Why they did that in Scala?

Question

Here is my discovery. I've found that in Scala I can simply write such expression:

val n: Int = (((2: Int): Int): Int): Int

and it'll 100% valid. But I have no idea why Scala creators did that? What is the point of this semantic? And are there some reasons for that kind of weirdness?

PS

I can assume that such approach is used for type inference to help compiler figure out what is going on. Or maybe there are some other cases?

This is valid C++ too: int n = ((int)((int)((int)((int)2)))); — Ross Bencina, Sep 26 '15 at 14:28
Yeah, the same stuff in Java `int n = (int) (int) (int) (int) 42;` — Finkelson, Sep 26 '15 at 14:33
Seems in Scala it is just a kind of Java legacy. And they just can not getting rid of it. — Finkelson, Sep 26 '15 at 14:34
I don't think it's a problem, it's just that type annotations can be attached to any expression. — Ross Bencina, Sep 26 '15 at 14:39
I don't understand why you think such an expression should be invalid. — Madoc, Sep 26 '15 at 15:48

score 6 · Accepted Answer · edited May 23 '17 at 12:22

6

Answer

Yes, the syntax is valid.
What you are doing is called "type ascription", telling the compiler what type you expect out of an expression. (see What is the purpose of type ascriptions in Scala? )
Where do you see any problems? AnInt (2 in your case) is an Int (and stays an Int). Even if you repeat that multiple times via ( (previous : ...) : Int ) it still stays an Int.

edited May 23 '17 at 12:22

Community

1
1

answered Sep 26 '15 at 14:39

Martin Senne

5,939
6
30
47

Why they did that in Scala?

1 Answers1

Answer