test/html-webhacc/cc.cgi

#!/usr/bin/perl
use strict;

use lib qw[/home/httpd/html/www/markup/html/whatpm
           /home/wakaba/work/manakai2/lib];
use CGI::Carp qw[fatalsToBrowser];

  require WebHACC::Input;

{
  require Message::CGI::HTTP;
  my $http = Message::CGI::HTTP->new;

  require WebHACC::Output;
  my $out = WebHACC::Output->new;
  $out->handle (*STDOUT);
  $out->set_utf8;

  if ($http->get_meta_variable ('PATH_INFO') ne '/') {
    $out->http_error (404);
    exit;
  }
  
  load_text_catalog ('en'); ## TODO: conneg

  $out->set_flush;
  $out->http_header;
  $out->html_header;
  $out->unset_flush;

  my $input = get_input_document ($http);
  $out->input ($input);

  require WebHACC::Result;
  my $result = WebHACC::Result->new;
  $result->output ($out);
  $result->{conforming_min} = 1;
  $result->{conforming_max} = 1;

  $out->html ('<script src="../cc-script.js"></script>');

  check_and_print ($input => $result => $out);
  
  $result->generate_result_section;

  $out->nav_list;

  exit;
}

sub check_and_print ($$$) {
  my ($input, $result, $out) = @_;
  my $original_input = $out->input;
  $out->input ($input);

  $input->generate_info_section ($result);

  $input->generate_transfer_sections ($result);

  unless (defined $input->{s}) {
    $result->{conforming_min} = 0;
    return;
  }

  my $checker_class = {
    'text/cache-manifest' => 'WebHACC::Language::CacheManifest',
    'text/css' => 'WebHACC::Language::CSS',
    'text/html' => 'WebHACC::Language::HTML',
    'text/x-webidl' => 'WebHACC::Language::WebIDL',

    'text/xml' => 'WebHACC::Language::XML',
    'application/atom+xml' => 'WebHACC::Language::XML',
    'application/rss+xml' => 'WebHACC::Language::XML',
    'image/svg+xml' => 'WebHACC::Language::XML',
    'application/xhtml+xml' => 'WebHACC::Language::XML',
    'application/xml' => 'WebHACC::Language::XML',
    ## TODO: Should we make all XML MIME Types fall
    ## into this category?

    ## NOTE: This type has different model from normal XML types.
    'application/rdf+xml' => 'WebHACC::Language::XML',
  }->{$input->{media_type}} || 'WebHACC::Language::Default';

  eval qq{ require $checker_class } or die "$0: Loading $checker_class: $@";
  my $checker = $checker_class->new;
  $checker->input ($input);
  $checker->output ($out);
  $checker->result ($result);

  ## TODO: A cache manifest MUST be text/cache-manifest
  ## TODO: WebIDL media type "text/x-webidl"

  $checker->generate_syntax_error_section;
  $checker->generate_source_string_section;

  my @subdoc;
  $checker->onsubdoc (sub {
    push @subdoc, shift;
  });

  $checker->generate_structure_dump_section;
  $checker->generate_structure_error_section;
  $checker->generate_additional_sections;

  my $id_prefix = 0;
  for my $_subinput (@subdoc) {
    my $subinput = WebHACC::Input::Subdocument->new (++$id_prefix);
    $subinput->{$_} = $_subinput->{$_} for keys %$_subinput;
    $subinput->{base_uri} = $subinput->{container_node}->base_uri
        unless defined $subinput->{base_uri};
    $subinput->{parent_input} = $input;

    $subinput->start_section ($result);
    check_and_print ($subinput => $result => $out);
    $subinput->end_section ($result);
  }

  $out->input ($original_input);
} # check_and_print


{
  my $Msg = {};

sub load_text_catalog ($) {
#  my $self = shift;
  my $lang = shift; # MUST be a canonical lang name
  open my $file, '<:utf8', "cc-msg.$lang.txt"
      or die "$0: cc-msg.$lang.txt: $!";
  while (<$file>) {
    if (s/^([^;]+);([^;]*);//) {
      my ($type, $cls, $msg) = ($1, $2, $_);
      $msg =~ tr/\x0D\x0A//d;
      $Msg->{$type} = [$cls, $msg];
    }
  }
} # load_text_catalog

sub get_text ($;$$) {
#  my $self = shift;
  my ($type, $level, $node) = @_;
  $type = $level . ':' . $type if defined $level;
  $level = 'm' unless defined $level;
  my @arg;
  {
    if (defined $Msg->{$type}) {
      my $msg = $Msg->{$type}->[1];
      $msg =~ s{<var>\$([0-9]+)</var>}{
        defined $arg[$1] ? ($arg[$1]) : '(undef)';
      }ge;                 ##BUG: ^ must be escaped
      $msg =~ s{<var>{\@([A-Za-z0-9:_.-]+)}</var>}{
        UNIVERSAL::can ($node, 'get_attribute_ns')
            ?  ($node->get_attribute_ns (undef, $1)) : ''
      }ge; ## BUG: ^ must be escaped
      $msg =~ s{<var>{\@}</var>}{        ## BUG: v must be escaped
        UNIVERSAL::can ($node, 'value') ? ($node->value) : ''
      }ge;
      $msg =~ s{<var>{local-name}</var>}{
        UNIVERSAL::can ($node, 'manakai_local_name')
          ? ($node->manakai_local_name) : ''
      }ge;  ## BUG: ^ must be escaped
      $msg =~ s{<var>{element-local-name}</var>}{
        (UNIVERSAL::can ($node, 'owner_element') and
         $node->owner_element)
          ?  ($node->owner_element->manakai_local_name)
          : '' ## BUG: ^ must be escaped
      }ge;
      return ($type, 'level-' . $level . ' ' . $Msg->{$type}->[0], $msg);
    } elsif ($type =~ s/:([^:]*)$//) {
      unshift @arg, $1;
      redo;
    }
  }
  return ($type, 'level-'.$level, ($_[0]));
                                 ## BUG: ^ must be escaped
} # get_text

}

sub get_input_document ($) {
  my $http = shift;

  require Message::DOM::DOMImplementation;
  my $dom = Message::DOM::DOMImplementation->new;

  require Encode;
  my $request_uri = Encode::decode ('utf-8', $http->get_parameter ('uri'));
  my $r = WebHACC::Input->new;
  if (defined $request_uri and length $request_uri) {
    my $uri = $dom->create_uri_reference ($request_uri);
    unless ({
             http => 1,
            }->{lc $uri->uri_scheme}) {
      $r = WebHACC::Input::Error->new;
      $r->{uri} = $request_uri;
      $r->{request_uri} = $request_uri;
      $r->{error_status_text} = 'URL scheme not allowed';
    }

    require Message::Util::HostPermit;
    my $host_permit = new Message::Util::HostPermit;
    $host_permit->add_rule (<<EOH);
Allow host=suika port=80
Deny host=suika
Allow host=suika.fam.cx port=80
Deny host=suika.fam.cx
Deny host=localhost
Deny host=*.localdomain
Deny ipv4=0.0.0.0/8
Deny ipv4=10.0.0.0/8
Deny ipv4=127.0.0.0/8
Deny ipv4=169.254.0.0/16
Deny ipv4=172.0.0.0/11
Deny ipv4=192.0.2.0/24
Deny ipv4=192.88.99.0/24
Deny ipv4=192.168.0.0/16
Deny ipv4=198.18.0.0/15
Deny ipv4=224.0.0.0/4
Deny ipv4=255.255.255.255/32
Deny ipv6=0::0/0
Allow host=*
EOH
    unless ($host_permit->check ($uri->uri_host, $uri->uri_port || 80)) {
      my $r = WebHACC::Input::Error->new;
      $r->{uri} = $request_uri;
      $r->{request_uri} = $request_uri;
      $r->{error_status_text} = 'Connection to the host is forbidden';
      return $r;
    }

    require LWP::UserAgent;
    my $ua = WDCC::LWPUA->new;
    $ua->{wdcc_dom} = $dom;
    $ua->{wdcc_host_permit} = $host_permit;
    $ua->agent ('Mozilla'); ## TODO: for now.
    $ua->parse_head (0);
    $ua->protocols_allowed ([qw/http/]);
    $ua->max_size (1000_000);
    my $req = HTTP::Request->new (GET => $request_uri);
    $req->header ('Accept-Encoding' => 'identity, *; q=0');
    my $res = $ua->request ($req);
    ## TODO: 401 sets |is_success| true.
    if ($res->is_success or $http->get_parameter ('error-page')) {
      $r->{base_uri} = $res->base; ## NOTE: It does check |Content-Base|, |Content-Location|, and <base>. ## TODO: Use our own code!
      $r->{uri} = $res->request->uri;
      $r->{request_uri} = $request_uri;

      ## TODO: More strict parsing...
      my $ct = $res->header ('Content-Type');
      if (defined $ct and $ct =~ /;\s*charset\s*=\s*"?([^\s;"]+)"?/i) {
        $r->{charset} = lc $1;
        $r->{charset} =~ tr/\\//d;
        $r->{official_charset} = $r->{charset};
      }

      my $input_charset = $http->get_parameter ('charset');
      if (defined $input_charset and length $input_charset) {
        $r->{charset_overridden}
            = (not defined $r->{charset} or $r->{charset} ne $input_charset);
        $r->{charset} = $input_charset;
      }

      ## TODO: Support for HTTP Content-Encoding

      $r->{s} = ''.$res->content;

      require Whatpm::ContentType;
      ($r->{official_type}, $r->{media_type})
          = Whatpm::ContentType->get_sniffed_type
              (get_file_head => sub {
                 return substr $r->{s}, 0, shift;
               },
               http_content_type_byte => $ct,
               has_http_content_encoding =>
                   defined $res->header ('Content-Encoding'),
               supported_image_types => {});
    } else {
      $r->{uri} = $res->request->uri;
      $r->{request_uri} = $request_uri;
      $r->{error_status_text} = $res->status_line;
    }

    $r->{header_field} = [];
    $res->scan (sub {
      push @{$r->{header_field}}, [$_[0], $_[1]];
    });
    $r->{header_status_code} = $res->code;
    $r->{header_status_text} = $res->message;
  } else {
    $r->{s} = ''.$http->get_parameter ('s');
    $r->{uri} = q<thismessage:/>;
    $r->{request_uri} = q<thismessage:/>;
    $r->{base_uri} = q<thismessage:/>;
    $r->{charset} = ''.$http->get_parameter ('_charset_');
    $r->{charset} =~ s/\s+//g;
    $r->{charset} = 'utf-8' if $r->{charset} eq '';
    $r->{official_charset} = $r->{charset};
    $r->{header_field} = [];

    require Whatpm::ContentType;
    ($r->{official_type}, $r->{media_type})
        = Whatpm::ContentType->get_sniffed_type
            (get_file_head => sub {
               return substr $r->{s}, 0, shift;
             },
             http_content_type_byte => undef,
             has_http_content_encoding => 0,
             supported_image_types => {});
  }

  my $input_format = $http->get_parameter ('i');
  if (defined $input_format and length $input_format) {
    $r->{media_type_overridden}
        = (not defined $r->{media_type} or $input_format ne $r->{media_type});
    $r->{media_type} = $input_format;
  }
  if (defined $r->{s} and not defined $r->{media_type}) {
    $r->{media_type} = 'text/html';
    $r->{media_type_overridden} = 1;
  }

  if ($r->{media_type} eq 'text/xml') {
    unless (defined $r->{charset}) {
      $r->{charset} = 'us-ascii';
      $r->{official_charset} = $r->{charset};
    } elsif ($r->{charset_overridden} and $r->{charset} eq 'us-ascii') {
      $r->{charset_overridden} = 0;
    }
  }

  if (length $r->{s} > 1000_000) {
    $r->{error_status_text} = 'Entity-body too large';
    delete $r->{s};
    return $r;
  }

  $r->{inner_html_element} = $http->get_parameter ('e');

  return $r;
} # get_input_document

package WDCC::LWPUA;
BEGIN { push our @ISA, 'LWP::UserAgent'; }

sub redirect_ok {
  my $ua = shift;
  unless ($ua->SUPER::redirect_ok (@_)) {
    return 0;
  }

  my $uris = $_[1]->header ('Location');
  return 0 unless $uris;
  my $uri = $ua->{wdcc_dom}->create_uri_reference ($uris);
  unless ({
           http => 1,
          }->{lc $uri->uri_scheme}) {
    return 0;
  }
  unless ($ua->{wdcc_host_permit}->check ($uri->uri_host, $uri->uri_port || 80)) {
    return 0;
  }
  return 1;
} # redirect_ok

=head1 AUTHOR

Wakaba <w@suika.fam.cx>.

=head1 LICENSE

Copyright 2007-2008 Wakaba <w@suika.fam.cx>

This library is free software; you can redistribute it
and/or modify it under the same terms as Perl itself.

=cut

## $Date: 2008/07/21 09:15:55 $
1	wakaba	1.1	#!/usr/bin/perl
2			use strict;
3
4			use lib qw[/home/httpd/html/www/markup/html/whatpm
5	wakaba	1.16	/home/wakaba/work/manakai2/lib];
6	wakaba	1.1	use CGI::Carp qw[fatalsToBrowser];
7
8	wakaba	1.53	require WebHACC::Input;
9	wakaba	1.2
10	wakaba	1.35	{
11	wakaba	1.58	require Message::CGI::HTTP;
12	wakaba	1.16	my $http = Message::CGI::HTTP->new;
13	wakaba	1.1
14	wakaba	1.58	require WebHACC::Output;
15			my $out = WebHACC::Output->new;
16			$out->handle (*STDOUT);
17			$out->set_utf8;
18
19	wakaba	1.16	if ($http->get_meta_variable ('PATH_INFO') ne '/') {
20	wakaba	1.58	$out->http_error (404);
21	wakaba	1.8	exit;
22			}
23	wakaba	1.53
24	wakaba	1.7	load_text_catalog ('en'); ## TODO: conneg
25
26	wakaba	1.53	$out->set_flush;
27	wakaba	1.58	$out->http_header;
28			$out->html_header;
29			$out->unset_flush;
30	wakaba	1.55
31	wakaba	1.58	my $input = get_input_document ($http);
32	wakaba	1.53	$out->input ($input);
33
34	wakaba	1.58	require WebHACC::Result;
35	wakaba	1.55	my $result = WebHACC::Result->new;
36			$result->output ($out);
37			$result->{conforming_min} = 1;
38			$result->{conforming_max} = 1;
39	wakaba	1.14
40	wakaba	1.55	$out->html ('<script src="../cc-script.js"></script>');
41	wakaba	1.54
42	wakaba	1.55	check_and_print ($input => $result => $out);
43
44			$result->generate_result_section;
45	wakaba	1.1
46	wakaba	1.53	$out->nav_list;
47	wakaba	1.16
48	wakaba	1.53	exit;
49	wakaba	1.35	}
50	wakaba	1.1
51	wakaba	1.53	sub check_and_print ($$$) {
52			my ($input, $result, $out) = @_;
53			my $original_input = $out->input;
54			$out->input ($input);
55	wakaba	1.31
56	wakaba	1.55	$input->generate_info_section ($result);
57
58	wakaba	1.54	$input->generate_transfer_sections ($result);
59	wakaba	1.31
60	wakaba	1.55	unless (defined $input->{s}) {
61			$result->{conforming_min} = 0;
62			return;
63			}
64	wakaba	1.31
65	wakaba	1.53	my $checker_class = {
66			'text/cache-manifest' => 'WebHACC::Language::CacheManifest',
67			'text/css' => 'WebHACC::Language::CSS',
68			'text/html' => 'WebHACC::Language::HTML',
69			'text/x-webidl' => 'WebHACC::Language::WebIDL',
70
71			'text/xml' => 'WebHACC::Language::XML',
72			'application/atom+xml' => 'WebHACC::Language::XML',
73			'application/rss+xml' => 'WebHACC::Language::XML',
74			'image/svg+xml' => 'WebHACC::Language::XML',
75			'application/xhtml+xml' => 'WebHACC::Language::XML',
76			'application/xml' => 'WebHACC::Language::XML',
77			## TODO: Should we make all XML MIME Types fall
78			## into this category?
79
80			## NOTE: This type has different model from normal XML types.
81			'application/rdf+xml' => 'WebHACC::Language::XML',
82			}->{$input->{media_type}} \|\| 'WebHACC::Language::Default';
83
84			eval qq{ require $checker_class } or die "$0: Loading $checker_class: $@";
85			my $checker = $checker_class->new;
86			$checker->input ($input);
87			$checker->output ($out);
88			$checker->result ($result);
89
90			## TODO: A cache manifest MUST be text/cache-manifest
91			## TODO: WebIDL media type "text/x-webidl"
92
93			$checker->generate_syntax_error_section;
94			$checker->generate_source_string_section;
95
96	wakaba	1.55	my @subdoc;
97	wakaba	1.53	$checker->onsubdoc (sub {
98			push @subdoc, shift;
99			});
100
101			$checker->generate_structure_dump_section;
102			$checker->generate_structure_error_section;
103			$checker->generate_additional_sections;
104
105	wakaba	1.34	my $id_prefix = 0;
106	wakaba	1.53	for my $_subinput (@subdoc) {
107	wakaba	1.55	my $subinput = WebHACC::Input::Subdocument->new (++$id_prefix);
108	wakaba	1.53	$subinput->{$_} = $_subinput->{$_} for keys %$_subinput;
109	wakaba	1.34	$subinput->{base_uri} = $subinput->{container_node}->base_uri
110			unless defined $subinput->{base_uri};
111	wakaba	1.55	$subinput->{parent_input} = $input;
112	wakaba	1.34
113	wakaba	1.55	$subinput->start_section ($result);
114	wakaba	1.53	check_and_print ($subinput => $result => $out);
115	wakaba	1.55	$subinput->end_section ($result);
116	wakaba	1.34	}
117	wakaba	1.53
118			$out->input ($original_input);
119	wakaba	1.31	} # check_and_print
120
121	wakaba	1.45
122	wakaba	1.7	{
123			my $Msg = {};
124
125			sub load_text_catalog ($) {
126	wakaba	1.53	# my $self = shift;
127	wakaba	1.7	my $lang = shift; # MUST be a canonical lang name
128	wakaba	1.26	open my $file, '<:utf8', "cc-msg.$lang.txt"
129			or die "$0: cc-msg.$lang.txt: $!";
130	wakaba	1.7	while (<$file>) {
131			if (s/^([^;]+);([^;]*);//) {
132			my ($type, $cls, $msg) = ($1, $2, $_);
133			$msg =~ tr/\x0D\x0A//d;
134			$Msg->{$type} = [$cls, $msg];
135			}
136			}
137			} # load_text_catalog
138
139	wakaba	1.53	sub get_text ($;$$) {
140			# my $self = shift;
141	wakaba	1.15	my ($type, $level, $node) = @_;
142	wakaba	1.7	$type = $level . ':' . $type if defined $level;
143	wakaba	1.29	$level = 'm' unless defined $level;
144	wakaba	1.7	my @arg;
145			{
146			if (defined $Msg->{$type}) {
147			my $msg = $Msg->{$type}->[1];
148	wakaba	1.10	$msg =~ s{<var>\$([0-9]+)</var>}{
149	wakaba	1.53	defined $arg[$1] ? ($arg[$1]) : '(undef)';
150			}ge; ##BUG: ^ must be escaped
151	wakaba	1.15	$msg =~ s{<var>{\@([A-Za-z0-9:_.-]+)}</var>}{
152			UNIVERSAL::can ($node, 'get_attribute_ns')
153	wakaba	1.53	? ($node->get_attribute_ns (undef, $1)) : ''
154			}ge; ## BUG: ^ must be escaped
155			$msg =~ s{<var>{\@}</var>}{ ## BUG: v must be escaped
156			UNIVERSAL::can ($node, 'value') ? ($node->value) : ''
157	wakaba	1.15	}ge;
158	wakaba	1.17	$msg =~ s{<var>{local-name}</var>}{
159			UNIVERSAL::can ($node, 'manakai_local_name')
160	wakaba	1.53	? ($node->manakai_local_name) : ''
161			}ge; ## BUG: ^ must be escaped
162	wakaba	1.17	$msg =~ s{<var>{element-local-name}</var>}{
163			(UNIVERSAL::can ($node, 'owner_element') and
164			$node->owner_element)
165	wakaba	1.53	? ($node->owner_element->manakai_local_name)
166			: '' ## BUG: ^ must be escaped
167	wakaba	1.17	}ge;
168	wakaba	1.29	return ($type, 'level-' . $level . ' ' . $Msg->{$type}->[0], $msg);
169	wakaba	1.7	} elsif ($type =~ s/:([^:]*)$//) {
170			unshift @arg, $1;
171			redo;
172			}
173			}
174	wakaba	1.53	return ($type, 'level-'.$level, ($_[0]));
175			## BUG: ^ must be escaped
176	wakaba	1.7	} # get_text
177
178			}
179
180	wakaba	1.58	sub get_input_document ($) {
181			my $http = shift;
182
183			require Message::DOM::DOMImplementation;
184			my $dom = Message::DOM::DOMImplementation->new;
185	wakaba	1.9
186	wakaba	1.54	require Encode;
187			my $request_uri = Encode::decode ('utf-8', $http->get_parameter ('uri'));
188	wakaba	1.53	my $r = WebHACC::Input->new;
189	wakaba	1.9	if (defined $request_uri and length $request_uri) {
190			my $uri = $dom->create_uri_reference ($request_uri);
191			unless ({
192			http => 1,
193			}->{lc $uri->uri_scheme}) {
194	wakaba	1.54	$r = WebHACC::Input::Error->new;
195			$r->{uri} = $request_uri;
196			$r->{request_uri} = $request_uri;
197			$r->{error_status_text} = 'URL scheme not allowed';
198	wakaba	1.9	}
199
200			require Message::Util::HostPermit;
201			my $host_permit = new Message::Util::HostPermit;
202			$host_permit->add_rule (<<EOH);
203			Allow host=suika port=80
204			Deny host=suika
205			Allow host=suika.fam.cx port=80
206			Deny host=suika.fam.cx
207			Deny host=localhost
208			Deny host=*.localdomain
209			Deny ipv4=0.0.0.0/8
210			Deny ipv4=10.0.0.0/8
211			Deny ipv4=127.0.0.0/8
212			Deny ipv4=169.254.0.0/16
213			Deny ipv4=172.0.0.0/11
214			Deny ipv4=192.0.2.0/24
215			Deny ipv4=192.88.99.0/24
216			Deny ipv4=192.168.0.0/16
217			Deny ipv4=198.18.0.0/15
218			Deny ipv4=224.0.0.0/4
219			Deny ipv4=255.255.255.255/32
220			Deny ipv6=0::0/0
221			Allow host=*
222			EOH
223			unless ($host_permit->check ($uri->uri_host, $uri->uri_port \|\| 80)) {
224	wakaba	1.54	my $r = WebHACC::Input::Error->new;
225			$r->{uri} = $request_uri;
226			$r->{request_uri} = $request_uri;
227			$r->{error_status_text} = 'Connection to the host is forbidden';
228			return $r;
229	wakaba	1.9	}
230
231			require LWP::UserAgent;
232			my $ua = WDCC::LWPUA->new;
233			$ua->{wdcc_dom} = $dom;
234			$ua->{wdcc_host_permit} = $host_permit;
235			$ua->agent ('Mozilla'); ## TODO: for now.
236			$ua->parse_head (0);
237			$ua->protocols_allowed ([qw/http/]);
238			$ua->max_size (1000_000);
239			my $req = HTTP::Request->new (GET => $request_uri);
240	wakaba	1.28	$req->header ('Accept-Encoding' => 'identity, *; q=0');
241	wakaba	1.9	my $res = $ua->request ($req);
242	wakaba	1.16	## TODO: 401 sets \|is_success\| true.
243			if ($res->is_success or $http->get_parameter ('error-page')) {
244	wakaba	1.9	$r->{base_uri} = $res->base; ## NOTE: It does check \|Content-Base\|, \|Content-Location\|, and <base>. ## TODO: Use our own code!
245			$r->{uri} = $res->request->uri;
246			$r->{request_uri} = $request_uri;
247
248			## TODO: More strict parsing...
249			my $ct = $res->header ('Content-Type');
250	wakaba	1.22	if (defined $ct and $ct =~ /;\scharset\s=\s*"?([^\s;"]+)"?/i) {
251	wakaba	1.9	$r->{charset} = lc $1;
252			$r->{charset} =~ tr/\\//d;
253	wakaba	1.26	$r->{official_charset} = $r->{charset};
254	wakaba	1.9	}
255
256	wakaba	1.16	my $input_charset = $http->get_parameter ('charset');
257	wakaba	1.9	if (defined $input_charset and length $input_charset) {
258			$r->{charset_overridden}
259			= (not defined $r->{charset} or $r->{charset} ne $input_charset);
260			$r->{charset} = $input_charset;
261	wakaba	1.25	}
262
263			## TODO: Support for HTTP Content-Encoding
264	wakaba	1.9
265			$r->{s} = ''.$res->content;
266	wakaba	1.25
267			require Whatpm::ContentType;
268			($r->{official_type}, $r->{media_type})
269			= Whatpm::ContentType->get_sniffed_type
270			(get_file_head => sub {
271			return substr $r->{s}, 0, shift;
272			},
273			http_content_type_byte => $ct,
274			has_http_content_encoding =>
275			defined $res->header ('Content-Encoding'),
276			supported_image_types => {});
277	wakaba	1.9	} else {
278			$r->{uri} = $res->request->uri;
279			$r->{request_uri} = $request_uri;
280			$r->{error_status_text} = $res->status_line;
281			}
282
283			$r->{header_field} = [];
284			$res->scan (sub {
285			push @{$r->{header_field}}, [$_[0], $_[1]];
286			});
287			$r->{header_status_code} = $res->code;
288			$r->{header_status_text} = $res->message;
289			} else {
290	wakaba	1.16	$r->{s} = ''.$http->get_parameter ('s');
291	wakaba	1.9	$r->{uri} = q<thismessage:/>;
292			$r->{request_uri} = q<thismessage:/>;
293			$r->{base_uri} = q<thismessage:/>;
294	wakaba	1.16	$r->{charset} = ''.$http->get_parameter ('_charset_');
295	wakaba	1.9	$r->{charset} =~ s/\s+//g;
296			$r->{charset} = 'utf-8' if $r->{charset} eq '';
297	wakaba	1.26	$r->{official_charset} = $r->{charset};
298	wakaba	1.9	$r->{header_field} = [];
299	wakaba	1.25
300			require Whatpm::ContentType;
301			($r->{official_type}, $r->{media_type})
302			= Whatpm::ContentType->get_sniffed_type
303			(get_file_head => sub {
304			return substr $r->{s}, 0, shift;
305			},
306			http_content_type_byte => undef,
307			has_http_content_encoding => 0,
308			supported_image_types => {});
309	wakaba	1.9	}
310
311	wakaba	1.16	my $input_format = $http->get_parameter ('i');
312	wakaba	1.9	if (defined $input_format and length $input_format) {
313			$r->{media_type_overridden}
314			= (not defined $r->{media_type} or $input_format ne $r->{media_type});
315			$r->{media_type} = $input_format;
316			}
317			if (defined $r->{s} and not defined $r->{media_type}) {
318			$r->{media_type} = 'text/html';
319			$r->{media_type_overridden} = 1;
320			}
321
322			if ($r->{media_type} eq 'text/xml') {
323			unless (defined $r->{charset}) {
324			$r->{charset} = 'us-ascii';
325	wakaba	1.26	$r->{official_charset} = $r->{charset};
326	wakaba	1.9	} elsif ($r->{charset_overridden} and $r->{charset} eq 'us-ascii') {
327			$r->{charset_overridden} = 0;
328			}
329			}
330
331			if (length $r->{s} > 1000_000) {
332			$r->{error_status_text} = 'Entity-body too large';
333			delete $r->{s};
334			return $r;
335			}
336
337	wakaba	1.35	$r->{inner_html_element} = $http->get_parameter ('e');
338
339	wakaba	1.9	return $r;
340			} # get_input_document
341
342			package WDCC::LWPUA;
343			BEGIN { push our @ISA, 'LWP::UserAgent'; }
344
345			sub redirect_ok {
346			my $ua = shift;
347			unless ($ua->SUPER::redirect_ok (@_)) {
348			return 0;
349			}
350
351			my $uris = $_[1]->header ('Location');
352			return 0 unless $uris;
353			my $uri = $ua->{wdcc_dom}->create_uri_reference ($uris);
354			unless ({
355			http => 1,
356			}->{lc $uri->uri_scheme}) {
357			return 0;
358			}
359			unless ($ua->{wdcc_host_permit}->check ($uri->uri_host, $uri->uri_port \|\| 80)) {
360			return 0;
361			}
362			return 1;
363			} # redirect_ok
364
365	wakaba	1.1	=head1 AUTHOR
366
367			Wakaba <w@suika.fam.cx>.
368
369			=head1 LICENSE
370
371	wakaba	1.35	Copyright 2007-2008 Wakaba <w@suika.fam.cx>
372	wakaba	1.1
373			This library is free software; you can redistribute it
374			and/or modify it under the same terms as Perl itself.
375
376			=cut
377
378	wakaba	1.58	## $Date: 2008/07/21 09:15:55 $