Jackson JSON pretty print using annotations - java

I'm using #JSONCreator and #JsonCreator to convert a response bean to JSON in Lagom Framework. But, the JSON is not formatted. How can I pretty print the JSON using the annotations (not ObjectMapper)? Here's my sample response bean:
#Immutable
#JsonDeserialize
public class foo {
private final List<Result> _result;
private final MetadataBean _meta;
#JsonCreator
public foo (List<Result> _result, MetadataBean _meta) {
this._result= _result;
this._meta = _meta;
}
}

It seems that pretty printing is controlled by the ObjectMapper and cannot be influenced by annotations. The Lagom documentation for negotiated serializers has this example:
public class JsonTextSerializer implements MessageSerializer.NegotiatedSerializer<String, ByteString> {
private final ObjectMapper mapper = new ObjectMapper();
#Override
public MessageProtocol protocol() {
return new MessageProtocol(Optional.of("application/json"), Optional.empty(), Optional.empty());
}
#Override
public ByteString serialize(String s) throws SerializationException {
try {
return ByteString.fromArray(mapper.writeValueAsBytes(s));
} catch (JsonProcessingException e) {
throw new SerializationException(e);
}
}
}
Pretty printing can then be enabled on the mapper (probably in a constructor):
mapper.enable(SerializationFeature.INDENT_OUTPUT);

Related

Stored Function Error: No converter found capable of converting from type

I created stored function in PostgreSQL. I am getting proper result in PostgreSQL. But when I am trying to call stored function in spring data jpa using native query. I am getting following error.
No converter found capable of converting from type [org.springframework.data.jpa.repository.query.AbstractJpaQuery$TupleConverter$TupleBackedMap] to type [com.spacestudy.model.ClientRT]"
Repository
#Repository
public interface ClientRoomTypeRepository extends JpaRepository<ClientRoomType, Integer> {
#Query(nativeQuery = true,value = "select * from roomtype(:int_inst_id)")
List<ClientRT> roomtype(#Param("int_inst_id")Integer int_inst_id);
}
Result class ClientRT
public class ClientRT {
public Integer res_nclient_room_type_id;
public String res_sclient_rt_desc;
public String res_sclient_rt_name;
public String res_sclient_rt_code;
//getter and setter
...
}
PostgrSQL Result
You can use a logic like this that I have used in my project that might be worked for you or may not. I have not used it with JPA object mapping.
public class ObjectSerializer {
private static ObjectMapper objectMapper;
#Autowired
public ObjectSerializer(ObjectMapper objectMapper) {
ObjectSerializer.objectMapper = objectMapper;
}
public static <T> T getObject(Object obj, Class<T> class1) {
String jsonObj = "";
T userDto = null;
try {
jsonObj = objectMapper.writeValueAsString(obj);
userDto = (T) objectMapper.readValue(jsonObj, class1);
System.out.println(jsonObj);
} catch (JsonProcessingException jpe) {
} catch (IOException e) {
e.printStackTrace();
}
return userDto;
}
}
Hope this will help.

ObjectMapper - How to send null value in JSON

According to third party API spec, I need to send null value in JSON using ObjectMapper if no value exists,
Expected results : "optional": null
If optional value exists, then send "optional": "value"
I didn't find such option in Jackson – Working with Maps and nulls
Code:
requestVO = new RequestVO(optional);
ObjectMapper mapper = new ObjectMapper();
String requestString = mapper.writeValueAsString(requestVO);
Class:
public class RequestVO {
String optional;
public RequestVO(String optional) {
this.optional = optional;
}
public String getOptional() {
return optional;
}
public void setOptional(String optional) {
this.optional= optional;
}
Add #JsonInclude(JsonInclude.Include.USE_DEFAULTS) annotation to your class.
#JsonInclude(JsonInclude.Include.USE_DEFAULTS)
class RequestVO {
String optional;
public RequestVO(String optional) {
this.optional = optional;
}
public String getOptional() {
return optional;
}
public void setOptional(String optional) {
this.optional = optional;
}
}
Example :
RequestVO requestVO = new RequestVO(null);
ObjectMapper mapper = new ObjectMapper();
try {
String requestString = mapper.writeValueAsString(requestVO);
System.out.println(requestString);
} catch (JsonProcessingException e) {
e.printStackTrace();
}
Output :
{"optional":null}
With value:
RequestVO requestVO = new RequestVO("test");
ObjectMapper mapper = new ObjectMapper();
try {
String requestString = mapper.writeValueAsString(requestVO);
System.out.println(requestString);
} catch (JsonProcessingException e) {
e.printStackTrace();
}
Output:
{"optional":"test"}
You can use #JsonInclude annotation on even properties. So, this way you can either serialize as null or ignore some of the properties while serializing.
You can configure your ObjectMapper this way:
ObjectMapper mapper = new ObjectMapper();
mapper.setSerializationInclusion(JsonInclude.Include.NON_NULL);
If no value is present on the JSON request your processed will have a null as you expected.
You can even configure a Spring bean for the ObjectMapper if you need.
EDIT:
I misunderstood the question, he was interested on the JSON response and not on the object parsed. The correct property is this case is JsonInclude.Include.USE_DEFAULTS.
Apologies for the confusion.

Dynamic serialization using Jackson - removing fields with specific annotations

Went down a path of creating an annotation that would dynamic determine whether a field should be serialized or not.
The annotation's implementation is as follows:
#Retention(RetentionPolicy.RUNTIME)
#JacksonAnnotationsInside
#JsonSerialize(using = HiddenFieldSerializer.class)
#Target(value = ElementType.FIELD)
public #interface Hidden {
}
Now the code for the Serializer:
public class HiddenFieldSerializer
extends StdSerializer<String>
implements ContextualSerializer {
public HiddenFieldSerializer() {
super(String.class);
}
#Override
public void serialize(String value,
JsonGenerator jgen,
SerializerProvider provider) {
try {
provider.defaultSerializeNull(jgen);
} catch (IOException e) {
}
}
#Override
public JsonSerializer<?> createContextual(SerializerProvider prov,
BeanProperty property) {
return shouldHide() ?
new HiddenFieldSerializer() : new StringSerializer();
}
public boolean shouldHide() {
/* Simplifying this */
return Boolean.TRUE;
}
}
A little bit of code to show how it works:
public class Test {
static final ObjectMapper mapper = new ObjectMapper()
.setSerializationInclusion(Include.NON_NULL)
.setSerializationInclusion(Include.NON_EMPTY);
static class User {
#JsonProperty
String username;
#Hidden
#JsonProperty
String pin;
}
public static void main(String... args)
throws JsonProcessingException {
final POC.User u = new POC.User();
u.username = "harry_potter";
u.pin = "1298";
System.out.println(mapper.writeValueAsString(u));
}
}
And the output is as follows:
{"username":"harry_potter","pin":null}
How do I get the field pin to be removed from the serialization instead of it being null? Obviously setting the mapper's properties was of very little user in such a context. Any suggestions? Thoughts? Maybe the whole thing is a bad idea?
Ideally I should be able to see the following:
{"username":"harry_potter"}
It's not clear whether you want to ignore a given property statically or dynamically. Anyways, looks like you have over-engineered it.
First of all, I want to make sure that you came across #JsonIgnore before. If it doesn't suit your needs, you could define your custom ignore annotation as following:
#Retention(RetentionPolicy.RUNTIME)
public #interface Hidden {
}
Then pick the approach that best suit your needs:
Approach #1
Extend JacksonAnnotationIntrospector and override the method that checks for the ignore marker:
public class CustomAnnotationIntrospector extends JacksonAnnotationIntrospector {
#Override
public boolean hasIgnoreMarker(AnnotatedMember m) {
return super.hasIgnoreMarker(m) || m.hasAnnotation(Hidden.class);
}
}
Configure ObjectMapper to use your annotation introspector:
ObjectMapper mapper = new ObjectMapper();
mapper.setAnnotationIntrospector(new CustomAnnotationIntrospector());
The annotation introspection occurs only once per class so you can not dynamically change the criteria you use (if any). A similar example can be seen in this answer.
Approach #2
Extend BeanSerializerModifier to modify the properties that will be serialized:
public class CustomBeanSerializerModifier extends BeanSerializerModifier {
#Override
public List<BeanPropertyWriter> changeProperties(SerializationConfig config,
BeanDescription beanDesc, List<BeanPropertyWriter> beanProperties) {
return beanProperties.stream()
.filter(property -> property.getAnnotation(Hidden.class) == null)
.collect(Collectors.toList());
}
}
Then add it to a Module and register it to your ObjectMapper:
ObjectMapper mapper = new ObjectMapper();
mapper.registerModule(new SimpleModule() {
#Override
public void setupModule(SetupContext context) {
super.setupModule(context);
context.addBeanSerializerModifier(new CustomBeanSerializerModifier());
}
});
This approach allows you to ignore properties dynamically.

Object Mapper mapping for a particular class during http calls

Is there a way where we can add ObjectMapper for a particular class through annotation.
#JsonRootName("employee")
public class Sample implements Serializable{
private String name;
private String id;
// Getters and setters
}
In the RestController i have RequestMapping and a method like:-
#ResponseBody
public Sample (#RequestBody Sample sample){
//some logic
return sample;
}
My input payload to this will be like
{
"employee":{
"name":"abcd",
"id":"1234"
}
}
My desired output would be
{
"name":"abcd",
"id":"1234"
}
1)Is there a way i can use the same class to fulfill the input and the output.
2) I have added #JsonRootName at the top of the class which requires ObjectMapper's Serialization feature enable to WRAP_ROOT_VALUE like :-
ObjectMapper mapper = new ObjectMapper();
mapper.enable(SerializationFeature.WRAP_ROOT_VALUE);
mapper.enable(DeserializationFeature.UNWRAP_ROOT_VALUE);
where this can be added to reflect in only this class.
Maybe just leave the default serialization behavior? Then, at deserialization you would still pull out the "employee" wrapper, but at serialization you would write it without the wrapper.
ObjectMapper mapper = new ObjectMapper();
//mapper.enable(SerializationFeature.WRAP_ROOT_VALUE);
mapper.enable(DeserializationFeature.UNWRAP_ROOT_VALUE);
With your input, I got the desired serialization output:
{"name":"abcd","id":"1234"}
EDIT
As for where to put this code, I'd recommend a singleton or class with static methods that handle your (de)serialization. You could have two different mappers than perform the "normal" or "wrapped" behavior. Here's an outline of the static method approach:
public class SerializationUtil {
private static ObjectMapper normalObjectMapper;
private static ObjectMapper wrappedObjectMapper;
static {
/* configure different (de)serialization strategies */
normalObjectMapper = new ObjectMapper();
wrappedObjectMapper = new ObjectMapper();
wrappedObjectMapper.enable(SerializationFeature.WRAP_ROOT_VALUE);
wrappedObjectMapper.enable(DeserializationFeature.UNWRAP_ROOT_VALUE);
}
public static <T> T normalDeserialize(String json, Class<T> clazz) throws Exception {
return normalObjectMapper.readValue(json, clazz);
}
public static String normalSerialize(Object bean) throws Exception {
return normalObjectMapper.writeValueAsString(bean);
}
public static <T> T deserializeWrappedObject(String json, Class<T> clazz) throws Exception {
return wrappedObjectMapper.readValue(json, clazz);
}
public static String serializeWrappedObject(Object bean) throws Exception {
return wrappedObjectMapper.writeValueAsString(bean);
}
}
The benefit of this method is it allows the caller to decide the serialization behavior. So if there are portions of your code where you need to handle it differently you can call another method. Note that the wrapping/unwrapping are both enabled. So to get your desired behavior, you would call these methods like so:
public static void main(String[] args) {
Bean bean = SerializationUtil.deserializeWrappedObject(jsonInput, Bean.class);
String jsonOutput = SerializationUtil.normalSerialize(bean);
}
If this does not appeal to you, you could alternatively detect the special case and handle it in the same method call:
public static <T> T deserialize(String json, Class<T> clazz) throws Exception {
if (clazz instanceof Bean) {
return wrappedObjectMapper.readValue(json, clazz);
} else {
return normalObjectMapper.readValue(json, clazz);
}
}

JSON to Java object deserialization with escaped properties

I need to convert the following JSON to Java object. The property providerResponse in the JSON contains map of properties but they are escaped and wrapped in doubleQuotes. As a result, it does not deserialize the property providerResponse into a Java object (it comes as String). I use objectMapper.readValue(msgStr, classType) to deserialize the JSON. The message is generated by AWS for SNS delivery status notifications and I don't have control to change the JSON message. Is it possible to configure ObjectMapper to unescape the property and deserialize into a Java object instead of String?
{
"delivery":{
"providerResponse":"{\"sqsRequestId\":\"308ee0c6-7d51-57b0-a472-af8e6c41be0b\",\"sqsMessageId\":\"88dd59eb-c34d-4e4d-bb27-7e0d226daa2a\"}"
}
}
#JsonProperty("providerResponse")
private String providerResponse;
There doesn't seem to be a way to configure ObjectMapper to handle this behavior by default. The solution is to create a custom JsonDeserializer:
public class Wrapper {
public Delivery delivery;
}
public class Delivery {
#JsonDeserialize(using = ProviderResponseDeserializer.class)
public ProviderResponse providerResponse;
}
public class ProviderResponse {
public String sqsRequestId;
public String sqsMessageId;
}
public class ProviderResponseDeserializer extends JsonDeserializer<ProviderResponse> {
private static final ObjectMapper mapper = new ObjectMapper();
#Override
public ProviderResponse deserialize(JsonParser jsonParser, DeserializationContext deserializationContext) throws IOException, JsonProcessingException {
return mapper.readValue(jsonParser.getText(), ProviderResponse.class);
}
}
Then you can deserialize the JSON by using your ObjectMapper:
ObjectMapper mapper = new ObjectMapper();
Wrapper wrapper = mapper.readValue(JSON, Wrapper.class);
I faced this similar issue. This gets resolved if we define a constructor in ProviderResponse which takes a single string argument (which is actually json) and then map the json in the constructor to the instance of ProviderResponse and use this temp instance to initialise the properties.
public class Wrapper {
public Delivery delivery;
}
public class Delivery {
public ProviderResponse providerResponse;
}
public class ProviderResponse {
public String sqsRequestId;
public String sqsMessageId;
private static ObjectMapper objMapper = new ObjectMapper();
public ProviderResponse(String json) {
ProviderResponse temp = objMapper.readValue(json, ProviderResponse.class);
this.sqsMessageId = temp.sqsMessageId;
this.sqsRequestId = temp.sqsRequestId;
}
}
The key is to keep the ObjectMapper instance and the its usage somewhere in your utility class and use it from there.

Categories

Resources