/[suikacvs]/test/html-webhacc/cc.cgi
Suika

Diff of /test/html-webhacc/cc.cgi

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.1 by wakaba, Wed Jun 27 11:08:03 2007 UTC revision 1.59 by wakaba, Mon Jul 21 12:56:33 2008 UTC
# Line 2  Line 2 
2  use strict;  use strict;
3    
4  use lib qw[/home/httpd/html/www/markup/html/whatpm  use lib qw[/home/httpd/html/www/markup/html/whatpm
5             /home/wakaba/work/manakai/lib             /home/wakaba/work/manakai2/lib];
            /home/wakaba/public_html/-temp/wiki/lib];  
6  use CGI::Carp qw[fatalsToBrowser];  use CGI::Carp qw[fatalsToBrowser];
 use Time::HiRes qw/time/;  
7    
8  use SuikaWiki::Input::HTTP; ## TODO: Use some better CGI module    require WebHACC::Input;
9    
10  my $http = SuikaWiki::Input::HTTP->new;  {
11      require Message::CGI::HTTP;
12      my $http = Message::CGI::HTTP->new;
13    
14      require WebHACC::Output;
15      my $out = WebHACC::Output->new;
16      $out->handle (*STDOUT);
17      $out->set_utf8;
18    
19  ## TODO: _charset_    if ($http->get_meta_variable ('PATH_INFO') ne '/') {
20        $out->http_error (404);
 my @mode = split m#/#, scalar $http->meta_variable ('PATH_INFO'), -1;  
 shift @mode if @mode and $mode[0] == '';  
 ## TODO: decode unreserved characters  
   
   my $s = $http->parameter ('s');  
   if (length $s > 1000_000) {  
     print STDOUT "Status: 400 Document Too Long\nContent-Type: text/plain; charset=us-ascii\n\nToo long";  
21      exit;      exit;
22    }    }
   my $char_length = length $s;  
   my %time;  
   my $time1;  
   my $time2;  
23    
24    require Message::DOM::DOMImplementation;    ## TODO: We need real conneg support...
25    my $dom = Message::DOM::DOMImplementation->____new;    my $primary_language = 'en';
26  #  $| = 1;    if ($ENV{HTTP_ACCEPT_LANGUAGE} =~ /ja/) {
27    my $doc;      $primary_language = 'ja';
28    my $el;    }
29      $out->load_text_catalog ($primary_language);
30  if (@mode == 3 and $mode[0] eq 'html' and    
31      ($mode[2] eq 'html' or $mode[2] eq 'test')) {    $out->set_flush;
32    print STDOUT "Content-Type: text/plain; charset=utf-8\n\n";    $out->http_header;
33      $out->html_header;
34      $out->unset_flush;
35    
36      my $input = get_input_document ($http);
37      $out->input ($input);
38    
39      require WebHACC::Result;
40      my $result = WebHACC::Result->new;
41      $result->output ($out);
42      $result->{conforming_min} = 1;
43      $result->{conforming_max} = 1;
44    
45    require Encode;    $out->html ('<script src="../cc-script.js"></script>');
   require Whatpm::HTML;  
46    
47    $time1 = time;    check_and_print ($input => $result => $out);
   $s = Encode::decode ('utf-8', $s);  
   $time2 = time;  
   $time{decode} = $time2 - $time1;  
48        
49      $result->generate_result_section;
50    
51    print STDOUT "#errors\n";    $out->nav_list;
52    
53    my $onerror = sub {    exit;
54      my (%opt) = @_;  }
55      print STDOUT "$opt{line},$opt{column},$opt{type}\n";  
56    };  sub check_and_print ($$$) {
57      my ($input, $result, $out) = @_;
58    $doc = $dom->create_document;    my $original_input = $out->input;
59    $time1 = time;    $out->input ($input);
60    if (length $mode[1]) {  
61      $el = $doc->create_element_ns    $input->generate_info_section ($result);
62          ('http://www.w3.org/1999/xhtml', [undef, $mode[1]]);  
63      Whatpm::HTML->set_inner_html ($el, $s, $onerror);    $input->generate_transfer_sections ($result);
64    } else {  
65      Whatpm::HTML->parse_string ($s => $doc, $onerror);    unless (defined $input->{s}) {
66        $result->{conforming_min} = 0;
67        return;
68      }
69    
70      my $checker_class = {
71        'text/cache-manifest' => 'WebHACC::Language::CacheManifest',
72        'text/css' => 'WebHACC::Language::CSS',
73        'text/html' => 'WebHACC::Language::HTML',
74        'text/x-webidl' => 'WebHACC::Language::WebIDL',
75    
76        'text/xml' => 'WebHACC::Language::XML',
77        'application/atom+xml' => 'WebHACC::Language::XML',
78        'application/rss+xml' => 'WebHACC::Language::XML',
79        'image/svg+xml' => 'WebHACC::Language::XML',
80        'application/xhtml+xml' => 'WebHACC::Language::XML',
81        'application/xml' => 'WebHACC::Language::XML',
82        ## TODO: Should we make all XML MIME Types fall
83        ## into this category?
84    
85        ## NOTE: This type has different model from normal XML types.
86        'application/rdf+xml' => 'WebHACC::Language::XML',
87      }->{$input->{media_type}} || 'WebHACC::Language::Default';
88    
89      eval qq{ require $checker_class } or die "$0: Loading $checker_class: $@";
90      my $checker = $checker_class->new;
91      $checker->input ($input);
92      $checker->output ($out);
93      $checker->result ($result);
94    
95      ## TODO: A cache manifest MUST be text/cache-manifest
96      ## TODO: WebIDL media type "text/x-webidl"
97    
98      $checker->generate_syntax_error_section;
99      $checker->generate_source_string_section;
100    
101      my @subdoc;
102      $checker->onsubdoc (sub {
103        push @subdoc, shift;
104      });
105    
106      $checker->generate_structure_dump_section;
107      $checker->generate_structure_error_section;
108      $checker->generate_additional_sections;
109    
110      my $id_prefix = 0;
111      for my $_subinput (@subdoc) {
112        my $subinput = WebHACC::Input::Subdocument->new (++$id_prefix);
113        $subinput->{$_} = $_subinput->{$_} for keys %$_subinput;
114        $subinput->{base_uri} = $subinput->{container_node}->base_uri
115            unless defined $subinput->{base_uri};
116        $subinput->{parent_input} = $input;
117    
118        $subinput->start_section ($result);
119        check_and_print ($subinput => $result => $out);
120        $subinput->end_section ($result);
121    }    }
   $time2 = time;  
   $time{parse} = $time2 - $time1;  
122    
123    print "#document\n";    $out->input ($original_input);
124    } # check_and_print
125    
126    my $out;  sub get_input_document ($) {
127    if ($mode[2] eq 'html') {    my $http = shift;
     $time1 = time;  
     $out = Whatpm::HTML->get_inner_html ($el || $doc);  
     $time2 = time;  
     $time{serialize_html} = $time2 - $time1;  
   } else { # test  
     $time1 = time;  
     $out = test_serialize ($el || $doc);  
     $time2 = time;  
     $time{serialize_test} = $time2 - $time1;  
   }  
   print STDOUT Encode::encode ('utf-8', $$out);  
   print STDOUT "\n";  
 } elsif (@mode == 3 and $mode[0] eq 'xhtml' and  
          ($mode[2] eq 'html' or $mode[2] eq 'test')) {  
   print STDOUT "Content-Type: text/plain; charset=utf-8\n\n";  
   
   require Message::DOM::XMLParserTemp;  
   print STDOUT "#errors\n";  
   
   my $onerror = sub {  
     my $err = shift;  
     print STDOUT $err->location->line_number, ",";  
     print STDOUT $err->location->column_number, ",";  
     print STDOUT $err->text, "\n";  
     return 1;  
   };  
   
   open my $fh, '<', \$s;  
   my $time1 = time;  
   $doc = Message::DOM::XMLParserTemp->parse_byte_stream  
       ($fh => $dom, $onerror, charset => 'utf-8');  
   my $time2 = time;  
   $time{parse_xml} = $time2 - $time1;  
   
   print "#document\n";  
   
   my $out;  
   if ($mode[2] eq 'html') {  
     ## TODO: Use XHTML serializer  
     #$out = Whatpm::HTML->get_inner_html ($doc);  
   } else { # test  
     $time1 = time;  
     $out = test_serialize ($doc);  
     $time2 = time;  
     $time{serialize_test} = $time2 - $time1;  
   }  
   print STDOUT Encode::encode ('utf-8', $$out);  
   print STDOUT "\n";  
 } else {  
   print STDOUT "Status: 404 Not Found\nContent-Type: text/plain; charset=us-ascii\n\n404";  
   exit;  
 }  
128    
129    if ($http->parameter ('dom5')) {    require Message::DOM::DOMImplementation;
130      require Whatpm::ContentChecker;    my $dom = Message::DOM::DOMImplementation->new;
131      my $onerror = sub {  
132        my %opt = @_;    require Encode;
133        print STDOUT get_node_path ($opt{node}) . ';' . $opt{type} . "\n";    my $request_uri = Encode::decode ('utf-8', $http->get_parameter ('uri'));
134      };    my $r = WebHACC::Input->new;
135      print STDOUT "#domerrors\n";    if (defined $request_uri and length $request_uri) {
136      $time1 = time;      my $uri = $dom->create_uri_reference ($request_uri);
137      if ($el) {      unless ({
138        Whatpm::ContentChecker->check_element ($el, $onerror);               http => 1,
139      } else {              }->{lc $uri->uri_scheme}) {
140        Whatpm::ContentChecker->check_document ($doc, $onerror);        $r = WebHACC::Input::Error->new;
141          $r->{uri} = $request_uri;
142          $r->{request_uri} = $request_uri;
143          $r->{error_status_text} = 'URL scheme not allowed';
144      }      }
     $time2 = time;  
     $time{check} = $time2 - $time1;  
   }  
145    
146    print STDOUT "#log\n";      require Message::Util::HostPermit;
147    for (qw/decode parse parse_xml serialize_html serialize_xml serialize_test      my $host_permit = new Message::Util::HostPermit;
148            check/) {      $host_permit->add_rule (<<EOH);
149      next unless defined $time{$_};  Allow host=suika port=80
150      print STDOUT {  Deny host=suika
151        decode => 'bytes->chars',  Allow host=suika.fam.cx port=80
152        parse => 'html5(chars)->dom5',  Deny host=suika.fam.cx
153        parse_xml => 'xml1(chars)->dom5',  Deny host=localhost
154        serialize_html => 'dom5->html5(char)',  Deny host=*.localdomain
155        serialize_xml => 'dom5->xml1(char)',  Deny ipv4=0.0.0.0/8
156        serialize_test => 'dom5->test(char)',  Deny ipv4=10.0.0.0/8
157        check => 'dom5 check',  Deny ipv4=127.0.0.0/8
158      }->{$_};  Deny ipv4=169.254.0.0/16
159      print STDOUT "\t", $time{$_}, "s\n";  Deny ipv4=172.0.0.0/11
160      open my $file, '>>', ".manakai-$_.txt" or die ".manakai-$_.txt: $!";  Deny ipv4=192.0.2.0/24
161      print $file $char_length, "\t", $time{$_}, "\n";  Deny ipv4=192.88.99.0/24
162    }  Deny ipv4=192.168.0.0/16
163    Deny ipv4=198.18.0.0/15
164  exit;  Deny ipv4=224.0.0.0/4
165    Deny ipv4=255.255.255.255/32
166  sub test_serialize ($) {  Deny ipv6=0::0/0
167    my $node = shift;  Allow host=*
168    my $r = '';  EOH
169        unless ($host_permit->check ($uri->uri_host, $uri->uri_port || 80)) {
170    my @node = map { [$_, ''] } @{$node->child_nodes};        my $r = WebHACC::Input::Error->new;
171    while (@node) {        $r->{uri} = $request_uri;
172      my $child = shift @node;        $r->{request_uri} = $request_uri;
173      my $nt = $child->[0]->node_type;        $r->{error_status_text} = 'Connection to the host is forbidden';
174      if ($nt == $child->[0]->ELEMENT_NODE) {        return $r;
175        $r .= '| ' . $child->[1] . '<' . $child->[0]->tag_name . ">\x0A"; ## ISSUE: case?      }
176    
177        for my $attr (sort {$a->[0] cmp $b->[0]} map { [$_->name, $_->value] }      require LWP::UserAgent;
178                      @{$child->[0]->attributes}) {      my $ua = WDCC::LWPUA->new;
179          $r .= '| ' . $child->[1] . '  ' . $attr->[0] . '="'; ## ISSUE: case?      $ua->{wdcc_dom} = $dom;
180          $r .= $attr->[1] . '"' . "\x0A";      $ua->{wdcc_host_permit} = $host_permit;
181        $ua->agent ('Mozilla'); ## TODO: for now.
182        $ua->parse_head (0);
183        $ua->protocols_allowed ([qw/http/]);
184        $ua->max_size (1000_000);
185        my $req = HTTP::Request->new (GET => $request_uri);
186        $req->header ('Accept-Encoding' => 'identity, *; q=0');
187        my $res = $ua->request ($req);
188        ## TODO: 401 sets |is_success| true.
189        if ($res->is_success or $http->get_parameter ('error-page')) {
190          $r->{base_uri} = $res->base; ## NOTE: It does check |Content-Base|, |Content-Location|, and <base>. ## TODO: Use our own code!
191          $r->{uri} = $res->request->uri;
192          $r->{request_uri} = $request_uri;
193    
194          ## TODO: More strict parsing...
195          my $ct = $res->header ('Content-Type');
196          if (defined $ct and $ct =~ /;\s*charset\s*=\s*"?([^\s;"]+)"?/i) {
197            $r->{charset} = lc $1;
198            $r->{charset} =~ tr/\\//d;
199            $r->{official_charset} = $r->{charset};
200        }        }
201          
202        unshift @node,        my $input_charset = $http->get_parameter ('charset');
203          map { [$_, $child->[1] . '  '] } @{$child->[0]->child_nodes};        if (defined $input_charset and length $input_charset) {
204      } elsif ($nt == $child->[0]->TEXT_NODE) {          $r->{charset_overridden}
205        $r .= '| ' . $child->[1] . '"' . $child->[0]->data . '"' . "\x0A";              = (not defined $r->{charset} or $r->{charset} ne $input_charset);
206      } elsif ($nt == $child->[0]->CDATA_SECTION_NODE) {          $r->{charset} = $input_charset;
207        $r .= '| ' . $child->[1] . '<![CDATA[' . $child->[0]->data . "]]>\x0A";        }
208      } elsif ($nt == $child->[0]->COMMENT_NODE) {  
209        $r .= '| ' . $child->[1] . '<!-- ' . $child->[0]->data . " -->\x0A";        ## TODO: Support for HTTP Content-Encoding
210      } elsif ($nt == $child->[0]->DOCUMENT_TYPE_NODE) {  
211        $r .= '| ' . $child->[1] . '<!DOCTYPE ' . $child->[0]->name . ">\x0A";        $r->{s} = ''.$res->content;
212      } elsif ($nt == $child->[0]->PROCESSING_INSTRUCTION_NODE) {  
213        $r .= '| ' . $child->[1] . '<?' . $child->[0]->target . ' ' .        require Whatpm::ContentType;
214            $child->[0]->data . "?>\x0A";        ($r->{official_type}, $r->{media_type})
215              = Whatpm::ContentType->get_sniffed_type
216                  (get_file_head => sub {
217                     return substr $r->{s}, 0, shift;
218                   },
219                   http_content_type_byte => $ct,
220                   has_http_content_encoding =>
221                       defined $res->header ('Content-Encoding'),
222                   supported_image_types => {});
223      } else {      } else {
224        $r .= '| ' . $child->[1] . $child->[0]->node_type . "\x0A"; # error        $r->{uri} = $res->request->uri;
225          $r->{request_uri} = $request_uri;
226          $r->{error_status_text} = $res->status_line;
227      }      }
228    
229        $r->{header_field} = [];
230        $res->scan (sub {
231          push @{$r->{header_field}}, [$_[0], $_[1]];
232        });
233        $r->{header_status_code} = $res->code;
234        $r->{header_status_text} = $res->message;
235      } else {
236        $r->{s} = ''.$http->get_parameter ('s');
237        $r->{uri} = q<thismessage:/>;
238        $r->{request_uri} = q<thismessage:/>;
239        $r->{base_uri} = q<thismessage:/>;
240        $r->{charset} = ''.$http->get_parameter ('_charset_');
241        $r->{charset} =~ s/\s+//g;
242        $r->{charset} = 'utf-8' if $r->{charset} eq '';
243        $r->{official_charset} = $r->{charset};
244        $r->{header_field} = [];
245    
246        require Whatpm::ContentType;
247        ($r->{official_type}, $r->{media_type})
248            = Whatpm::ContentType->get_sniffed_type
249                (get_file_head => sub {
250                   return substr $r->{s}, 0, shift;
251                 },
252                 http_content_type_byte => undef,
253                 has_http_content_encoding => 0,
254                 supported_image_types => {});
255    }    }
     
   return \$r;  
 } # test_serialize  
256    
257  sub get_node_path ($) {    my $input_format = $http->get_parameter ('i');
258    my $node = shift;    if (defined $input_format and length $input_format) {
259    my @r;      $r->{media_type_overridden}
260    while (defined $node) {          = (not defined $r->{media_type} or $input_format ne $r->{media_type});
261      my $rs;      $r->{media_type} = $input_format;
262      if ($node->node_type == 1) {    }
263        $rs = $node->manakai_local_name;    if (defined $r->{s} and not defined $r->{media_type}) {
264        $node = $node->parent_node;      $r->{media_type} = 'text/html';
265      } elsif ($node->node_type == 2) {      $r->{media_type_overridden} = 1;
266        $rs = '@' . $node->manakai_local_name;    }
267        $node = $node->owner_element;  
268      } elsif ($node->node_type == 3) {    if ($r->{media_type} eq 'text/xml') {
269        $rs = '"' . $node->data . '"';      unless (defined $r->{charset}) {
270        $node = $node->parent_node;        $r->{charset} = 'us-ascii';
271      } elsif ($node->node_type == 9) {        $r->{official_charset} = $r->{charset};
272        $rs = '';      } elsif ($r->{charset_overridden} and $r->{charset} eq 'us-ascii') {
273        $node = $node->parent_node;        $r->{charset_overridden} = 0;
     } else {  
       $rs = '#' . $node->node_type;  
       $node = $node->parent_node;  
274      }      }
     unshift @r, $rs;  
275    }    }
276    return join '/', @r;  
277  } # get_node_path    if (length $r->{s} > 1000_000) {
278        $r->{error_status_text} = 'Entity-body too large';
279        delete $r->{s};
280        return $r;
281      }
282    
283      $r->{inner_html_element} = $http->get_parameter ('e');
284    
285      return $r;
286    } # get_input_document
287    
288    package WDCC::LWPUA;
289    BEGIN { push our @ISA, 'LWP::UserAgent'; }
290    
291    sub redirect_ok {
292      my $ua = shift;
293      unless ($ua->SUPER::redirect_ok (@_)) {
294        return 0;
295      }
296    
297      my $uris = $_[1]->header ('Location');
298      return 0 unless $uris;
299      my $uri = $ua->{wdcc_dom}->create_uri_reference ($uris);
300      unless ({
301               http => 1,
302              }->{lc $uri->uri_scheme}) {
303        return 0;
304      }
305      unless ($ua->{wdcc_host_permit}->check ($uri->uri_host, $uri->uri_port || 80)) {
306        return 0;
307      }
308      return 1;
309    } # redirect_ok
310    
311  =head1 AUTHOR  =head1 AUTHOR
312    
# Line 229  Wakaba <w@suika.fam.cx>. Line 314  Wakaba <w@suika.fam.cx>.
314    
315  =head1 LICENSE  =head1 LICENSE
316    
317  Copyright 2007 Wakaba <w@suika.fam.cx>  Copyright 2007-2008 Wakaba <w@suika.fam.cx>
318    
319  This library is free software; you can redistribute it  This library is free software; you can redistribute it
320  and/or modify it under the same terms as Perl itself.  and/or modify it under the same terms as Perl itself.
Legend:
Removed from v.1.1  
changed lines
  Added in v.1.59
admin@suikawiki.org
 ViewVC Help
Powered by ViewVC 1.1.24