Module that adds support for serialization/deserialization of Kotlin classes and data classes. Previously a default constructor must have existed on the Kotlin object for Jackson to deserialize into the object. With this module, single constructor classes can be used automatically, and those with secondary constructors or static factories are also supported.
- release
2.13.2
(for Jackson2.13.x
) - release
2.12.6
(for Jackson2.12.x
) - release
2.11.4
(for Jackson2.11.x
) - release
2.10.5
(for Jackson2.10.x
)
Releases require that you have included Kotlin stdlib and reflect libraries already.
Gradle:
implementation "com.fasterxml.jackson.module:jackson-module-kotlin:2.13.+"
Maven:
<dependency>
<groupId>com.fasterxml.jackson.module</groupId>
<artifactId>jackson-module-kotlin</artifactId>
<version>2.13.3</version>
</dependency>
For any Kotlin class or data class constructor, the JSON property names will be inferred from the parameters using Kotlin runtime type information.
To use, just register the Kotlin module with your ObjectMapper instance:
// With Jackson 2.12 and later
import com.fasterxml.jackson.module.kotlin.jacksonObjectMapper
...
val mapper = jacksonObjectMapper()
// or
import com.fasterxml.jackson.module.kotlin.registerKotlinModule
...
val mapper = ObjectMapper().registerKotlinModule()
// or
import com.fasterxml.jackson.module.kotlin.jsonMapper
import com.fasterxml.jackson.module.kotlin.kotlinModule
...
val mapper = jsonMapper {
addModule(kotlinModule())
}
Jackson versions prior to 2.10–2.11
import com.fasterxml.jackson.databind.json.JsonMapper
import com.fasterxml.jackson.module.kotlin.KotlinModule
...
val mapper = JsonMapper.builder().addModule(KotlinModule()).build()
Jackson versions prior to 2.10
import com.fasterxml.jackson.databind.ObjectMapper
import com.fasterxml.jackson.module.kotlin.KotlinModule
...
val mapper = ObjectMapper().registerModule(KotlinModule())
A simple data class example:
import com.fasterxml.jackson.module.kotlin.jacksonObjectMapper
import com.fasterxml.jackson.module.kotlin.readValue
data class MyStateObject(val name: String, val age: Int)
...
val mapper = jacksonObjectMapper()
val state = mapper.readValue<MyStateObject>(json)
// or
val state: MyStateObject = mapper.readValue(json)
// or
myMemberWithType = mapper.readValue(json)
All inferred types for the extension functions carry in full generic information (reified generics).
Therefore, using readValue()
extension without the Class
parameter will reify the type and automatically create a TypeReference
for Jackson.
Also, there are some convenient operator overloading extension functions for JsonNode inheritors.
import com.fasterxml.jackson.databind.node.ArrayNode
import com.fasterxml.jackson.databind.node.ObjectNode
import com.fasterxml.jackson.databind.node.JsonNodeFactory
import com.fasterxml.jackson.module.kotlin.*
// ...
val objectNode: ObjectNode = JsonNodeFactory.instance.objectNode()
objectNode.put("foo1", "bar").put("foo2", "baz").put("foo3", "bax")
objectNode -= "foo1"
objectNode -= listOf("foo2")
println(objectNode.toString()) // {"foo3":"bax"}
// ...
val arrayNode: ArrayNode = JsonNodeFactory.instance.arrayNode()
arrayNode += "foo"
arrayNode += true
arrayNode += 1
arrayNode += 1.0
arrayNode += "bar".toByteArray()
println(arrayNode.toString()) // ["foo",true,1,1.0,"YmFy"]
You can intermix non-field values in the constructor and JsonProperty
annotation in the constructor.
Any fields not present in the constructor will be set after the constructor call.
An example of these concepts:
@JsonInclude(JsonInclude.Include.NON_EMPTY)
class StateObjectWithPartialFieldsInConstructor(val name: String, @JsonProperty("age") val years: Int) {
@JsonProperty("address") lateinit var primaryAddress: String // set after construction
var createdDt: DateTime by Delegates.notNull() // set after construction
var neverSetProperty: String? = null // not in JSON so must be nullable with default
}
Note that using lateinit
or Delegates.notNull()
will ensure that the value is never null
when read, while letting it be instantiated after the construction of the class.
- The
@JsonCreator
annotation is optional unless you have more than one constructor that is valid, or you want to use a static factory method (which also must haveplatformStatic
annotation, e.g.@JvmStatic
). In these cases, annotate only one method asJsonCreator
. - Serializing a member or top-level Kotlin class that implements Iterator requires a workaround, see Issue #4 for easy workarounds.
- If using proguard:
kotlin.Metadata
annotations may be stripped, preventing deserialization. Add a proguard rule to keep thekotlin.Metadata
class:-keep class kotlin.Metadata { *; }
- If you're getting
java.lang.ExceptionInInitializerError
, you may also need:-keep class kotlin.reflect.** { *; }
- If you're still running into problems, you might also need to add a proguard keep rule for the specific classes you want to (de-)serialize. For example, if all your models are inside the package
com.example.models
, you could add the rule-keep class com.example.models.** { *; }
These Kotlin classes are supported with the following fields for serialization/deserialization (and other fields are hidden that are not relevant):
- Pair (first, second)
- Triple (first, second, third)
- IntRange (start, end)
- CharRange (start, end)
- LongRange (start, end)
(others are likely to work, but may not be tuned for Jackson)
Subclasses can be detected automatically for sealed classes, since all possible subclasses are known
at compile-time to Kotlin. This makes com.fasterxml.jackson.annotation.JsonSubTypes
redundant.
A com.fasterxml.jackson.annotation.@JsonTypeInfo
annotation at the base-class is still necessary.
@JsonTypeInfo(use = JsonTypeInfo.Id.NAME)
sealed class SuperClass{
class A: SuperClass()
class B: SuperClass()
}
...
val mapper = jacksonObjectMapper()
val root: SuperClass = mapper.readValue(json)
when(root){
is A -> "It's A"
is B -> "It's B"
}
The Kotlin module may be given a few configuration parameters at construction time; see the inline documentation for details on what options are available and what they do.
val mapper = JsonMapper.builder()
.addModule(KotlinModule(strictNullChecks = true))
.build()
If your ObjectMapper
is constructed in Java, there is a builder method
provided for configuring these options:
KotlinModule kotlinModule = new KotlinModule.Builder()
.strictNullChecks(true)
.build();
ObjectMapper objectMapper = JsonMapper.builder()
.addModule(kotlinModule)
.build();
Following developers have committer access to this project.
- Author: Jayson Minard (@apatrida) wrote this module; still helps issues from time to time
- Active Maintainers:
- Drew Stephens (@dinomite)
- Vyacheslav Artemyev (@viartemev)
- Co-maintainers:
- Tatu Saloranta (@cowtowncoder)
You may at-reference them as necessary but please keep in mind that all maintenance work is strictly voluntary (no one gets paid to work on this or any other Jackson components) so there is no guarantee for timeliness of responses.
All Pull Requests should be reviewed by at least one of active maintainers; bigger architectural/design questions should be agreed upon by majority of active maintainers (at this point meaning both Drew and Vyacheslav :) ).
This module follows the release schedule of the rest of Jackson—the current version is consistent across all Jackson components & modules. See the jackson-databind README for details.
We welcome any contributions—reports of issues, ideas for enhancements, and pull requests related to either of those.
See the main Jackson contribution guidlines for more details.
If you are going to write code, choose the appropriate base branch:
2.12
for bugfixes against the current stable version2.13
for additive functionality & features or minor, backwards compatible changes to existing behavior to be included in the next minor version releasemaster
for significant changes to existing behavior, which will be part of Jackson 3.0
There are a number of tests for functionality that is broken, mostly in the failing
package but a few as part of other test suites. Instead of ignoring these tests (with JUnit's @Ignore
annotation)
or excluding them from being run as part of automated testing, the tests are written to demonstrate the failure
(either making a call that throws an exception or with an assertion that fails) but not fail the build, except if the
underlying issue is fixed. This allows us to know when the tested functionality has been incidentally fixed by
unrelated code changes.
See the tests readme for more information.