• "Bent Rasmussen" <exo bent-rasmussen.info> Jul 14 2004
• Stewart Gordon <smjg_1998 yahoo.com> Jul 15 2004
• "Bent Rasmussen" <exo bent-rasmussen.info> Jul 15 2004

"Bent Rasmussen" <exo bent-rasmussen.info> writes:

dmd test.d




struct test
{
    bit x;
    bit opIndexAssign(bit v, uint i)
    {
        return x = v;
    }
    void ex()
    {
        this[0] = true; // oops
    }
}

The error does not manifest if test is a class.

DMD 0.95 XP SP1

Stewart Gordon <smjg_1998 yahoo.com> writes:

Bent Rasmussen wrote:

 dmd test.d


 test.d(15): cannot implicitly convert int to test


There aren't 15 lines in your code.  So we have yet another line 
numbering bug?

 struct test
 {
     bit x;
     bit opIndexAssign(bit v, uint i)
     {
         return x = v;
     }
     void ex()
     {
         this[0] = true; // oops
     }
 }
 
 The error does not manifest if test is a class.


I think it's an issue with this in struct actually.

In a struct, this means the address of the current object.  Don't ask me 
why.

	(*this)[0] = true;

should work.

-- 
My e-mail is valid but not my primary mailbox, aside from its being the 
unfortunate victim of intensive mail-bombing at the moment.  Please keep 
replies on the 'group where everyone may benefit.

"Bent Rasmussen" <exo bent-rasmussen.info> writes:

 There aren't 15 lines in your code.  So we have yet another line
 numbering bug?


No. I just sliced the relevant piece of the code to make the case shorter,
and inserted a comment on the interesting line. I didn't figure this was
relevant

module test;

...

void main
{
}

 I think it's an issue with this in struct actually.

 In a struct, this means the address of the current object.  Don't ask me
 why.

 (*this)[0] = true;

 should work.


Thanks. Excuse the noise. :-)