Carl MÃsak writes:
> On Fri, 28 Jan 2005 11:38:51 -0700, Luke Palmer <[EMAIL PROTECTED]> wrote:
> > I don't think it's the cleanest solution, but it works.
>
> Just out of curiosity, what do you think would be a cleaner solution?
> And why would one not want to implement such a solution instead?
I suppose I was a bit terse. I believe it to be a dirtier part of the
the overall clean solution we currently have. To make that cleaner
would involve making something else dirtier, such as requiring too much
information at compile time.
Ultimately, $_ should work because it's referring to an outer lexical
pad; i.e. the compiler knows that $_ belongs to some outer block.
However, because of our dynamism, that would mean all of our closures
would have to be pointy or use $^vars for parameters, which outlaws such
very nice constructs as these:
@names = map { .name } @fields;
I don't think there's anything we can do to clean it, and really we're
just sweeping that strange default behavior under the rug, and nobody
will be the wiser to what's actually going on. It looks like it knows
what you're talking about even though it really doesn't.
Luke
> // Carl
>
> On Fri, 28 Jan 2005 11:38:51 -0700, Luke Palmer <[EMAIL PROTECTED]> wrote:
> > Juerd writes:
> > > Luke Palmer skribis 2005-01-28 9:31 (-0700):
> > > > > And can $_ be used instead of $^n?
> > > > Of course it can. You know that.
> > >
> > > I do?
> > >
> > > Can't say I understand well when a topic is implicitly defined and when
> > > not. It's obvious for for-loops and given, but everything else is
> > > blurry to me.
> >
> > Okay, I'll explain then. If you use $_ inside a closure, then it is
> > assumed to mean an argument to that closure. If the closure isn't given
> > an argument (or the block is declared to have zero arguments at compile
> > time) then $_ defaults to the outer lexical $_, which fixes this
> > problem:
> >
> > for @stuff {
> > if something() {
> > print; # prints the argument to if's block
> > }
> > }
> >
> > I don't think it's the cleanest solution, but it works.
> >
> > Luke
> >
>
