test/html-webhacc/cc.cgi

#!/usr/bin/perl
use strict;
use utf8;

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

  require WebHACC::Input;
  require WebHACC::Result;
  require WebHACC::Output;

my $out;

  require Message::DOM::DOMImplementation;
  my $dom = Message::DOM::DOMImplementation->new;
{
  use Message::CGI::HTTP;
  my $http = Message::CGI::HTTP->new;

  if ($http->get_meta_variable ('PATH_INFO') ne '/') {
    print STDOUT "Status: 404 Not Found\nContent-Type: text/plain; charset=us-ascii\n\n400";
    exit;
  }
  
  load_text_catalog ('en'); ## TODO: conneg

  $out = WebHACC::Output->new;
  $out->handle (*STDOUT);
  $out->set_utf8;
  $out->set_flush;
  $out->html (qq[Content-Type: text/html; charset=utf-8

<!DOCTYPE html>
<html lang="en">
<head>
<title>Web Document Conformance Checker (BETA)</title>
<link rel="stylesheet" href="../cc-style.css" type="text/css">
</head>
<body>
<h1><a href="../cc-interface">Web Document Conformance Checker</a> 
(<em>beta</em>)</h1>
]);

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

  my $char_length = 0;

  $out->start_section (id => 'document-info', title => 'Information');
  $out->html (qq[<dl>
<dt>Request URL</dt>
    <dd>]);
  $out->url ($input->{request_uri});
  $out->html (q[<dt>Document URL<!-- HTML5 document's address? -->
    <dd>]);
  $out->url ($input->{uri}, id => 'anchor-document-url');
  $out->html (q[
    <script>
      document.title = '<'
          + document.getElementById ('anchor-document-url').href + '> \\u2014 '
          + document.title;
    </script>]);
  ## NOTE: no </dl> yet

  if (defined $input->{s}) {
    $char_length = length $input->{s};

    $out->html (qq[<dt>Base URI<dd>]);
    $out->url ($input->{base_uri});
    $out->html (qq[<dt>Internet Media Type</dt>
    <dd><code class="MIME" lang="en">]);
    $out->text ($input->{media_type});
    $out->html (qq[</code> ]);
    if ($input->{media_type_overridden}) {
      $out->html ('<em>(overridden)</em>');
    } elsif (defined $input->{official_type}) {
      if ($input->{media_type} eq $input->{official_type}) {
        #
      } else {
        $out->html ('<em>(sniffed; official type is: <code class=MIME lang=en>');
        $out->text ($input->{official_type});
        $out->html ('</code>)');
      }
    } else {
      $out->html ('<em>(sniffed)</em>');
    }
    $out->html (q[<dt>Character Encoding<dd>]);
    if (defined $input->{charset}) {
      $out->html ('<code class="charset" lang="en">');
      $out->text ($input->{charset});
      $out->html ('</code>');
    } else {
      $out->text ('(none)');
    }
    $out->html (' <em>overridden</em>') if $input->{charset_overridden};
    $out->html (qq[
<dt>Length</dt>
    <dd>$char_length byte@{[$char_length == 1 ? '' : 's']}</dd>
</dl>

<script src="../cc-script.js"></script>
]);
    $out->end_section;

    my $result = WebHACC::Result->new;
    $result->output ($out);
    $result->{conforming_min} = 1;
    $result->{conforming_max} = 1;
    check_and_print ($input => $result => $out);
    $result->generate_result_section;
  } else {
    $out->html ('</dl>');
    $out->end_section;

    my $result = WebHACC::Result->new;
    $result->output ($out);
    $result->{conforming_min} = 0;
    $result->{conforming_max} = 1;

    $input->generate_transfer_sections ($result);
    $result->generate_result_section;
  }

  $out->nav_list;

  exit;
}

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

  $input->generate_transfer_sections ($result);

  my @subdoc;

  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;

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

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

=pod

  if (defined $doc or defined $el) {

    print_table_section ($input, $elements->{table}) if @{$elements->{table}};
    print_listing_section ({
      id => 'identifiers', label => 'IDs', heading => 'Identifiers',
    }, $input, $elements->{id}) if keys %{$elements->{id}};
    print_listing_section ({
      id => 'terms', label => 'Terms', heading => 'Terms',
    }, $input, $elements->{term}) if keys %{$elements->{term}};
    print_listing_section ({
      id => 'classes', label => 'Classes', heading => 'Classes',
    }, $input, $elements->{class}) if keys %{$elements->{class}};
   
    print_rdf_section ($input, $elements->{rdf}) if @{$elements->{rdf}};
  }

=cut

  my $id_prefix = 0;
  for my $_subinput (@subdoc) {
    my $subinput = WebHACC::Input->new;
    $subinput->{$_} = $_subinput->{$_} for keys %$_subinput;
    $subinput->id_prefix ('subdoc-' . ++$id_prefix);
    $subinput->nested (1);
    $subinput->{base_uri} = $subinput->{container_node}->base_uri
        unless defined $subinput->{base_uri};
    my $ebaseuri = htescape ($subinput->{base_uri});
    $out->start_section (id => $subinput->id_prefix,
                         title => qq[Subdocument #$id_prefix]);
    print STDOUT qq[
      <dl>
      <dt>Internet Media Type</dt>
        <dd><code class="MIME" lang="en">@{[htescape $subinput->{media_type}]}</code>
      <dt>Container Node</dt>
        <dd>@{[get_node_link ($input, $subinput->{container_node})]}</dd>
      <dt>Base <abbr title="Uniform Resource Identifiers">URI</abbr></dt>
        <dd><code class=URI>&lt;<a href="$ebaseuri">$ebaseuri</a>></code></dd>
      </dl>];              

    $subinput->{id_prefix} .= '-';
    check_and_print ($subinput => $result => $out);

    $out->end_section;
  }

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

sub print_table_section ($$) {
  my ($input, $tables) = @_;
  
#  push @nav, [qq[#$input->{id_prefix}tables] => 'Tables']
#      unless $input->{nested};
  print STDOUT qq[
<div id="$input->{id_prefix}tables" class="section">
<h2>Tables</h2>

<!--[if IE]><script type="text/javascript" src="../excanvas.js"></script><![endif]-->
<script src="../table-script.js" type="text/javascript"></script>
<noscript>
<p><em>Structure of tables are visualized here if scripting is enabled.</em></p>
</noscript>
];
  
  require JSON;
  
  my $i = 0;
  for my $table (@$tables) {
    $i++;
    print STDOUT qq[<div class="section" id="$input->{id_prefix}table-$i"><h3>] .
        get_node_link ($input, $table->{element}) . q[</h3>];

    delete $table->{element};

    for (@{$table->{column_group}}, @{$table->{column}}, $table->{caption},
         @{$table->{row}}) {
      next unless $_;
      delete $_->{element};
    }
    
    for (@{$table->{row_group}}) {
      next unless $_;
      next unless $_->{element};
      $_->{type} = $_->{element}->manakai_local_name;
      delete $_->{element};
    }
    
    for (@{$table->{cell}}) {
      next unless $_;
      for (@{$_}) {
        next unless $_;
        for (@$_) {
          $_->{id} = refaddr $_->{element} if defined $_->{element};
          delete $_->{element};
          $_->{is_header} = $_->{is_header} ? 1 : 0;
        }
      }
    }
        
    print STDOUT '</div><script type="text/javascript">tableToCanvas (';
    print STDOUT JSON::objToJson ($table);
    print STDOUT qq[, document.getElementById ('$input->{id_prefix}table-$i')];
    print STDOUT qq[, '$input->{id_prefix}');</script>];
  }
  
  print STDOUT qq[</div>];
} # print_table_section

sub print_listing_section ($$$) {
  my ($opt, $input, $ids) = @_;
  
#  push @nav, ['#' . $input->{id_prefix} . $opt->{id} => $opt->{label}]
#      unless $input->{nested};
  print STDOUT qq[
<div id="$input->{id_prefix}$opt->{id}" class="section">
<h2>$opt->{heading}</h2>

<dl>
];
  for my $id (sort {$a cmp $b} keys %$ids) {
    print STDOUT qq[<dt><code>@{[htescape $id]}</code></dt>];
    for (@{$ids->{$id}}) {
      print STDOUT qq[<dd>].get_node_link ($input, $_).qq[</dd>];
    }
  }
  print STDOUT qq[</dl></div>];
} # print_listing_section


sub print_rdf_section ($$$) {
  my ($input, $rdfs) = @_;
  
#  push @nav, ['#' . $input->{id_prefix} . 'rdf' => 'RDF']
#      unless $input->{nested};
  print STDOUT qq[
<div id="$input->{id_prefix}rdf" class="section">
<h2>RDF Triples</h2>

<dl>];
  my $i = 0;
  for my $rdf (@$rdfs) {
    print STDOUT qq[<dt id="$input->{id_prefix}rdf-@{[$i++]}">];
    print STDOUT get_node_link ($input, $rdf->[0]);
    print STDOUT qq[<dd><dl>];
    for my $triple (@{$rdf->[1]}) {
      print STDOUT '<dt>' . get_node_link ($input, $triple->[0]) . '<dd>';
      print STDOUT get_rdf_resource_html ($triple->[1]);
      print STDOUT ' ';
      print STDOUT get_rdf_resource_html ($triple->[2]);
      print STDOUT ' ';
      print STDOUT get_rdf_resource_html ($triple->[3]);
    }
    print STDOUT qq[</dl>];
  }
  print STDOUT qq[</dl></div>];
} # print_rdf_section

sub get_rdf_resource_html ($) {
  my $resource = shift;
  if (defined $resource->{uri}) {
    my $euri = htescape ($resource->{uri});
    return '<code class=uri>&lt;<a href="' . $euri . '">' . $euri .
        '</a>></code>';
  } elsif (defined $resource->{bnodeid}) {
    return htescape ('_:' . $resource->{bnodeid});
  } elsif ($resource->{nodes}) {
    return '(rdf:XMLLiteral)';
  } elsif (defined $resource->{value}) {
    my $elang = htescape (defined $resource->{language}
                              ? $resource->{language} : '');
    my $r = qq[<q lang="$elang">] . htescape ($resource->{value}) . '</q>';
    if (defined $resource->{datatype}) {
      my $euri = htescape ($resource->{datatype});
      $r .= '^^<code class=uri>&lt;<a href="' . $euri . '">' . $euri .
          '</a>></code>';
    } elsif (length $resource->{language}) {
      $r .= '@' . htescape ($resource->{language});
    }
    return $r;
  } else {
    return '??';
  }
} # get_rdf_resource_html

{
  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, $dom) = @_;

  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: 2008/07/18 14:44:16 $
1	wakaba	1.1	#!/usr/bin/perl
2			use strict;
3	wakaba	1.23	use utf8;
4	wakaba	1.1
5			use lib qw[/home/httpd/html/www/markup/html/whatpm
6	wakaba	1.16	/home/wakaba/work/manakai2/lib];
7	wakaba	1.1	use CGI::Carp qw[fatalsToBrowser];
8	wakaba	1.2	use Scalar::Util qw[refaddr];
9	wakaba	1.1
10	wakaba	1.53	require WebHACC::Input;
11			require WebHACC::Result;
12			require WebHACC::Output;
13
14			my $out;
15	wakaba	1.2
16	wakaba	1.35	require Message::DOM::DOMImplementation;
17			my $dom = Message::DOM::DOMImplementation->new;
18			{
19	wakaba	1.16	use Message::CGI::HTTP;
20			my $http = Message::CGI::HTTP->new;
21	wakaba	1.1
22	wakaba	1.16	if ($http->get_meta_variable ('PATH_INFO') ne '/') {
23	wakaba	1.8	print STDOUT "Status: 404 Not Found\nContent-Type: text/plain; charset=us-ascii\n\n400";
24			exit;
25			}
26	wakaba	1.53
27	wakaba	1.7	load_text_catalog ('en'); ## TODO: conneg
28
29	wakaba	1.53	$out = WebHACC::Output->new;
30			$out->handle (*STDOUT);
31			$out->set_utf8;
32			$out->set_flush;
33			$out->html (qq[Content-Type: text/html; charset=utf-8
34	wakaba	1.2
35			<!DOCTYPE html>
36			<html lang="en">
37			<head>
38			<title>Web Document Conformance Checker (BETA)</title>
39	wakaba	1.3	<link rel="stylesheet" href="../cc-style.css" type="text/css">
40	wakaba	1.2	</head>
41			<body>
42	wakaba	1.13	<h1><a href="../cc-interface">Web Document Conformance Checker</a>
43			(<em>beta</em>)</h1>
44	wakaba	1.53	]);
45	wakaba	1.2
46	wakaba	1.14	my $input = get_input_document ($http, $dom);
47	wakaba	1.53	$out->input ($input);
48			$out->unset_flush;
49
50	wakaba	1.16	my $char_length = 0;
51	wakaba	1.14
52	wakaba	1.53	$out->start_section (id => 'document-info', title => 'Information');
53			$out->html (qq[<dl>
54			<dt>Request URL</dt>
55			<dd>]);
56			$out->url ($input->{request_uri});
57			$out->html (q[<dt>Document URL<!-- HTML5 document's address? -->
58			<dd>]);
59			$out->url ($input->{uri}, id => 'anchor-document-url');
60			$out->html (q[
61	wakaba	1.25	<script>
62			document.title = '<'
63	wakaba	1.53	+ document.getElementById ('anchor-document-url').href + '> \\u2014 '
64	wakaba	1.25	+ document.title;
65	wakaba	1.53	</script>]);
66			## NOTE: no </dl> yet
67	wakaba	1.1
68	wakaba	1.53	if (defined $input->{s}) {
69			$char_length = length $input->{s};
70	wakaba	1.9
71	wakaba	1.53	$out->html (qq[<dt>Base URI<dd>]);
72			$out->url ($input->{base_uri});
73			$out->html (qq[<dt>Internet Media Type</dt>
74			<dd><code class="MIME" lang="en">]);
75			$out->text ($input->{media_type});
76			$out->html (qq[</code> ]);
77			if ($input->{media_type_overridden}) {
78			$out->html ('<em>(overridden)</em>');
79			} elsif (defined $input->{official_type}) {
80			if ($input->{media_type} eq $input->{official_type}) {
81			#
82			} else {
83			$out->html ('<em>(sniffed; official type is: <code class=MIME lang=en>');
84			$out->text ($input->{official_type});
85			$out->html ('</code>)');
86			}
87			} else {
88			$out->html ('<em>(sniffed)</em>');
89			}
90			$out->html (q[<dt>Character Encoding<dd>]);
91			if (defined $input->{charset}) {
92			$out->html ('<code class="charset" lang="en">');
93			$out->text ($input->{charset});
94			$out->html ('</code>');
95			} else {
96			$out->text ('(none)');
97			}
98			$out->html (' <em>overridden</em>') if $input->{charset_overridden};
99			$out->html (qq[
100	wakaba	1.16	<dt>Length</dt>
101			<dd>$char_length byte@{[$char_length == 1 ? '' : 's']}</dd>
102	wakaba	1.9	</dl>
103	wakaba	1.39
104			<script src="../cc-script.js"></script>
105	wakaba	1.53	]);
106			$out->end_section;
107	wakaba	1.9
108	wakaba	1.53	my $result = WebHACC::Result->new;
109	wakaba	1.54	$result->output ($out);
110	wakaba	1.53	$result->{conforming_min} = 1;
111			$result->{conforming_max} = 1;
112			check_and_print ($input => $result => $out);
113	wakaba	1.54	$result->generate_result_section;
114	wakaba	1.53	} else {
115			$out->html ('</dl>');
116			$out->end_section;
117	wakaba	1.54
118			my $result = WebHACC::Result->new;
119			$result->output ($out);
120			$result->{conforming_min} = 0;
121			$result->{conforming_max} = 1;
122
123			$input->generate_transfer_sections ($result);
124			$result->generate_result_section;
125	wakaba	1.3	}
126	wakaba	1.1
127	wakaba	1.53	$out->nav_list;
128	wakaba	1.16
129	wakaba	1.53	exit;
130	wakaba	1.35	}
131	wakaba	1.1
132	wakaba	1.53	sub check_and_print ($$$) {
133			my ($input, $result, $out) = @_;
134			my $original_input = $out->input;
135			$out->input ($input);
136	wakaba	1.31
137	wakaba	1.54	$input->generate_transfer_sections ($result);
138	wakaba	1.31
139	wakaba	1.34	my @subdoc;
140	wakaba	1.31
141	wakaba	1.53	my $checker_class = {
142			'text/cache-manifest' => 'WebHACC::Language::CacheManifest',
143			'text/css' => 'WebHACC::Language::CSS',
144			'text/html' => 'WebHACC::Language::HTML',
145			'text/x-webidl' => 'WebHACC::Language::WebIDL',
146
147			'text/xml' => 'WebHACC::Language::XML',
148			'application/atom+xml' => 'WebHACC::Language::XML',
149			'application/rss+xml' => 'WebHACC::Language::XML',
150			'image/svg+xml' => 'WebHACC::Language::XML',
151			'application/xhtml+xml' => 'WebHACC::Language::XML',
152			'application/xml' => 'WebHACC::Language::XML',
153			## TODO: Should we make all XML MIME Types fall
154			## into this category?
155
156			## NOTE: This type has different model from normal XML types.
157			'application/rdf+xml' => 'WebHACC::Language::XML',
158			}->{$input->{media_type}} \|\| 'WebHACC::Language::Default';
159
160			eval qq{ require $checker_class } or die "$0: Loading $checker_class: $@";
161			my $checker = $checker_class->new;
162			$checker->input ($input);
163			$checker->output ($out);
164			$checker->result ($result);
165
166			## TODO: A cache manifest MUST be text/cache-manifest
167			## TODO: WebIDL media type "text/x-webidl"
168
169			$checker->generate_syntax_error_section;
170			$checker->generate_source_string_section;
171
172			$checker->onsubdoc (sub {
173			push @subdoc, shift;
174			});
175
176			$checker->generate_structure_dump_section;
177			$checker->generate_structure_error_section;
178			$checker->generate_additional_sections;
179
180			=pod
181	wakaba	1.31
182			if (defined $doc or defined $el) {
183	wakaba	1.53
184	wakaba	1.32	print_table_section ($input, $elements->{table}) if @{$elements->{table}};
185	wakaba	1.33	print_listing_section ({
186			id => 'identifiers', label => 'IDs', heading => 'Identifiers',
187			}, $input, $elements->{id}) if keys %{$elements->{id}};
188			print_listing_section ({
189			id => 'terms', label => 'Terms', heading => 'Terms',
190			}, $input, $elements->{term}) if keys %{$elements->{term}};
191			print_listing_section ({
192			id => 'classes', label => 'Classes', heading => 'Classes',
193			}, $input, $elements->{class}) if keys %{$elements->{class}};
194	wakaba	1.53
195	wakaba	1.45	print_rdf_section ($input, $elements->{rdf}) if @{$elements->{rdf}};
196	wakaba	1.31	}
197	wakaba	1.34
198	wakaba	1.53	=cut
199
200	wakaba	1.34	my $id_prefix = 0;
201	wakaba	1.53	for my $_subinput (@subdoc) {
202			my $subinput = WebHACC::Input->new;
203			$subinput->{$_} = $_subinput->{$_} for keys %$_subinput;
204			$subinput->id_prefix ('subdoc-' . ++$id_prefix);
205			$subinput->nested (1);
206	wakaba	1.34	$subinput->{base_uri} = $subinput->{container_node}->base_uri
207			unless defined $subinput->{base_uri};
208			my $ebaseuri = htescape ($subinput->{base_uri});
209	wakaba	1.53	$out->start_section (id => $subinput->id_prefix,
210			title => qq[Subdocument #$id_prefix]);
211			print STDOUT qq[
212	wakaba	1.34	<dl>
213			<dt>Internet Media Type</dt>
214			<dd><code class="MIME" lang="en">@{[htescape $subinput->{media_type}]}</code>
215			<dt>Container Node</dt>
216			<dd>@{[get_node_link ($input, $subinput->{container_node})]}</dd>
217			<dt>Base <abbr title="Uniform Resource Identifiers">URI</abbr></dt>
218			<dd><code class=URI><<a href="$ebaseuri">$ebaseuri</a>></code></dd>
219			</dl>];
220
221	wakaba	1.35	$subinput->{id_prefix} .= '-';
222	wakaba	1.53	check_and_print ($subinput => $result => $out);
223	wakaba	1.34
224	wakaba	1.53	$out->end_section;
225	wakaba	1.34	}
226	wakaba	1.53
227			$out->input ($original_input);
228	wakaba	1.31	} # check_and_print
229
230	wakaba	1.32	sub print_table_section ($$) {
231			my ($input, $tables) = @_;
232	wakaba	1.18
233	wakaba	1.53	# push @nav, [qq[#$input->{id_prefix}tables] => 'Tables']
234			# unless $input->{nested};
235	wakaba	1.18	print STDOUT qq[
236	wakaba	1.32	<div id="$input->{id_prefix}tables" class="section">
237	wakaba	1.18	<h2>Tables</h2>
238
239			<!--[if IE]><script type="text/javascript" src="../excanvas.js"></script><![endif]-->
240			<script src="../table-script.js" type="text/javascript"></script>
241			<noscript>
242			<p><em>Structure of tables are visualized here if scripting is enabled.</em></p>
243			</noscript>
244			];
245
246			require JSON;
247
248			my $i = 0;
249	wakaba	1.50	for my $table (@$tables) {
250	wakaba	1.18	$i++;
251	wakaba	1.32	print STDOUT qq[<div class="section" id="$input->{id_prefix}table-$i"><h3>] .
252	wakaba	1.50	get_node_link ($input, $table->{element}) . q[</h3>];
253
254			delete $table->{element};
255	wakaba	1.18
256	wakaba	1.49	for (@{$table->{column_group}}, @{$table->{column}}, $table->{caption},
257			@{$table->{row}}) {
258	wakaba	1.18	next unless $_;
259			delete $_->{element};
260			}
261
262			for (@{$table->{row_group}}) {
263			next unless $_;
264			next unless $_->{element};
265			$_->{type} = $_->{element}->manakai_local_name;
266			delete $_->{element};
267			}
268
269			for (@{$table->{cell}}) {
270			next unless $_;
271			for (@{$_}) {
272			next unless $_;
273			for (@$_) {
274			$_->{id} = refaddr $_->{element} if defined $_->{element};
275			delete $_->{element};
276			$_->{is_header} = $_->{is_header} ? 1 : 0;
277			}
278			}
279			}
280
281			print STDOUT '</div><script type="text/javascript">tableToCanvas (';
282			print STDOUT JSON::objToJson ($table);
283	wakaba	1.32	print STDOUT qq[, document.getElementById ('$input->{id_prefix}table-$i')];
284			print STDOUT qq[, '$input->{id_prefix}');</script>];
285	wakaba	1.18	}
286
287			print STDOUT qq[</div>];
288			} # print_table_section
289
290	wakaba	1.33	sub print_listing_section ($$$) {
291			my ($opt, $input, $ids) = @_;
292	wakaba	1.18
293	wakaba	1.53	# push @nav, ['#' . $input->{id_prefix} . $opt->{id} => $opt->{label}]
294			# unless $input->{nested};
295	wakaba	1.18	print STDOUT qq[
296	wakaba	1.33	<div id="$input->{id_prefix}$opt->{id}" class="section">
297			<h2>$opt->{heading}</h2>
298	wakaba	1.18
299			<dl>
300			];
301			for my $id (sort {$a cmp $b} keys %$ids) {
302			print STDOUT qq[<dt><code>@{[htescape $id]}</code></dt>];
303			for (@{$ids->{$id}}) {
304	wakaba	1.32	print STDOUT qq[<dd>].get_node_link ($input, $_).qq[</dd>];
305	wakaba	1.18	}
306			}
307			print STDOUT qq[</dl></div>];
308	wakaba	1.33	} # print_listing_section
309	wakaba	1.18
310	wakaba	1.48
311	wakaba	1.45	sub print_rdf_section ($$$) {
312			my ($input, $rdfs) = @_;
313
314	wakaba	1.53	# push @nav, ['#' . $input->{id_prefix} . 'rdf' => 'RDF']
315			# unless $input->{nested};
316	wakaba	1.45	print STDOUT qq[
317			<div id="$input->{id_prefix}rdf" class="section">
318			<h2>RDF Triples</h2>
319
320			<dl>];
321			my $i = 0;
322			for my $rdf (@$rdfs) {
323			print STDOUT qq[<dt id="$input->{id_prefix}rdf-@{[$i++]}">];
324			print STDOUT get_node_link ($input, $rdf->[0]);
325			print STDOUT qq[<dd><dl>];
326			for my $triple (@{$rdf->[1]}) {
327			print STDOUT '<dt>' . get_node_link ($input, $triple->[0]) . '<dd>';
328			print STDOUT get_rdf_resource_html ($triple->[1]);
329			print STDOUT ' ';
330			print STDOUT get_rdf_resource_html ($triple->[2]);
331			print STDOUT ' ';
332			print STDOUT get_rdf_resource_html ($triple->[3]);
333			}
334			print STDOUT qq[</dl>];
335			}
336			print STDOUT qq[</dl></div>];
337			} # print_rdf_section
338
339			sub get_rdf_resource_html ($) {
340			my $resource = shift;
341	wakaba	1.46	if (defined $resource->{uri}) {
342	wakaba	1.45	my $euri = htescape ($resource->{uri});
343			return '<code class=uri><<a href="' . $euri . '">' . $euri .
344			'</a>></code>';
345	wakaba	1.46	} elsif (defined $resource->{bnodeid}) {
346	wakaba	1.45	return htescape ('_:' . $resource->{bnodeid});
347			} elsif ($resource->{nodes}) {
348			return '(rdf:XMLLiteral)';
349			} elsif (defined $resource->{value}) {
350			my $elang = htescape (defined $resource->{language}
351			? $resource->{language} : '');
352			my $r = qq[<q lang="$elang">] . htescape ($resource->{value}) . '</q>';
353			if (defined $resource->{datatype}) {
354			my $euri = htescape ($resource->{datatype});
355			$r .= '^^<code class=uri><<a href="' . $euri . '">' . $euri .
356			'</a>></code>';
357			} elsif (length $resource->{language}) {
358			$r .= '@' . htescape ($resource->{language});
359			}
360			return $r;
361			} else {
362			return '??';
363			}
364			} # get_rdf_resource_html
365
366	wakaba	1.7	{
367			my $Msg = {};
368
369			sub load_text_catalog ($) {
370	wakaba	1.53	# my $self = shift;
371	wakaba	1.7	my $lang = shift; # MUST be a canonical lang name
372	wakaba	1.26	open my $file, '<:utf8', "cc-msg.$lang.txt"
373			or die "$0: cc-msg.$lang.txt: $!";
374	wakaba	1.7	while (<$file>) {
375			if (s/^([^;]+);([^;]*);//) {
376			my ($type, $cls, $msg) = ($1, $2, $_);
377			$msg =~ tr/\x0D\x0A//d;
378			$Msg->{$type} = [$cls, $msg];
379			}
380			}
381			} # load_text_catalog
382
383	wakaba	1.53	sub get_text ($;$$) {
384			# my $self = shift;
385	wakaba	1.15	my ($type, $level, $node) = @_;
386	wakaba	1.7	$type = $level . ':' . $type if defined $level;
387	wakaba	1.29	$level = 'm' unless defined $level;
388	wakaba	1.7	my @arg;
389			{
390			if (defined $Msg->{$type}) {
391			my $msg = $Msg->{$type}->[1];
392	wakaba	1.10	$msg =~ s{<var>\$([0-9]+)</var>}{
393	wakaba	1.53	defined $arg[$1] ? ($arg[$1]) : '(undef)';
394			}ge; ##BUG: ^ must be escaped
395	wakaba	1.15	$msg =~ s{<var>{\@([A-Za-z0-9:_.-]+)}</var>}{
396			UNIVERSAL::can ($node, 'get_attribute_ns')
397	wakaba	1.53	? ($node->get_attribute_ns (undef, $1)) : ''
398			}ge; ## BUG: ^ must be escaped
399			$msg =~ s{<var>{\@}</var>}{ ## BUG: v must be escaped
400			UNIVERSAL::can ($node, 'value') ? ($node->value) : ''
401	wakaba	1.15	}ge;
402	wakaba	1.17	$msg =~ s{<var>{local-name}</var>}{
403			UNIVERSAL::can ($node, 'manakai_local_name')
404	wakaba	1.53	? ($node->manakai_local_name) : ''
405			}ge; ## BUG: ^ must be escaped
406	wakaba	1.17	$msg =~ s{<var>{element-local-name}</var>}{
407			(UNIVERSAL::can ($node, 'owner_element') and
408			$node->owner_element)
409	wakaba	1.53	? ($node->owner_element->manakai_local_name)
410			: '' ## BUG: ^ must be escaped
411	wakaba	1.17	}ge;
412	wakaba	1.29	return ($type, 'level-' . $level . ' ' . $Msg->{$type}->[0], $msg);
413	wakaba	1.7	} elsif ($type =~ s/:([^:]*)$//) {
414			unshift @arg, $1;
415			redo;
416			}
417			}
418	wakaba	1.53	return ($type, 'level-'.$level, ($_[0]));
419			## BUG: ^ must be escaped
420	wakaba	1.7	} # get_text
421
422			}
423
424	wakaba	1.9	sub get_input_document ($$) {
425			my ($http, $dom) = @_;
426
427	wakaba	1.54	require Encode;
428			my $request_uri = Encode::decode ('utf-8', $http->get_parameter ('uri'));
429	wakaba	1.53	my $r = WebHACC::Input->new;
430	wakaba	1.9	if (defined $request_uri and length $request_uri) {
431			my $uri = $dom->create_uri_reference ($request_uri);
432			unless ({
433			http => 1,
434			}->{lc $uri->uri_scheme}) {
435	wakaba	1.54	$r = WebHACC::Input::Error->new;
436			$r->{uri} = $request_uri;
437			$r->{request_uri} = $request_uri;
438			$r->{error_status_text} = 'URL scheme not allowed';
439	wakaba	1.9	}
440
441			require Message::Util::HostPermit;
442			my $host_permit = new Message::Util::HostPermit;
443			$host_permit->add_rule (<<EOH);
444			Allow host=suika port=80
445			Deny host=suika
446			Allow host=suika.fam.cx port=80
447			Deny host=suika.fam.cx
448			Deny host=localhost
449			Deny host=*.localdomain
450			Deny ipv4=0.0.0.0/8
451			Deny ipv4=10.0.0.0/8
452			Deny ipv4=127.0.0.0/8
453			Deny ipv4=169.254.0.0/16
454			Deny ipv4=172.0.0.0/11
455			Deny ipv4=192.0.2.0/24
456			Deny ipv4=192.88.99.0/24
457			Deny ipv4=192.168.0.0/16
458			Deny ipv4=198.18.0.0/15
459			Deny ipv4=224.0.0.0/4
460			Deny ipv4=255.255.255.255/32
461			Deny ipv6=0::0/0
462			Allow host=*
463			EOH
464			unless ($host_permit->check ($uri->uri_host, $uri->uri_port \|\| 80)) {
465	wakaba	1.54	my $r = WebHACC::Input::Error->new;
466			$r->{uri} = $request_uri;
467			$r->{request_uri} = $request_uri;
468			$r->{error_status_text} = 'Connection to the host is forbidden';
469			return $r;
470	wakaba	1.9	}
471
472			require LWP::UserAgent;
473			my $ua = WDCC::LWPUA->new;
474			$ua->{wdcc_dom} = $dom;
475			$ua->{wdcc_host_permit} = $host_permit;
476			$ua->agent ('Mozilla'); ## TODO: for now.
477			$ua->parse_head (0);
478			$ua->protocols_allowed ([qw/http/]);
479			$ua->max_size (1000_000);
480			my $req = HTTP::Request->new (GET => $request_uri);
481	wakaba	1.28	$req->header ('Accept-Encoding' => 'identity, *; q=0');
482	wakaba	1.9	my $res = $ua->request ($req);
483	wakaba	1.16	## TODO: 401 sets \|is_success\| true.
484			if ($res->is_success or $http->get_parameter ('error-page')) {
485	wakaba	1.9	$r->{base_uri} = $res->base; ## NOTE: It does check \|Content-Base\|, \|Content-Location\|, and <base>. ## TODO: Use our own code!
486			$r->{uri} = $res->request->uri;
487			$r->{request_uri} = $request_uri;
488
489			## TODO: More strict parsing...
490			my $ct = $res->header ('Content-Type');
491	wakaba	1.22	if (defined $ct and $ct =~ /;\scharset\s=\s*"?([^\s;"]+)"?/i) {
492	wakaba	1.9	$r->{charset} = lc $1;
493			$r->{charset} =~ tr/\\//d;
494	wakaba	1.26	$r->{official_charset} = $r->{charset};
495	wakaba	1.9	}
496
497	wakaba	1.16	my $input_charset = $http->get_parameter ('charset');
498	wakaba	1.9	if (defined $input_charset and length $input_charset) {
499			$r->{charset_overridden}
500			= (not defined $r->{charset} or $r->{charset} ne $input_charset);
501			$r->{charset} = $input_charset;
502	wakaba	1.25	}
503
504			## TODO: Support for HTTP Content-Encoding
505	wakaba	1.9
506			$r->{s} = ''.$res->content;
507	wakaba	1.25
508			require Whatpm::ContentType;
509			($r->{official_type}, $r->{media_type})
510			= Whatpm::ContentType->get_sniffed_type
511			(get_file_head => sub {
512			return substr $r->{s}, 0, shift;
513			},
514			http_content_type_byte => $ct,
515			has_http_content_encoding =>
516			defined $res->header ('Content-Encoding'),
517			supported_image_types => {});
518	wakaba	1.9	} else {
519			$r->{uri} = $res->request->uri;
520			$r->{request_uri} = $request_uri;
521			$r->{error_status_text} = $res->status_line;
522			}
523
524			$r->{header_field} = [];
525			$res->scan (sub {
526			push @{$r->{header_field}}, [$_[0], $_[1]];
527			});
528			$r->{header_status_code} = $res->code;
529			$r->{header_status_text} = $res->message;
530			} else {
531	wakaba	1.16	$r->{s} = ''.$http->get_parameter ('s');
532	wakaba	1.9	$r->{uri} = q<thismessage:/>;
533			$r->{request_uri} = q<thismessage:/>;
534			$r->{base_uri} = q<thismessage:/>;
535	wakaba	1.16	$r->{charset} = ''.$http->get_parameter ('_charset_');
536	wakaba	1.9	$r->{charset} =~ s/\s+//g;
537			$r->{charset} = 'utf-8' if $r->{charset} eq '';
538	wakaba	1.26	$r->{official_charset} = $r->{charset};
539	wakaba	1.9	$r->{header_field} = [];
540	wakaba	1.25
541			require Whatpm::ContentType;
542			($r->{official_type}, $r->{media_type})
543			= Whatpm::ContentType->get_sniffed_type
544			(get_file_head => sub {
545			return substr $r->{s}, 0, shift;
546			},
547			http_content_type_byte => undef,
548			has_http_content_encoding => 0,
549			supported_image_types => {});
550	wakaba	1.9	}
551
552	wakaba	1.16	my $input_format = $http->get_parameter ('i');
553	wakaba	1.9	if (defined $input_format and length $input_format) {
554			$r->{media_type_overridden}
555			= (not defined $r->{media_type} or $input_format ne $r->{media_type});
556			$r->{media_type} = $input_format;
557			}
558			if (defined $r->{s} and not defined $r->{media_type}) {
559			$r->{media_type} = 'text/html';
560			$r->{media_type_overridden} = 1;
561			}
562
563			if ($r->{media_type} eq 'text/xml') {
564			unless (defined $r->{charset}) {
565			$r->{charset} = 'us-ascii';
566	wakaba	1.26	$r->{official_charset} = $r->{charset};
567	wakaba	1.9	} elsif ($r->{charset_overridden} and $r->{charset} eq 'us-ascii') {
568			$r->{charset_overridden} = 0;
569			}
570			}
571
572			if (length $r->{s} > 1000_000) {
573			$r->{error_status_text} = 'Entity-body too large';
574			delete $r->{s};
575			return $r;
576			}
577
578	wakaba	1.35	$r->{inner_html_element} = $http->get_parameter ('e');
579
580	wakaba	1.9	return $r;
581			} # get_input_document
582
583			package WDCC::LWPUA;
584			BEGIN { push our @ISA, 'LWP::UserAgent'; }
585
586			sub redirect_ok {
587			my $ua = shift;
588			unless ($ua->SUPER::redirect_ok (@_)) {
589			return 0;
590			}
591
592			my $uris = $_[1]->header ('Location');
593			return 0 unless $uris;
594			my $uri = $ua->{wdcc_dom}->create_uri_reference ($uris);
595			unless ({
596			http => 1,
597			}->{lc $uri->uri_scheme}) {
598			return 0;
599			}
600			unless ($ua->{wdcc_host_permit}->check ($uri->uri_host, $uri->uri_port \|\| 80)) {
601			return 0;
602			}
603			return 1;
604			} # redirect_ok
605
606	wakaba	1.1	=head1 AUTHOR
607
608			Wakaba <w@suika.fam.cx>.
609
610			=head1 LICENSE
611
612	wakaba	1.35	Copyright 2007-2008 Wakaba <w@suika.fam.cx>
613	wakaba	1.1
614			This library is free software; you can redistribute it
615			and/or modify it under the same terms as Perl itself.
616
617			=cut
618
619	wakaba	1.53	## $Date: 2008/07/18 14:44:16 $