#!/usr/bin/perl
#load libwwwperl
use HTML::TokeParser;
use LWP::UserAgent;
$ua = LWP::UserAgent->new;
$ua->agent('Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)');
use HTTP::Cookies;
$ua->cookie_jar(HTTP::Cookies->new(file => "lwpcookies.txt",
autosave => 1));
#send initial request for username/pass form (this is where we get the nice authtoken key)
$req = HTTP::Request->new(GET => 'http://news.bbc.co.uk/register_audience?url=/text_only.stm&audience_name=International&bandwidth=low');
$req->header('Accept' => 'text/html');
$res = $ua->request($req);
# check the outcome
if ($res->is_success) {
$content= $res->content;
#print $res->content;
$stream = HTML::TokeParser->new(\$res->content);
while ($token = $stream->get_tag("font")){
if ($token->[1]{size} eq "2") {
$ham = $stream->get_trimmed_text('/font');
$ham =~ /^(.* GMT).*$/;
$date = $1;
}
}
$content =~ / NEWS FRONT PAGE<\/h3>(.*)AFRICA<\/b>/s;
$news = $1;
$news =~ s/\n{2,}/\n/g;
#$news =~ s/\s{2,}/\s/g;
$stream2 = HTML::TokeParser->new(\$news);
$stream3 = HTML::TokeParser->new(\$news);
while ($stream3->get_tag("div")) {
push @synopsis, $stream3->get_trimmed_text("/div");
}
while (my $token = $stream2->get_tag("a")) {
my $url = $token->[1]{href} || "-";
my $text = $stream2->get_trimmed_text("/a");
# print "$text
\n";
$url =~ s/\/low\//\/hi\//;
push @url, $url;
push @text, $text;
}
#print $news;
}
#print "$date
\n\n";
$synopsis[0] =~ s/$text[0]//;
print "$text[0]
\n$synopsis[0]
\n";
$synopsis[1] =~ s/$text[1]//;
print "$text[1]
\n$synopsis[1]
\n";
$synopsis[2] =~ s/$text[2]//;
print "$text[2]
\n$synopsis[2]
\n";
$synopsis[3] =~ s/$text[3]//;
print "$text[3]
\n$synopsis[3]
\n";
$synopsis[4] =~ s/$text[4]//;
print "$text[4]
\n$synopsis[4]
\n";