Consider I have a base of an open hierarchy where all the properties are delegated to a single map:
interface JsonExternalizable {
fun serialize(): MutableMap<String, JsonNode>
}
abstract class AbstractJsonExternalizable protected constructor(private val map: Map<String, Any?>) : JsonExternalizable {
override fun serialize(): MutableMap<String, JsonNode> {
// ...
}
}
Now a sample immutable implementation (I want to have an additional secondary constructor where the list of arguments would "mirror" the type's corresponding properties):
open class Bean1 protected constructor(map: Map<String, Any?>) : AbstractJsonExternalizable(map) {
constructor(property1: String,
property2: Int,
property3: Boolean) : this(mapOf(Bean1::property1 to property1,
Bean1::property2 to property2,
Bean1::property3 to property3).mapKeys { it.key.name })
val property1: String by map
val property2: Int by map
val property3: Boolean by map
}
Things get worse for more "distant" descendants, as I have to list every supertype property (along with the own properties) in the secondary constructor:
class Bean3 private constructor(map: Map<String, Any?>) : Bean1(map) {
val propertyX: Array<Byte> by map
constructor(property1: String,
property2: Int,
property3: Boolean,
propertyX: Array<Byte>) : this(mapOf(Bean3::property1 to property1,
Bean3::property2 to property2,
Bean3::property3 to property3,
Bean3::propertyX to propertyX).mapKeys { it.key.name })
}
Is there a more elegant way to express the same idea, with less boilerplate code?
Related
Not sure if it is possible yet but for the life of me I cannot figure out how to serialize this.
sealed class ServiceResult<out T : Any> {
data class Success<out T : Any>(val data: T) : ServiceResult<T>()
data class Error(val exception: Exception) : ServiceResult<Nothing>()
}
Everything that is stuff into T is using #Serializable ex:
#Serializable
data class GalleryDTO(
override val id: Int,
override val dateCreated: Long,
override val dateUpdated: Long,
val name:String,
val description:String,
val photos:List<DTOMin>
) : DTO
As Animesh Sahu already mentioned there is an issue for this topic that is still open, but the solution using a surrogate suggested by Михаил Нафталь for serialization of Error can actually be used also to serialize the polymorphic ServiceResult, by creating a surrogate that mixes the fields of Success and Error. For the sake of simplicity in the example I only represent the exception message.
#Serializable(with = ServiceResultSerializer::class)
sealed class ServiceResult<out T : Any> {
data class Success<out T : Any>(val data: T) : ServiceResult<T>()
data class Error(val exceptionMessage: String?) : ServiceResult<Nothing>()
}
class ServiceResultSerializer<T : Any>(
tSerializer: KSerializer<T>
) : KSerializer<ServiceResult<T>> {
#Serializable
#SerialName("ServiceResult")
data class ServiceResultSurrogate<T : Any>(
val type: Type,
// The annotation is not necessary, but it avoids serializing "data = null"
// for "Error" results.
#EncodeDefault(EncodeDefault.Mode.NEVER)
val data: T? = null,
#EncodeDefault(EncodeDefault.Mode.NEVER)
val exceptionMessage: String? = null
) {
enum class Type { SUCCESS, ERROR }
}
private val surrogateSerializer = ServiceResultSurrogate.serializer(tSerializer)
override val descriptor: SerialDescriptor = surrogateSerializer.descriptor
override fun deserialize(decoder: Decoder): ServiceResult<T> {
val surrogate = surrogateSerializer.deserialize(decoder)
return when (surrogate.type) {
ServiceResultSurrogate.Type.SUCCESS ->
if (surrogate.data != null)
ServiceResult.Success(surrogate.data)
else
throw SerializationException("Missing data for successful result")
ServiceResultSurrogate.Type.ERROR ->
ServiceResult.Error(surrogate.exceptionMessage)
}
}
override fun serialize(encoder: Encoder, value: ServiceResult<T>) {
val surrogate = when (value) {
is ServiceResult.Error -> ServiceResultSurrogate(
ServiceResultSurrogate.Type.ERROR,
exceptionMessage = value.exceptionMessage
)
is ServiceResult.Success -> ServiceResultSurrogate(
ServiceResultSurrogate.Type.SUCCESS,
data = value.data
)
}
surrogateSerializer.serialize(encoder, surrogate)
}
}
This solution can also be easily extended to support nullable Ts. In this case when deserializing you will also have to check if null is a valid value for T (it can be done by checking descriptor.isNullable on tSerializer) and you will also have to cast data as T.
Polymorphic serialization will be a mess in this case (you will have to manually register all possible types passed as a generic parameter to ServiceResult<T>), and will have several limitations (it would be impossible to register primitive types (including Nothing and String) as generic parameters, for instance).
If you only need serialization (aka encoding), I'd recommend to serialize both subtypes independently (for convenience, wrap subtype determination into auxilary function):
inline fun <reified T : Any> serializeServiceResult(x: ServiceResult<T>) = when (x) {
is ServiceResult.Success -> Json.encodeToString(x)
is ServiceResult.Error -> Json.encodeToString(x)
}
To serialize ServiceResult.Success you need just to mark it with #Serializable annotation. The tricky part here is serialization of ServiceResult.Error, or more precisely, serialization of its exception: Exception field. I'd suggest to serialize only its message (via surrogate):
sealed class ServiceResult<out T : Any> {
#Serializable
data class Success<out T : Any>(val data: T) : ServiceResult<T>()
#Serializable(with = ErrorSerializer::class)
data class Error(val exception: Exception) : ServiceResult<Nothing>()
}
#Serializable
private data class ErrorSurrogate(val error: String)
class ErrorSerializer : KSerializer<ServiceResult.Error> {
override val descriptor: SerialDescriptor = ErrorSurrogate.serializer().descriptor
override fun deserialize(decoder: Decoder): ServiceResult.Error {
val surrogate = decoder.decodeSerializableValue(ErrorSurrogate.serializer())
return ServiceResult.Error(Exception(surrogate.error))
}
override fun serialize(encoder: Encoder, value: ServiceResult.Error) {
val surrogate = ErrorSurrogate(value.exception.toString())
encoder.encodeSerializableValue(ErrorSurrogate.serializer(), surrogate)
}
}
I have the following data classes:
sealed class ExampleDto
object Type1ExampleDto : ExampleDto()
object Type2ExampleDto : ExampleDto()
data class Type3ExampleDto(val name: Int, val age: Int) : ExampleDto()
data class Type4ExampleDto(val name: Int, val age: Int) : ExampleDto()
data class Type5ExampleDto(val email: String) : ExampleDto()
data class Type6ExampleDto(val name: Int, val age: Int, val email: String) : ExampleDto()
In particular, Type3ExampleDto, Type4ExampleDto and Type6ExampleDto share some common fields but it's important for my business logic to distinguish between types (i.e. even if Type3ExampleDto and Type4ExampleDto are identical, I have to know if I'm in the type3 or type4 case).
In one of my method I have the following call:
when (type) {
is Type3ExampleDto -> myMethod(type.vote, type.text)
is Type4ExampleDto -> myMethod(type.vote, type.text)
is Type6ExampleDto -> myMethod(type.vote, type.text)
else -> null
}
I find very ugly that I'm doing the same operation in all 3 cases and repeating the same line...
It makes sense to made Type3ExampleDto, Type4ExampleDto and Type6ExampleDto an implementation of some kind of interface just because only in this point I'm doing this kind of ugly repetition?
If all three dtos implement the following interface
interface MyInterface{
fun getVote() : Int
fun getText() : String
}
I can write:
if (type is MyInterface) {
myMethod(type.getVote(), type.getText())
}
So, it's acceptable to create this interface just to solve this isolated repetition?
Thanks
Note you can do it much more cleanly like this:
interface NameAndAgeDto {
val name: Int
val age: Int
}
data class Type3ExampleDto(override val name: Int, override val age: Int) : ExampleDto(), NameAndAgeDto
if (type is NameAndAgeDto) {
myMethod(type.name, type.age)
}
Whether it's "acceptable" is opinion. Looks fine to me.
You may change your model to have your logic based on behaviour instead of inheritance.
This way of modelling is based on principles of (but ain't exactly) Strategy Design Pattern.
interface HasName {
val name: String
}
interface HasAge {
val age: Int
}
interface HasEmail {
val email: String
}
object Type1
object Type2
data class Type3(
override val name: String,
override val age: Int
) : HasName, HasAge
data class Type4(
override val name: String,
override val age: Int
) : HasName, HasAge
data class Type5(
override val email: String
) : HasEmail
data class Type6(
override val name: String,
override val age: Int,
override val email: String
) : HasName, HasAge, HasEmail
// Then you can pass any object to it.
fun main(obj: Any) {
// Koltin type-casts it nicely to both interfaces.
if (obj is HasName && obj is HasAge) {
myMethod(text = obj.name, vote = obj.age)
}
}
fun myMethod(vote: Int, text: String) {
}
If you still want all the types to belong to some parent type, you can use marker interface (without any methods).
interface DTO
interface HasName {
val name: String
}
interface HasAge {
val age: Int
}
interface HasEmail {
val email: String
}
object Type1 : DTO
object Type2 : DTO
data class Type3(
override val name: String,
override val age: Int
) : HasName, HasAge, DTO
data class Type4(
override val name: String,
override val age: Int
) : HasName, HasAge, DTO
data class Type5(
override val email: String
) : HasEmail, DTO
data class Type6(
override val name: String,
override val age: Int,
override val email: String
) : HasName, HasAge, HasEmail, DTO
// Here, it is DTO instead of Any
fun main(obj: DTO) {
if (obj is HasName && obj is HasAge) {
myMethod(text = obj.name, vote = obj.age)
}
}
And use sealed class instead of marker interface if you need classes as enum.
In that case, when over sealed class is exhaustive with all options without null ->.
I'm got a situation where I have a common property that must be defined on each of the subclasses of a sealed class.
I'd like the ability to be able to access the set/list of these values without 'duplicating' the list (by hard coding it)
Hopefully the below code conveys what I mean
sealed class S {
companion object {
// want to avoid typing: listOf("these", "values", please")
// instead grab it from the classes themselves
val properties = S::class.sealedSubclasses.map { /* What to do here? */ }
}
abstract val property: String
}
class A(val d: String) : S() {
override val property: String = "these"
}
class B(val e: String) : S() {
override val property: String = "values"
}
class C(val f: String) : S() {
override val property: String = "please"
}
I'm aware of fun <T : Any> KClass<T>.createInstance(): T from kotlin.reflect.full, but my constructors have non optional parameters.
You can create a createInstance(vararg) extension function for that:
fun <T : Any> KClass<T>.createInstance(vararg args: Any): T =
java.constructors.first().newInstance(*args) as T
S::class.sealedSubclasses.map { it.createInstance("the string") }
As mentioned in the official tutorial, we can store properties in a Map and delegate a class to it:
class User(val map: Map<String, Any?>) {
val name: String by map
val age: Int by map
}
However, sometimes we store non-trivial structures in a map, like another class (this is usual when working with complicated jsons). To better elaborate my idea, I came up with a pseudo-code like this:
class User(val map: Map<String, Any?>) {
val name: String by map
val otherType: OtherType by map
}
class OtherType {}
Is it possible to delegate such nested structure?
No problem, you can do this. It works:
fun main(args: Array<String>) {
val user = User(mapOf("name" to OtherType(1)))
println(user)
}
data class User(val map: Map<String?, Any?>) {
val name: String by map
val otherType: OtherType by map
}
data class OtherType(val something:Int) {}
You can delegate any type you want.
I have the following abstract class:
abstract class AbstractBook {
abstract val type: String
abstract val privateData: Any
abstract val publicData: Any
}
and the following class which inherits the AbstactBook class:
data class FantasyBook (
override val type: String = "FANTASY",
override val privateData: FantasyBookPrivateData,
override val publicData: FantasyBookPublicData
) : AbstractBook()
And then there is this class which should include data from any type of AbstractBook:
data class BookState(
val owner: String,
val bookData: AbstractBook,
val status: String
)
If I have an instance of BookState, how do I check which type of Book it is and then access the according FantasyBookPrivateData, and FantasyBookPublicData variables?
I hope I described my issue well & thanks in advance for any help!
What you describe is a sealed class:
sealed class Book<T, K> {
abstract val type: String
abstract val privateData: T
abstract val publicData: K
data class FantasyBook(
override val type: String = "FANTASY",
override val privateData: String,
override val publicData: Int) : Book<String, Int>()
}
and in your data class you can do pattern matching like this:
data class BookState(
val owner: String,
val bookData: Book<out Any, out Any>,
val status: String) {
init {
when(bookData) {
is Book.FantasyBook -> {
val privateData: String = bookData.privateData
}
}
}
}
to access your data in a type-safe manner. This solution also makes type redundant since you have that information in the class itself.
I agree with #Marko Topolnik that this seems like a code smell, so you might want to rethink your design.
interface AbstractBook<T , U> {
val privateData: T
val publicData: U
}
data class FantasyBook (
override val privateData: FantasyBookPrivateData,
override val publicData: FantasyBookPublicData
) : AbstractBook<FantasyBookPrivateData , FantasyBookPublicData>
data class BookState(
val owner: String,
val bookData: AbstractBook<*, *>,
val status: String
)
if(bookState.bookData is FantasyBook) {
// Do stuff
}
Creating a type variable is a weak type language writing style. You should use generic class.