use Perl

• return when the body = 1000? (Score:2)

  by merlyn (47)

  No need to gather a larger body, only to truncate it...

  --
  

  • Randal L. Schwartz
  • Stonehenge
• abstraction is always nice (Score:1)

  by markjugg (792)

  How about: my $compressed_body = compress_body($html);

  My attempt at humor is meant to indicate the upside here:

  At least your annoying code is in a routine, so you have made the functionality less annoying for everything that uses it.
• use tt2 (Score:1)

  by kennyg (3143)

  [% FILTER truncate(1000) %] really long text from example... [% END %]
• Re: (Score:1)

  by Aristotle (5147)

  I’m not sure I understand what your code is doing, so I’ll code to your specification instead…

  sub compress_body {
      my ( $html ) = @_;
      my $parser = HTML::TokeParser::Simple->new( \$html );
      my $text = '';
      my $in_body;
  
      1 while $_ = $parser->get_token and not $_->is_start_tag( 'body' );
  
      while ( my $token = $parser->get_token ) {
          if( $token->is_text ) {
       
  • Re: (Score:1)

    by Aristotle (5147)

    That is, minus the $in_body I accidentally left in there.

    Oh, and my own preferred solution would be to load the thing with libxml’s HTML parsing mode, then do $dom->findvalue( 'substring( /html/body, 0, 1000 )' ). XPath rocks.
  • Re: (Score:2)

    by Ovid (2709)

    The problem is, it gets passed an entire HTML document and has to preserve everything up to the first body tag (inclusive) and after the last body tag (also inclusive).
    • Re: (Score:1)

      by Aristotle (5147)

      Ah, now suddenly all your contortions make sense.

      sub compress_body {
          my ( $html ) = @_;
          my $parser = HTML::TokeParser::Simple->new(\$html);
          my $out = '';
          my $body = '';
      
          while ( my $token = $parser->get_token ) {
              if( $token->is_start_tag('body') .. $token->is_end_tag('body') ) {
                  $out .= $token->as_is if $token->is_tag( 'body' );
             
      • Re: (Score:1)

        by Aristotle (5147)

        Hmm, that has a subtle bug. There has to be a last if not $token; before undef $in_body;.