--- test/html-webhacc/cc.cgi 2007/07/01 10:02:24 1.8 +++ test/html-webhacc/cc.cgi 2008/07/21 09:15:55 1.57 @@ -1,47 +1,36 @@ #!/usr/bin/perl use strict; +use utf8; 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 + require WebHACC::Input; + require WebHACC::Result; + require WebHACC::Output; -sub htescape ($) { - my $s = $_[0]; - $s =~ s/&/&/g; - $s =~ s//>/g; - $s =~ s/"/"/g; - $s =~ s!([\x00-\x09\x0B-\x1F\x7F-\x80])!sprintf 'U+%04X', ord $1!ge; - return $s; -} # htescape +my $out; -my $http = SuikaWiki::Input::HTTP->new; - -## TODO: _charset_ + require Message::DOM::DOMImplementation; + my $dom = Message::DOM::DOMImplementation->new; +{ + use Message::CGI::HTTP; + my $http = Message::CGI::HTTP->new; - if ($http->meta_variable ('PATH_INFO') ne '/') { + if ($http->get_meta_variable ('PATH_INFO') ne '/') { print STDOUT "Status: 404 Not Found\nContent-Type: text/plain; charset=us-ascii\n\n400"; exit; } - - my $input_format = $http->parameter ('i') || 'text/html'; - my $inner_html_element = $http->parameter ('e'); - my $input_uri = 'thismessage:/'; - - 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"; - exit; - } - + load_text_catalog ('en'); ## TODO: conneg - my @nav; - print STDOUT qq[Content-Type: text/html; charset=utf-8 + $out = WebHACC::Output->new; + $out->handle (*STDOUT); + $out->set_utf8; + $out->set_flush; + $out->html (qq[Content-Type: text/html; charset=utf-8 @@ -50,382 +39,110 @@ -

Web Document Conformance Checker (beta)

+

Web Document Conformance Checker +(beta)

+]); -
-
-
Document URI
-
<@{[htescape $input_uri]}>
-
Internet Media Type
-
@{[htescape $input_format]}
-]; # no
yet - push @nav, ['#document-info' => 'Information']; + my $input = get_input_document ($http, $dom); - require Message::DOM::DOMImplementation; - my $dom = Message::DOM::DOMImplementation->____new; - my $doc; - my $el; - - if ($input_format eq 'text/html') { - require Encode; - require Whatpm::HTML; - - $s = Encode::decode ('utf-8', $s); - - print STDOUT qq[ -
Character Encoding
-
(none)
- -
- -
-

Document Source

-]; - push @nav, ['#source-string' => 'Source']; - print_source_string (\$s); - print STDOUT qq[ -
- -
-

Parse Errors

- -
-]; - push @nav, ['#parse-errors' => 'Parse Error']; - - my $onerror = sub { - my (%opt) = @_; - my ($cls, $msg) = get_text ($opt{type}, $opt{level}); - if ($opt{column} > 0) { - print STDOUT qq[
Line $opt{line} column $opt{column}
\n]; - } else { - $opt{line} = $opt{line} - 1 || 1; - print STDOUT qq[
Line $opt{line}
\n]; - } - $opt{type} =~ tr/ /-/; - $opt{type} =~ s/\|/%7C/g; - $msg .= qq[ [Description]]; - print STDOUT qq[
$msg
\n]; - }; - - $doc = $dom->create_document; - if (defined $inner_html_element and length $inner_html_element) { - $el = $doc->create_element_ns - ('http://www.w3.org/1999/xhtml', [undef, $inner_html_element]); - Whatpm::HTML->set_inner_html ($el, $s, $onerror); - } else { - Whatpm::HTML->parse_string ($s => $doc, $onerror); - } + $out->input ($input); + $out->unset_flush; - print STDOUT qq[ -
-
-]; - } elsif ($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)
- - - -
-

Document Source

-]; - push @nav, ['#source-string' => 'Source']; - print_source_string (\$t); - print STDOUT qq[ -
- -
-

Parse Errors

- -
]; - push @nav, ['#parse-errors' => 'Parse Error']; - - my $onerror = sub { - my $err = shift; - my $line = $err->location->line_number; - print STDOUT qq[
Line $line column ]; - print STDOUT $err->location->column_number, "
"; - print STDOUT htescape $err->text, "
\n"; - return 1; - }; - - open my $fh, '<', \$s; - $doc = Message::DOM::XMLParserTemp->parse_byte_stream - ($fh => $dom, $onerror, charset => 'utf-8'); - - print STDOUT qq[
-
-]; - } else { - print STDOUT qq[ - - - -
-

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

-
-]; - push @nav, ['#result-summary' => 'Result']; - } + my $result = WebHACC::Result->new; + $result->output ($out); + $result->{conforming_min} = 1; + $result->{conforming_max} = 1; + $out->html (''); - if (defined $doc or defined $el) { - print STDOUT qq[ -
-

Document Tree

-]; - push @nav, ['#document-tree' => 'Tree']; - - print_document_tree ($el || $doc); - - print STDOUT qq[ -
- -
-

Document Errors

- -
]; - push @nav, ['#document-errors' => 'Document Error']; - - require Whatpm::ContentChecker; - my $onerror = sub { - my %opt = @_; - my ($cls, $msg) = get_text ($opt{type}, $opt{level}); - $opt{type} = $opt{level} . ':' . $opt{type} if defined $opt{level}; - $opt{type} =~ tr/ /-/; - $opt{type} =~ s/\|/%7C/g; - $msg .= qq[ [Description]]; - print STDOUT qq[
] . get_node_link ($opt{node}) . - qq[
\n
], $msg, "
\n"; - }; - - my $elements; - if ($el) { - $elements = Whatpm::ContentChecker->check_element ($el, $onerror); - } else { - $elements = Whatpm::ContentChecker->check_document ($doc, $onerror); - } + check_and_print ($input => $result => $out); + + $result->generate_result_section; - print STDOUT qq[
-
-]; - - if (@{$elements->{table}}) { - require JSON; - - print STDOUT qq[ -
-

Tables

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

] . - get_node_link ($table_el) . q[

]; - - 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}; - } - } - } - - print STDOUT '
]; - } - - print STDOUT qq[
]; - } + $out->nav_list; - if (keys %{$elements->{term}}) { - print STDOUT qq[ -
-

Terms

- -
-]; - for my $term (sort {$a cmp $b} keys %{$elements->{term}}) { - print STDOUT qq[
@{[htescape $term]}
]; - for (@{$elements->{term}->{$term}}) { - print STDOUT qq[
].get_node_link ($_).qq[
]; - } - } - print STDOUT qq[
]; - } - } - - ## TODO: Show result - - print STDOUT qq[ - - - -]; - -exit; - -sub print_source_string ($) { - my $s = $_[0]; - my $i = 1; - print STDOUT qq[
    \n]; - if (length $$s) { - while ($$s =~ /\G([^\x0A]*?)\x0D?\x0A/gc) { - print STDOUT qq[
  1. ], htescape $1, "
    2. \n"; - $i++; - } - if ($$s =~ /\G([^\x0A]+)/gc) { - print STDOUT qq[
  2. ], htescape $1, "
    3. \n"; - } - } else { - print STDOUT q[
  3. ]; - } - print STDOUT "
"; -} # print_input_string - -sub print_document_tree ($) { - my $node = shift; - my $r = '
    '; - - my @node = ($node); - while (@node) { - my $child = shift @node; - unless (ref $child) { - $r .= $child; - next; - } - - my $node_id = 'node-'.refaddr $child; - my $nt = $child->node_type; - if ($nt == $child->ELEMENT_NODE) { - my $child_nsuri = $child->namespace_uri; - $r .= qq[
  1. ] . htescape ($child->tag_name) . - ''; ## ISSUE: case - - if ($child->has_attributes) { - $r .= ''; - } + exit; +} - if ($child->has_child_nodes) { - $r .= '
      '; - unshift @node, @{$child->child_nodes}, '
    2. '; - } else { - $r .= ''; - } - } elsif ($nt == $child->TEXT_NODE) { - $r .= qq'
  2. ' . htescape ($child->data) . '
    3. '; - } elsif ($nt == $child->CDATA_SECTION_NODE) { - $r .= qq'
  3. <[CDATA[' . htescape ($child->data) . ']]>
    4. '; - } elsif ($nt == $child->COMMENT_NODE) { - $r .= qq'
  4. <!--' . htescape ($child->data) . '-->
    5. '; - } elsif ($nt == $child->DOCUMENT_NODE) { - $r .= qq'
  5. Document'; - $r .= qq[]; - if ($child->has_child_nodes) { - $r .= '
      '; - unshift @node, @{$child->child_nodes}, '
    6. '; - } - } elsif ($nt == $child->DOCUMENT_TYPE_NODE) { - $r .= qq'
  6. <!DOCTYPE>
    7. '; - } elsif ($nt == $child->PROCESSING_INSTRUCTION_NODE) { - $r .= qq'
  7. <?@{[htescape ($child->target)]} @{[htescape ($child->data)]}?>
    8. '; - } else { - $r .= qq'
  8. @{[$child->node_type]} @{[htescape ($child->node_name)]}
    9. '; # error - } +sub check_and_print ($$$) { + 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); } - $r .= '
'; - print STDOUT $r; -} # print_document_tree - -sub get_node_path ($) { - my $node = shift; - my @r; - while (defined $node) { - my $rs; - if ($node->node_type == 1) { - $rs = $node->manakai_local_name; - $node = $node->parent_node; - } elsif ($node->node_type == 2) { - $rs = '@' . $node->manakai_local_name; - $node = $node->owner_element; - } elsif ($node->node_type == 3) { - $rs = '"' . $node->data . '"'; - $node = $node->parent_node; - } elsif ($node->node_type == 9) { - $rs = ''; - $node = $node->parent_node; - } else { - $rs = '#' . $node->node_type; - $node = $node->parent_node; - } - unshift @r, $rs; - } - return join '/', @r; -} # get_node_path + $out->input ($original_input); +} # check_and_print -sub get_node_link ($) { - return qq[] . - htescape (get_node_path ($_[0])) . qq[]; -} # get_node_link { my $Msg = {}; sub load_text_catalog ($) { +# my $self = shift; my $lang = shift; # MUST be a canonical lang name - open my $file, '<', "cc-msg.$lang.txt" or die "$0: cc-msg.$lang.txt: $!"; + 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, $_); @@ -435,36 +152,240 @@ } } # load_text_catalog -sub get_text ($) { - my ($type, $level) = @_; +sub get_text ($;$$) { +# my $self = shift; + my ($type, $level, $node) = @_; $type = $level . ':' . $type if defined $level; + $level = 'm' unless defined $level; my @arg; { if (defined $Msg->{$type}) { my $msg = $Msg->{$type}->[1]; - $msg =~ s/\$([0-9]+)/defined $arg[$1] ? htescape ($arg[$1]) : '(undef)'/ge; - return ($Msg->{$type}->[0], $msg); + $msg =~ s{\$([0-9]+)}{ + defined $arg[$1] ? ($arg[$1]) : '(undef)'; + }ge; ##BUG: ^ must be escaped + $msg =~ s{{\@([A-Za-z0-9:_.-]+)}}{ + UNIVERSAL::can ($node, 'get_attribute_ns') + ? ($node->get_attribute_ns (undef, $1)) : '' + }ge; ## BUG: ^ must be escaped + $msg =~ s{{\@}}{ ## BUG: v must be escaped + UNIVERSAL::can ($node, 'value') ? ($node->value) : '' + }ge; + $msg =~ s{{local-name}}{ + UNIVERSAL::can ($node, 'manakai_local_name') + ? ($node->manakai_local_name) : '' + }ge; ## BUG: ^ must be escaped + $msg =~ s{{element-local-name}}{ + (UNIVERSAL::can ($node, 'owner_element') and + $node->owner_element) + ? ($node->owner_element->manakai_local_name) + : '' ## BUG: ^ must be escaped + }ge; + return ($type, 'level-' . $level . ' ' . $Msg->{$type}->[0], $msg); } elsif ($type =~ s/:([^:]*)$//) { unshift @arg, $1; redo; } } - return ('', htescape ($_[0])); + return ($type, 'level-'.$level, ($_[0])); + ## BUG: ^ must be escaped } # get_text } +sub get_input_document ($$) { + 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; + $host_permit->add_rule (<check ($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 . ## 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; + $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->{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 . =head1 LICENSE -Copyright 2007 Wakaba +Copyright 2007-2008 Wakaba This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself. =cut -## $Date: 2007/07/01 10:02:24 $ +## $Date: 2008/07/21 09:15:55 $