Write a "peek" function that will print the next $count items in the stream and uses the unget_token method to restore your parser to its original state. Here's my take on it. It should be easy enough to munge this for your needs.

#!/usr/bin/perl

use strict;
use warnings;

use HTML::TokeParser::Simple 3.13;

my $url = 'http://www.perlmonks.com/';

my $parser = HTML::TokeParser::Simple->new( url => $url );
$parser->get_tag("ul");
print peek($parser);

sub peek {
    my $parser = shift;
    my $count  = shift || 5;

    my $items = 0;
    my $html  = '';
    my @tokens;
    while ( ( my $token = $parser->get_token ) && $items++ < $count ) 
+{
        $html .= $token->as_is;
        push @tokens, $token;
    }
    $parser->unget_token(@tokens);
    return $html;
}
[download]

You know what? I like this so much I should probably add it to HTML::TokeParser::Simple.