#!/usr/bin/perl -w             

use HTML::LinkExtor;           
use LWP::Simple;
use HTML::TreeBuilder::XPath;  
use Term::ProgressBar;


my $url = "http://www.totalpolitics.com/blog/463546/senior-medics-call-on-uk-to-stay-but-brexiteers-have-more-to-cheer.thtml";
my $content = get $url;

my $tree = HTML::TreeBuilder::XPath->new_from_content($content); 

my $title = $tree->findvalue(q{//div[@id="article"]/h1});
my $body = $tree->findvalue(q{//div[@class="article-body"]});
my $author = $tree->findnodes(q{//div[@class="article-body"]/p/strong});
$author = $author->[0]->getValue;

$body =~ s/$author//;

my $xml .= '<?xml version="1.0" encoding="UTF-8" ?>';
$xml .= '<nodes>';
$xml .= '<node>';
$xml .= '<url>';
$xml .= $url;
$xml .= '</url>';
$xml .= '<title>';
$xml .= $title;
$xml .= '</title>';
$xml .= '<description>';
$xml .= "<![CDATA[$body]]>";
$xml .= '</description>';
$xml .= '<author>';
$xml .= $author;
$xml .= '</author>';
$xml .= "</node>\n";
$xml .= "</nodes>";

print $xml;