Jackson - Required property? - java

I'm using Jackson's readValue() method on an object mapper to read from a JSON file and convert it into my java object.
eg.
mapperObject.readValue( node, MyTargetClass.class )
Are there any annotations that I can set on MyTargetClass to enforce required attributes? For example, if I have a JSON object with properties ABC,DEF and GHI, and my Json is the following
{
"ABC" : "somevalue"
"DEF" : "someothervalue"
}
I want it to fail somehow, and only succeed on the readValue if it contained ABC, DEF and GHI.

You can mark a property as required with the #JsonProperty(required = true) annotation, and it will throw a JsonMappingException during deserialization if the property is missing or null.
Edit: I received a downvote for this without comment. I'd love to know why, since it does exactly the right thing.

Jackson does not include validation functionality, and this is by design (i.e. that is considered out-of-scope). But what is usually used is Bean Validation API implementation.
The nice thing about this is decoupling between data format handling, and validation logic.
This is what frameworks like DropWizard use; and it's the direction JAX-RS (like Jersey) are taking things for JAX-RS 2.0.

If you want to make sure a json field is provided, you have to use the #JsonProperty(value = "fieldName", required = true) annotation as a parameter to the constructor. But this is not enough, also the Constructor should have #JsonCreator annotation.
For example, if you have a field named 'endPoint' and you want o make sure it is provided in the JSON file, then the following code will throw an exception if it is not provided.
#JsonCreator
public QuerySettings(#JsonProperty(value = "endPoint", required = true) String endPoint) {
this.endPoint = endPoint;
}
I found this link helpful to understand the Jackson annotations. It also well explains why required=true is not enough and counter-intuitive to its name.

If you are neither satisfied with using #JsonProperty(required = true) as it works only with #JsonCreator nor with the use of bean validation then one more way of tackling it would be to catch this in your setter methods for the relevant variables.
You can simply check if the variable is null before setting it and throw an IllegalArgumentException or NullPointerException (as preferred by few people)
Note: It depends on how your POJO is defined too, so please make sure that it is going the setter method route for this solution to work.

Related

org.springframework.web.bind.MethodArgumentNotValidException

I am using spring boot and I have a RestController with method
addBook(#Validated #RequestBody BookDto bookDto)
BookDto has a field annotated with write only
#JsonProperty(access = Access.WRITE_ONLY)
#NotNull(message = "cannot be empty")
#Size(min = 1, max = 20)
private String isdnNo;
This works as expected and I have to provide isdnNo while i do a post request.
And when I get it don't bring the isdnNo in reponse which is fine.
Issue here is in JUnit test case. I am using stand alone setup to test controller.
MockMvcBuilders.standaloneSetup(bookController).build();
MockHttpServletRequestBuilder postRequest = MockMvcRequestBuilders.post("/books/").contentType(JSON)
.accept(JSON);
postRequest.content(asJsonString(bookDto));
ResultActions result = mvc.perform(postRequest);
This JUnit test case gives exception org.springframework.web.bind.MethodArgumentNotValidException and says idnNo cannot be empty even though I have provided isdnNo in bookDto on performing post. What can be the issue here.
if I remove #JsonProperty(access = Access.WRITE_ONLY) the test case works fine.
Thanks in advance !
You want to keep Access.WRITE_ONLY removed because, according to the documentation, during serialization (ie writing it to a string) it will not read the value for serialization. You're method asJsonString is serializing the object thus removing the value.
AUTO: Access setting which means that visibility rules are to be used to automatically determine read- and/or write-access of this property.
READ_ONLY:
Access setting that means that the property may only be read for serialization, but not written (set) during deserialization.
READ_WRITE
Access setting that means that the property will be accessed for both serialization (writing out values as external representation) and deserialization (reading values from external representation), regardless of visibility rules.
WRITE_ONLY
Access setting that means that the property may only be written (set) for deserialization, but will not be read (get) on serialization, that is, the value of the property is not included in serialization.
https://fasterxml.github.io/jackson-annotations/javadoc/2.6/index.html?com/fasterxml/jackson/annotation/JsonProperty.Access.html

graphql-java-tools: null parameter vs not present parameter

How can I distinguish in java graphQL if a parameter was explicitly set to null, or if it was not provided at all?
The use case that I try to achieve is the following: I have a mutation like this
updateUser(id:Int!, changes:UserChanges!)
#where UserChanges is defined as:
type UserChanges {
login:String
email:String
#etc
}
The idea here is that the user provides only the fields that he wants to change (like react setState for example).
So if email is ommited, I want to leave it unchanged.
But what if he/she wants to explicitly set email to null?
Is there any way figure this out from my resolver method and act accordingly?
(I use graphql-java-tools library)
I found the answer. In case somebody needs it:
An instance of graphql.schema.DataFetchingEnvironment is available to every resolver method.
This provides methods like getArguments(), hasArgument() etc.
Using those methods, we can find out if an argument was an explicit set to null, or if it was not provided at all.
Looks like deserialization from query/variables is handled by fasterxml Jackson, and that's proper place to deal with the issue, otherwise it becomes too complex: check every field? nested?
So: UserChanges.java should look like this:
class UserChanges {
// SHOULD NOT HAVE ALL ARGUMENT CONSTRUCTOR!
Optional<String> login;
Optional<String> email;
... getters & setters
}
in this case deserializer will use setters, ONLY FOR PROVIDED FIELDS!
And {"login":null} will become:
UserChanges.login = Optional.empty
UserChanges.email = null

Is there some way for a Jackson Delegate-based Creator to access the raw Json String?

Is there some way for a Jackson Delegate-based creator to access the raw Json String?
#JsonCreator
private static MyClass createFromJson(Map<String, Object> jsonProperties) {
return new MyClass(rawJson);
}
I am able to get the raw input as a Map of Strings to Objects in the code above, but I want to be able to access the json as a string. I tried the code below (based off of http://www.cowtowncoder.com/blog/archives/2011/07/entry_457.html) but that code as written is never invoked.
#JsonCreator
private static MyClass createFromJson(String rawJson) {
return new MyClass(rawJson);
}
Note: This is a spring boot application (1.3.1.RELEASE) that uses Jackson 2.6.4.
Looks like this type of functionality would not make sense in this context. In fact, it appears to me now that requesting the JSON string in this instance defeats the purpose of using jackson in the first place. However if anyone finds themselves here, then the comments from Sotirios Delimanolis may be useful:
"Hack: you can receive a JsonNode as the parameter type and use its toString method to get the corresponding JSON."
"It looks like you want a JsonDeserializer"

Java annotations: pass value of annotation attribute to another annotation

I have interface Resource and several classes implementing it, for example Audio, Video... Further, I have created custom annotation MyAnnotation with Class type param:
#MyAnnotation(type = Audio.class)
class Audio {
...
}
#MyAnnotation(type = Video.class)
class Video{
...
}
In some other place in code I have to use Interface Resource as a returned type:
public class Operations<T extends Resource> {
....
#OtherAnnotation(type = Audio.class (if audio), type = Video.class (if video) )
T getResource();
....
}
The question is how to appropriatelly annotate annotation #OtherAnnotation depending of what kind of Resource type will be returned ?
What you are asking is for dynamic values for annotation attributes.
However annotations can only be set at compile time which is the reason why their values can only be compile time constants. You may only read them at runtime.
There was a similar question in which someone tried to generate the annotation value , it's answer explains why there is no way to dynamically generate a value used in annotation in a bit more detail. In that question there was an attempt to use a final class variable generated with a static method.
There are annotation processors which offer a bit more flexibility by handling placeholders. However i don't think this fits your case, as you want the dynamic values at runtime.
This answer refers to spring's use of the expression language for the Value annotation in which the placeholder (#Value("#{systemProperties.dbName})") gets overrided with the data from one of the property sources defined ( example in spring boot )
In any case, you will have to rethink your architecture a bit.

Bean Validation - constructor/factory parameter

I have a dto object which keeps an IP Range using first and last fields. Simple CRUD operations are made with this class using dropwizard (jersey-jackson-hibernate validator)
public class IpRangeDto {
#JsonCreator
public static IpRangeDto fromCidr(#JsonProperty("cidr") String cidr) {
//Resolve CIDR and assign first and last fields
}
#NotNull
#IpAddress // My custom validator
private String first;
#NotNull
#IpAddress
private String last;
}
For the sake of user-friendliness I had decided to add an alternative way to create this object, which is by using CIDR. So the client could send either first and last fields in JSON or only the cidr field. So the way to do it is as above, using #JsonCreator. And it works just fine.
"ipRange":{
"first": "15.0.0.1",
"last": "15.0.0.255",
}
"ipRange":{
"cidr": "15.0.0.0/24"
}
I want to validate this CIDR value that it's the right format so I can return 422 with proper error message. If I throw exception in the constructor/factory method then jersey-jackson returns 400 directly (even if I throw ConstraintViolationException, it's encapsulated by JsonProcessingException).
I could simply ignore the exceptions, and leave the fields empty which will return 422 because of #NotNull constraints but then the error message will not be as clear as it should be.
I tried adding my #Cidr validator next to the #JsonProperty parameter but that doesn't seem to be effective. My understanding is that validation occurs after Jackson is finished with creating Dtos, so with my #JsonCreator approach there might not be any solution to this problem. So I'm open to refactoring suggestions as well.
I am not an expert on the exact integration of Bean Validation into jackson, but I think it is just doing actual property validation. This means as you already pointed out, the entities are created first and then the properties are validated.
Bean Validation (as of version 1.1) also offers so called method validation, in which case you could place your Cidr constraint onto the string parameter of the method, but as said, I don't think that there is an integration in jackson for that.
And one more thing ;-) - static methods and properties are generally excluded from validation in Bean Validation (see also http://beanvalidation.org/1.1/spec/#d0e2815).
Regarding a workaround, one thing comes to mind (even though it feels a bit complicated). Write a custom class level IpRange constraint. In a class constraint you would get passed a IpRangeDto instance and it is up to you to validate the whole object and select the right error message for any violations. Provided you would add a cidr property to the dto which gets set when fromCidr is called, you would have then all information you need for the validation and selection of a proper error message.

Categories