On Tue, 10 Jul 2001, Danny Yoo wrote:

>  Is there a way to get XML::Twig to not read certain elements?  I'm working
>  with some biological XMLs, and each data file has a really huge element
>  approximately 50 MB long.  Let's call its tag "FOOBAR".  I'm trying to get
>  XML::Twig to ignore FOOBAR, because it takes at least a few hours for
>  XML::Twig to run through and construct it.
>  
>  Since this element is always at the end, I'm using the hacky solution of
>  calling twig->finish() right before I hit FOOBAR; however, this feels
>  unsatisfactory.  I'd like a more robust solution where I can say something
>  like:
>  
>  ###
>  my $twig = XML::Twig->new(TwigHandlers => $handlers,
>                            IgnoreTags => 'FOOBAR');
>  ###
>  
>  and have XML::Twig just skip over certain elements, avoiding the cost of
>  constructing a PCDATA node.  I haven't found such an 'IgnoreTags' option
>  yet, so I'm wondering if anyone here could give any suggestions.

Hi,

That's what I get for going on vacation for 2 weeks, people keep working
in the meantime ;--)

I had been thinking about such an option but postponed implementing it as
long as nobody actually required it, as there will be a (hopefully small)
performance hit for all other users.

I can think of 2 useful ways to implement that feature:

IgnoreTags =>  ['FOOBAR'];

and a method that could be called from a start_tag_handler:

sub foobar # declared as a start_tag_handler
  { my( $t, $foobar)= @_;
    if( $foobar-> att( 'baz') eq 'whatever value')
      { $foobar-> skip; }
  }

I will have a look at it and get back to you about it later this week
hopefully.

Michel Rodriguez
Perl & XML
http://www.xmltwig.com
Toulouse Perl Mongers: http://toulouse.pm.org


_______________________________________________
Perl-XML mailing list
Perl-XML@[...].com
http://listserv.ActiveState.com/mailman/listinfo/perl-xml