#Usage: archive.pl USERNAME
#
#Description:  Saves all entries of USERNAME's xanga to "archive.html" in the working directory
use LWP::UserAgent;

$end = 'http://www.xanga.com/';

if ($a=shift) {
  $uid = $a;
} else {
  print "What is your username?  ";
  $uid = <STDIN>;
  chop $uid; 
}


$first_page = 'http://www.xanga.com/home.aspx?user=' . $uid;

print "Connecting to $uid's Xanga...\n";
grab($first_page);
$next_page = save();  #save() returns the url to Next 5
print "\$next_page is $next_page\n";

until ($finished) {
  grab($next_page);
  $next_page = save();
  print "\$next_page is $next_page\n";

  last if $next_page =~ /$end$/; 
}

print "\n\n\nCompleted Archiving\n\n\n";

#Usage: grab(url)
#
#Description: 
sub grab{
  open TMP, ">tmp.html" or die;
  $url = shift;
  print "grabbing $url\n";
  $ua = LWP::UserAgent->new;
  $ua->agent("MyApp/0.1 ");
  
  # Be nice to Xanga servers ;-)
  sleep 5;

  # Create a request
  my $req = HTTP::Request->new(GET => $url);
  $req->content_type('application/x-www-form-urlencoded');
  $req->content('query=libwww-perl&mode=dist');

  # Pass request to the user agent and get a response back
  my $res = $ua->request($req);

  # Check the outcome of the response
  if ($res->is_success) {
     print TMP $res->content;
     close TMP;
     print "Successfully grabbed html...\n";
  }
  else {
      print $res->status_line, "\n";
  }
}

#Useage: save(url);
#
#Description:   sub save parses through a given URL and appends all found entries of that page to
#		"archive.html"  It also finds the url of the next page to grab

sub save {
  open IN, "tmp.html" or die;
  open OUT, ">>archive.html" or die;
  print "Saving...\n";
  while ($line = <IN>) {
    if ($line =~ /<div class="blogheader">/) { last; }
  }

  print OUT $line;
  print "Wrote out \$line\n";
  REST: while($line = <IN>) {  
    print OUT $line;
    last REST if $line =~ /Next 5 &gt;&gt;/;
  }
   
  print "Saved\n";
  $line = reverse($line);
  $line =~ /"(.*?)"/;
  close IN;
  close OUT;
  $a = 'http://www.xanga.com/' . reverse($1); #home.aspx?user=....  
}