Consider the following code:
class A { def print = println("A") }
class B extends A { override def print = println("B") }
def foo(implicit a: A) = a.print
def bar(implicit a: A) = {
implicit val b = new B
foo
}
bar(new A) // B
I am wondering why calling foo
in bar
isn't raising an ambiguous implicit values
error. Of course
implicit val b: A = new B
will raise that error. Why does foo
pick the implicit b
and not the implicit a
? Or even more general: What are the rules what will be picked?
EDIT:
Due to my comment-conversation with Ivan I want to clarify: I would know the answer to my question if I named the local implicit val the same way as the implicit method parameter.
def bar(implicit a: A) = {
implicit val a = new B
foo
}
Then only the local val a
is in scope which scope-overrides the method parameter because they have the same name.
Note: I'm probably vastly oversimplying things, but in testing it seemed like the following.
It's because the 2nd one is in an inner scope, so it has precedence. Its the same thing that happens with
object test {
val a = 5
def test(i: Int) = {
val a = 6
i + a
}
}
In this case you would expect a
to be 6 inside the function. The following is similar.
object test {
implicit val i = 5;
{
implicit val b = 6;
test
}
def test(implicit ii:Int) = println(ii)
}
Updated from comment.
scala> def test(a: Int) = {val a = 5; a }
test: (a: Int)Int
scala> test(6)
res1: Int = 5
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With