--- Kurt Starsinic <[EMAIL PROTECTED]> wrote:
> > similar to this solved a problem I had with Class::Trait
> > whereby I was exporting methods which should not have
> > been exported.
>
> Do you mean "calling methods which should not have been
> called?"
No. I did mean "which should not have been exported", but a thorough
explanation of traits and what they do won't fit into this margin.
> What I'm asking for, if you're interested in feedback,
> is an example of a situation where youvre imported a
> function into a class, from other code you've called
> can() on said class before invoking a method call, it
> did The Wrong Thing,
This is actually very easy to do in capability based systems. For
example, let's say I want the "clone" method from Clone.pm, but I want
to call it as a function:
package Foo;
use Clone 'clone';
sub new { bless {}, shift }
sub frobnicate {
my $self = shift;
my $args = clone([EMAIL PROTECTED]); # don't mess with their refs!
...
}
I've imported the clone function quite a few times but it's quite
reasonable for someone to do this:
my $foo = Foo->new;
# later
if ( $foo->can('clone') ) { # whoops
Yes, they can just call the clone() method directly and watch things
blow up but if they're testing an objects capabilities first, they get
a bit of sanity.
> You *are* suggesting making this a pragma under the very
> well-known "strict" namespace and putting it on CPAN, so I'm
> expressing my personal concern that it may not be as generally
> useful and safe as the nomenclature might suggest.
I agree that the name is ugly. I only used that after I saw Sean
Burke's "strict::ModuleName".
Namespace suggestions are quite welcome.
Cheers,
Ovid
--
If this message is a response to a question on a mailing list, please send
follow up questions to the list.
Web Programming with Perl -- http://users.easystreet.com/ovid/cgi_course/
