value);
jsonGen.writeFieldName("type");
jsonGen.writeString("String");
jsonGen.writeEndObject();
}
else if (value instanceof Id){
Id id = (Id) value;
jsonGen.writeStartObject();
jsonGen.writeFieldName("id");
jsonGen.writeString(id.getStuff());
jsonGen.writeFieldName("type");
jsonGen.writeString("Id");
jsonGen.writeEndObject();
}
else{
throw new IllegalArgumentException();
}
}
}
In your deserializer, you can parse this 'type' field and return an Object of the proper
type.

Custom Jackson Serializer for Wrapper Object

I have the provider interface
interface IProvider<T> {
T locate();
}
and a class containing a field of type IProvider (can be another type for other fields).
class MyObject {
MyLocator<String> field;
}
I need to serialize instances of MyObject to JSON using Jackson 1.7. The output must be the same as if MyObject.field had been a String (i.e. no reference to ILocator).
I can't figure out how to build the custom serializer required to achieve this. Here is the structure I am trying to use for this task:
class MyLocatorSerializer extends SerializerBase<MyLocator<?>> {
public MyLocatorSerializer() {
super(MyLocator.class, false);
}
#Override
public void serialize(MyLocator<?> a_value, JsonGenerator a_jgen,
SerializerProvider a_provider) throws IOException, JsonGenerationException {
// Insert code here to serialize a_value.locate(), whatever its type
}
#Override
public JsonNode getSchema(SerializerProvider a_provider, Type a_typeHint)
throws JsonMappingException {
// What should I return here? I can't find documentation regarding the different schema types...
}
}
The custom serializer would be registered using
SimpleModule module = new SimpleModule("MyModule", new Version(1, 0, 0, null));
module.addSerializer(new MyLocatorSerializer());
objectMapper.registerModule(module);

Another answer using mix-in annotations following the comment from Staxman.
static class JacksonCustomModule extends SimpleModule {
public JacksonCustomModule() {
super("JacksonCustomModule", new Version(1, 0, 0, null));
}
#Override
public void setupModule(SetupContext context) {
context.setMixInAnnotations(IProvider.class, IProviderMixIn.class);
super.setupModule(context);
}
interface IProviderMixIn<T> {
#JsonValue
T locate();
}
}
Activate the module with:
objectMapper.registerModule(new JacksonCustomModule());

Apologies if I misunderstand the question, but would this be as simple as just using #JsonValue on 'Locate' method, instead of writing a custom serializer?
What #JsonValue does is take value of a property as is, and use it instead of creating a JSON Object: often this is used for serializing a POJO as a simple String or number, like so:
public class StringWrapper {
#JsonValue public String value;
}
so that for class like:
public class POJO {
public StringWrapper wrapped;
}
we would get serialization like:
{
"wrapper" : "string value of 'value'"
}
instead of what would otherwise be seen:
{
"wrapper" : {
"value" : "... string value ... "
}
}
Annotation can be used for any types of values obviously.

Following StaxMan's answer, I inspected the workings of #JsonValue and got the following serializer:
// Based on JsonValueSerializer
private static class ProviderSerializer extends SerializerBase<IProvider<?>> {
public ProviderSerializer() {
super(IProvider.class, false);
}
#Override
public void serialize(IProvider<?> value, JsonGenerator jgen, SerializerProvider provider)
throws IOException, JsonGenerationException {
Object object = value.locate();
// and if we got null, can also just write it directly
if (object == null) {
provider.defaultSerializeNull(jgen);
return;
}
Class<?> c = object.getClass();
JsonSerializer<Object> ser = provider.findTypedValueSerializer(c, true, null);
// note: now we have bundled type serializer, so should NOT call with typed version
ser.serialize(object, jgen, provider);
}
#Override
public JsonNode getSchema(SerializerProvider provider, Type typeHint)
throws JsonMappingException {
// is this right??
return JsonSchema.getDefaultSchemaNode();
}
}
After some tests, this does what I need. However, I don't fully really understand the purpose of the getSchema method, so maybe I'm doing something wrong...