1 |
wakaba |
1.1 |
package WebHACC::Language::HTML; |
2 |
|
|
use strict; |
3 |
|
|
require WebHACC::Language::DOM; |
4 |
|
|
push our @ISA, 'WebHACC::Language::DOM'; |
5 |
|
|
|
6 |
|
|
sub new ($) { |
7 |
|
|
return bless {}, shift; |
8 |
|
|
} # new |
9 |
|
|
|
10 |
|
|
sub generate_syntax_error_section ($) { |
11 |
|
|
my $self = shift; |
12 |
|
|
|
13 |
wakaba |
1.4 |
require Message::DOM::DOMImplementation; |
14 |
wakaba |
1.1 |
require Encode; |
15 |
|
|
require Whatpm::HTML; |
16 |
|
|
|
17 |
|
|
my $out = $self->output; |
18 |
wakaba |
1.3 |
$out->start_section (role => 'parse-errors'); |
19 |
|
|
$out->start_error_list (role => 'parse-errors'); |
20 |
wakaba |
1.5 |
$self->result->layer_applicable ('syntax'); |
21 |
wakaba |
1.1 |
|
22 |
|
|
my $input = $self->input; |
23 |
|
|
my $result = $self->result; |
24 |
|
|
|
25 |
|
|
my $onerror = sub { |
26 |
wakaba |
1.7 |
my %opt = @_; |
27 |
|
|
$result->add_error (layer => 'syntax', %opt); |
28 |
|
|
|
29 |
|
|
if ($opt{type} eq 'chardecode:no error') { |
30 |
|
|
$self->result->layer_uncertain ('encode'); |
31 |
|
|
} elsif ($opt{type} eq 'chardecode:fallback') { |
32 |
|
|
$self->result->layer_uncertain ('charset'); |
33 |
|
|
$self->result->layer_uncertain ('syntax'); |
34 |
|
|
$self->result->layer_uncertain ('structure'); |
35 |
|
|
$self->result->layer_uncertain ('semantics'); |
36 |
|
|
} |
37 |
wakaba |
1.1 |
}; |
38 |
|
|
|
39 |
wakaba |
1.10 |
$self->result->layer_applicable ('charset'); |
40 |
|
|
my $char_checker = sub ($) { |
41 |
|
|
require Whatpm::Charset::UnicodeChecker; |
42 |
wakaba |
1.11 |
return Whatpm::Charset::UnicodeChecker->new_handle ($_[0], 'html5'); |
43 |
wakaba |
1.10 |
}; # $char_checker |
44 |
|
|
|
45 |
wakaba |
1.1 |
my $dom = Message::DOM::DOMImplementation->new; |
46 |
|
|
my $doc = $dom->create_document; |
47 |
|
|
my $el; |
48 |
|
|
my $inner_html_element = $input->{inner_html_element}; |
49 |
|
|
if (defined $inner_html_element and length $inner_html_element) { |
50 |
wakaba |
1.7 |
$input->{charset} ||= 'utf-8'; |
51 |
wakaba |
1.1 |
my $t = \($input->{s}); |
52 |
|
|
unless ($input->{is_char_string}) { |
53 |
|
|
$t = \(Encode::decode ($input->{charset}, $$t)); |
54 |
wakaba |
1.7 |
$self->result->layer_applicable ('encode'); |
55 |
wakaba |
1.1 |
} |
56 |
|
|
|
57 |
|
|
$el = $doc->create_element_ns |
58 |
|
|
('http://www.w3.org/1999/xhtml', [undef, $inner_html_element]); |
59 |
wakaba |
1.10 |
Whatpm::HTML->set_inner_html ($el, $$t, $onerror, $char_checker); |
60 |
wakaba |
1.1 |
|
61 |
|
|
$self->{structure} = $el; |
62 |
wakaba |
1.7 |
$self->{_structure_root} = $doc; |
63 |
|
|
## NOTE: This is necessary, otherwise it would be garbage collected |
64 |
|
|
## before $el is useless, since $el->owner_document is only a weak |
65 |
|
|
## reference. |
66 |
wakaba |
1.1 |
} else { |
67 |
|
|
if ($input->{is_char_string}) { |
68 |
wakaba |
1.10 |
Whatpm::HTML->parse_char_string ($input->{s} => $doc, |
69 |
|
|
$onerror, $char_checker); |
70 |
wakaba |
1.1 |
} else { |
71 |
wakaba |
1.7 |
$self->result->layer_applicable ('encode'); |
72 |
wakaba |
1.1 |
Whatpm::HTML->parse_byte_string |
73 |
wakaba |
1.10 |
($input->{charset}, $input->{s} => $doc, $onerror, $char_checker); |
74 |
wakaba |
1.1 |
} |
75 |
|
|
|
76 |
|
|
$self->{structure} = $doc; |
77 |
|
|
} |
78 |
|
|
$doc->manakai_charset ($input->{official_charset}) |
79 |
|
|
if defined $input->{official_charset}; |
80 |
|
|
|
81 |
wakaba |
1.9 |
$doc->document_uri ($input->url); |
82 |
wakaba |
1.1 |
$doc->manakai_entity_base_uri ($input->{base_uri}); |
83 |
|
|
|
84 |
wakaba |
1.8 |
$doc->input_encoding (undef) if $input->isa ('WebHACC::Input::Text'); |
85 |
|
|
|
86 |
wakaba |
1.3 |
$out->end_error_list (role => 'parse-errors'); |
87 |
wakaba |
1.1 |
$out->end_section; |
88 |
|
|
} # generate_syntax_error_section |
89 |
|
|
|
90 |
|
|
sub source_charset ($) { |
91 |
|
|
my $self = shift; |
92 |
|
|
return $self->input->{charset} || ($self->{structure}->owner_document || $self->{structure})->input_encoding; |
93 |
|
|
## TODO: Can we always use input_encoding? |
94 |
|
|
} # source_charset |
95 |
|
|
|
96 |
|
|
1; |