/[suikacvs]/markup/html/whatpm/Whatpm/HTML.pm.src
Suika

Diff of /markup/html/whatpm/Whatpm/HTML.pm.src

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.210 by wakaba, Tue Oct 14 13:24:52 2008 UTC revision 1.215 by wakaba, Thu Jul 2 22:11:51 2009 UTC
# Line 525  sub parse_byte_stream ($$$$;$$) { Line 525  sub parse_byte_stream ($$$$;$$) {
525            
526      if ($char_stream) { # if supported      if ($char_stream) { # if supported
527        ## "Change the encoding" algorithm:        ## "Change the encoding" algorithm:
   
       ## Step 1      
       if ($charset->{category} &  
           Message::Charset::Info::CHARSET_CATEGORY_UTF16 ()) {  
         $charset = Message::Charset::Info->get_by_html_name ('utf-8');  
         ($char_stream, $e_status) = $charset->get_decode_handle  
             ($byte_stream,  
              byte_buffer => \ $buffer->{buffer});  
       }  
       $charset_name = $charset->get_iana_name;  
528                
529        ## Step 2        ## Step 1
530        if (defined $self->{input_encoding} and        if (defined $self->{input_encoding} and
531            $self->{input_encoding} eq $charset_name) {            $self->{input_encoding} eq $charset_name) {
532          !!!parse-error (type => 'charset label:matching',          !!!parse-error (type => 'charset label:matching',
# Line 546  sub parse_byte_stream ($$$$;$$) { Line 536  sub parse_byte_stream ($$$$;$$) {
536          return;          return;
537        }        }
538    
539          ## Step 2 (HTML5 revision 3205)
540          if (defined $self->{input_encoding} and
541              Message::Charset::Info->get_by_html_name ($self->{input_encoding})
542              ->{category} & Message::Charset::Info::CHARSET_CATEGORY_UTF16 ()) {
543            $self->{confident} = 1;
544            return;
545          }
546    
547          ## Step 3
548          if ($charset->{category} &
549              Message::Charset::Info::CHARSET_CATEGORY_UTF16 ()) {
550            $charset = Message::Charset::Info->get_by_html_name ('utf-8');
551            ($char_stream, $e_status) = $charset->get_decode_handle
552                ($byte_stream,
553                 byte_buffer => \ $buffer->{buffer});
554          }
555          $charset_name = $charset->get_iana_name;
556    
557        !!!parse-error (type => 'charset label detected',        !!!parse-error (type => 'charset label detected',
558                        text => $self->{input_encoding},                        text => $self->{input_encoding},
559                        value => $charset_name,                        value => $charset_name,
560                        level => $self->{level}->{warn},                        level => $self->{level}->{warn},
561                        token => $token);                        token => $token);
562                
563        ## Step 3        ## Step 4
564        # if (can) {        # if (can) {
565          ## change the encoding on the fly.          ## change the encoding on the fly.
566          #$self->{confident} = 1;          #$self->{confident} = 1;
567          #return;          #return;
568        # }        # }
569                
570        ## Step 4        ## Step 5
571        throw Whatpm::HTML::RestartParser ();        throw Whatpm::HTML::RestartParser ();
572      }      }
573    }; # $self->{change_encoding}    }; # $self->{change_encoding}
# Line 936  sub _tree_construction_initial ($) { Line 944  sub _tree_construction_initial ($) {
944        $doctype->public_id ($token->{pubid}) if defined $token->{pubid};        $doctype->public_id ($token->{pubid}) if defined $token->{pubid};
945        $doctype->system_id ($token->{sysid}) if defined $token->{sysid};        $doctype->system_id ($token->{sysid}) if defined $token->{sysid};
946        ## NOTE: Other DocumentType attributes are null or empty lists.        ## NOTE: Other DocumentType attributes are null or empty lists.
947        ## ISSUE: internalSubset = null??        ## In Firefox3, |internalSubset| attribute is set to the empty
948          ## string, while |null| is an allowed value for the attribute
949          ## according to DOM3 Core.
950        $self->{document}->append_child ($doctype);        $self->{document}->append_child ($doctype);
951                
952        if ($token->{quirks} or $doctype_name ne 'HTML') {        if ($token->{quirks} or $doctype_name ne 'HTML') {
# Line 2500  sub _tree_construction_main ($) { Line 2510  sub _tree_construction_main ($) {
2510                  ## Reprocess in the "after head" insertion mode...                  ## Reprocess in the "after head" insertion mode...
2511                } elsif ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {                } elsif ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
2512                  !!!cp ('t143.3');                  !!!cp ('t143.3');
2513                  ## ISSUE: Two parse errors for <head><noscript></br>                  ## NOTE: Two parse errors for <head><noscript></br>
2514                  !!!parse-error (type => 'unmatched end tag',                  !!!parse-error (type => 'unmatched end tag',
2515                                  text => 'br', token => $token);                                  text => 'br', token => $token);
2516                  ## As if </noscript>                  ## As if </noscript>
# Line 4790  sub _tree_construction_main ($) { Line 4800  sub _tree_construction_main ($) {
4800                           line => $token->{line}, column => $token->{column}},                           line => $token->{line}, column => $token->{column}},
4801                          {type => START_TAG_TOKEN, tag_name => 'hr',                          {type => START_TAG_TOKEN, tag_name => 'hr',
4802                           line => $token->{line}, column => $token->{column}},                           line => $token->{line}, column => $token->{column}},
                         {type => START_TAG_TOKEN, tag_name => 'p',  
                          line => $token->{line}, column => $token->{column}},  
4803                          {type => START_TAG_TOKEN, tag_name => 'label',                          {type => START_TAG_TOKEN, tag_name => 'label',
4804                           line => $token->{line}, column => $token->{column}},                           line => $token->{line}, column => $token->{column}},
4805                         );                         );
# Line 4814  sub _tree_construction_main ($) { Line 4822  sub _tree_construction_main ($) {
4822                          #{type => CHARACTER_TOKEN, data => ''}, # SHOULD                          #{type => CHARACTER_TOKEN, data => ''}, # SHOULD
4823                          {type => END_TAG_TOKEN, tag_name => 'label',                          {type => END_TAG_TOKEN, tag_name => 'label',
4824                           line => $token->{line}, column => $token->{column}},                           line => $token->{line}, column => $token->{column}},
                         {type => END_TAG_TOKEN, tag_name => 'p',  
                          line => $token->{line}, column => $token->{column}},  
4825                          {type => START_TAG_TOKEN, tag_name => 'hr',                          {type => START_TAG_TOKEN, tag_name => 'hr',
4826                           line => $token->{line}, column => $token->{column}},                           line => $token->{line}, column => $token->{column}},
4827                          {type => END_TAG_TOKEN, tag_name => 'form',                          {type => END_TAG_TOKEN, tag_name => 'form',
# Line 4899  sub _tree_construction_main ($) { Line 4905  sub _tree_construction_main ($) {
4905              last INSCOPE;              last INSCOPE;
4906            }            }
4907          } # INSCOPE          } # INSCOPE
4908              
4909            ## TODO: <non-ruby><rt> is not allowed.
4910    
4911          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
4912    
# Line 5613  sub set_inner_html ($$$$;$) { Line 5621  sub set_inner_html ($$$$;$) {
5621      }->{$node_ln};      }->{$node_ln};
5622      $p->{content_model} = PCDATA_CONTENT_MODEL      $p->{content_model} = PCDATA_CONTENT_MODEL
5623          unless defined $p->{content_model};          unless defined $p->{content_model};
         ## ISSUE: What is "the name of the element"? local name?  
5624    
5625      $p->{inner_html_node} = [$node, $el_category->{$node_ln}];      $p->{inner_html_node} = [$node, $el_category->{$node_ln}];
5626        ## TODO: Foreign element OK?        ## TODO: Foreign element OK?
Legend:
Removed from v.1.210  
changed lines
  Added in v.1.215
[email protected]
 ViewVC Help
Powered by ViewVC 1.1.24