On Mon May 12 09:00:35 2008, pmichaud wrote:
> On Mon, May 12, 2008 at 08:07:27AM -0700, Will Coleda via RT wrote:
> > On Sat May 10 20:42:00 2008, tene wrote:
> > > 19:13 <@pmichaud> so, set $PX[$PX], $PX   is calling set_pmc_keyed 
even
> > > when [$PX] is an Integer
> > > 19:31 <@pmichaud> $P1[$P2]  always invokes  get_pmc_keyed and 
never
> > > get_pmc_keyed_int
> > > 19:31 <@pmichaud> that's a bit of a problem for aggregates that 
have
> > > both hash keys and int keys
> > > 
> > > There is some inline PIR in rakudo that needs to be removed when 
this is
> > > fixed.  Check r27433.
> > 
> > I think it's arguable that the current behavior is correct; If you 
want
> > to use the keyed int syntax, isn't:
> > 
> > $I0 = $P1
> > $P2[$I0]
> > 
> > the proper way to trigger that?
> 
> Yes.  But this also means that we have to have separate PAST node
> types for keyed versus keyed_int.  That's not necessarily a bad thing,
> but it does increase the amount of code a bit.
> 
> (And yes, this is the approach I'm planning to adopt for the time 
being.)
> 
> Pm
> 

It looks like you have a workaround here (box to int, keyed access with 
that), so closing the original RT.

If you have a new/related issue or an RFC on how to improve the syntax, 
please open a new trac at https://trac.parrot.org/

Thanks.
-- 
Will "Coke" Coleda
_______________________________________________
http://lists.parrot.org/mailman/listinfo/parrot-dev

Reply via email to