Why does the following code compile?

	import std.stdio;
	int f(ref int x) {
		return x++;
	}
	class A {
		int x=123;
		int g() const {
			return f(x);
		}
	}
	void main() {
		auto a = new A;
		writeln(a.g());
		writeln(a.g());
	}

Shouldn't the const member g() be prohibited from passing a ref to a
member variable to f()?

But this code not only compiles, it outputs:

	123
	124

So I've managed to call a const member of A to alter the value of A.x? Am I misunderstanding the meaning of const when applied to a member function, or is this a compiler bug?

P.S. I'm using gdc-4.6; does dmd also have this behaviour?


T

-- 
Lawyer: (n.) An innocence-vending machine, the effectiveness of which depends on how much money is inserted.

On Fri, 13 Jan 2012 15:27:56 -0500, H. S. Teoh <hsteoh@quickfur.ath.cx> wrote:

> Why does the following code compile?
>
> 	import std.stdio;
> 	int f(ref int x) {
> 		return x++;
> 	}
> 	class A {
> 		int x=123;
> 		int g() const {
> 			return f(x);
> 		}
> 	}
> 	void main() {
> 		auto a = new A;
> 		writeln(a.g());
> 		writeln(a.g());
> 	}
>
> Shouldn't the const member g() be prohibited from passing a ref to a
> member variable to f()?

Yes, it's a bug (I think already reported, let me check... Yep: http://d.puremagic.com/issues/show_bug.cgi?id=5493)

> But this code not only compiles, it outputs:
>
> 	123
> 	124
>
> So I've managed to call a const member of A to alter the value of A.x?
> Am I misunderstanding the meaning of const when applied to a member
> function, or is this a compiler bug?
>
> P.S. I'm using gdc-4.6; does dmd also have this behaviour?

Yes, the pull request hasn't been merged yet.

BTW, I tested with 2.057, which I downloaded from github.  Holy CRAP, downloaded in less than 1 second!!!

Awesome :)

Can we archive all the old DMD downloads there?

-Steve