--- test/html-webhacc/cc.cgi 2007/06/30 14:51:10 1.6 +++ test/html-webhacc/cc.cgi 2007/07/17 13:52:54 1.13 @@ -15,7 +15,9 @@ $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 @@ -23,16 +25,21 @@ ## TODO: _charset_ - 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 $s > 1000_000) { - print STDOUT "Status: 400 Document Too Long\nContent-Type: text/plain; charset=us-ascii\n\nToo long"; + if ($http->meta_variable ('PATH_INFO') ne '/') { + print STDOUT "Status: 404 Not Found\nContent-Type: text/plain; charset=us-ascii\n\n400"; exit; } + binmode STDOUT, ':utf8'; + + require Message::DOM::DOMImplementation; + my $dom = Message::DOM::DOMImplementation->new; + + my $input = get_input_document ($http, $dom); + my $inner_html_element = $http->parameter ('e'); + + load_text_catalog ('en'); ## TODO: conneg + my @nav; print STDOUT qq[Content-Type: text/html; charset=utf-8 @@ -43,98 +50,94 @@ -

Web Document Conformance Checker (beta)

+

Web Document Conformance Checker +(beta)

+
Request URI
+
<@{[htescape $input->{request_uri}]}>
Document URI
-
<@{[htescape $input_uri]}>
-
Internet Media Type
-
@{[htescape $input_format]}
+
<@{[htescape $input->{uri}]}>
]; # no
yet push @nav, ['#document-info' => 'Information']; - require Message::DOM::DOMImplementation; - my $dom = Message::DOM::DOMImplementation->____new; +if (defined $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)']} + @{[$input->{charset_overridden} ? '(overridden)' : '']}
+ +
+]; + + print_http_header_section ($input); + 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. + + my $t = Encode::decode ($input->{charset}, $input->{s}); -
-

Document Source

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

Parse Errors

-
-]; +
]; push @nav, ['#parse-errors' => 'Parse Error']; my $onerror = sub { my (%opt) = @_; + my ($type, $cls, $msg) = get_text ($opt{type}, $opt{level}); if ($opt{column} > 0) { - print STDOUT qq[
Line $opt{line} column $opt{column}
\n]; + print STDOUT qq[
Line $opt{line} column $opt{column}
\n]; } else { - $opt{line}--; - print STDOUT qq[
Line $opt{line}
\n]; + $opt{line} = $opt{line} - 1 || 1; + print STDOUT qq[
Line $opt{line}
\n]; } - print STDOUT qq[
@{[htescape $opt{type}]}
\n]; + $type =~ tr/ /-/; + $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, $onerror); + Whatpm::HTML->set_inner_html ($el, $t, $onerror); } else { - Whatpm::HTML->parse_string ($s => $doc, $onerror); + Whatpm::HTML->parse_string ($t => $doc, $onerror); } - print STDOUT qq[ -
+ print STDOUT qq[
]; - } elsif ($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}), $input->{charset}); + } elsif ({ + 'text/xml' => 1, + 'application/xhtml+xml' => 1, + 'application/xml' => 1, + }->{$input->{media_type}}) { + require Message::DOM::XMLParserTemp; -
-

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 { @@ -146,21 +149,20 @@ return 1; }; - open my $fh, '<', \$s; + open my $fh, '<', \($input->{s}); $doc = Message::DOM::XMLParserTemp->parse_byte_stream - ($fh => $dom, $onerror, charset => 'utf-8'); + ($fh => $dom, $onerror, charset => $input->{charset}); - print STDOUT qq[ -
+ 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']; @@ -182,15 +184,18 @@

Document Errors

-
-]; +
]; push @nav, ['#document-errors' => 'Document Error']; require Whatpm::ContentChecker; my $onerror = sub { my %opt = @_; - print STDOUT qq[
] . get_node_link ($opt{node}) . - "
\n
", htescape $opt{type}, "
\n"; + my ($type, $cls, $msg) = get_text ($opt{type}, $opt{level}); + $type =~ tr/ /-/; + $type =~ s/\|/%7C/g; + $msg .= qq[ [Description]]; + print STDOUT qq[
] . get_node_link ($opt{node}) . + qq[
\n
], $msg, "
\n"; }; my $elements; @@ -200,8 +205,7 @@ $elements = Whatpm::ContentChecker->check_document ($doc, $onerror); } - print STDOUT qq[ -
+ print STDOUT qq[
]; @@ -224,7 +228,9 @@ $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}}, $table->{caption}) { @@ -246,6 +252,7 @@ for (@$_) { $_->{id} = refaddr $_->{element} if defined $_->{element}; delete $_->{element}; + $_->{is_header} = $_->{is_header} ? 1 : 0; } } } @@ -258,6 +265,22 @@ print STDOUT qq[
]; } + if (keys %{$elements->{id}}) { + print STDOUT qq[ +
+

Identifiers

+ +
+]; + for my $id (sort {$a cmp $b} keys %{$elements->{id}}) { + print STDOUT qq[
@{[htescape $id]}
]; + for (@{$elements->{id}->{$id}}) { + print STDOUT qq[
].get_node_link ($_).qq[
]; + } + } + print STDOUT qq[
]; + } + if (keys %{$elements->{term}}) { print STDOUT qq[
@@ -276,6 +299,18 @@ } ## TODO: Show result +} else { + print STDOUT qq[ + +
+ +
+

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

+
+]; + push @nav, ['#result-summary' => 'Result']; + +} print STDOUT qq[ '; } - if ($node->has_child_nodes) { + if ($child->has_child_nodes) { $r .= '
    '; unshift @node, @{$child->child_nodes}, '
'; } else { @@ -348,8 +428,21 @@ $r .= qq'
  • <!--' . htescape ($child->data) . '-->
    • '; } elsif ($nt == $child->DOCUMENT_NODE) { $r .= qq'
  • Document'; + $r .= qq[]; if ($child->has_child_nodes) { - $r .= '
      '; + $r .= '
        '; unshift @node, @{$child->child_nodes}, '
      '; } } elsif ($nt == $child->DOCUMENT_TYPE_NODE) { @@ -384,6 +477,7 @@ $rs = '"' . $node->data . '"'; $node = $node->parent_node; } elsif ($node->node_type == 9) { + @r = ('') unless @r; $rs = ''; $node = $node->parent_node; } else { @@ -400,6 +494,191 @@ htescape (get_node_path ($_[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) = @_; + $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[$1]) : '(undef)'; + }ge; + return ($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->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); + if ($res->is_success or $http->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->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->parameter ('s'); + $r->{uri} = q; + $r->{request_uri} = q; + $r->{base_uri} = q; + $r->{charset} = ''.$http->parameter ('_charset_'); + $r->{charset} =~ s/\s+//g; + $r->{charset} = 'utf-8' if $r->{charset} eq ''; + $r->{header_field} = []; + } + + my $input_format = $http->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 . @@ -413,4 +692,4 @@ =cut -## $Date: 2007/06/30 14:51:10 $ +## $Date: 2007/07/17 13:52:54 $