#!/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];
use Time::HiRes qw/time/;
sub htescape ($) {
my $s = $_[0];
$s =~ s/&/&/g;
$s =~ s/</g;
$s =~ s/>/>/g;
$s =~ s/"/"/g;
$s =~ s{([\x00-\x09\x0B-\x1F\x7F-\xA0\x{FEFF}\x{FFFC}-\x{FFFF}])}{
sprintf 'U+%04X ', ord $1;
}ge;
return $s;
} # htescape
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;
}
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)
];
$| = 0;
my $input = get_input_document ($http, $dom);
my $char_length = 0;
my %time;
print qq[
Request URI
<@{[htescape $input->{request_uri}]} >
Document URI
<@{[htescape $input->{uri}]} >
]; # no 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) ' : defined $input->{official_type} ? $input->{media_type} eq $input->{official_type} ? '' : '(sniffed; official type is: '.htescape ($input->{official_type}).'
)' : '(sniffed) ']}
Character Encoding
@{[defined $input->{charset} ? ''.htescape ($input->{charset}).'
' : '(none)']}
@{[$input->{charset_overridden} ? '(overridden) ' : '']}
Length
$char_length byte@{[$char_length == 1 ? '' : 's']}
];
my $result = {conforming_min => 1, conforming_max => 1};
print_http_header_section ($input, $result);
my $doc;
my $el;
my $manifest;
if ($input->{media_type} eq 'text/html') {
($doc, $el) = print_syntax_error_html_section ($input, $result);
print_source_string_section
(\($input->{s}), $input->{charset} || $doc->input_encoding);
} elsif ({
'text/xml' => 1,
'application/atom+xml' => 1,
'application/rss+xml' => 1,
'application/svg+xml' => 1,
'application/xhtml+xml' => 1,
'application/xml' => 1,
}->{$input->{media_type}}) {
($doc, $el) = print_syntax_error_xml_section ($input, $result);
print_source_string_section (\($input->{s}), $doc->input_encoding);
} elsif ($input->{media_type} eq 'text/cache-manifest') {
## TODO: MUST be text/cache-manifest
$manifest = print_syntax_error_manifest_section ($input, $result);
print_source_string_section (\($input->{s}), 'utf-8');
} else {
## TODO: Change HTTP status code??
print_result_unknown_type_section ($input, $result);
}
if (defined $doc or defined $el) {
print_structure_dump_dom_section ($doc, $el);
my $elements = print_structure_error_dom_section ($doc, $el, $result);
print_table_section ($elements->{table}) if @{$elements->{table}};
print_id_section ($elements->{id}) if keys %{$elements->{id}};
print_term_section ($elements->{term}) if keys %{$elements->{term}};
print_class_section ($elements->{class}) if keys %{$elements->{class}};
} elsif (defined $manifest) {
print_structure_dump_manifest_section ($manifest);
print_structure_error_manifest_section ($manifest, $result);
}
print_result_section ($result);
} else {
print STDOUT qq[];
print_result_input_error_section ($input);
}
print STDOUT qq[
];
for (@nav) {
print STDOUT qq[$_->[1] ];
}
print STDOUT qq[
];
for (qw/decode parse parse_html parse_xml parse_manifest
check check_manifest/) {
next unless defined $time{$_};
open my $file, '>>', ".cc-$_.txt" or die ".cc-$_.txt: $!";
print $file $char_length, "\t", $time{$_}, "\n";
}
exit;
sub add_error ($$$) {
my ($layer, $err, $result) = @_;
if (defined $err->{level}) {
if ($err->{level} eq 's') {
$result->{$layer}->{should}++;
$result->{$layer}->{score_min} -= 2;
$result->{conforming_min} = 0;
} elsif ($err->{level} eq 'w' or $err->{level} eq 'g') {
$result->{$layer}->{warning}++;
} elsif ($err->{level} eq 'unsupported') {
$result->{$layer}->{unsupported}++;
$result->{unsupported} = 1;
} else {
$result->{$layer}->{must}++;
$result->{$layer}->{score_max} -= 2;
$result->{$layer}->{score_min} -= 2;
$result->{conforming_min} = 0;
$result->{conforming_max} = 0;
}
} else {
$result->{$layer}->{must}++;
$result->{$layer}->{score_max} -= 2;
$result->{$layer}->{score_min} -= 2;
$result->{conforming_min} = 0;
$result->{conforming_max} = 0;
}
} # add_error
sub print_http_header_section ($$) {
my ($input, $result) = @_;
return unless defined $input->{header_status_code} or
defined $input->{header_status_text} or
@{$input->{header_field}};
push @nav, ['#source-header' => 'HTTP Header'];
print STDOUT qq[];
} # print_http_header_section
sub print_syntax_error_html_section ($$) {
my ($input, $result) = @_;
require Encode;
require Whatpm::HTML;
print STDOUT qq[
Parse Errors
];
push @nav, ['#parse-errors' => 'Parse Error'];
my $onerror = sub {
my (%opt) = @_;
my ($type, $cls, $msg) = get_text ($opt{type}, $opt{level});
if ($opt{column} > 0) {
print STDOUT qq[Line $opt{line} column $opt{column} \n];
} else {
$opt{line} = $opt{line} - 1 || 1;
print STDOUT qq[Line $opt{line} \n];
}
$type =~ tr/ /-/;
$type =~ s/\|/%7C/g;
$msg .= qq[ [Description ]];
print STDOUT qq[], get_error_level_label (\%opt);
print STDOUT qq[$msg \n];
add_error ('syntax', \%opt => $result);
};
my $doc = $dom->create_document;
my $el;
my $inner_html_element = $http->get_parameter ('e');
if (defined $inner_html_element and length $inner_html_element) {
$input->{charset} ||= 'windows-1252'; ## TODO: for now.
my $time1 = time;
my $t = Encode::decode ($input->{charset}, $input->{s});
$time{decode} = time - $time1;
$el = $doc->create_element_ns
('http://www.w3.org/1999/xhtml', [undef, $inner_html_element]);
$time1 = time;
Whatpm::HTML->set_inner_html ($el, $t, $onerror);
$time{parse} = time - $time1;
} else {
my $time1 = time;
Whatpm::HTML->parse_byte_string
($input->{charset}, $input->{s} => $doc, $onerror);
$time{parse_html} = time - $time1;
}
$doc->manakai_charset ($input->{official_charset})
if defined $input->{official_charset};
print STDOUT qq[ ];
return ($doc, $el);
} # print_syntax_error_html_section
sub print_syntax_error_xml_section ($$) {
my ($input, $result) = @_;
require Message::DOM::XMLParserTemp;
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";
add_error ('syntax', {type => $err->text,
level => [
$err->SEVERITY_FATAL_ERROR => 'm',
$err->SEVERITY_ERROR => 'm',
$err->SEVERITY_WARNING => 's',
]->[$err->severity]} => $result);
return 1;
};
my $time1 = time;
open my $fh, '<', \($input->{s});
my $doc = Message::DOM::XMLParserTemp->parse_byte_stream
($fh => $dom, $onerror, charset => $input->{charset});
$time{parse_xml} = time - $time1;
$doc->manakai_charset ($input->{official_charset})
if defined $input->{official_charset};
print STDOUT qq[ ];
return ($doc, undef);
} # print_syntax_error_xml_section
sub print_syntax_error_manifest_section ($$) {
my ($input, $result) = @_;
require Whatpm::CacheManifest;
print STDOUT qq[
Parse Errors
];
push @nav, ['#parse-errors' => 'Parse Error'];
my $onerror = sub {
my (%opt) = @_;
my ($type, $cls, $msg) = get_text ($opt{type}, $opt{level});
print STDOUT qq[], get_error_label (\%opt), qq[ ];
$type =~ tr/ /-/;
$type =~ s/\|/%7C/g;
$msg .= qq[ [Description ]];
print STDOUT qq[], get_error_level_label (\%opt);
print STDOUT qq[$msg \n];
add_error ('syntax', \%opt => $result);
};
my $time1 = time;
my $manifest = Whatpm::CacheManifest->parse_byte_string
($input->{s}, $input->{uri}, $input->{base_uri}, $onerror);
$time{parse_manifest} = time - $time1;
print STDOUT qq[ ];
return $manifest;
} # print_syntax_error_manifest_section
sub print_source_string_section ($$) {
require Encode;
my $enc = Encode::find_encoding ($_[1]); ## TODO: charset name -> Perl name
return unless $enc;
my $s = \($enc->decode (${$_[0]}));
my $i = 1;
push @nav, ['#source-string' => 'Source'];
print STDOUT qq[
Document Source
\n];
if (length $$s) {
while ($$s =~ /\G([^\x0A]*?)\x0D?\x0A/gc) {
print STDOUT qq[], htescape $1, " \n";
$i++;
}
if ($$s =~ /\G([^\x0A]+)/gc) {
print STDOUT qq[], htescape $1, " \n";
}
} else {
print STDOUT q[ ];
}
print STDOUT " ";
} # print_input_string_section
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[] . htescape ($child->tag_name) .
'
'; ## ISSUE: case
if ($child->has_attributes) {
$r .= '';
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 .= '' . htescape ($attr->[1]) . ' '; ## TODO: children
}
$r .= ' ';
}
if ($child->has_child_nodes) {
$r .= '';
unshift @node, @{$child->child_nodes}, ' ';
} else {
$r .= '';
}
} elsif ($nt == $child->TEXT_NODE) {
$r .= qq'' . htescape ($child->data) . ' ';
} elsif ($nt == $child->CDATA_SECTION_NODE) {
$r .= qq'<[CDATA[
' . htescape ($child->data) . ' ]]>
';
} elsif ($nt == $child->COMMENT_NODE) {
$r .= qq'';
} elsif ($nt == $child->DOCUMENT_NODE) {
$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}, ' ';
}
} elsif ($nt == $child->DOCUMENT_TYPE_NODE) {
$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)]}
@{[htescape ($child->data)]} ?>
';
} else {
$r .= qq'@{[$child->node_type]} @{[htescape ($child->node_name)]} '; # error
}
}
$r .= ' ';
print STDOUT $r;
} # print_document_tree
sub print_structure_dump_dom_section ($$) {
my ($doc, $el) = @_;
print STDOUT qq[
Document Tree
];
push @nav, ['#document-tree' => 'Tree'];
print_document_tree ($el || $doc);
print STDOUT qq[];
} # print_structure_dump_dom_section
sub print_structure_dump_manifest_section ($) {
my $manifest = shift;
print STDOUT qq[
Cache Manifest
];
push @nav, ['#dump-manifest' => 'Caceh Manifest'];
print STDOUT qq[
Explicit entries ];
for my $uri (@{$manifest->[0]}) {
my $euri = htescape ($uri);
print STDOUT qq[<$euri >
];
}
print STDOUT qq[Fallback entries
Oppotunistic Caching Namespace
Fallback Entry ];
for my $uri (sort {$a cmp $b} keys %{$manifest->[1]}) {
my $euri = htescape ($uri);
my $euri2 = htescape ($manifest->[1]->{$uri});
print STDOUT qq[<$euri >
<$euri2 >
];
}
print STDOUT qq[
Online whitelist ];
for my $uri (@{$manifest->[2]}) {
my $euri = htescape ($uri);
print STDOUT qq[<$euri >
];
}
print STDOUT qq[ ];
} # print_structure_dump_manifest_section
sub print_structure_error_dom_section ($$$) {
my ($doc, $el, $result) = @_;
print STDOUT qq[
Document Errors
];
push @nav, ['#document-errors' => 'Document Error'];
require Whatpm::ContentChecker;
my $onerror = sub {
my %opt = @_;
my ($type, $cls, $msg) = get_text ($opt{type}, $opt{level}, $opt{node});
$type =~ tr/ /-/;
$type =~ s/\|/%7C/g;
$msg .= qq[ [Description ]];
print STDOUT qq[] . get_error_label (\%opt) .
qq[ \n], get_error_level_label (\%opt);
print STDOUT $msg, " \n";
add_error ('structure', \%opt => $result);
};
my $elements;
my $time1 = time;
if ($el) {
$elements = Whatpm::ContentChecker->check_element ($el, $onerror);
} else {
$elements = Whatpm::ContentChecker->check_document ($doc, $onerror);
}
$time{check} = time - $time1;
print STDOUT qq[ ];
return $elements;
} # print_structure_error_dom_section
sub print_structure_error_manifest_section ($$$) {
my ($manifest, $result) = @_;
print STDOUT qq[
Document Errors
];
push @nav, ['#document-errors' => 'Document Error'];
require Whatpm::CacheManifest;
Whatpm::CacheManifest->check_manifest ($manifest, sub {
my %opt = @_;
my ($type, $cls, $msg) = get_text ($opt{type}, $opt{level}, $opt{node});
$type =~ tr/ /-/;
$type =~ s/\|/%7C/g;
$msg .= qq[ [Description ]];
print STDOUT qq[] . get_error_label (\%opt) .
qq[ \n], $msg, " \n";
add_error ('structure', \%opt => $result);
});
print STDOUT qq[ ];
} # print_structure_error_manifest_section
sub print_table_section ($) {
my $tables = shift;
push @nav, ['#tables' => 'Tables'];
print STDOUT qq[
Tables
Structure of tables are visualized here if scripting is enabled.
];
require JSON;
my $i = 0;
for my $table_el (@$tables) {
$i++;
print STDOUT qq[
] .
get_node_link ($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} = $_->{is_header} ? 1 : 0;
}
}
}
print STDOUT '];
}
print STDOUT qq[
];
} # print_table_section
sub print_id_section ($) {
my $ids = shift;
push @nav, ['#identifiers' => 'IDs'];
print STDOUT qq[
Identifiers
];
for my $id (sort {$a cmp $b} keys %$ids) {
print STDOUT qq[@{[htescape $id]}
];
for (@{$ids->{$id}}) {
print STDOUT qq[].get_node_link ($_).qq[ ];
}
}
print STDOUT qq[ ];
} # print_id_section
sub print_term_section ($) {
my $terms = shift;
push @nav, ['#terms' => 'Terms'];
print STDOUT qq[
Terms
];
for my $term (sort {$a cmp $b} keys %$terms) {
print STDOUT qq[@{[htescape $term]} ];
for (@{$terms->{$term}}) {
print STDOUT qq[].get_node_link ($_).qq[ ];
}
}
print STDOUT qq[ ];
} # print_term_section
sub print_class_section ($) {
my $classes = shift;
push @nav, ['#classes' => 'Classes'];
print STDOUT qq[
Classes
];
for my $class (sort {$a cmp $b} keys %$classes) {
print STDOUT qq[@{[htescape $class]}
];
for (@{$classes->{$class}}) {
print STDOUT qq[].get_node_link ($_).qq[ ];
}
}
print STDOUT qq[ ];
} # print_class_section
sub print_result_section ($) {
my $result = shift;
print STDOUT qq[
Result ];
if ($result->{unsupported} and $result->{conforming_max}) {
print STDOUT qq[
The conformance
checker cannot decide whether the document is conforming or
not, since the document contains one or more unsupported
features. The document might or might not be conforming.
];
} elsif ($result->{conforming_min}) {
print STDOUT qq[
No conformance-error is
found in this document.
];
} elsif ($result->{conforming_max}) {
print STDOUT qq[
This document
is likely non -conforming , but in rare case
it might be conforming.
];
} else {
print STDOUT qq[
This document is
non -conforming .
];
}
print STDOUT qq[
MUST ‐level
Errors
SHOULD ‐level
Errors
Warnings
Score ];
my $must_error = 0;
my $should_error = 0;
my $warning = 0;
my $score_min = 0;
my $score_max = 0;
my $score_base = 20;
my $score_unit = $score_base / 100;
for (
[Transfer => 'transfer', ''],
[Character => 'char', ''],
[Syntax => 'syntax', '#parse-errors'],
[Structure => 'structure', '#document-errors'],
) {
$must_error += ($result->{$_->[1]}->{must} += 0);
$should_error += ($result->{$_->[1]}->{should} += 0);
$warning += ($result->{$_->[1]}->{warning} += 0);
$score_min += (($result->{$_->[1]}->{score_min} *= $score_unit) += $score_base);
$score_max += (($result->{$_->[1]}->{score_max} *= $score_unit) += $score_base);
my $uncertain = $result->{$_->[1]}->{unsupported} ? '?' : '';
my $label = $_->[0];
if ($result->{$_->[1]}->{must} or
$result->{$_->[1]}->{should} or
$result->{$_->[1]}->{warning} or
$result->{$_->[1]}->{unsupported}) {
$label = qq[$label ];
}
print STDOUT qq[$label $result->{$_->[1]}->{must}$uncertain $result->{$_->[1]}->{should}$uncertain $result->{$_->[1]}->{warning}$uncertain ];
if ($uncertain) {
print qq[−∞..$result->{$_->[1]}->{score_max} ];
} elsif ($result->{$_->[1]}->{score_min} != $result->{$_->[1]}->{score_max}) {
print qq[$result->{$_->[1]}->{score_min}..$result->{$_->[1]}->{score_max} ];
} else {
print qq[$result->{$_->[1]}->{score_min} ];
}
}
$score_max += $score_base;
print STDOUT qq[
Semantics 0? 0? 0? −∞..$score_base
Total
$must_error?
$should_error?
$warning?
−∞..$score_max
Important : This conformance checking service
is under development . The result above might be wrong .
];
push @nav, ['#result-summary' => 'Result'];
} # print_result_section
sub print_result_unknown_type_section ($$) {
my ($input, $result) = @_;
my $euri = htescape ($input->{uri});
print STDOUT qq[
Errors
<$euri >
Not
supported :
Media type
@{[htescape $input->{media_type}]}
is not supported.
];
push @nav, ['#parse-errors' => 'Errors'];
add_error (char => {level => 'unsupported'} => $result);
add_error (syntax => {level => 'unsupported'} => $result);
add_error (structure => {level => 'unsupported'} => $result);
} # print_result_unknown_type_section
sub print_result_input_error_section ($) {
my $input = shift;
print STDOUT qq[
Input Error : @{[htescape ($input->{error_status_text})]}
];
push @nav, ['#result-summary' => 'Result'];
} # print_Result_input_error_section
sub get_error_label ($) {
my $err = shift;
my $r = '';
if (defined $err->{line}) {
if ($err->{column} > 0) {
$r = qq[Line $err->{line} column $err->{column}];
} else {
$err->{line} = $err->{line} - 1 || 1;
$r = qq[Line $err->{line} ];
}
}
if (defined $err->{node}) {
$r .= ' ' if length $r;
$r = get_node_link ($err->{node});
}
if (defined $err->{index}) {
$r .= ' ' if length $r;
$r .= 'Index ' . (0+$err->{index});
}
if (defined $err->{value}) {
$r .= ' ' if length $r;
$r .= '' . htescape ($err->{value}) . '
';
}
return $r;
} # get_error_label
sub get_error_level_label ($) {
my $err = shift;
my $r = '';
if (not defined $err->{level} or $err->{level} eq 'm') {
$r = qq[MUST ‐level
error : ];
} elsif ($err->{level} eq 's') {
$r = qq[SHOULD ‐level
error : ];
} elsif ($err->{level} eq 'w') {
$r = qq[Warning :
];
} elsif ($err->{level} eq 'unsupported') {
$r = qq[Not
supported : ];
} else {
my $elevel = htescape ($err->{level});
$r = qq[$elevel :
];
}
return $r;
} # get_error_level_label
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) {
@r = ('') unless @r;
$rs = '';
$node = $node->parent_node;
} else {
$rs = '#' . $node->node_type;
$node = $node->parent_node;
}
unshift @r, $rs;
}
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, '<: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 ($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 =~ /;\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;
}
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
This library is free software; you can redistribute it
and/or modify it under the same terms as Perl itself.
=cut
## $Date: 2007/11/18 11:05:12 $