Just met a behaviour that I don’t understand. I have a multi-module project, let’s say: A, B, C. The dependency between them is:  • A - doesn’t depend on other modules, • B - depends on A • C - depends on A and B In A and B modules I use some lib, for ex: Koin. To add this lib, I should add the next lines to each module’s gradle file:

repositories {
   jcenter()
   maven { url = uri("<https://dl.bintray.com/ekito/koin>") }
}

and 

dependencies {
      implementation("org.koin:koin-core:3.0.0-alpha-4")
}

I’m ok with that, but the thing I don’t understand, why I have to add a repository block to C module as well, while I don’t have a dependency on that lib in C module? Thank you in advance.

We're getting the following interesting error:

java.lang.VerifyError: Bad type on operand stack
Exception Details:
  Location:
    com/example/Foo.bar(Lkotlin/jvm/functions/Function3;)Lcom/example/Other$ThingConsumer; @13: invokespecial
  Reason:
    Type 'kotlin/jvm/functions/Function3' (current frame, stack[3]) is not assignable to 'kotlin/jvm/internal/Ref$ObjectRef'

when attempting to use:

runBlocking { consumer(connection, event) }

where

consumer: suspend (Connection, Event) -> Unit

is passed into the function calling

runBlocking

. We can hack around it by replacing it with something like

GlobalScope.async { consumer(connection, event) }.asCompletableFuture().get()

but it's less than ideal 😬 This is using Kotlin 1.4.10 and kotlinx.coroutines 1.4.0. Any thoughts?

how to create a (Mutable)Collection<String!> value?

Is it possible to exclude paths from

explicitApi

? I want to enable it for a library but that library also exposes generated classes by sqldelight and I'd like to not fail the compilation on generated classes.

Can someone explain to me why the

reduce

signature is like:

inline fun <S, T : S> Array<out T>.reduce(
    operation: (acc: S, T) -> S
): S

While the

fold

signature is like:

inline fun <T, R> Array<out T>.fold(
    initial: R,
    operation: (acc: R, T) -> R
): R

Shouldn't

reduce

be able to take two independent types instead? Something like

inline fun <S, T> Array<out T>.reduce(
    operation: (acc: S, T) -> S
): S

For context, I am trying to use

reduce

with a

long

accumulator and a domain object I am using. Something like:

myCollection.reduce { acc, myObject ->
    acc + myObject.myField
}

That doesn't work but this does:

myCollection.fold(0) { acc, myObject ->
    acc + myObject.myField
}

This is counterintuitive and confusing.

I've recently upgraded a project from kotlin 1.3.72 to 1.4.10. Suddenly I am experiencing error of the following kind:

val deviceMap = mutableMapOf(
                "organisationId" to authenticationData.organisationId,
                "userId" to authenticationData.userId,
                "deviceTemplateUuid" to AuthenticationService.DeviceTemplate.findById(authenticationData.deviceTemplateId).clientId,
                "dateInstalled" to dateInstalled
        ).also {
            it.putAll(extraProps)
        }

The variable "extraProp" is defined as Map<String, Any>. In the past, I'm certain kotlin would have interpreted the return from "mutableMapOf" as MutableMap<String, Any>, but now I'm suddenly getting the following compiler error: Kotlin: Type mismatch: inferred type is Map<String, Any> but Map<out String, {Comparable<{Instant & String & UUID}> & java.io.Serializable}> was expected Is it mandatory now that I give a type hint to the function as "mutableMapOf<String,Any>(...)" to avoid this Union type nightmare?

Is there a way to write code in such a way that the kotlin analyzer wouldn't complain about these variables being nullable? I'm paraphrasing my code here, but it looks like this

val a: String? = ...
    val b: String? = ...

    if (listOf(a,b).any { it == null }) {
      return
    }
    
    val l = a.length // compiler complains here

Basically I'm doing null checks in for a bunch of variables before doing something with them. Shouldn't the compiler know about

a

on the last line, not being null?

Do these functions already exist in the stdlib? I’m failing to find them, but bitter experience suggests that this may just be me being stupid…

fun <T> populate(i: Int, f: (Int) -> T): List<T> = (0 until i).map(f)
fun <T> populate(i: Int, t: T): List<T> = populate(i) { t }

populate(4, "Hello") == listOf("Hello", "Hello", "Hello", "Hello")
populate(4) { "Hello $it" } == listOf("Hello 0", "Hello 1", "Hello 2", "Hello 3")

That feeling when you are not on latest kotlin so you can’t use trailing comma.

Hi all, do anyone knows if Jetbrains is planning any port of Jetpack Compose to iOS and the Web ?

Is it possible to hide fields from the

toString

function on data classes? Can be with annotation or anything. I want to achieve something like this: https://pl.kotl.in/Ld0EBnHM1

I am so confused by kotlix.serialization it seems it is meant to serialize custom objects only, annotated with @Serializable I cannot figure out how to serialize a simple Map<K,V>

I have a syntetic example of a problem I'm trying to solve.

class MyDelegatedProperty {
    var innerStr: String = ""

    operator fun getValue(thisRef: Any, property: KProperty<*>): String {
        return innerStr
    }
    operator fun setValue(thisRef: Any, property: KProperty<*>, value: String) {
        innerStr = value
    }

    fun doSomething() {
        println(innerStr)
    }
}

class Example {
    val str: String by MyDelegatedProperty()

    fun main() {
        // use the doSomething function from str
    }
}

Is this possible somehow? Can I access the object that is created by a delegated property?

Hi, I need an immutable sorted map. Which type should I use? •

mapOf()

is not sorted •

sortedMapOf()

is mutable •

hashMapOf()

is mutable and not sorted •

linkedMapOf()

is mutable Any other suggestions?

Hi, I'm trying to modify the compiler plugin

kotlin-serialization

, but I failed to test the modified one in local. How can I use a local modified one when building?

is there a way for me to ignore experimental warnings for the entire project through the top level

build.gradle

?

Any doc/article/reference for the new Kotlin's IR (intermediate representation)? The sources are so wide (even though well documented more than half of those), its hard to follow a path i.e. where to see and understand the things.

Have somebody else noticed that github seems to not be syntax-highlighting Kotlin now? Could be a bug in their highlighters?

Off topic question: does anyone who is using jetbrains spaces know if it supports real time collaborative source code editing ?

What is the benefit of putting values like this into a private object? Is the only purpose that we have the

PreferencesKey.

prefix?

private object PreferencesKeys {
        val SORT_ORDER = preferencesKey<String>("sort_order")
        val HIDE_COMPLETED = preferencesKey<Boolean>("hide_completed")
    }

Evening, all. I'm converting a project from Gradle-Kotlin to Maven, and in the process of migration have kicked up a worrying warning message. I'm not sure where to report it, suggestions, please?

[INFO] --- kotlin-maven-plugin:1.4.20-RC:compile (compile) @ app ---
WARNING: An illegal reflective access operation has occurred
WARNING: Illegal reflective access by com.intellij.util.ReflectionUtil (file:/Users/alecm/.m2/repository/org/jetbrains/kotlin/kotlin-compiler/1.4.20-RC/kotlin-compiler-1.4.20-RC.jar) to method java.util.ResourceBundle.setParent(java.util.ResourceBundle)
WARNING: Please consider reporting this to the maintainers of com.intellij.util.ReflectionUtil
WARNING: Use --illegal-access=warn to enable warnings of further illegal reflective access operations
WARNING: All illegal access operations will be denied in a future release

Sorry for russian. народ, кто в тонкостях потоков хорошо разбирается. Задача на оптимизацию. Голову ломаю уже неделю. Решение на кончике но никак не обернется в готово. /Есть большой поток (стрим). Скажем CSV на гиков 100, (те его читать по кругу оч не эффективно). В нем содержатся скажем объекты

data claaa D(vtype: string, value: int, dim1 ... dimX)

где dimX - некоторые "классификации, дайменшены) Собственно я беру этот CSV - прогоняю через скажем 5 фильтров и остается у меня от общем массы по 1% в каждом из этих 5 флоу. Вопрос. Как эффективно также продолжая 1 проход отфильровать из 5 отдельных фору группу где di1 и dim2 и dim3 у элементов совпадают). Те сделать группировку по разрезу и отсеять где нет полного совпадения.... Завис Пока только смог сделать промежуточную фильтрацию "Явно не войдет" и дальше второй проход уже по (флоу в буфер/канал) каналам в который результат первого цикла. Есть идеи? Как можно не дожидаясь окончание потока получить на текущем результате пересечение и сразу же по нему отдельный поток/корутин .

Hi, is there a given pattern on when (and when not) to use

check

functions? We have a use-case where we'd like to throw a concrete custom Exception. What would be the ideal way to do it? Wrap it later, but use

check

try {
  check(<exp>){ "check failed"}
}  catch(e: IllegalStateException) {
  throw CustomException(e)
}

or throw it directly,

if (<exp>) {
  throw CustomException("check failed)
}

Do you think there is any performance difference between these two implementations?

Does anyone know why this doesn't compile? If I understand correctly: * T is

String

*

in T

means that I can pass T or any of its super classes *

Any

is superclass of

String

* so

Container<in T>

should accept

Container<Any>

right? Instead I get:

Type mismatch.
Required :String
Found: Any

Similar question to the above, but with covariance this time. If I undestand correctly: *

T

is String *

out T

means that String or any of its subtypes can fit this * Since

String

in

methodB

matches that, why I can't pass

methodB

into

methodA

?

inline fun SearchView.onQueryTextChanged(crossinline listener: (String) -> Unit) {
    this.setOnQueryTextListener(object : SearchView.OnQueryTextListener {
        override fun onQueryTextSubmit(query: String?): Boolean {
            return true
        }

        override fun onQueryTextChange(newText: String?): Boolean {
            listener(newText.orEmpty())
            return true
        }
    })
}

Do I understand that correctly, the

crossinline

modifier assures that we can't call return in the lambda we are passing to this function? Because if we don't call

return

here, we would leave the

onQueryTextChange

method.

Is there a channel I can go to for compiler plugin development (knowing that yes, it is an unsupported api)?

why does (0..0).forEach { ... } iterate once instead of zero times?