#!/usr/bin/perl
package Metabot;

use strict;
use WWW::SimpleRobot;
use HTML::Entities;
require HTML::Parser;
use HTML::Summary;
use HTML::TreeBuilder;
use Lingua::EN::Keywords; 

@Metabot::ISA = qw(HTML::Parser);

my $url = $ARGV[0];

my $parser = Metabot->new;

my $robot = WWW::SimpleRobot->new(
    URLS            => [ $url ],
    FOLLOW_REGEX    => "^$url",
    DEPTH           => 2,
    TRAVERSAL       => 'depth',
    VISIT_CALLBACK  => 
        sub { 
            my ( $url, $depth, $html, $links ) = @_;
            print "$url - depth $depth\n"; 
            $html = decode_entities($html);
            $html =~ s/document\.write\(.+?\)\;//g;
            $html =~ s/\&\#.+?\;//g;
            
            my $tree = new HTML::TreeBuilder;
            $tree->parse($html);
            
            my $summarizer = new HTML::Summary(
               LENGTH      => 250,
               USE_META    => 1,
            );
            
            my $summary = $summarizer->generate( $tree );
            $summary =~ s/\s+/ /gs;
            
            print "Summary: $summary\n";
            
            $parser->parse($html);
            my $text = $parser->{TEXT};
            my @keywords = keywords($summary . $text);
            print "Keywords: " . join(", ", @keywords) . "\n\n";      
+      
        }
    ,
    BROKEN_LINK_CALLBACK  => 
        sub { 
            my ( $url, $linked_from, $depth ) = @_;
            print STDERR "$url looks like a broken link on $linked_fro
+m\n"; 
            print STDERR "Depth = $depth\n"; 
        }
);

$robot->traverse;

my @urls = @{$robot->urls};

my @pages = @{$robot->pages};

for my $page ( @pages )    {
    my $url = $page->{url};
    my $depth = $page->{depth};
    my $modification_time = $page->{modification_time};
}

sub text {
    my ($self,$text) = @_;
    $self->{TEXT} .= $text;
}
[download]

[Ganesha:~/Desktop] davistv% perl metabot4.pl http://cincypg.org
http://cincypg.org/ - depth 0
Summary: The Cincinnati Programmers' Guild is founded on the premise t
+hat the art of software design is best practiced with a sense of craf
+tsmanship and personal responsibility. 2003 February 11 - Member Tom 
+Wulf going on Safari CPG Member Tom Wulf has volunt
Keywords: art of software design, member tom wulf, guild, tom wulf, sa
+fari cpg member tom wulf, cincinnati programmers

http://cincypg.org/legal/Bylaws.html - depth 1
Summary: BYLAWS OF THE CINCINNATI PROGRAMMERS GUILD April 16, 2002 ART
+ICLE I. ARTICLE II. Section 1. ARTICLE III. Section 1. Section 2. Sec
+tion 3. Section 4. Section 5. ARTICLE IV. Section 1. Section 2. Secti
+on 3. Section 4. Section 5. Section 6. Section 7. S
Keywords: section, meetings, council, member, offices, councilors

http://cincypg.org/contact.shtml - depth 1
Summary: Cincinnati Programmers' Guild: Contacts General Guild Informa
+tion Troy Davis Foundertroy@glyss.com Secretary Jason Paul Secretaryj
+ason@adaptiveinfosystems.com Webmaster Jeremy Phelps Guild Webmasterw
+ebmaster@cincyp
Keywords: meetings, section, council, member, offices, guild

http://cincypg.org/cgi-bin/links.pl - depth 1
Summary: Submit a link for inclusion on this page: Title: URL: Select 
+a categoryTutorials and Online DocumentationOther Computer GroupsInfo
+rmation Technology NewsOtherUSENET NewsgroupsAlgorithm sitesComputer 
+humorOpinionated tripeVendor WebsitesOther Guilds
Keywords: meeting, section, council, guild, member, offices

http://cincypg.org/directions.shtml - depth 1
Summary: Cincinnati Programmers' Guild This page has moved.
Keywords: meeting, section, council, member, guild, offices

http://cincypg.org/events.shtml - depth 1
Summary: Many thanks to our host: Future Events Event TypeWhenWhere(Cl
+ick for directions.)TopicPresenter(s) Monthly MeetingJune 17th, 2003 
+18:30 (6:30PM)KiZAN TechnologiesInvitation to CVSMr. Possible Future 
+Topics: ActionScript: Flash isn't just for designer
Keywords: meeting, monthly meeting, section, ), 6:00pm, council

http://cincypg.org/subscribe.shtml - depth 1
Summary: Enter your e-mail address below. Check here if you are unsubs
+cribing.
Keywords: meeting, monthly meeting, section, ), council, 6:00pm

http://cincypg.org/join.shtml - depth 1
Summary: Join the Cincinnati Programmers' Guild We are not currently k
+eeping an Official Membership List, nor is there a formal definition 
+of who is and is not a Guild member. Just show up at our next meeting
+.
Keywords: meeting, monthly meeting, members, section, ), guild
[download]