Scala, why when I pass a lambda function as argument to another function i cannot execute it - scala

Question

I am trying to create a concurrent method that saves in two different variables the result of two blocks of code task1 and task2:
def parallel[A, B](fun_a: => A, fun_b: => B): (A, B) = {
var res_a: A = 1.asInstanceOf[A]
var res_b: B = 1.asInstanceOf[B]
val hilo = new Thread {
override def run(): Unit = {
res_a = fun_a
}
}
val hilo1 = new Thread() {
override def run(): Unit = {
res_b = fun_b
}
}
hilo.start()
hilo1.start()
hilo.join()
hilo1.join()
(res_a, res_b)
}
val task1 = {3+4}
val task2 = {3*1}
println(parallel(task1, task2))
This is the correct answer in theory but I do not understand if val task1/task2 are functions or just the result of the block inside brackets. I tried to edit the method parallel to accept only generic values:
def parallel[A, B](fun_a: A, fun_b: B): (A, B)
and it still works fine. In this case, The computation of task1 and task2 is made before they pass as parameter to the function, or is made when the result is assigned to the variables res_a, res_b in each thread?
I tried to pass task1 and task2 as a method, it works fine but when I pass it as function, in the variable is saved the serialization of the function.
def task1():Int = {3+4} //method
val task2: Int => Int = (x:Int) => {3*x} //anonymous function
println(parallel(task1, task2))
i changed the code so the anonymous function gets executed but now i get an exception:
def parallel[A, B](fun_a: => A, fun_b: => B): (A, B) = {
var res_a: A = 1.asInstanceOf[A]
var res_b: B = 1.asInstanceOf[B]
val hilo = new Thread {
override def run(): Unit = {
res_a = fun_a
}
}
val hilo1 = new Thread() {
override def run(): Unit = {
res_b = fun_b(3) // here, because i think this is a function {val task2: Int => Int = (x:Int) => {3*x} i try to pass its argument
}
}
hilo.start()
hilo1.start()
hilo.join()
hilo1.join()
(res_a, res_b)
}
it says that B does not take parameters but i defined it before as a lambda that takes one parameter.
I do not understand why i defined the method as one that takes as arguments two functions but it does not let me pass arguments to that functions

Answer 1

The following code works:
def parallel[A, B](fun_a: => A, fun_b: Int => B): (A, B) = {
var res_a: A = 1.asInstanceOf[A]
var res_b: B = 1.asInstanceOf[B]
val hilo = new Thread {
override def run(): Unit = {
res_a = fun_a
}
}
val hilo1 = new Thread() {
override def run(): Unit = {
res_b = fun_b(3)
}
}
hilo.start()
hilo1.start()
hilo.join()
hilo1.join()
(res_a, res_b)
}
println(
parallel(
{ 3 * 3 },
(x: Int) => { 3 + x }
)
)
Note how the type declaration of fun_b is changed to fun_b: Int => B. In this case, fun_a doesn't do much, though.
And admittedly, the generic type doesn't make much sense, if Int is hard coded in there. So a more sensible solution would be something like this:
def parallel[A, B](fun_a: A => A, fun_b: B => B, a: A, b: B): (A, B) = {
var res_a: A = 1.asInstanceOf[A]
var res_b: B = 1.asInstanceOf[B]
val hilo = new Thread {
override def run(): Unit = {
res_a = fun_a(a)
}
}
val hilo1 = new Thread() {
override def run(): Unit = {
res_b = fun_b(b)
}
}
hilo.start()
hilo1.start()
hilo.join()
hilo1.join()
(res_a, res_b)
}
println(
parallel(
(y: Int) => { y * 3 },
(x: Int) => { x + 3 },
3,
3
)
)
println(
parallel(
(y: Int) => { y * 3 },
(x: String) => { x + 3 },
3,
"3"
)
)
println(
parallel(
(y: String) => { y * 3 },
(x: Int) => { x + 3 },
"3",
3
)
)
This will print:
(9,6)
(9,33)
(333,6)