I'm creating a builder (for Java compat), where context is both a private property and public method.
private lateinit var context: Context
fun context(appContext: Context) = apply {
context = appContext
}
fun build(): MySdk {
// this::context fails to compile because it cannot differentiate between the
// method `context()` vs property `context`
require(this::context.isInitialized) {
"context == null"
}
But I get a compilation issue for ::context.isInitialized, because it cannot differentiate between the method context() vs property context
Does Kotlin have a workaround for this? or am I forced to use unique property/method names?
This is a case of overload resolution ambiguity and the kotlin compiler is unable to identify whether you are using the property or the method.
This is because of callable references (::) . Internally when you are using the callable references it calls a method.
Callable references : References to functions, properties, and
constructors, apart from introspecting the program structure, can also
be called or used as instances of function types.
The common supertype for all callable references is KCallable, where R is the return value type, which is the property type for properties, and the constructed type for constructors.
KCallable<out R> // supertype for all callable references
So, for function the type is KFunction and for properties the type is KProperty
interface KFunction<out R> : KCallable<R>, Function<R> (source)
interface KProperty<out R> : KCallable<R> (source)
When you use a function like :
fun context(appContext: Context) = apply {
context = appContext
}
It can be used as a Function reference
::context // This is a Function reference i.e. KFunction
When you use a property reference, like
private lateinit var context: Context
fun something(){
::context // this is a property reference, KProperty
}
A property reference can be used where a function with one parameter is expected:
val strs = listOf("a", "bc", "def")
println(strs.map(String::length))
So, its not that Kotlin forces you to use different property and function names("although it is not recommended"). Its just that its unable to differentiate in this case as
Both are KCallable and have the same name
A property reference can be used where a function with one parameter is expected
You can resolve the ambiguity between the property and the method by giving the expected type:
val prop: kotlin.reflect.KProperty0<*> = this::context
Alas, prop.isInitialized then gives a compilation error:
This declaration can only be called on a property literal (e.g. 'Foo::bar')
So this doesn't appear to be possible currently. OTOH, since the error shows isInitialized is already handled specially by the compiler, it's likely possible to fix; I suggest reporting it on http://youtrack.jetbrains.com/ (after searching for duplicates).
Related
I'm trying to use inline classes in Kotlin to create a class inlining the String class, such that if I have an instance of my class that it will always be true for the contained string that s == s.trim().
I was initially expecting there to be a straightforward way to do this, like perhaps:
#JvmInline
value class Trimmed private constructor(val str: String) : {
constructor(s : String) : super(s.trim())
}
but that doesn't work, and neither do the other direct approaches I considered ("this(s.trim())", etc.).
This problem has turned out to be surprisingly tricky:
Kotlin seems to provide no easy way to have the primary constructor filter or modify the data that is passed to the constructor of the contained String object.
Even if I make the primary constructor private, I can't declare another constructor with the same signature (taking a single String as a parameter).
If this were a normal (non-inlined) class, I could just set the value after superclass class construction (e.g. "init { str = str.trim() }", but since it's an inline class, I can't do that. ("this=this.trim()" doesn't work either, and String objects themselves are immutable so I can't change the contents of 'str'.)
I tried making the class constructor private and creating a factory function in the same file with the same name as the class, but then I couldn't call the class constructor from within the factory function due to access restrictions.
I then tried making the factory function within the class's companion object, but then Kotlin tried to make that function call itself recursively instead of calling the class's constructor. I wasn't able to find a way to syntactially disambiguate this. I managed to work around this by creating a file-private typealias to give another name for the class so I could call the constructor from within the factory function. (Annoyingly, I couldn't declare the typealias in the companion object next to the factory function: I had to declare it outside.)
This worked, but seemed ugly:
typealias Trimmed2 = Trimmed
#JvmInline
value class Trimmed private constructor(val str: String) {
init { assert(str == str.trim()) }
companion object {
// Kotlin won't let me put the typealias here. :-(
fun Trimmed(s: String): Trimmed = Trimmed2(s.trim()) // Don't want recursion here!
}
}
Another working solution is here, using a private constructor with a dummy argument. Of course Kotlin complained that the dummy argument was unused and so I had to put in a big (why is it so big?) annotation suppressing the warning, which is, again, ugly:
#JvmInline
value class Trimmed private constructor(val str: String) {
private constructor (untrimmed: String, #Suppress("UNUSED_PARAMETER") dummy: Unit) : this(untrimmed.trim())
init { assert(str == str.trim()) }
companion object {
fun Trimmed(s: String): Trimmed = Trimmed(s, Unit)
}
}
Is there a simpler, cleaner way to do this? For instance, a syntactic way to clarify to Kotlin that the companion function is trying to call the class constructor and not itself and so avoid the need for a dummy parameter?
Goals:
Code to construct instances of the class from outside this file should look like constructing an instance of a normal class: 'Trimmed("abc")', not using some factory function with a different name (e.g. "of" or "trimmedOf") or other alternate syntax.
It should be impossible to construct the object containing an untrimmed string. Outside code, and the Trimmed class itself, should be able to trust that if a Trimmed instance exists, that its contained str will be a trimmed string.
Suppose I have the following code to simulate a state machine in Kotlin:
sealed interface State {
object A : State
object B: State
object C: State
object D: State
}
interface StateMachine<Self: StateMachine<Self, *>, T: State>
fun <S : StateMachine<S, State.A>> S.transitionX() = object : StateMachine<S, State.B> {}
fun <S: StateMachine<S, State.B>> S.transitionQ() = object : StateMachine<S, State.B> {}
object Start: StateMachine<Start, State.A>
fun main() {
val stateMachine = Start.transitionX().transitionQ()
}
However, this doesn't compile because
Unresolved reference. None of the following candidates is applicable because of receiver type mismatch:
public fun <S : StateMachine<TypeVariable(S), State.B>> TypeVariable(S).transitionQ(): StateMachine<TypeVariable(S), State.B> defined in root package in file Main.kt
which is probably because of the Self generic constraint.
Ideally, stateMachine should have a type StateMachine<StateMachine<Start, State.A>, State.B.
I was wondering if there's any way to fix the generic constraints so that this does compile? Note: I am aware that the Self generic parameter isn't actually needed for this state machine, but I'm just interested to see if this is actually possible.
I have tried a few different changes to the generic type bounds, but the closest I could get resulted in stateMachine just having a type of StateMachine<Start, State.B>, which isn't quite what I want. Other changes I've made have just caused the Kotlin Finite Bound Restriction error.
Any help is appreciated, thanks!
I don't know what you're trying to do with the self type, so it's hard to say whether these solutions will actually work for your use case.
You don't need to involve new generics in your function itself, only within its receiver and return type. So you can use * types to represent Self. This of course assumes that the Self type isn't needed outside its own private implementation, like if you had a fun copy(): Self. It's impossible to define an implementation of your interface using an anonymous object, since it has to have a class name to be able to describe its own self type. So you either need to define it with a named object outside the function, or by defining a class inside the function and returning an instance of it.
fun StateMachine<*, State.A>.transitionX(): StateMachine<*, State.B> {
class Impl: StateMachine<Impl, State.B>{
}
return Impl()
}
You could define explicit interfaces for all the possible children and use those. Since State is sealed, this is possible.
interface AStateMachine: StateMachine<AStateMachine, State.A>
interface BStateMachine: StateMachine<BStateMachine, State.B>
interface CStateMachine: StateMachine<CStateMachine, State.C>
interface DStateMachine: StateMachine<DStateMachine, State.D>
fun AStateMachine.transitionX() = object : BStateMachine {}
fun BStateMachine.transitionQ() = object : CStateMachine {}
This SO post outlines how to test if a lateinit var has been initialized. However, in the example, the lateinit var is conveniently located within the same class.
How do you do the same thing from outside the class? This is the situation I have:
Foo.kt
class Foo {
lateinit var foo: String
}
Bar.kt
class Bar {
fun doSomething() {
val foo = Foo().foo
if (::foo.isInitialized) { // Unsupported [reference to variables aren't supported yet]
Log.i("TAG", "do something")
}
}
}
What's the workaround for this?
If this was going to work, you'd need to do
val foo = Foo()
if (foo::foo.isInitialized)
//...
The way you're doing it, you're trying to get a property reference of your local variable, which isn't a property. That's why the error says "reference to variables aren't supported yet" rather than "backing field not accessible at this point". Also, you'd be accessing the getter of the lateinit property when assigning the local variable, so it would fail if it weren't initialized yet.
But it doesn't work because of compiler limitations. You could simply add a getter
val fooReady: Boolean get() = ::foo.isInitialized
But I would say the design has very poor encapsulation if outside classes need to check whether a particular public property is initialized yet. In my opinion, any use of isInitialized is a code smell to begin with. If you need to guard calls to the getter with isInitialized, you might as well make the property nullable instead. Then you can use the familiar idioms of null checks instead of resorting to reflection, and it will work in a familiar way even for external classes that access it.
If object of another class has to make a decision based on whether or not the property is initialised, then having this property initialised - or answering whether or not it has already been initialised - is a public business capacity of your object and therefore I would recommend you to simply make it a part of your public API via public fun isFooInitialised(): Boolean function that utilises the fact that the object itself can inspect the state of its lateinit properties.
I have a class for binding a viewholder
class ViewHolderBinder(val onBind: (ViewHolder) -> Unit, val onClick: () -> Unit)
There is a list of these items in presenter
val items: MutableList<ViewHolderBinder> = mutableListOf()
ViewHolderBinder contains 2 non-null functions, but this call causes compile-time error
items.getOrNull(position)?.onClick()
However this call compiles as expected
items.getOrNull(position)?.let { it.onClick() }
Maybe i have missed something, but 2 these constructions are fully equivalent and i prefer to use first one, but it is not compiling.
I am using kotlin 1.3.10
Here's the actual error message you're getting:
Reference has a nullable type '(() -> Unit)?', use explicit '?.invoke()' to make a function-like call instead
Suppose you had this code:
val binder: ViewHolderBinder? = getBinder()
binder?.onClick()
onClick() is not a function you can invoke on the ViewHolderBinder instance. It's a property that holds a callback object. The full syntax to invoke that callback's function is
binder?.onClick?.invoke()
Kotlin also offers a special shorthand syntax that would work on a non-nullable binder:
binder.onClick()
If you apply it to a nullable binder,
binder?.onClick()
it expands to
binder?.onClick.invoke()
The type of the expression binder?.onClick is (() -> Unit)?, just like the error says. You aren't allowed to apply the . operator to a nullable type.
I'm playing with reflection and I came out with this problem. When using bound class reference via the ::class syntax, I get a covariant KClass type:
fun <T> foo(entry: T) {
with(entry::class) {
this // is instance of KClass<out T>
}
}
As I could learn from the docs, this will return the exact type of the object, in case it is instance of a subtype of T, hence the variance modifier.
However this prevents retrieving properties declared in the T class and getting their value (which is what I'm trying to do)
fun <T> foo(entry: T) {
with(entry::class) {
for (prop in memberProperties) {
val v = prop.get(entry) //compile error: I can't consume T
}
}
}
I found that a solution is using javaClass.kotlin extension function on the object reference, to get instead the invariant type:
fun <T> foo(entry: T) {
with(entry.javaClass.kotlin) {
this // is instance of KClass<T>
}
}
This way, I get both the exact type at runtime and the possibility to consume the type.
Interestingly, if I use a supertype instead of a generic, with the latter method I still get access to the correct type, without the need of variance:
class Derived: Base()
fun foo(entry: Base) {
with(entry.javaClass.kotlin) {
println(this == Derived::class)
}
}
fun main(args: Array<String>) {
val derived = Derived()
foo(derived) // prints 'true'
}
If I got it correct, ::class is equal to calling the java getClass, which returns a variant type with a wildcard, while javaClass is a getClass with a cast to the specific type.
Still, I don't get why would I ever need a covariant KClass, when it limits me to only produce the type, given that there are other ways to access the exact class at runtime and use it freely, and I wonder if the more immediate ::class should return an invariant type by design.
The reason for covariance in bound ::class references is, the actual runtime type of an object the expression is evaluated to might differ from the declared or inferred type of the expression.
Example:
open class Base
class Derived : Base()
fun someBase(): Base = Derived()
val kClass = someBase()::class
The expression someBase() is typed as Base, but at runtime it's a Derived object that it gets evaluated to.
Typing someBase()::class as invariant KClass<Base> is simply incorrect, in fact, the actuall result of evaluating this expression is KClass<Derived>.
To solve this possible inconsistency (that would lead to broken type-safety), all bound class references are covariant: someBase()::class is KClass<out Base>, meaning that at runtime someBase() might be a subtype of Base, and therefore this might be a class token of a subtype of Base.
This is, of course, not the case with unbound class references: when you take Base::class, you know for sure that it's the class token of Base and not of some of its subtypes, so it's invariant KClass<Base>.