Scala example source code file (ApplyUsage.scala)

This example Scala source code file (ApplyUsage.scala) is included in the alvinalexander.com "Java Source Code Warehouse" project. The intent of this project is to help you "Learn Scala by Example" ^TM.

Learn more about this Scala project at its project page.

Java - Scala tags/keywords

apply, applyusage, dlist, int, list, logged, none, some, string, vector, writer\,dlist

The ApplyUsage.scala Scala example source code

package scalaz.example

object ApplyUsage extends App {
  import scalaz.Apply
  import scalaz.std.option._
  import scalaz.std.list._
  import scalaz.std.string._
  import scalaz.std.anyVal._
  import scalaz.std.vector._
  import scalaz.std.tuple._
  import scalaz.syntax.equal._
  import scalaz.syntax.std.option._

  // Apply extends the (hopefully familar) Functor Typeclass by adding
  // a method named "ap" which is similar to "map" from Functor in
  // that it applies a function to values in a context, however with
  // ap, the function is also in the same context. Here are some
  // examples, contrasted with map

  val intToString: Int => String = _.toString
  val double: Int => Int = _ * 2
  val addTwo: Int => Int = _ + 2

  // map
  assert(Apply[Option].map(1.some)(intToString) === "1".some)
  assert(Apply[Option].map(1.some)(double) === 2.some)
  assert(Apply[Option].map(none)(double) === none)

  // ap
  assert(Apply[Option].ap(1.some)(some(intToString)) === "1".some)
  assert(Apply[Option].ap(1.some)(some(double)) === 2.some)
  assert(Apply[Option].ap(none)(some(double)) === none)
  assert(Apply[Option].ap(1.some)(none[Int => Int]) === none[Int])
  assert(Apply[Option].ap(none)(none[Int => Int]) === none[Int])
  assert(Apply[List].ap(List(1,2,3))(List(double, addTwo)) === List(2,4,6,3,4,5))

  // from these two methods (map and ap) we are able to derive some
  // very useful methods which allow us to "lift" a function of
  // multiple arguments into a context. There are methods named
  // by how many parameters they take
  val add2 = ((_:Int) + (_:Int))
  val add3 = ((_:Int) + (_:Int) + (_:Int))
  val add4 = ((_:Int) + (_:Int) + (_:Int) + (_:Int))
  assert(Apply[Option].apply2(some(1), some(2))(add2) === some(3))
  assert(Apply[Option].apply3(some(1), some(2), some(3))(add3) === some(6))
  assert(Apply[Option].apply4(some(1), some(2), some(3), some(4))(add4) === some(10))

  // the effects from the context we are operating on are carried
  // through the computation, so, for example, in the case of the
  // Option Apply instance here, if any of the arguments are None, the
  // result of the entire computation is None:
  assert(Apply[Option].apply3(some(1), none, some(3))(add3) === None)

  // tuple2, tuple3 etc, will construct tuples from values from the
  // provided contexts
  assert(Apply[List].tuple3(List(1,2,3), List("a", "b"), List(())) ===
           List((1,"a",()),(1,"b",()),(2,"a",()),
                (2,"b",()),(3,"a",()),(3,"b",())))


  // There some helpful syntax available for the Apply typeclass:
  import scalaz.syntax.apply._

  // <*> is syntax for the "ap" method
  val plus1: Int => Int = _ + 1
  val plus2: Int => Int = _ + 2
  assert(List(1,2,3) <*> List(plus1, plus2) === List(2,3,4,3,4,5))

  // |@| is refered to as "applicative builder", it allows you to
  // evaluate a function of multiple arguments in a context, similar
  // to apply2, apply3, apply4, etc:
  assert((some(1) |@| some(2) |@| some(3))(_ + _ + _) === Some(6))
  assert((some(1) |@| none[Int] |@| some(3))(_ + _ + _) === None)

  def add8(a: Int, b: Int, c: Int, d: Int, e: Int, f: Int, g: Int, h: Int) = a+b+c+d+e+f+g+h
  val someOf8Options = (1.some |@| 2.some |@| 3.some |@| 4.some |@|
                        5.some |@| 6.some |@| 7.some |@| 8.some)(add8 _)
  assert(someOf8Options === 36.some)

  // the applicative builder created by |@| also has a "tupled" method
  // which will tuple the arguments
  assert((List(1,2,3) |@| List("a","b","c")).tupled ===
           List(1 -> "a", 1 -> "b", 1 -> "c",
                2 -> "a", 2 -> "b", 2 -> "c",
                3 -> "a", 3 -> "b", 3 -> "c"))


  // there are ^, ^^, ^^^, etc methods which correspond respectively
  // to apply2, apply, apply4, etc.
  assert(^(1.some, 2.some)(_ + _) === 3.some)
  assert(^^^(1.some, 2.some, 3.some, 4.some)(_ + _ + _ + _) === 10.some)
  assert(^^^(1.some, 2.some, none[Int], 4.some)(_ + _ + _ + _) === none)

  // sometimes we will want to apply a function in a context in order
  // to preserve the effects of running the function in a context, but
  // are not interested in the result of the function, for this there
  // are two combinators, *> which discards the value on the left, and
  // <* which discards the value on the right.

  // as an example we'll use "Writer", which a context which performs
  // a computation while accumulating a "side-value". Very commonly,
  // writer is used to emit log messages along with the value being
  // computed.

  import scalaz.{Writer,DList}
  import scalaz.syntax.writer._
  type Logged[A] = Writer[DList[String], A]

  // log a message, return no results (hence Unit)
  def log(message: String): Logged[Unit] = DList(message).tell

  // log that we are adding, and return the results of adding x and y
  def compute(x: Int, y: Int): Logged[Int] =
    log("adding " + x + " and " + y) as (x+y)

  // we log a message "begin", we add two numbers, we log "end",
  // neither calls to "log" compute a value, they are only evaluated
  // for the "effect" of logging a message, so we can use *> and <* to
  // discard the computed value (which in this case is Unit), while
  // preserving the value computed in the call to "compute"
  def addAndLog(x: Int, y: Int): Logged[Int] =
    log("begin") *> compute(x,y) <* log("end")

  val (written,result) = addAndLog(1,2).run
  assert(written === DList("begin", "adding 1 and 2", "end"))
  assert(result === 3)

  val (written2,result2) = addAndLog(1, 10).run
  assert(written2 === DList("begin", "adding 1 and 10", "end"))
  assert(result2 === 11)

  // Apply instances can be composed, which allows us to lift a
  // function into a computation in multiple nested contexts, while
  // applying the effects of all contexts:
  val applyVLO = Apply[Vector] compose Apply[List] compose Apply[Option]

  val deepResult =
    applyVLO.apply2(Vector(List(1.some, none[Int]),
                           List(2.some, 3.some)),
                    Vector(List("a".some, "b".some, "c".some)))(_.toString + _)

  val expectedDeep = Vector(List(Some("1a"), Some("1b"), Some("1c"),
                                 None, None, None),
                            List(Some("2a"), Some("2b"), Some("2c"),
                                 Some("3a"), Some("3b"), Some("3c")))

  assert(deepResult === expectedDeep)

}