Λrrow Fx

Project Reactor

Arrow aims to enhance the user experience when using Project Reactor. While providing other datatypes that are capable of handling effects, like IO, the style of programming encouraged by the library allows users to generify behavior for any existing abstractions.

One of these abstractions is Project Reactor, a library that, like RxJava, offers reactive streams.

val flux = Flux.just(7, 4, 11 ,3)
    .map { it + 1 }
    .filter { it % 2 == 0 }
    .scan { acc, value -> acc + value }
    .collectList()
    .subscribeOn(Schedulers.parallel())
    .block()
//[8, 20, 24]

Integration with your existing Flux chains

The largest quality of life improvement when using Flux streams in Arrow is the introduction of the Monad Comprehension. This library construct allows expressing asynchronous Flux sequences as synchronous code using binding/bind.

Arrow Wrapper

To wrap any existing Flux in its Arrow Wrapper counterpart, you can use the extension function k().

import arrow.fx.reactor.*
import reactor.core.publisher.*

val flux = Flux.just(1, 2, 3, 4, 5).k()
flux
// FluxK(flux=FluxArray)

val mono = Mono.just(1).k()
mono
// MonoK(mono=MonoJust)

You can return to their regular forms using the function value().

flux.value()
// FluxArray

mono.value()
// MonoJust

Observable comprehensions

The library provides instances of MonadError and MonadDefer.

Async allows you to generify over datatypes that can run asynchronous code. You can use it with FluxK or MonoK.

fun <F> getSongUrlAsync(MS: MonadDefer<F>) =
  MS { getSongUrl() }

val songFlux: FluxKOf<Url> = getSongUrlAsync(FluxK.monadDefer())
val songMono: MonoKOf<Url> = getSongUrlAsync(MonoK.monadDefer())

MonadThrow can be used to start a Monad Comprehension using the method fx.monadThrow, with all its benefits.

Let’s take an example and convert it to a comprehension. We’ll create an observable that loads a song from a remote location, and then reports the current play % every 100 milliseconds until the percentage reaches 100%:

getSongUrlAsync()
  .map { songUrl -> MediaPlayer.load(songUrl) }
  .flatMap {
    val totalTime = musicPlayer.getTotaltime()
    Flux.interval(Duration.ofMillis(100))
      .flatMap {
        Flux.create { musicPlayer.getCurrentTime() }
          .map { tick -> (tick / totalTime * 100).toInt() }
      }
      .takeUntil { percent -> percent >= 100 }
  }

When rewritten using fx.monadThrow, it becomes:

import arrow.fx.reactor.*
import arrow.typeclasses.*
import arrow.fx.reactor.extensions.fluxk.monadThrow.monadThrow

FluxK.monadThrow().fx.monadThrow {
  val (songUrl) = getSongUrlAsync()
  val musicPlayer = MediaPlayer.load(songUrl)
  val totalTime = musicPlayer.getTotaltime()

  val end = DirectProcessor.create<Unit>()
  Flux.interval(Duration.ofMillis(100)).takeUntilOther(end).bind()

  val (tick) = musicPlayer.getCurrentTime()
  val percent = (tick / totalTime * 100).toInt()
  if (percent >= 100) {
    end.onNext(Unit)
  }
  percent
}

Note that any unexpected exception, like ArithmeticException when totalTime is 0, is automatically caught and wrapped inside the flux.

Subscription and cancellation

Flux streams created with comprehensions like fx.monadThrow behave the same way regular flux streams do, including cancellation by disposing the subscription.

val disposable =
  songFlux.value()
    .subscribe({ println("Song $it") }, { System.err.println("Error $it") })

disposable.dispose()

Stack safety

While MonadDefer usually guarantees stack safety, this does not apply for the reactor wrapper types. This is a limitation on reactor’s side. See the corresponding GitHub issue.

To overcome this limitation and run code in a stack in a safe way, one can make use of fx.stackSafe, which is provided for every instance of Monad when you have arrow-free included.

import arrow.Kind
import arrow.fx.reactor.MonoK
import arrow.fx.reactor.ForMonoK
import arrow.fx.reactor.fix
import arrow.fx.reactor.extensions.monok.monad.monad
import arrow.free.stackSafe

fun main() {
  //sampleStart
  // This will not result in a stack overflow
  val result = MonoK.monad().fx.stackSafe {
    (1..50000).fold(just(0)) { acc: Kind<ForMonoK, Int>, x: Int ->
      just(acc.bind() + 1)
    }.bind()
  }.run(MonoK.monad())
  //sampleEnd
  println(result.fix().mono.block()!!)
}

import arrow.fx.IO

// This will result in a stack overflow
IO {
  MonoK.monad().fx.monad {
    (1..50000).fold(just(0)) { acc: Kind<ForMonoK, Int>, x: Int ->
      just(acc.bind() + 1)
    }.bind()
  }.fix().mono.block()
}.attempt().unsafeRunSync()
// Left(java.lang.StackOverflowError)