Re: [Vala] Boxed type generics

[Tal Hadad <tal_hd hotmail com>]

Even known it's doesn't sound like that, and it's work in .Net, double? is always
a pointer type. It's the same as double*. This explains the first example error.
In the second one, you convert some type to double* and then to double(in the
deceleration "double tmp"). The compiler seems to ignore the first conversion,
and skip to the second.

In conclusion, use (double) instead of (double?). If you want to achieve a null
passing in by value(i.e. not pointer), pass another Boolean which tells if it's
empty.

Yours,
Tal

> Date: Mon, 19 Mar 2012 13:59:33 +0100
> From: stephan ecshi net
> To: vala-list gnome org
> Subject: [Vala] Boxed type generics
>
> Hello,
>
> I'd like to do something like the following (somewhat oversimplified):
>
> public T increment<T> (T number) {
>         return (double?) number + 1;
> }
> public static void main () {
>         stdout.printf("%f\n", increment<double?> (4.7));
> }
>
> Unfortunately the C-compiler complains about not being able to convert
> the double to a pointer type in the return statement.
> I have found out this works as expected:
>
> public T increment<T> (T number) {
>         double tmp = (double?) number;
>         tmp += 1;
>         return &tmp;
> }
>
> This, however, seems not to be the right thing to do. Could someone
> explain how to handle those boxed types correctly without resorting to
> the address-of-operator?
>
> Thanks,
>
> Stephan
>
> _______________________________________________
> vala-list mailing list
> vala-list gnome org
> http://mail.gnome.org/mailman/listinfo/vala-list