"Michael A. Chase" <[EMAIL PROTECTED]> writes:
> Here's a patch that will let HTML::Treebuilder work for now.
Good to see that somebody is awake when I'm asleep :-)
> It looks like we need to add an initializer to HTML::Parser or HTML::Element
> to allow one to initialize objects of the other's class. Adding the
> initializer to HTML::Element would probably be easier.
Yes, separating the object allocation from the initialization is
probably a good idea in general. Especially when multiple inheritance
comes to play.
package A; # and similar for B
sub new
{
my $class = shift;
my $self = bless {}, $class;
return $self->init(@_);
}
sub init
{
my($self, %arg) = @_;
$self->{some_field} = "foo";
#...
$self;
}
package AB;
use base qw(A B);
sub init
{
my($self, %arg) = @_;
$self->A::init(%arg);
$self->B::init(%arg);
$self;
}
Given that we have not done it before, I am not sure we really want to
do it now. Copying the fields like both your and mine patch did works
good enough I think.
> Do we have a memory leak when a reblessed reference to a HTML::Parser object
> is reassigned?
I am not sure what you are thinking of here.
We get a memory leak if the HTML::Parser destructor is not called
somehow. This might be caused by a subclass overriding DESTROY, but
not calling SUPER::DESTROY or by reblessing the object to some class
that don't @ISA HTML::Parser. This should probably have been
explained in the SUBCLASSING section of the HTML::Parser POD.
Regards,
Gisle
