DMD 1.033:

import tango.io.Stdout;

void a(void delegate() dg)
{
	dg();
}

void b(char[][] bb)
{
	Stdout(bb);
}

void main(char[][] argv)
{
//why:
	{Stdout(argv);}.a(); //this is not working
	argv.b(); //and this is working?
}

Regards, bobef

Ultimately I want to be able to do something like this:

bool silent(T,ARGS...)(T delegate(ARGS) dgTry,ARGS args)
{
	try {
		dgTry(args);
		return true;
	}
	catch(Object o) return false;
}

delegate(whatever W){W.dosomething();}.silent(somevar);

Prototype like programming is very nice :)

Regards, bobef

"bobef" <bobef@nosmap-abv.bg> wrote in message news:g5o6c5$2is4$1@digitalmars.com...
> DMD 1.033:
>
> import tango.io.Stdout;
>
> void a(void delegate() dg)
> {
> dg();
> }
>
> void b(char[][] bb)
> {
> Stdout(bb);
> }
>
> void main(char[][] argv)
> {
> //why:
> {Stdout(argv);}.a(); //this is not working
> argv.b(); //and this is working?
> }

Because the "a.f() == f(a)" only works for arrays and associative arrays and not delegates.

Jarrett Billingsley Wrote:

> "bobef" <bobef@nosmap-abv.bg> wrote in message news:g5o6c5$2is4$1@digitalmars.com...
> > DMD 1.033:
> >
> > import tango.io.Stdout;
> >
> > void a(void delegate() dg)
> > {
> > dg();
> > }
> >
> > void b(char[][] bb)
> > {
> > Stdout(bb);
> > }
> >
> > void main(char[][] argv)
> > {
> > //why:
> > {Stdout(argv);}.a(); //this is not working
> > argv.b(); //and this is working?
> > }
> 
> Because the "a.f() == f(a)" only works for arrays and associative arrays and not delegates.
> 
> 

Obviously. And why is that? If it is working for one primitive type I see no reason it shouldn't work for the rest.

bobef wrote:
> Jarrett Billingsley Wrote:
> 
>> "bobef" <bobef@nosmap-abv.bg> wrote in message news:g5o6c5$2is4$1@digitalmars.com...
>>> DMD 1.033:
>>>
>>> import tango.io.Stdout;
>>>
>>> void a(void delegate() dg)
>>> {
>>> dg();
>>> }
>>>
>>> void b(char[][] bb)
>>> {
>>> Stdout(bb);
>>> }
>>>
>>> void main(char[][] argv)
>>> {
>>> //why:
>>> {Stdout(argv);}.a(); //this is not working
>>> argv.b(); //and this is working?
>>> }
>> Because the "a.f() == f(a)" only works for arrays and associative arrays and not delegates. 
>>
>>
> 
> Obviously. And why is that? If it is working for one primitive type I see no reason it shouldn't work for the rest.

It was originally just an unadvertised easter egg that it worked for any type at all.  It was not in the spec at all for a long time, but I guess since the experiment seemed to be a success, Walter decided to put it in.

Making it work for all types is on the to-do list for D2.0.  See WalterAndrei.pdf from the 2007 conference.  Or watch the video (!).

--bb

Grrr. Too bad things come down to Tango or D2, which is almost the choice to use any D library at all or D2, since most of them use Tango. This is bad situation.

Regards, bobef

Bill Baxter Wrote:

> bobef wrote:
> > Jarrett Billingsley Wrote:
> > 
> >> "bobef" <bobef@nosmap-abv.bg> wrote in message news:g5o6c5$2is4$1@digitalmars.com...
> >>> DMD 1.033:
> >>>
> >>> import tango.io.Stdout;
> >>>
> >>> void a(void delegate() dg)
> >>> {
> >>> dg();
> >>> }
> >>>
> >>> void b(char[][] bb)
> >>> {
> >>> Stdout(bb);
> >>> }
> >>>
> >>> void main(char[][] argv)
> >>> {
> >>> //why:
> >>> {Stdout(argv);}.a(); //this is not working
> >>> argv.b(); //and this is working?
> >>> }
> >> Because the "a.f() == f(a)" only works for arrays and associative arrays and not delegates.
> >>
> >>
> > 
> > Obviously. And why is that? If it is working for one primitive type I see no reason it shouldn't work for the rest.
> 
> It was originally just an unadvertised easter egg that it worked for any type at all.  It was not in the spec at all for a long time, but I guess since the experiment seemed to be a success, Walter decided to put it in.
> 
> Making it work for all types is on the to-do list for D2.0.  See WalterAndrei.pdf from the 2007 conference.  Or watch the video (!).
> 
> --bb

Some more problems on this matter. This tries to invoke TST.ASD instead of .ASD :

void ASD(char[] a,int b)
{
}

class TST
{
	this(){"asd".ASD(5);}
	void ASD(){}
}

"bobef" <bobef@nosmap-abv.bg> wrote in message news:g64mei$1vu4$1@digitalmars.com...
> Some more problems on this matter. This tries to invoke TST.ASD instead of .ASD :
>
> void ASD(char[] a,int b)
> {
> }
>
> class TST
> {
> this(){"asd".ASD(5);}
> void ASD(){}
> }

This is behaving correctly.  Remember that "asd".ASD(5) is just sugar for ASD("asd", 5).  Name lookup in D proceeds from the innermost scope, so the first ASD it finds, it tries to use.  It's exactly the same as when you try to do something like:

char[] toString()
{
    return toString(someMember) ~ ", " ~ toString(otherMember);
}

and the compiler complains about not using the right params to toString.