Scala FRP

This library replaces the idea of Publishers and Subscribers with EventStreams. An EventStream can be treated like a collection of events; it can be transformed similarly to any other Scala collection, and instead of needing to call publisher.addEventListener(new EventListener(){...}), you simply call eventStream.foreach{...}

Background

Scala FRP (stands for Functional Reactive Programming) is a library inspired by Ingo Maier's paper, Deprecating the Observer Pattern. Ingo Maier made an implementation of his "scala.react" framework which is available on Github in its original form. I also made a version of the same library that works with SBT to manage its dependencies, available here.

Getting it

Add the following to your build.sbt file:

libraryDependencies += "io.dylemma" %% "scala-frp" % "1.2"

Example Usage

import io.dylemma.frp._

// Mix in `Observer` for free memory management.
object Example extends App with Observer {

    // Create a source of events.
    val ints = EventSource[Int]

    // You can derive new event streams any other event stream.
    val evenInts = ints.filter{ _ % 2 == 0 }
    val intsWithIndices = ints.zipWithIndex
    val soonInts = ints.before(2 seconds fromNow)

    // Attach event handlers.
    ints.foreach{ x => println(s"An int: $x") }
    evenInts.foreach { x => println(s"Even int: $x") }
    intsWithIndices.foreach { case (x, i) => println(s"$ith int was $x") }
    soonInts.foreach { x => println(s"$x came soon enough") }

    // Fire events!
    ints fire 1
    ints fire 2
    ints fire 3

}

FRP-101

Attaching event handlers requires an implicit Observer. The observer helps make sure that no cyclical references are made between the EventStream and the handler function; it keeps a weak reference to the handler so that the handler function may be garbage-collected once the Observer is garbage-collected. You can get an implicit Observer one of two ways:

Make one yourself: implicit object myObserver extends Observer
Mix it into the containing class, like in the example above. The Observer trait has an implicit reference to itself.

With that requirement out of the way, you're ready to start! There are two main classes that you will want to interact with.

EventStream is a read-only class that you can attach event handlers to, and create mappings and combinations with.
EventSource is a concrete implementation of EventStream that also exposes fire and stop methods.

FRP-102

EventStreams are finite. At a low level, they emit events as an Event[A], which can either be a Fire(item) or a Stop. When you call foreach on a stream, it will receive all Fire events until the stream emits a Stop. You can attach to these lower-level events with the sink(handler: Event[A] => Boolean) method. The handler will remain attached until it returns false in response to an event.

Because of the concept of finite EventStreams, streams are able to be concatenated; handlers may be attached that simply wait for the last event from the stream, or the eventual Stop.

For further details and a full list of capabilities, check out the docs

dylemma / scala.frp

readme