Front page | perl.perl6.language |
Postings from May 2003
On Fri, May 02, 2003 at 07:09:02AM -0600, Luke Palmer wrote:
> After reading Dan's recent blog entry, I realized how (semantically)
> ugly coroutines are, with all the different ways of calling them, and
> dealing with parameters, etc.
>
> There's a couple problems as far as what's easy with the approaches
> described there. For instance, you can't really have a coroutine that
> everyone can call with the same state, because it would either start
> over, or create a new one.
>
> It's very hard to maintain several coroutine states of the same
> coroutine in the same scope. You have to do tricky stuff with
> closures, or some other such workaround.
>
> So, since they are generally used as iterators, why not steal the
> iterator semantics for them? This clears up most of the issues:
>
> sub range($start, $end) is coroutine {
> for $start .. $end { yield $_ }
> }
>
> my $iter = range(1,10); # Doesn't actually run range's code.
> for <$iter> {
> print "Got $_"
> }
>
> So calling a function declared as a coroutine wouldn't run anything at
> all. It would return an iterator, that when advanced would run until
> the first C<yield>.
>
> A problem that I see is that this breaks the "encapsulation" of
> continuing a coroutine being equivalent to just calling a function.
> However, coroutines are stateful, so it should be explicit that you're
> using an existing state rather than starting a new one (as in the case
> of a function call).
I don't like this. Iterators can be implemented in terms of
continuations which can be implemented in terms of coroutines.
(Actually, what Dan has described in his blog entry is a continuation in
my book; I'd be glad if someone show me that I am wrong)
So what you suggest is essentially a needless limitation. Even perl5
coroutines (Coro and Coro::*) can do more at this point!
\Anton.
--
Perl is strongly typed, it just has very few types. -- Dan Sugalski
Thread Previous
|
Thread Next