--- test/html-webhacc/cc.cgi 2007/06/27 12:35:24 1.2 +++ test/html-webhacc/cc.cgi 2007/09/02 07:59:01 1.17 @@ -2,12 +2,10 @@ use strict; use lib qw[/home/httpd/html/www/markup/html/whatpm - /home/wakaba/work/manakai/lib - /home/wakaba/public_html/-temp/wiki/lib]; + /home/wakaba/work/manakai2/lib]; use CGI::Carp qw[fatalsToBrowser]; use Scalar::Util qw[refaddr]; - -use SuikaWiki::Input::HTTP; ## TODO: Use some better CGI module +use Time::HiRes qw/time/; sub htescape ($) { my $s = $_[0]; @@ -15,147 +13,179 @@ $s =~ s//>/g; $s =~ s/"/"/g; - $s =~ s!([\x00-\x09\x0B-\x1F\x7F-\x80])!sprintf 'U+%04X', ord $1!ge; + $s =~ s{([\x00-\x09\x0B-\x1F\x7F-\xA0\x{FEFF}\x{FFFC}-\x{FFFF}])}{ + sprintf 'U+%04X', ord $1; + }ge; return $s; } # htescape -my $http = SuikaWiki::Input::HTTP->new; - -## TODO: _charset_ - - my $input_format = $http->parameter ('i') || 'text/html'; - my $inner_html_element = $http->parameter ('e'); - my $input_uri = 'thismessage:/'; + use Message::CGI::HTTP; + my $http = Message::CGI::HTTP->new; - my $s = $http->parameter ('s'); - if (length $s > 1000_000) { - print STDOUT "Status: 400 Document Too Long\nContent-Type: text/plain; charset=us-ascii\n\nToo long"; + if ($http->get_meta_variable ('PATH_INFO') ne '/') { + print STDOUT "Status: 404 Not Found\nContent-Type: text/plain; charset=us-ascii\n\n400"; exit; } + binmode STDOUT, ':utf8'; + $| = 1; + + require Message::DOM::DOMImplementation; + my $dom = Message::DOM::DOMImplementation->new; + + load_text_catalog ('en'); ## TODO: conneg + + my @nav; print STDOUT qq[Content-Type: text/html; charset=utf-8 Web Document Conformance Checker (BETA) - - + -

Web Document Conformance Checker (beta)

Web Document Conformance Checker +(beta)

+]; + $| = 0; + my $input = get_input_document ($http, $dom); + my $inner_html_element = $http->get_parameter ('e'); + my $char_length = 0; + my %time; + my $time1; + my $time2; + + print qq[ +

Request URI: <@{[htescape $input->{request_uri}]}>
Document URI: <@{[htescape $input_uri]}>
Internet Media Type: @{[htescape $input_format]}; <@{[htescape $input->{uri}]}>

yet + push @nav, ['#document-info' => 'Information']; + +if (defined $input->{s}) { + $char_length = length $input->{s}; + + print STDOUT qq[ +

Base URI

<@{[htescape $input->{base_uri}]}>

Internet Media Type

@{[htescape $input->{media_type}]} + @{[$input->{media_type_overridden} ? '(overridden)' : '']}

Character Encoding

@{[defined $input->{charset} ? ''.htescape ($input->{charset}).'' : '(none)']} + @{[$input->{charset_overridden} ? '(overridden)' : '']}

Length

$char_length byte@{[$char_length == 1 ? '' : 's']}

+ +

+]; + + print_http_header_section ($input); - require Message::DOM::DOMImplementation; - my $dom = Message::DOM::DOMImplementation->____new; my $doc; my $el; - if ($input_format eq 'text/html') { + if ($input->{media_type} eq 'text/html') { require Encode; require Whatpm::HTML; - - $s = Encode::decode ('utf-8', $s); - print STDOUT qq[ -

Character Encoding

(none)

- + $input->{charset} ||= 'ISO-8859-1'; ## TODO: for now. -

-]; - print_source_string (\$s); - print STDOUT qq[ -

+ $time1 = time; + my $t = Encode::decode ($input->{charset}, $input->{s}); + $time2 = time; + $time{decode} = $time2 - $time1; + print STDOUT qq[

Parse Errors

Line $opt{line} column $opt{column}
Line $opt{line}: $msg

+ print STDOUT qq[

]; - } elsif ($input_format eq 'application/xhtml+xml') { - require Message::DOM::XMLParserTemp; - require Encode; - - my $t = Encode::decode ('utf-8', $s); - print STDOUT qq[ -

Character Encoding

(none)

- + print_source_string_section (\($input->{s}), $input->{charset}); + } elsif ({ + 'text/xml' => 1, + 'application/atom+xml' => 1, + 'application/rss+xml' => 1, + 'application/svg+xml' => 1, + 'application/xhtml+xml' => 1, + 'application/xml' => 1, + }->{$input->{media_type}}) { + require Message::DOM::XMLParserTemp; -

-]; - print_source_string (\$t); print STDOUT qq[ -

Parse Errors

Line $line column ]; + print STDOUT $err->location->column_number, ": "; + print STDOUT htescape $err->text, "

+ print STDOUT qq[

+ ]; + print_source_string_section (\($input->{s}), $doc->input_encoding); } else { + ## TODO: Change HTTP status code?? print STDOUT qq[ - - -

Media type @{[htescape $input_format]} is not supported!

Media type @{[htescape $input->{media_type}]} is not supported!

]; + push @nav, ['#result-summary' => 'Result']; } @@ -164,6 +194,7 @@

Document Tree

]; + push @nav, ['#document-tree' => 'Tree']; print_document_tree ($el || $doc); @@ -173,50 +204,235 @@

Document Errors

] . get_node_link ($opt{node}) . + qq[: ], $msg, "

+ print STDOUT qq[

]; + + if (@{$elements->{table}}) { + require JSON; + + push @nav, ['#tables' => 'Tables']; + print STDOUT qq[ +

Tables

+ + + + +]; + + my $i = 0; + for my $table_el (@{$elements->{table}}) { + $i++; + print STDOUT qq[

] . + get_node_link ($table_el) . q[

]; + + ## TODO: Make |ContentChecker| return |form_table| result + ## so that this script don't have to run the algorithm twice. + my $table = Whatpm::HTMLTable->form_table ($table_el); + + for (@{$table->{column_group}}, @{$table->{column}}, $table->{caption}) { + 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 '

]; + } + + print STDOUT qq[

]; + } + + if (keys %{$elements->{id}}) { + push @nav, ['#identifiers' => 'IDs']; + print STDOUT qq[ +

Identifiers

+ +

@{[htescape $id]}: ].get_node_link ($_).qq[

]; + } + + if (keys %{$elements->{term}}) { + push @nav, ['#terms' => 'Terms']; + print STDOUT qq[ +

Terms

+ +

@{[htescape $term]}: ].get_node_link ($_).qq[

]; + } + + if (keys %{$elements->{class}}) { + push @nav, ['#classes' => 'Classes']; + print STDOUT qq[ +

Classes

+ +

@{[htescape $class]}: ].get_node_link ($_).qq[

]; + } } ## TODO: Show result +} else { + print STDOUT qq[ + +

+ +

Input Error: @{[htescape ($input->{error_status_text})]}

+]; + push @nav, ['#result-summary' => 'Result']; + +} + + print STDOUT qq[ + ]; + for (qw/decode parse parse_xml check/) { + next unless defined $time{$_}; + open my $file, '>>', ".cc-$_.txt" or die ".cc-$_.txt: $!"; + print $file $char_length, "\t", $time{$_}, "\n"; + } + exit; -sub print_source_string ($) { - my $s = $_[0]; - my $i = 1; - print STDOUT qq[

], htescape $1, "

HTTP Header

+ +

Note: Due to the limitation of the +network library in use, the content of this section might +not be the real header.

+ + +]; + + if (defined $input->{header_status_code}) { + print STDOUT qq[]; + print STDOUT qq[]; + } + if (defined $input->{header_status_text}) { + print STDOUT qq[]; + print STDOUT qq[]; } - if ($$s =~ /\G([^\x0A]+)/gc) { - print STDOUT qq[

], htescape $1, "

\n"; + + for (@{$input->{header_field}}) { + print STDOUT qq[]; + print STDOUT qq[]; + } + + print STDOUT qq[

Status code	`@{[htescape ($input->{header_status_code})]}`
Status text	`@{[htescape ($input->{header_status_text})]}`
`@{[htescape ($_->[0])]}`	`@{[htescape ($_->[1])]}`

Document Source

], htescape $1, "
], htescape $1, "

"; -} # print_input_string + print STDOUT "

"; +} # print_input_string_section sub print_document_tree ($) { my $node = shift; @@ -233,46 +449,61 @@ my $node_id = 'node-'.refaddr $child; my $nt = $child->node_type; if ($nt == $child->ELEMENT_NODE) { - $r .= qq'

' . htescape ($child->tag_name) .
+      my $child_nsuri = $child->namespace_uri;
+      $r .= qq[

] . htescape ($child->tag_name) .
           ''; ## ISSUE: case
 
       if ($child->has_attributes) {
         $r .= '';
-        for my $attr (sort {$a->[0] cmp $b->[0]} map { [$_->name, $_->value, 'node-'.refaddr $_] }
+        for my $attr (sort {$a->[0] cmp $b->[0]} map { [$_->name, $_->value, $_->namespace_uri, 'node-'.refaddr $_] }
                       @{$child->attributes}) {
-          $r .= qq'' . htescape ($attr->[0]) . ' = '; ## ISSUE: case?
+          $r .= qq[
] . htescape ($attr->[0]) . ' = '; ## ISSUE: case?
           $r .= '' . htescape ($attr->[1]) . ''; ## TODO: children
         }
         $r .= '';
       }
 
-      if ($node->has_child_nodes) {
+      if ($child->has_child_nodes) {
         $r .= '';
-        unshift @node, @{$child->child_nodes}, '';
+        unshift @node, @{$child->child_nodes}, '

';
+      } else {
+        $r .= '';
       }
     } elsif ($nt == $child->TEXT_NODE) {
-      $r .= qq'' . htescape ($child->data) . '';
+      $r .= qq'' . htescape ($child->data) . '';
     } elsif ($nt == $child->CDATA_SECTION_NODE) {
-      $r .= qq'<[CDATA[' . htescape ($child->data) . ']]>';
+      $r .= qq'<[CDATA[' . htescape ($child->data) . ']]>';
     } elsif ($nt == $child->COMMENT_NODE) {
-      $r .= qq'<!--' . htescape ($child->data) . '-->';
+      $r .= qq'<!--' . htescape ($child->data) . '-->';
     } elsif ($nt == $child->DOCUMENT_NODE) {
-      $r .= qq'Document';
+      $r .= qq'Document';
+      $r .= qq[];
+      $r .= qq[@{[scalar get_text ('manakaiIsHTML:'.($child->manakai_is_html?1:0))]}];
+      $r .= qq[@{[scalar get_text ('manakaiCompatMode:'.$child->manakai_compat_mode)]}];
+      unless ($child->manakai_is_html) {
+        $r .= qq[XML version = @{[htescape ($child->xml_version)]}];
+        if (defined $child->xml_encoding) {
+          $r .= qq[XML encoding = @{[htescape ($child->xml_encoding)]}];
+        } else {
+          $r .= qq[XML encoding = (null)];
+        }
+        $r .= qq[XML standalone = @{[$child->xml_standalone ? 'true' : 'false']}];
+      }
+      $r .= qq[];
       if ($child->has_child_nodes) {
-        $r .= '';
-        unshift @node, @{$child->child_nodes}, '';
+        $r .= '';
+        unshift @node, @{$child->child_nodes}, '';
       }
     } elsif ($nt == $child->DOCUMENT_TYPE_NODE) {
-      $r .= qq'<!DOCTYPE>';
-      $r .= 'Name = @{[htescape ($child->name)]}';
-      $r .= 'Public identifier = @{[htescape ($child->public_id)]}';
-      $r .= 'System identifier = @{[htescape ($child->system_id)]}';
+      $r .= qq'<!DOCTYPE>';
+      $r .= qq[Name = @{[htescape ($child->name)]}];
+      $r .= qq[Public identifier = @{[htescape ($child->public_id)]}];
+      $r .= qq[System identifier = @{[htescape ($child->system_id)]}];
       $r .= '';
     } elsif ($nt == $child->PROCESSING_INSTRUCTION_NODE) {
-      $r .= qq'<?@{[htescape ($child->target)]}?>';
-      $r .= '@{[htescape ($child->data)]}';
+      $r .= qq'<?@{[htescape ($child->target)]} @{[htescape ($child->data)]}?>';
     } else {
-      $r .= qq'@{[$child->node_type]} @{[htescape ($child->node_name)]}'; # error
+      $r .= qq'@{[$child->node_type]} @{[htescape ($child->node_name)]}'; # error
     }
   }
 
@@ -295,6 +526,7 @@
       $rs = '"' . $node->data . '"';
       $node = $node->parent_node;
     } elsif ($node->node_type == 9) {
+      @r = ('') unless @r;
       $rs = '';
       $node = $node->parent_node;
     } else {
@@ -306,6 +538,214 @@
   return join '/', @r;
 } # get_node_path
 
+sub get_node_link ($) {
+  return qq[] .
+      htescape (get_node_path ($_[0])) . qq[];
+} # get_node_link
+
+{
+  my $Msg = {};
+
+sub load_text_catalog ($) {
+  my $lang = shift; # MUST be a canonical lang name
+  open my $file, '<', "cc-msg.$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 ($type, $level, $node) = @_;
+  $type = $level . ':' . $type if defined $level;
+  my @arg;
+  {
+    if (defined $Msg->{$type}) {
+      my $msg = $Msg->{$type}->[1];
+      $msg =~ s{\$([0-9]+)}{
+        defined $arg[$1] ? htescape ($arg[$1]) : '(undef)';
+      }ge;
+      $msg =~ s{{\@([A-Za-z0-9:_.-]+)}}{
+        UNIVERSAL::can ($node, 'get_attribute_ns')
+            ? htescape ($node->get_attribute_ns (undef, $1)) : ''
+      }ge;
+      $msg =~ s{{\@}}{
+        UNIVERSAL::can ($node, 'value') ? htescape ($node->value) : ''
+      }ge;
+      $msg =~ s{{local-name}}{
+        UNIVERSAL::can ($node, 'manakai_local_name')
+          ? htescape ($node->manakai_local_name) : ''
+      }ge;
+      $msg =~ s{{element-local-name}}{
+        (UNIVERSAL::can ($node, 'owner_element') and
+         $node->owner_element)
+          ? htescape ($node->owner_element->manakai_local_name)
+          : ''
+      }ge;
+      return ($type, $Msg->{$type}->[0], $msg);
+    } elsif ($type =~ s/:([^:]*)$//) {
+      unshift @arg, $1;
+      redo;
+    }
+  }
+  return ($type, '', htescape ($_[0]));
+} # get_text
+
+}
+
+sub get_input_document ($$) {
+  my ($http, $dom) = @_;
+
+  my $request_uri = $http->get_parameter ('uri');
+  my $r = {};
+  if (defined $request_uri and length $request_uri) {
+    my $uri = $dom->create_uri_reference ($request_uri);
+    unless ({
+             http => 1,
+            }->{lc $uri->uri_scheme}) {
+      return {uri => $request_uri, request_uri => $request_uri,
+              error_status_text => 'URI scheme not allowed'};
+    }
+
+    require Message::Util::HostPermit;
+    my $host_permit = new Message::Util::HostPermit;
+    $host_permit->add_rule (<check ($uri->uri_host, $uri->uri_port || 80)) {
+      return {uri => $request_uri, request_uri => $request_uri,
+              error_status_text => 'Connection to the host is forbidden'};
+    }
+
+    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);
+    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 . ## 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 =~ m#^([0-9A-Za-z._+-]+/[0-9A-Za-z._+-]+)#) {
+        $r->{media_type} = lc $1;
+      }
+      if (defined $ct and $ct =~ /;\s*charset\s*=\s*"?(\S+)"?/i) {
+        $r->{charset} = lc $1;
+        $r->{charset} =~ tr/\\//d;
+      }
+
+      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;
+      } 
+
+      $r->{s} = ''.$res->content;
+    } 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;
+    $r->{request_uri} = q;
+    $r->{base_uri} = q;
+    $r->{charset} = ''.$http->get_parameter ('_charset_');
+    $r->{charset} =~ s/\s+//g;
+    $r->{charset} = 'utf-8' if $r->{charset} eq '';
+    $r->{header_field} = [];
+  }
+
+  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}) {
+      $r->{charset} = 'us-ascii';
+    } elsif ($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;
+  }
+
+  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 .
@@ -319,4 +759,4 @@
 
 =cut
 
-## $Date: 2007/06/27 12:35:24 $
+## $Date: 2007/09/02 07:59:01 $