Wrote a blog about the problems with FP today: https://blog.jdriven.com/2022/10/the-problem-with-functional-programming/ (dont worry 😛 its not that negative)

Is there a way to match on two boolean variables at the same time with a single

when

statement? Aka

when (Pair(x, y)) {
  Pair(true, false) -> ...
}

Is there a way to only match part of a

Pair

match?

val mimes = rec.mimeType.split("/")
when (mimes[0] to mimes[1]) {
  "text" to "vcard" -> {}
  "video" to _ -> {}
}

^ the second one isn't really working out.

In Rust I can do this:

struct Circle {
    radius: u32,
}

impl Circle {
    fn new(radius: u32) -> Circle {
        Circle { radius }
    }
}

trait Shape {
    fn area(&self) -> f64;
    fn perimeter(&self) -> f64;
}

impl Shape for Circle {
    fn area(&self) -> f64 {
        std::f64::consts::PI * (self.radius as f64).powi(2)
    }

    fn perimeter(&self) -> f64 {
        2.0 * std::f64::consts::PI * (self.radius as f64)
    }
}

fn pretty_print<T: Shape>(shape: T) {
    println!("Area: {}", shape.area());
    println!("Perimeter: {}", shape.perimeter());
}

fn main() {
    let circle = Circle::new(1);

    pretty_print(circle);
}

So the trait is like a Kotlin Interface. And now my function pretty_print can take any type(struct) that implements Shape. This seems like a very powerful feature to me as it means I can extend types. How can I achieve the same thing in Kotlin?

Is there a name for this patter in functional programming? What would you name the function?

fun ((A, B) -> Unit).whatShouldThisBeNamed(b: B): (A) -> Unit {
        return { a -> this(a, b) }
    }

(A function that returns a simpler function by providing one of the parameters to the first function)

Hi, lets say I have a sequence of objects (for simplicity I use strings):

val s = sequenceOf("Lorem", "ipsum", "dolor", "sit", "amet")

What’s the best way to limit elements so total size is smaller than desired. For example:

val s2 = s.someFunction(maxSize = 17) { it.length }
println(s2.toList()) // [Lorem, ipsum, dolor]

@elizarov Testing using context receivers for type classes (as claimed in the KEEP should be possible?) Why isn't this working?:

interface ServiceLocator<I> {
//    fun getRealImpl(): I
    fun getEmulatedImpl(): I
//    fun getStubbedImpl(): I
}

context(ServiceLocator<I>)
inline fun <reified I> locateEmulated(): I = getEmulatedImpl()

object StringServiceLocator : ServiceLocator<String> {
    override fun getEmulatedImpl(): String = "test"
}

object LongServiceLocator : ServiceLocator<Long> {
    override fun getEmulatedImpl(): Long = 5
}

object StringLongLocatorFactory: ServiceLocatorFactory {
    // In the future (scope properties): with val ...
    inline fun <reified I> locate(): I {
        with(LongServiceLocator) {
            with(StringServiceLocator) {
                return locateEmulated<I>()
            }
        }
    }
}

No required context receiver found: Cxt { context(ServiceLocator<I>) public inline fun <reified I> locateEmulated(): I defined in [...] in file ServiceLocatorFactory.kt[SimpleFunctionDescriptorImpl@1aa4cbc9] }

Working through category theory for programmers and doing the exercises in Kotlin. What is wrong with this memoize function?

fun <A, B> memoize (f: (A) -> B) = { a: A ->
    val cache = mutableMapOf<A, B>()
    cache.computeIfAbsent(a) { f(a) }
}

I've got a problem statement where it would be really helpful to be able to match on about 4 variables (incoming states from different subsystems, and a requested state, and possibly even previous states), are there some compiler plugins / good workarounds for that?

I started a Discord server for folks working through the book FP Made Easier by C Scalfani. Here's an invite (good for 7 days).

Hi buddies. Do you have any sample code about apply FP/Arrow to CleanArchitect Android app ?

I’m thinking about creating a tool that would promote a more functional approach in certain contexts, like forbid using

var

or

for

loops (mutability in general), with a way to opt out in certain cases. Think: a step towards more pure FP. I’m wondering if you’d be keen to use it and, if yes, what approach you’d suggest to implement it. I’m thinking about a detekt rule or a compiler plugin that would emit an error or a warning (configurable)

Hi all! I'm reading a huge file/object from a remote source in bunches... send fragments of the file thru a Flow and wanna process complete lines one by one. I got below implementation... but I was wondering if there is a way to get rid of the var and get a more functional solution. Thanks in advance.

fun main() = runBlocking {
    var lastline = ""
    readHugeFile().collect { value ->
        val lines = (lastline+value).lines()
        lines.dropLast(1).forEach {
            println(it)
        }
        lastline = lines.last()
    }
    println(lastline)
}

fun readHugeFile(): Flow<String> = flowOf(
    """line 1
        |line 2
        |li""".trimMargin(),
    """ne 3
        |line 4
        |lin""".trimMargin(),
    """e 5
        |line 6
        """.trimMargin()
)

Somebody was telling me that ArrowKt changes a lot... (maybe that was a conclusion he made having used it in previous, less stable versions?) does that mean there are often breaking changes? Is Arrow now stable in it's api? I know that there's always a will to improve as long as the language itself doesn't support all of the features a functional language would need...

Just watched your video on the Ktor + Arrow example @simon.vergauwen, it was one of the clearest explanations I saw! What got me really afraid was all the complex utility functions in that project... like TestResource and the Kotest integration... I was wondering if all that could be pulled out into some libraries that could be re-used? Also, wondering why if anyways you're using interfaces (when not using context receivers), why not just use the

class Foo : IFoo

instead of

fun Foo() = object : Foo

?

Little thought that probably wouldn't happen... if Kotlin ever decided to let

?.

and

?:

be overridable operators for a non-null type... then it could avoid all the `bind()`s and `flatMap`s and hacks that Arrow is doing... did someone ever propose such a thing? (or maybe some similar operator if people might find it too confusing and think that they're dealing with nullables.)

I have a function which makes an external call and returns nothing, but I need to do something based on if it succeeded or not. Is returning

Result<Unit>

the best thing to do here or is there anything else in the std library?

@simon.vergauwen It seems like Ktor uses

Runtime.getRuntime().addShutdownHook(hook)

, is that only SIGINT? It seems like in your video it was clearly using SIGINT, but in my current version not...

how can i get the stacktrace of an error in functional style? throwing exception is not a right thing in FP.

Recently I've become a fan of explicit error handling using Result<V, E> (using https://github.com/michaelbull/kotlin-result). Yesterday I was shocked by the thought that handling errors this way somehow reminds me of Java's checked exceptions, with different syntax 😐 Is this analogy really correct? EDIT: I know that in Java some things are not possible like nice chaining of function calls. But in Kotlin it would be possible because of the language's "almost everything is an expression" paradigm.

I'm experimenting with Arrow's

raise()

functionality and I try to find a library function similar to the following hypothetical

orElse()

:

context(Raise<String>)
fun f1() = 1

context(Raise<String>)
fun f2(): Nothing = raise("error")

suspend inline fun <R, A> Effect<R, A>.orElse(noinline block: suspend (R) -> A) =
    fold(
        recover = block,
        transform = ::identity
    )

val v1 = effect { f1() }.orElse { 33 }
assertEquals(1, v1)

val v2 = effect { f2() }.orElse { 33 }
assertEquals(33, v2)

I use arrow 1.1.6-alpha.36. Thanks.

I have been writing a lot of Scala 3 recently and it really has made me wish Kotlin had a bit less brackets! Kotlin feels so noisy now. Especially when writing in a functional style

Anyone here doing functional Kotlin on Android? I am interested to hear what wider architecture you are using.

Is there a preferred way to approach companion function with a functional style ?

data class SomeData(val someValue: String) {
  companion object {
    fun someFunction() {}
  }
}

// OR

data class SomeData(val someValue: String) {
  companion object
}

fun SomeData.Companion.someFunction() {}

I have 2 functions which return Results.

functionOne: Result<Something>

functionTwo: Result<Unit>

I’ll only call functionTwo if the first one returns a Success, and return a

Result<Something>

in case both of them return a Success. What’s the best way to do it using std lib? I thought the nested folds and maps and ifs turned out a bit ugly

I finally had some time to experiment with my very old 🙂 idea of mutating immutable object graphs using Kotlin (with KSP help). I've created a short readme about its usage, I'd like to hear your opinions about the idea itself (eg. whether it is worth to continue development or not): https://github.com/kotlinw/kotlinw/tree/main/kotlinw-immutator

Hey, wondering if anyone has best practices tips / examples for implementing a "Tool" (i.e. asynchronous user workflow with a specific start / end, as opposed to something with an open-ended lifecycle like a view) in a MVU/Elm style using Kotlin coroutines. We currently have a sealed class "action" interface for representing actions the user can perform to interact with the tool, and a reducer function

(State, Action) -> State

, but my question was more about how to actually action this reducer to do stuff with the UI. Currently we are taking in a Flow of Actions, an initial state, and updating that state (as a StateFlow) by launching a collect block on the action flow to update the state. The body of the collect statement on the flow is synchronized with a mutex so actions should (I was hoping anyway) get executed in order sequentially. In practice, this seems to work. However, I've noticed in tests this sometimes fails, which makes me wonder if my setup of this is actually optimal. I was thinking maybe a channel of user actions might be more robust, but open to ideas!

My book on functional Kotln and DDD is now available on Amazon: https://www.amazon.co.uk/Objects-Functions-Software-Functional-Programming/dp/1680508458

Does kotlin have a

reduce

where I can pass the initial

acc

?

... where does the default

S

of the result even come from? O.o