Melvin Smith <[EMAIL PROTECTED]> writes:
> At 09:23 AM 4/10/2002 +0100, Piers Cawley wrote:
>>Okay, this is the beginnings of Scheme in Perl6. I'm sure there's
>>stuff I'm getting wrong. I've not written the parser yet for instance
>
> Very nice! Quite a sample, maybe Larry/Damian can use this
> in one of the next $(A,E)'s
>
>
>> my SchemeExpr $.value;
>
> I haven't been keeping up in the back, I've a wedding bearing down on me.
>
> What is the significance of the . in the declaration?
class Class {
my $class_variable;
my $.instance_variable;
...
}
Easy eh?
> I think I paid attention enough to know a little about the unary dot
> but I'm still confused. We are able to use .foo to mean self.foo,
> but I would assume foo would be declared with my Foo $foo, not my
> Foo $.foo ?
>
>> method car { .value.key }
>> method cdr { .value.value }
>
> Maybe its the C++ in me but why the use of the unary . inside methods
> of the current class who's scope includes C<value> already?
Consider
class SpecializedPair is SchemePair {
method value {...}
}
If you've written 'cdr' without the unary . it will attempt to
dispatch to a *subroutine* in the same context, ie
SchemePair::value. Even assuming that we play nice and allow confusion
between methods and subroutines (which I'm personally not keen on),
it's still apparent that, in the case of a SpecializedPair, car and
cdr would use the wrong &value.
> Isn't this like using C<this> in C++ from inside a non-static method?
Don't ask me. I know nothing about C++ -- Objective C (Looong ago),
Perl 5, Smalltalk and Ruby for me.
> I'll await your ruler on my knuckles, but overall; very impressed
> here.
Thanks. Wait for the next version though, I'm busy implementing
lexical scopes at the moment.
--
Piers
"It is a truth universally acknowledged that a language in
possession of a rich syntax must be in need of a rewrite."
-- Jane Austen?
