0x0000
  
  !Ø318×Ãç >   !<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>403 Forbidden</title>
</head><body>
<h1>Forbidden</h1>
<p>You don't have permission to access this resource.</p>
<p>Additionally, a 403 Forbidden
error was encountered while trying to use an ErrorDocument to handle the request.</p>
</body></html>
  
  #Ø318×Ãç >   #<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>403 Forbidden</title>
</head><body>
<h1>Forbidden</h1>
<p>You don't have permission to access this resource.</p>
<p>Additionally, a 403 Forbidden
error was encountered while trying to use an ErrorDocument to handle the request.</p>
</body></html>
  "  %ˆl–äY>”Ji?u@¹q–n	µ1hßë40êÄè  (   %‹     Ó+ÎWÈMÌ3ÔO*ÍÌ)ÉÌ+Ö3ä ÏÇ     /  'ˆl–Ü4ý(©„±¨ J»€À>¦-ÿìƒuÐ³_‹uÐb&=LV[(Æê !   'package LWP::RobotUA;

require LWP::UserAgent;
@ISA = qw(LWP::UserAgent);
$VERSION = "6.03";

require WWW::RobotRules;
require HTTP::Request;
require HTTP::Response;

use Carp ();
use HTTP::Status ();
use HTTP::Date qw(time2str);
use strict;


#
# Additional attributes in addition to those found in LWP::UserAgent:
#
# $self->{'delay'}    Required delay between request to the same
#                     server in minutes.
#
# $self->{'rules'}     A WWW::RobotRules object
#

sub new
{
    my $class = shift;
    my %cnf;
    if (@_ < 4) {
	# legacy args
	@cnf{qw(agent from rules)} = @_;
    }
    else {
	%cnf = @_;
    }

    Carp::croak('LWP::RobotUA agent required') unless $cnf{agent};
    Carp::croak('LWP::RobotUA from address required')
	unless $cnf{from} && $cnf{from} =~ m/\@/;

    my $delay = delete $cnf{delay} || 1;
    my $use_sleep = delete $cnf{use_sleep};
    $use_sleep = 1 unless defined($use_sleep);
    my $rules = delete $cnf{rules};

    my $self = LWP::UserAgent->new(%cnf);
    $self = bless $self, $class;

    $self->{'delay'} = $delay;   # minutes
    $self->{'use_sleep'} = $use_sleep;

    if ($rules) {
	$rules->agent($cnf{agent});
	$self->{'rules'} = $rules;
    }
    else {
	$self->{'rules'} = WWW::RobotRules->new($cnf{agent});
    }

    $self;
}


sub delay     { shift->_elem('delay',     @_); }
sub use_sleep { shift->_elem('use_sleep', @_); }


sub agent
{
    my $self = shift;
    my $old = $self->SUPER::agent(@_);
    if (@_) {
	# Changing our name means to start fresh
	$self->{'rules'}->agent($self->{'agent'}); 
    }
    $old;
}


sub rules {
    my $self = shift;
    my $old = $self->_elem('rules', @_);
    $self->{'rules'}->agent($self->{'agent'}) if @_;
    $old;
}


sub no_visits
{
    my($self, $netloc) = @_;
    $self->{'rules'}->no_visits($netloc) || 0;
}

*host_count = \&no_visits;  # backwards compatibility with LWP-5.02


sub host_wait
{
    my($self, $netloc) = @_;
    return undef unless defined $netloc;
    my $last = $self->{'rules'}->last_visit($netloc);
    if ($last) {
	my $wait = int($self->{'delay'} * 60 - (time - $last));
	$wait = 0 if $wait < 0;
	return $wait;
    }
    return 0;
}


sub simple_request
{
    my($self, $request, $arg, $size) = @_;

    # Do we try to access a new server?
    my $allowed = $self->{'rules'}->allowed($request->uri);

    if ($allowed < 0) {
	# Host is not visited before, or robots.txt expired; fetch "robots.txt"
	my $robot_url = $request->uri->clone;
	$robot_url->path("robots.txt");
	$robot_url->query(undef);

	# make access to robot.txt legal since this will be a recursive call
	$self->{'rules'}->parse($robot_url, ""); 

	my $robot_req = HTTP::Request->new('GET', $robot_url);
	my $parse_head = $self->parse_head(0);
	my $robot_res = $self->request($robot_req);
	$self->parse_head($parse_head);
	my $fresh_until = $robot_res->fresh_until;
	my $content = "";
	if ($robot_res->is_success && $robot_res->content_is_text) {
	    $content = $robot_res->decoded_content;
	    $content = "" unless $content && $content =~ /^\s*Disallow\s*:/mi;
	}
	$self->{'rules'}->parse($robot_url, $content, $fresh_until);

	# recalculate allowed...
	$allowed = $self->{'rules'}->allowed($request->uri);
    }

    # Check rules
    unles