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

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

Parent Directory | Revision Log | View Patch Patch

-revision 1.156 by wakaba,
Sat Aug 30 13:43:50 2008 UTC
+revision 1.208 by wakaba,
Tue Oct 14 02:27:55 2008 UTC
 Line 3 
 use strict;
  our $VERSION=do{my @r=(q$Revision$=~/\d+/g);sprintf "%d."."%02d" x $#r,@r};
  use Error qw(:try);
+ use Whatpm::HTML::Tokenizer;
+ ## NOTE: This module don't check all HTML5 parse errors; character
+ ## encoding related parse errors are expected to be handled by relevant
+ ## modules.
+ ## Parse errors for control characters that are not allowed in HTML5
+ ## documents, for surrogate code points, and for noncharacter code
+ ## points, as well as U+FFFD substitions for characters whose code points
+ ## is higher than U+10FFFF may be detected by combining the parser with
+ ## the checker implemented by Whatpm::Charset::UnicodeChecker (for its
+ ## usage example, see |t/HTML-tree.t| in the Whatpm package or the
+ ## WebHACC::Language::HTML module in the WebHACC package).
  ## ISSUE:
  ## var doc = implementation.createDocument (null, null, null);
  ## doc.write ('');
-Line 10 
 use Error qw(:try);
+Line 23 
 use Error qw(:try);
  require IO::Handle;
+ ## Namespace URLs
  my $HTML_NS = q<http://www.w3.org/1999/xhtml>;
  my $MML_NS = q<http://www.w3.org/1998/Math/MathML>;
  my $SVG_NS = q<http://www.w3.org/2000/svg>;
-Line 17 
 my $XLINK_NS = q<http://www.w3.org/1999/
+Line 32 
 my $XLINK_NS = q<http://www.w3.org/1999/
  my $XML_NS = q<http://www.w3.org/XML/1998/namespace>;
  my $XMLNS_NS = q<http://www.w3.org/2000/xmlns/>;
- sub A_EL () { 0b1 }
+ ## Element categories
- sub ADDRESS_EL () { 0b10 }
- sub BODY_EL () { 0b100 }
- sub BUTTON_EL () { 0b1000 }
- sub CAPTION_EL () { 0b10000 }
- sub DD_EL () { 0b100000 }
- sub DIV_EL () { 0b1000000 }
- sub DT_EL () { 0b10000000 }
- sub FORM_EL () { 0b100000000 }
- sub FORMATTING_EL () { 0b1000000000 }
- sub FRAMESET_EL () { 0b10000000000 }
- sub HEADING_EL () { 0b100000000000 }
- sub HTML_EL () { 0b1000000000000 }
- sub LI_EL () { 0b10000000000000 }
- sub NOBR_EL () { 0b100000000000000 }
- sub OPTION_EL () { 0b1000000000000000 }
- sub OPTGROUP_EL () { 0b10000000000000000 }
- sub P_EL () { 0b100000000000000000 }
- sub SELECT_EL () { 0b1000000000000000000 }
- sub TABLE_EL () { 0b10000000000000000000 }
- sub TABLE_CELL_EL () { 0b100000000000000000000 }
- sub TABLE_ROW_EL () { 0b1000000000000000000000 }
- sub TABLE_ROW_GROUP_EL () { 0b10000000000000000000000 }
- sub MISC_SCOPING_EL () { 0b100000000000000000000000 }
- sub MISC_SPECIAL_EL () { 0b1000000000000000000000000 }
- sub FOREIGN_EL () { 0b10000000000000000000000000 }
- sub FOREIGN_FLOW_CONTENT_EL () { 0b100000000000000000000000000 }
- sub MML_AXML_EL () { 0b1000000000000000000000000000 }
- sub RUBY_EL () { 0b10000000000000000000000000000 }
- sub RUBY_COMPONENT_EL () { 0b100000000000000000000000000000 }
- sub TABLE_ROWS_EL () {
-   TABLE_EL |
-   TABLE_ROW_EL |
-   TABLE_ROW_GROUP_EL
- }
- ## NOTE: Used in "generate implied end tags" algorithm.
+ ## Bits 12-15
- ## NOTE: There is a code where a modified version of END_TAG_OPTIONAL_EL
+ sub SPECIAL_EL () { 0b1_000000000000000 }
- ## is used in "generate implied end tags" implementation (search for the
+ sub SCOPING_EL () { 0b1_00000000000000 }
- ## function mae).
+ sub FORMATTING_EL () { 0b1_0000000000000 }
- sub END_TAG_OPTIONAL_EL () {
+ sub PHRASING_EL () { 0b1_000000000000 }
-   DD_EL |
-   DT_EL |
+ ## Bits 10-11
-   LI_EL |
+ #sub FOREIGN_EL () { 0b1_00000000000 } # see Whatpm::HTML::Tokenizer
-   P_EL |
+ sub FOREIGN_FLOW_CONTENT_EL () { 0b1_0000000000 }
-   RUBY_COMPONENT_EL
- }
+ ## Bits 6-9
+ sub TABLE_SCOPING_EL () { 0b1_000000000 }
+ sub TABLE_ROWS_SCOPING_EL () { 0b1_00000000 }
+ sub TABLE_ROW_SCOPING_EL () { 0b1_0000000 }
+ sub TABLE_ROWS_EL () { 0b1_000000 }
+ ## Bit 5
+ sub ADDRESS_DIV_P_EL () { 0b1_00000 }
  ## NOTE: Used in </body> and EOF algorithms.
- sub ALL_END_TAG_OPTIONAL_EL () {
+ ## Bit 4
-   DD_EL |
+ sub ALL_END_TAG_OPTIONAL_EL () { 0b1_0000 }
-   DT_EL |
-   LI_EL |
-   P_EL |
-   BODY_EL |
-   HTML_EL |
-   TABLE_CELL_EL |
-   TABLE_ROW_EL |
-   TABLE_ROW_GROUP_EL
- }
- sub SCOPING_EL () {
+ ## NOTE: Used in "generate implied end tags" algorithm.
-   BUTTON_EL |
+ ## NOTE: There is a code where a modified version of
-   CAPTION_EL |
+ ## END_TAG_OPTIONAL_EL is used in "generate implied end tags"
-   HTML_EL |
+ ## implementation (search for the algorithm name).
-   TABLE_EL |
+ ## Bit 3
-   TABLE_CELL_EL |
+ sub END_TAG_OPTIONAL_EL () { 0b1_000 }
-   MISC_SCOPING_EL
+ ## Bits 0-2
+ sub MISC_SPECIAL_EL () { SPECIAL_EL | 0b000 }
+ sub FORM_EL () { SPECIAL_EL | 0b001 }
+ sub FRAMESET_EL () { SPECIAL_EL | 0b010 }
+ sub HEADING_EL () { SPECIAL_EL | 0b011 }
+ sub SELECT_EL () { SPECIAL_EL | 0b100 }
+ sub SCRIPT_EL () { SPECIAL_EL | 0b101 }
+ sub ADDRESS_DIV_EL () { SPECIAL_EL | ADDRESS_DIV_P_EL | 0b001 }
+ sub BODY_EL () { SPECIAL_EL | ALL_END_TAG_OPTIONAL_EL | 0b001 }
+ sub DTDD_EL () {
+   SPECIAL_EL |
+   END_TAG_OPTIONAL_EL |
+   ALL_END_TAG_OPTIONAL_EL |
+b010
  }
+ sub LI_EL () {
- sub TABLE_SCOPING_EL () {
+   SPECIAL_EL |
-   HTML_EL |
+   END_TAG_OPTIONAL_EL |
-   TABLE_EL
+   ALL_END_TAG_OPTIONAL_EL |
+b100
  }
+ sub P_EL () {
- sub TABLE_ROWS_SCOPING_EL () {
+   SPECIAL_EL |
-   HTML_EL |
+   ADDRESS_DIV_P_EL |
-   TABLE_ROW_GROUP_EL
+   END_TAG_OPTIONAL_EL |
+   ALL_END_TAG_OPTIONAL_EL |
+b001
  }
- sub TABLE_ROW_SCOPING_EL () {
+ sub TABLE_ROW_EL () {
-   HTML_EL |
+   SPECIAL_EL |
-   TABLE_ROW_EL
+   TABLE_ROWS_EL |
+   TABLE_ROW_SCOPING_EL |
+   ALL_END_TAG_OPTIONAL_EL |
+b001
+ }
+ sub TABLE_ROW_GROUP_EL () {
+   SPECIAL_EL |
+   TABLE_ROWS_EL |
+   TABLE_ROWS_SCOPING_EL |
+   ALL_END_TAG_OPTIONAL_EL |
+b001
  }
- sub SPECIAL_EL () {
+ sub MISC_SCOPING_EL () { SCOPING_EL | 0b000 }
-   ADDRESS_EL |
+ sub BUTTON_EL () { SCOPING_EL | 0b001 }
-   BODY_EL |
+ sub CAPTION_EL () { SCOPING_EL | 0b010 }
-   DIV_EL |
+ sub HTML_EL () {
+   SCOPING_EL |
-   DD_EL |
+   TABLE_SCOPING_EL |
-   DT_EL |
+   TABLE_ROWS_SCOPING_EL |
-   LI_EL |
+   TABLE_ROW_SCOPING_EL |
-   P_EL |
+   ALL_END_TAG_OPTIONAL_EL |
+b001
-   FORM_EL |
+ }
-   FRAMESET_EL |
+ sub TABLE_EL () {
-   HEADING_EL |
+   SCOPING_EL |
-   OPTION_EL |
+   TABLE_ROWS_EL |
-   OPTGROUP_EL |
+   TABLE_SCOPING_EL |
-   SELECT_EL |
+b001
-   TABLE_ROW_EL |
+ }
-   TABLE_ROW_GROUP_EL |
+ sub TABLE_CELL_EL () {
-   MISC_SPECIAL_EL
+   SCOPING_EL |
+   TABLE_ROW_SCOPING_EL |
+   ALL_END_TAG_OPTIONAL_EL |
+b001
  }
+ sub MISC_FORMATTING_EL () { FORMATTING_EL | 0b000 }
+ sub A_EL () { FORMATTING_EL | 0b001 }
+ sub NOBR_EL () { FORMATTING_EL | 0b010 }
+ sub RUBY_EL () { PHRASING_EL | 0b001 }
+ ## ISSUE: ALL_END_TAG_OPTIONAL_EL?
+ sub OPTGROUP_EL () { PHRASING_EL | END_TAG_OPTIONAL_EL | 0b001 }
+ sub OPTION_EL () { PHRASING_EL | END_TAG_OPTIONAL_EL | 0b010 }
+ sub RUBY_COMPONENT_EL () { PHRASING_EL | END_TAG_OPTIONAL_EL | 0b100 }
+ sub MML_AXML_EL () { PHRASING_EL | FOREIGN_EL | 0b001 }
  my $el_category = {
-   a => A_EL | FORMATTING_EL,
+   a => A_EL,
-   address => ADDRESS_EL,
+   address => ADDRESS_DIV_EL,
    applet => MISC_SCOPING_EL,
    area => MISC_SPECIAL_EL,
+   article => MISC_SPECIAL_EL,
+   aside => MISC_SPECIAL_EL,
    b => FORMATTING_EL,
    base => MISC_SPECIAL_EL,
    basefont => MISC_SPECIAL_EL,
-Line 143 
 my $el_category = {
+Line 168 
 my $el_category = {
    center => MISC_SPECIAL_EL,
    col => MISC_SPECIAL_EL,
    colgroup => MISC_SPECIAL_EL,
-   dd => DD_EL,
+   command => MISC_SPECIAL_EL,
+   datagrid => MISC_SPECIAL_EL,
+   dd => DTDD_EL,
+   details => MISC_SPECIAL_EL,
+   dialog => MISC_SPECIAL_EL,
    dir => MISC_SPECIAL_EL,
-   div => DIV_EL,
+   div => ADDRESS_DIV_EL,
    dl => MISC_SPECIAL_EL,
-   dt => DT_EL,
+   dt => DTDD_EL,
    em => FORMATTING_EL,
    embed => MISC_SPECIAL_EL,
+   eventsource => MISC_SPECIAL_EL,
    fieldset => MISC_SPECIAL_EL,
+   figure => MISC_SPECIAL_EL,
    font => FORMATTING_EL,
+   footer => MISC_SPECIAL_EL,
    form => FORM_EL,
    frame => MISC_SPECIAL_EL,
    frameset => FRAMESET_EL,
-Line 162 
 my $el_category = {
+Line 194 
 my $el_category = {
    h5 => HEADING_EL,
    h6 => HEADING_EL,
    head => MISC_SPECIAL_EL,
+   header => MISC_SPECIAL_EL,
    hr => MISC_SPECIAL_EL,
    html => HTML_EL,
    i => FORMATTING_EL,
    iframe => MISC_SPECIAL_EL,
    img => MISC_SPECIAL_EL,
+   #image => MISC_SPECIAL_EL, ## NOTE: Commented out in the spec.
    input => MISC_SPECIAL_EL,
    isindex => MISC_SPECIAL_EL,
    li => LI_EL,
-Line 175 
 my $el_category = {
+Line 209 
 my $el_category = {
    marquee => MISC_SCOPING_EL,
    menu => MISC_SPECIAL_EL,
    meta => MISC_SPECIAL_EL,
-   nobr => NOBR_EL | FORMATTING_EL,
+   nav => MISC_SPECIAL_EL,
+   nobr => NOBR_EL,
    noembed => MISC_SPECIAL_EL,
    noframes => MISC_SPECIAL_EL,
    noscript => MISC_SPECIAL_EL,
-Line 193 
 my $el_category = {
+Line 228 
 my $el_category = {
    s => FORMATTING_EL,
    script => MISC_SPECIAL_EL,
    select => SELECT_EL,
+   section => MISC_SPECIAL_EL,
    small => FORMATTING_EL,
    spacer => MISC_SPECIAL_EL,
    strike => FORMATTING_EL,
-Line 216 
 my $el_category = {
+Line 252 
 my $el_category = {
  my $el_category_f = {
    $MML_NS => {
      'annotation-xml' => MML_AXML_EL,
-     mi => FOREIGN_FLOW_CONTENT_EL,
+     mi => FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
-     mo => FOREIGN_FLOW_CONTENT_EL,
+     mo => FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
-     mn => FOREIGN_FLOW_CONTENT_EL,
+     mn => FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
-     ms => FOREIGN_FLOW_CONTENT_EL,
+     ms => FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
-     mtext => FOREIGN_FLOW_CONTENT_EL,
+     mtext => FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
    },
    $SVG_NS => {
-     foreignObject => FOREIGN_FLOW_CONTENT_EL,
+     foreignObject => SCOPING_EL | FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
-     desc => FOREIGN_FLOW_CONTENT_EL,
+     desc => FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
-     title => FOREIGN_FLOW_CONTENT_EL,
+     title => FOREIGN_EL | FOREIGN_FLOW_CONTENT_EL,
    },
    ## NOTE: In addition, FOREIGN_EL is set to non-HTML elements.
  };
-Line 312 
 my $foreign_attr_xname = {
+Line 348 
 my $foreign_attr_xname = {
  ## ISSUE: xmlns:xlink="non-xlink-ns" is not an error.
- my $c1_entity_char = {
+ ## TODO: Invoke the reset algorithm when a resettable element is
-x80 => 0x20AC,
+ ## created (cf. HTML5 revision 2259).
-x81 => 0xFFFD,
-x82 => 0x201A,
-x83 => 0x0192,
-x84 => 0x201E,
-x85 => 0x2026,
-x86 => 0x2020,
-x87 => 0x2021,
-x88 => 0x02C6,
-x89 => 0x2030,
-x8A => 0x0160,
-x8B => 0x2039,
-x8C => 0x0152,
-x8D => 0xFFFD,
-x8E => 0x017D,
-x8F => 0xFFFD,
-x90 => 0xFFFD,
-x91 => 0x2018,
-x92 => 0x2019,
-x93 => 0x201C,
-x94 => 0x201D,
-x95 => 0x2022,
-x96 => 0x2013,
-x97 => 0x2014,
-x98 => 0x02DC,
-x99 => 0x2122,
-x9A => 0x0161,
-x9B => 0x203A,
-x9C => 0x0153,
-x9D => 0xFFFD,
-x9E => 0x017E,
-x9F => 0x0178,
- }; # $c1_entity_char
  sub parse_byte_string ($$$$;$) {
    my $self = shift;
-Line 354 
 sub parse_byte_string ($$$$;$) {
+Line 358 
 sub parse_byte_string ($$$$;$) {
    return $self->parse_byte_stream ($charset_name, $input, @_[1..$#_]);
  } # parse_byte_string
- sub parse_byte_stream ($$$$;$) {
+ sub parse_byte_stream ($$$$;$$) {
+   # my ($self, $charset_name, $byte_stream, $doc, $onerror, $get_wrapper) = @_;
    my $self = ref $_[0] ? shift : shift->new;
    my $charset_name = shift;
    my $byte_stream = $_[0];
-Line 365 
 sub parse_byte_stream ($$$$;$) {
+Line 370 
 sub parse_byte_stream ($$$$;$) {
    };
    $self->{parse_error} = $onerror; # updated later by parse_char_string
+   my $get_wrapper = $_[3] || sub ($) {
+     return $_[0]; # $_[0] = byte stream handle, returned = arg to char handle
+   };
    ## HTML5 encoding sniffing algorithm
    require Message::Charset::Info;
    my $charset;
-Line 372 
 sub parse_byte_stream ($$$$;$) {
+Line 381 
 sub parse_byte_stream ($$$$;$) {
    my ($char_stream, $e_status);
    SNIFFING: {
+     ## NOTE: By setting |allow_fallback| option true when the
+     ## |get_decode_handle| method is invoked, we ignore what the HTML5
+     ## spec requires, i.e. unsupported encoding should be ignored.
+       ## TODO: We should not do this unless the parser is invoked
+       ## in the conformance checking mode, in which this behavior
+       ## would be useful.
      ## Step 1
      if (defined $charset_name) {
-       $charset = Message::Charset::Info->get_by_iana_name ($charset_name);
+       $charset = Message::Charset::Info->get_by_html_name ($charset_name);
+           ## TODO: Is this ok?  Transfer protocol's parameter should be
+           ## interpreted in its semantics?
-       ## ISSUE: Unsupported encoding is not ignored according to the spec.
        ($char_stream, $e_status) = $charset->get_decode_handle
            ($byte_stream, allow_error_reporting => 1,
             allow_fallback => 1);
-Line 385 
 sub parse_byte_stream ($$$$;$) {
+Line 401 
 sub parse_byte_stream ($$$$;$) {
          $self->{confident} = 1;
          last SNIFFING;
        } else {
-         ## TODO: unsupported error
+         !!!parse-error (type => 'charset:not supported',
+                         layer => 'encode',
+                         line => 1, column => 1,
+                         value => $charset_name,
+                         level => $self->{level}->{uncertain});
        }
      }
-Line 399 
 sub parse_byte_stream ($$$$;$) {
+Line 419 
 sub parse_byte_stream ($$$$;$) {
      ## Step 3
      if ($byte_buffer =~ /^\xFE\xFF/) {
-       $charset = Message::Charset::Info->get_by_iana_name ('utf-16be');
+       $charset = Message::Charset::Info->get_by_html_name ('utf-16be');
        ($char_stream, $e_status) = $charset->get_decode_handle
            ($byte_stream, allow_error_reporting => 1,
             allow_fallback => 1, byte_buffer => \$byte_buffer);
        $self->{confident} = 1;
        last SNIFFING;
      } elsif ($byte_buffer =~ /^\xFF\xFE/) {
-       $charset = Message::Charset::Info->get_by_iana_name ('utf-16le');
+       $charset = Message::Charset::Info->get_by_html_name ('utf-16le');
        ($char_stream, $e_status) = $charset->get_decode_handle
            ($byte_stream, allow_error_reporting => 1,
             allow_fallback => 1, byte_buffer => \$byte_buffer);
        $self->{confident} = 1;
        last SNIFFING;
      } elsif ($byte_buffer =~ /^\xEF\xBB\xBF/) {
-       $charset = Message::Charset::Info->get_by_iana_name ('utf-8');
+       $charset = Message::Charset::Info->get_by_html_name ('utf-8');
        ($char_stream, $e_status) = $charset->get_decode_handle
            ($byte_stream, allow_error_reporting => 1,
             allow_fallback => 1, byte_buffer => \$byte_buffer);
-Line 432 
 sub parse_byte_stream ($$$$;$) {
+Line 452 
 sub parse_byte_stream ($$$$;$) {
      $charset_name = Whatpm::Charset::UniversalCharDet->detect_byte_string
          ($byte_buffer);
      if (defined $charset_name) {
-       $charset = Message::Charset::Info->get_by_iana_name ($charset_name);
+       $charset = Message::Charset::Info->get_by_html_name ($charset_name);
-       ## ISSUE: Unsupported encoding is not ignored according to the spec.
        require Whatpm::Charset::DecodeHandle;
        $buffer = Whatpm::Charset::DecodeHandle::ByteBuffer->new
            ($byte_stream);
-Line 455 
 sub parse_byte_stream ($$$$;$) {
+Line 474 
 sub parse_byte_stream ($$$$;$) {
      ## Step 7: default
      ## TODO: Make this configurable.
-     $charset = Message::Charset::Info->get_by_iana_name ('windows-1252');
+     $charset = Message::Charset::Info->get_by_html_name ('windows-1252');
          ## NOTE: We choose |windows-1252| here, since |utf-8| should be
          ## detectable in the step 6.
      require Whatpm::Charset::DecodeHandle;
-Line 475 
 sub parse_byte_stream ($$$$;$) {
+Line 494 
 sub parse_byte_stream ($$$$;$) {
      $self->{confident} = 0;
    } # SNIFFING
-   $self->{input_encoding} = $charset->get_iana_name;
    if ($e_status & Message::Charset::Info::FALLBACK_ENCODING_IMPL ()) {
+     $self->{input_encoding} = $charset->get_iana_name; ## TODO: Should we set actual charset decoder's encoding name?
      !!!parse-error (type => 'chardecode:fallback',
-                     text => $self->{input_encoding},
+                     #text => $self->{input_encoding},
                      level => $self->{level}->{uncertain},
                      line => 1, column => 1,
                      layer => 'encode');
    } elsif (not ($e_status &
-                 Message::Charset::Info::ERROR_REPORTING_ENCODING_IMPL())) {
+                 Message::Charset::Info::ERROR_REPORTING_ENCODING_IMPL ())) {
+     $self->{input_encoding} = $charset->get_iana_name;
      !!!parse-error (type => 'chardecode:no error',
                      text => $self->{input_encoding},
                      level => $self->{level}->{uncertain},
                      line => 1, column => 1,
                      layer => 'encode');
+   } else {
+     $self->{input_encoding} = $charset->get_iana_name;
    }
    $self->{change_encoding} = sub {
-Line 496 
 sub parse_byte_stream ($$$$;$) {
+Line 518 
 sub parse_byte_stream ($$$$;$) {
      $charset_name = shift;
      my $token = shift;
-     $charset = Message::Charset::Info->get_by_iana_name ($charset_name);
+     $charset = Message::Charset::Info->get_by_html_name ($charset_name);
      ($char_stream, $e_status) = $charset->get_decode_handle
          ($byte_stream, allow_error_reporting => 1, allow_fallback => 1,
           byte_buffer => \ $buffer->{buffer});
-Line 507 
 sub parse_byte_stream ($$$$;$) {
+Line 529 
 sub parse_byte_stream ($$$$;$) {
        ## Step 1
        if ($charset->{category} &
            Message::Charset::Info::CHARSET_CATEGORY_UTF16 ()) {
-         $charset = Message::Charset::Info->get_by_iana_name ('utf-8');
+         $charset = Message::Charset::Info->get_by_html_name ('utf-8');
          ($char_stream, $e_status) = $charset->get_decode_handle
              ($byte_stream,
               byte_buffer => \ $buffer->{buffer});
-Line 545 
 sub parse_byte_stream ($$$$;$) {
+Line 567 
 sub parse_byte_stream ($$$$;$) {
    my $char_onerror = sub {
      my (undef, $type, %opt) = @_;
      !!!parse-error (layer => 'encode',
-                     %opt, type => $type,
+                     line => $self->{line}, column => $self->{column} + 1,
-                     line => $self->{line}, column => $self->{column} + 1);
+                     %opt, type => $type);
      if ($opt{octets}) {
        ${$opt{octets}} = "\x{FFFD}"; # relacement character
      }
    };
-   $char_stream->onerror ($char_onerror);
-   my @args = @_; shift @args; # $s
+   my $wrapped_char_stream = $get_wrapper->($char_stream);
+   $wrapped_char_stream->onerror ($char_onerror);
+   my @args = ($_[1], $_[2]); # $doc, $onerror - $get_wrapper = undef;
    my $return;
    try {
-     $return = $self->parse_char_stream ($char_stream, @args);
+     $return = $self->parse_char_stream ($wrapped_char_stream, @args);
    } catch Whatpm::HTML::RestartParser with {
      ## NOTE: Invoked after {change_encoding}.
-     $self->{input_encoding} = $charset->get_iana_name;
      if ($e_status & Message::Charset::Info::FALLBACK_ENCODING_IMPL ()) {
+       $self->{input_encoding} = $charset->get_iana_name; ## TODO: Should we set actual charset decoder's encoding name?
        !!!parse-error (type => 'chardecode:fallback',
-                       text => $self->{input_encoding},
                        level => $self->{level}->{uncertain},
+                       #text => $self->{input_encoding},
                        line => 1, column => 1,
                        layer => 'encode');
      } elsif (not ($e_status &
-                   Message::Charset::Info::ERROR_REPORTING_ENCODING_IMPL())) {
+                   Message::Charset::Info::ERROR_REPORTING_ENCODING_IMPL ())) {
+       $self->{input_encoding} = $charset->get_iana_name;
        !!!parse-error (type => 'chardecode:no error',
                        text => $self->{input_encoding},
                        level => $self->{level}->{uncertain},
                        line => 1, column => 1,
                        layer => 'encode');
+     } else {
+       $self->{input_encoding} = $charset->get_iana_name;
      }
      $self->{confident} = 1;
-     $char_stream->onerror ($char_onerror);
-     $return = $self->parse_char_stream ($char_stream, @args);
+     $wrapped_char_stream = $get_wrapper->($char_stream);
+     $wrapped_char_stream->onerror ($char_onerror);
+     $return = $self->parse_char_stream ($wrapped_char_stream, @args);
    };
    return $return;
  } # parse_byte_stream
-Line 591 
 sub parse_byte_stream ($$$$;$) {
+Line 621 
 sub parse_byte_stream ($$$$;$) {
  ## such as |parse_byte_string| in this module, must ensure that it does
  ## strip the BOM and never strip any ZWNBSP.
- sub parse_char_string ($$$;$) {
+ sub parse_char_string ($$$;$$) {
+   #my ($self, $s, $doc, $onerror, $get_wrapper) = @_;
    my $self = shift;
-   require utf8;
    my $s = ref $_[0] ? $_[0] : \($_[0]);
-   open my $input, '<' . (utf8::is_utf8 ($$s) ? ':utf8' : ''), $s;
+   require Whatpm::Charset::DecodeHandle;
+   my $input = Whatpm::Charset::DecodeHandle::CharString->new ($s);
    return $self->parse_char_stream ($input, @_[1..$#_]);
  } # parse_char_string
- *parse_string = \&parse_char_string;
+ *parse_string = \&parse_char_string; ## NOTE: Alias for backward compatibility.
- sub parse_char_stream ($$$;$) {
+ sub parse_char_stream ($$$;$$) {
    my $self = ref $_[0] ? shift : shift->new;
    my $input = $_[0];
    $self->{document} = $_[1];
-Line 611 
 sub parse_char_stream ($$$;$) {
+Line 642 
 sub parse_char_stream ($$$;$) {
    $self->{confident} = 1 unless exists $self->{confident};
    $self->{document}->input_encoding ($self->{input_encoding})
        if defined $self->{input_encoding};
+ ## TODO: |{input_encoding}| is needless?
-   my $i = 0;
    $self->{line_prev} = $self->{line} = 1;
-   $self->{column_prev} = $self->{column} = 0;
+   $self->{column_prev} = -1;
-   $self->{set_next_char} = sub {
+   $self->{column} = 0;
+   $self->{set_nc} = sub {
      my $self = shift;
-     pop @{$self->{prev_char}};
+     my $char = '';
-     unshift @{$self->{prev_char}}, $self->{next_char};
+     if (defined $self->{next_nc}) {
+       $char = $self->{next_nc};
-     my $char;
+       delete $self->{next_nc};
-     if (defined $self->{next_next_char}) {
+       $self->{nc} = ord $char;
-       $char = $self->{next_next_char};
-       delete $self->{next_next_char};
      } else {
-       $char = $input->getc;
+       $self->{char_buffer} = '';
+       $self->{char_buffer_pos} = 0;
+       my $count = $input->manakai_read_until
+          ($self->{char_buffer}, qr/[^\x00\x0A\x0D]/, $self->{char_buffer_pos});
+       if ($count) {
+         $self->{line_prev} = $self->{line};
+         $self->{column_prev} = $self->{column};
+         $self->{column}++;
+         $self->{nc}
+             = ord substr ($self->{char_buffer}, $self->{char_buffer_pos}++, 1);
+         return;
+       }
+       if ($input->read ($char, 1)) {
+         $self->{nc} = ord $char;
+       } else {
+         $self->{nc} = -1;
+         return;
+       }
      }
-     $self->{next_char} = -1 and return unless defined $char;
-     $self->{next_char} = ord $char;
      ($self->{line_prev}, $self->{column_prev})
          = ($self->{line}, $self->{column});
      $self->{column}++;
-     if ($self->{next_char} == 0x000A) { # LF
+     if ($self->{nc} == 0x000A) { # LF
        !!!cp ('j1');
        $self->{line}++;
        $self->{column} = 0;
-     } elsif ($self->{next_char} == 0x000D) { # CR
+     } elsif ($self->{nc} == 0x000D) { # CR
        !!!cp ('j2');
-       my $next = $input->getc;
+ ## TODO: support for abort/streaming
-       if (defined $next and $next ne "\x0A") {
+       my $next = '';
-         $self->{next_next_char} = $next;
+       if ($input->read ($next, 1) and $next ne "\x0A") {
+         $self->{next_nc} = $next;
        }
-       $self->{next_char} = 0x000A; # LF # MUST
+       $self->{nc} = 0x000A; # LF # MUST
        $self->{line}++;
        $self->{column} = 0;
-     } elsif ($self->{next_char} > 0x10FFFF) {
+     } elsif ($self->{nc} == 0x0000) { # NULL
-       !!!cp ('j3');
-       $self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
-     } elsif ($self->{next_char} == 0x0000) { # NULL
        !!!cp ('j4');
        !!!parse-error (type => 'NULL');
-       $self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
+       $self->{nc} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
-     } elsif ($self->{next_char} <= 0x0008 or
+     }
-              (0x000E <= $self->{next_char} and $self->{next_char} <= 0x001F) or
+   };
-              (0x007F <= $self->{next_char} and $self->{next_char} <= 0x009F) or
-              (0xD800 <= $self->{next_char} and $self->{next_char} <= 0xDFFF) or
+   $self->{read_until} = sub {
-              (0xFDD0 <= $self->{next_char} and $self->{next_char} <= 0xFDDF) or
+     #my ($scalar, $specials_range, $offset) = @_;
-              {
+     return 0 if defined $self->{next_nc};
-xFFFE => 1, 0xFFFF => 1, 0x1FFFE => 1, 0x1FFFF => 1,
-x2FFFE => 1, 0x2FFFF => 1, 0x3FFFE => 1, 0x3FFFF => 1,
+     my $pattern = qr/[^$_[1]\x00\x0A\x0D]/;
-x4FFFE => 1, 0x4FFFF => 1, 0x5FFFE => 1, 0x5FFFF => 1,
+     my $offset = $_[2] || 0;
-x6FFFE => 1, 0x6FFFF => 1, 0x7FFFE => 1, 0x7FFFF => 1,
-x8FFFE => 1, 0x8FFFF => 1, 0x9FFFE => 1, 0x9FFFF => 1,
+     if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
-xAFFFE => 1, 0xAFFFF => 1, 0xBFFFE => 1, 0xBFFFF => 1,
+       pos ($self->{char_buffer}) = $self->{char_buffer_pos};
-xCFFFE => 1, 0xCFFFF => 1, 0xDFFFE => 1, 0xDFFFF => 1,
+       if ($self->{char_buffer} =~ /\G(?>$pattern)+/) {
-xEFFFE => 1, 0xEFFFF => 1, 0xFFFFE => 1, 0xFFFFF => 1,
+         substr ($_[0], $offset)
-x10FFFE => 1, 0x10FFFF => 1,
+             = substr ($self->{char_buffer}, $-[0], $+[0] - $-[0]);
-              }->{$self->{next_char}}) {
+         my $count = $+[0] - $-[0];
-       !!!cp ('j5');
+         if ($count) {
-       if ($self->{next_char} < 0x10000) {
+           $self->{column} += $count;
-         !!!parse-error (type => 'control char',
+           $self->{char_buffer_pos} += $count;
-                         text => (sprintf 'U+%04X', $self->{next_char}));
+           $self->{line_prev} = $self->{line};
+           $self->{column_prev} = $self->{column} - 1;
+           $self->{nc} = -1;
+         }
+         return $count;
        } else {
-         !!!parse-error (type => 'control char',
+         return 0;
-                         text => (sprintf 'U-%08X', $self->{next_char}));
        }
+     } else {
+       my $count = $input->manakai_read_until ($_[0], $pattern, $_[2]);
+       if ($count) {
+         $self->{column} += $count;
+         $self->{line_prev} = $self->{line};
+         $self->{column_prev} = $self->{column} - 1;
+         $self->{nc} = -1;
+       }
+       return $count;
      }
-   };
+   }; # $self->{read_until}
-   $self->{prev_char} = [-1, -1, -1];
-   $self->{next_char} = -1;
    my $onerror = $_[2] || sub {
      my (%opt) = @_;
-Line 694 
 sub parse_char_stream ($$$;$) {
+Line 749 
 sub parse_char_stream ($$$;$) {
      $onerror->(line => $self->{line}, column => $self->{column}, @_);
    };
+   my $char_onerror = sub {
+     my (undef, $type, %opt) = @_;
+     !!!parse-error (layer => 'encode',
+                     line => $self->{line}, column => $self->{column} + 1,
+                     %opt, type => $type);
+   }; # $char_onerror
+   if ($_[3]) {
+     $input = $_[3]->($input);
+     $input->onerror ($char_onerror);
+   } else {
+     $input->onerror ($char_onerror) unless defined $input->onerror;
+   }
    $self->_initialize_tokenizer;
    $self->_initialize_tree_constructor;
    $self->_construct_tree;
-Line 708 
 sub new ($) {
+Line 777 
 sub new ($) {
    my $class = shift;
    my $self = bless {
      level => {must => 'm',
+               should => 's',
                warn => 'w',
                info => 'i',
                uncertain => 'u'},
    }, $class;
-   $self->{set_next_char} = sub {
+   $self->{set_nc} = sub {
-     $self->{next_char} = -1;
+     $self->{nc} = -1;
    };
    $self->{parse_error} = sub {
      #
-Line 730 
 sub new ($) {
+Line 800 
 sub new ($) {
    return $self;
  } # new
- sub CM_ENTITY () { 0b001 } # & markup in data
+ ## Insertion modes
- sub CM_LIMITED_MARKUP () { 0b010 } # < markup in data (limited)
- sub CM_FULL_MARKUP () { 0b100 } # < markup in data (any)
- sub PLAINTEXT_CONTENT_MODEL () { 0 }
- sub CDATA_CONTENT_MODEL () { CM_LIMITED_MARKUP }
- sub RCDATA_CONTENT_MODEL () { CM_ENTITY | CM_LIMITED_MARKUP }
- sub PCDATA_CONTENT_MODEL () { CM_ENTITY | CM_FULL_MARKUP }
- sub DATA_STATE () { 0 }
- sub ENTITY_DATA_STATE () { 1 }
- sub TAG_OPEN_STATE () { 2 }
- sub CLOSE_TAG_OPEN_STATE () { 3 }
- sub TAG_NAME_STATE () { 4 }
- sub BEFORE_ATTRIBUTE_NAME_STATE () { 5 }
- sub ATTRIBUTE_NAME_STATE () { 6 }
- sub AFTER_ATTRIBUTE_NAME_STATE () { 7 }
- sub BEFORE_ATTRIBUTE_VALUE_STATE () { 8 }
- sub ATTRIBUTE_VALUE_DOUBLE_QUOTED_STATE () { 9 }
- sub ATTRIBUTE_VALUE_SINGLE_QUOTED_STATE () { 10 }
- sub ATTRIBUTE_VALUE_UNQUOTED_STATE () { 11 }
- sub ENTITY_IN_ATTRIBUTE_VALUE_STATE () { 12 }
- sub MARKUP_DECLARATION_OPEN_STATE () { 13 }
- sub COMMENT_START_STATE () { 14 }
- sub COMMENT_START_DASH_STATE () { 15 }
- sub COMMENT_STATE () { 16 }
- sub COMMENT_END_STATE () { 17 }
- sub COMMENT_END_DASH_STATE () { 18 }
- sub BOGUS_COMMENT_STATE () { 19 }
- sub DOCTYPE_STATE () { 20 }
- sub BEFORE_DOCTYPE_NAME_STATE () { 21 }
- sub DOCTYPE_NAME_STATE () { 22 }
- sub AFTER_DOCTYPE_NAME_STATE () { 23 }
- sub BEFORE_DOCTYPE_PUBLIC_IDENTIFIER_STATE () { 24 }
- sub DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED_STATE () { 25 }
- sub DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED_STATE () { 26 }
- sub AFTER_DOCTYPE_PUBLIC_IDENTIFIER_STATE () { 27 }
- sub BEFORE_DOCTYPE_SYSTEM_IDENTIFIER_STATE () { 28 }
- sub DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED_STATE () { 29 }
- sub DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED_STATE () { 30 }
- sub AFTER_DOCTYPE_SYSTEM_IDENTIFIER_STATE () { 31 }
- sub BOGUS_DOCTYPE_STATE () { 32 }
- sub AFTER_ATTRIBUTE_VALUE_QUOTED_STATE () { 33 }
- sub SELF_CLOSING_START_TAG_STATE () { 34 }
- sub CDATA_BLOCK_STATE () { 35 }
- sub DOCTYPE_TOKEN () { 1 }
- sub COMMENT_TOKEN () { 2 }
- sub START_TAG_TOKEN () { 3 }
- sub END_TAG_TOKEN () { 4 }
- sub END_OF_FILE_TOKEN () { 5 }
- sub CHARACTER_TOKEN () { 6 }
  sub AFTER_HTML_IMS () { 0b100 }
  sub HEAD_IMS ()       { 0b1000 }
-Line 792 
 sub ROW_IMS ()        { 0b10000000 }
+Line 811 
 sub ROW_IMS ()        { 0b10000000 }
  sub BODY_AFTER_IMS () { 0b100000000 }
  sub FRAME_IMS ()      { 0b1000000000 }
  sub SELECT_IMS ()     { 0b10000000000 }
- sub IN_FOREIGN_CONTENT_IM () { 0b100000000000 }
+ #sub IN_FOREIGN_CONTENT_IM () { 0b100000000000 } # see Whatpm::HTML::Tokenizer
      ## NOTE: "in foreign content" insertion mode is special; it is combined
      ## with the secondary insertion mode.  In this parser, they are stored
      ## together in the bit-or'ed form.
+ sub IN_CDATA_RCDATA_IM () { 0b1000000000000 }
+     ## NOTE: "in CDATA/RCDATA" insertion mode is also special; it is
+     ## combined with the original insertion mode.  In thie parser,
+     ## they are stored together in the bit-or'ed form.
  ## NOTE: "initial" and "before html" insertion modes have no constants.
-Line 822 
 sub IN_SELECT_IM () { SELECT_IMS | 0b01
+Line 845 
 sub IN_SELECT_IM () { SELECT_IMS | 0b01
  sub IN_SELECT_IN_TABLE_IM () { SELECT_IMS | 0b10 }
  sub IN_COLUMN_GROUP_IM () { 0b10 }
- ## Implementations MUST act as if state machine in the spec
- sub _initialize_tokenizer ($) {
-   my $self = shift;
-   $self->{state} = DATA_STATE; # MUST
-   $self->{content_model} = PCDATA_CONTENT_MODEL; # be
-   undef $self->{current_token}; # start tag, end tag, comment, or DOCTYPE
-   undef $self->{current_attribute};
-   undef $self->{last_emitted_start_tag_name};
-   undef $self->{last_attribute_value_state};
-   delete $self->{self_closing};
-   $self->{char} = [];
-   # $self->{next_char}
-   !!!next-input-character;
-   $self->{token} = [];
-   # $self->{escape}
- } # _initialize_tokenizer
- ## A token has:
- ##   ->{type} == DOCTYPE_TOKEN, START_TAG_TOKEN, END_TAG_TOKEN, COMMENT_TOKEN,
- ##       CHARACTER_TOKEN, or END_OF_FILE_TOKEN
- ##   ->{name} (DOCTYPE_TOKEN)
- ##   ->{tag_name} (START_TAG_TOKEN, END_TAG_TOKEN)
- ##   ->{public_identifier} (DOCTYPE_TOKEN)
- ##   ->{system_identifier} (DOCTYPE_TOKEN)
- ##   ->{quirks} == 1 or 0 (DOCTYPE_TOKEN): "force-quirks" flag
- ##   ->{attributes} isa HASH (START_TAG_TOKEN, END_TAG_TOKEN)
- ##        ->{name}
- ##        ->{value}
- ##        ->{has_reference} == 1 or 0
- ##   ->{data} (COMMENT_TOKEN, CHARACTER_TOKEN)
- ## NOTE: The "self-closing flag" is hold as |$self->{self_closing}|.
- ##     |->{self_closing}| is used to save the value of |$self->{self_closing}|
- ##     while the token is pushed back to the stack.
- ## Emitted token MUST immediately be handled by the tree construction state.
- ## Before each step, UA MAY check to see if either one of the scripts in
- ## "list of scripts that will execute as soon as possible" or the first
- ## script in the "list of scripts that will execute asynchronously",
- ## has completed loading.  If one has, then it MUST be executed
- ## and removed from the list.
- ## NOTE: HTML5 "Writing HTML documents" section, applied to
- ## documents and not to user agents and conformance checkers,
- ## contains some requirements that are not detected by the
- ## parsing algorithm:
- ## - Some requirements on character encoding declarations. ## TODO
- ## - "Elements MUST NOT contain content that their content model disallows."
- ##   ... Some are parse error, some are not (will be reported by c.c.).
- ## - Polytheistic slash SHOULD NOT be used. (Applied only to atheists.) ## TODO
- ## - Text (in elements, attributes, and comments) SHOULD NOT contain
- ##   control characters other than space characters. ## TODO: (what is control character? C0, C1 and DEL?  Unicode control character?)
- ## TODO: HTML5 poses authors two SHOULD-level requirements that cannot
- ## be detected by the HTML5 parsing algorithm:
- ## - Text,
- sub _get_next_token ($) {
-   my $self = shift;
-   if ($self->{self_closing}) {
-     !!!parse-error (type => 'nestc', token => $self->{current_token});
-     ## NOTE: The |self_closing| flag is only set by start tag token.
-     ## In addition, when a start tag token is emitted, it is always set to
-     ## |current_token|.
-     delete $self->{self_closing};
-   }
-   if (@{$self->{token}}) {
-     $self->{self_closing} = $self->{token}->[0]->{self_closing};
-     return shift @{$self->{token}};
-   }
-   A: {
-     if ($self->{state} == DATA_STATE) {
-       if ($self->{next_char} == 0x0026) { # &
-         if ($self->{content_model} & CM_ENTITY and # PCDATA | RCDATA
-             not $self->{escape}) {
-           !!!cp (1);
-           $self->{state} = ENTITY_DATA_STATE;
-           !!!next-input-character;
-           redo A;
-         } else {
-           !!!cp (2);
-           #
-         }
-       } elsif ($self->{next_char} == 0x002D) { # -
-         if ($self->{content_model} & CM_LIMITED_MARKUP) { # RCDATA | CDATA
-           unless ($self->{escape}) {
-             if ($self->{prev_char}->[0] == 0x002D and # -
-                 $self->{prev_char}->[1] == 0x0021 and # !
-                 $self->{prev_char}->[2] == 0x003C) { # <
-               !!!cp (3);
-               $self->{escape} = 1;
-             } else {
-               !!!cp (4);
-             }
-           } else {
-             !!!cp (5);
-           }
-         }
-         #
-       } elsif ($self->{next_char} == 0x003C) { # <
-         if ($self->{content_model} & CM_FULL_MARKUP or # PCDATA
-             (($self->{content_model} & CM_LIMITED_MARKUP) and # CDATA | RCDATA
-              not $self->{escape})) {
-           !!!cp (6);
-           $self->{state} = TAG_OPEN_STATE;
-           !!!next-input-character;
-           redo A;
-         } else {
-           !!!cp (7);
-           #
-         }
-       } elsif ($self->{next_char} == 0x003E) { # >
-         if ($self->{escape} and
-             ($self->{content_model} & CM_LIMITED_MARKUP)) { # RCDATA | CDATA
-           if ($self->{prev_char}->[0] == 0x002D and # -
-               $self->{prev_char}->[1] == 0x002D) { # -
-             !!!cp (8);
-             delete $self->{escape};
-           } else {
-             !!!cp (9);
-           }
-         } else {
-           !!!cp (10);
-         }
-         #
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (11);
-         !!!emit ({type => END_OF_FILE_TOKEN,
-                   line => $self->{line}, column => $self->{column}});
-         last A; ## TODO: ok?
-       } else {
-         !!!cp (12);
-       }
-       # Anything else
-       my $token = {type => CHARACTER_TOKEN,
-                    data => chr $self->{next_char},
-                    line => $self->{line}, column => $self->{column},
-                   };
-       ## Stay in the data state
-       !!!next-input-character;
-       !!!emit ($token);
-       redo A;
-     } elsif ($self->{state} == ENTITY_DATA_STATE) {
-       ## (cannot happen in CDATA state)
-       my ($l, $c) = ($self->{line_prev}, $self->{column_prev});
-       my $token = $self->_tokenize_attempt_to_consume_an_entity (0, -1);
-       $self->{state} = DATA_STATE;
-       # next-input-character is already done
-       unless (defined $token) {
-         !!!cp (13);
-         !!!emit ({type => CHARACTER_TOKEN, data => '&',
-                   line => $l, column => $c,
-                  });
-       } else {
-         !!!cp (14);
-         !!!emit ($token);
-       }
-       redo A;
-     } elsif ($self->{state} == TAG_OPEN_STATE) {
-       if ($self->{content_model} & CM_LIMITED_MARKUP) { # RCDATA | CDATA
-         if ($self->{next_char} == 0x002F) { # /
-           !!!cp (15);
-           !!!next-input-character;
-           $self->{state} = CLOSE_TAG_OPEN_STATE;
-           redo A;
-         } else {
-           !!!cp (16);
-           ## reconsume
-           $self->{state} = DATA_STATE;
-           !!!emit ({type => CHARACTER_TOKEN, data => '<',
-                     line => $self->{line_prev},
-                     column => $self->{column_prev},
-                    });
-           redo A;
-         }
-       } elsif ($self->{content_model} & CM_FULL_MARKUP) { # PCDATA
-         if ($self->{next_char} == 0x0021) { # !
-           !!!cp (17);
-           $self->{state} = MARKUP_DECLARATION_OPEN_STATE;
-           !!!next-input-character;
-           redo A;
-         } elsif ($self->{next_char} == 0x002F) { # /
-           !!!cp (18);
-           $self->{state} = CLOSE_TAG_OPEN_STATE;
-           !!!next-input-character;
-           redo A;
-         } elsif (0x0041 <= $self->{next_char} and
-                  $self->{next_char} <= 0x005A) { # A..Z
-           !!!cp (19);
-           $self->{current_token}
-             = {type => START_TAG_TOKEN,
-                tag_name => chr ($self->{next_char} + 0x0020),
-                line => $self->{line_prev},
-                column => $self->{column_prev}};
-           $self->{state} = TAG_NAME_STATE;
-           !!!next-input-character;
-           redo A;
-         } elsif (0x0061 <= $self->{next_char} and
-                  $self->{next_char} <= 0x007A) { # a..z
-           !!!cp (20);
-           $self->{current_token} = {type => START_TAG_TOKEN,
-                                     tag_name => chr ($self->{next_char}),
-                                     line => $self->{line_prev},
-                                     column => $self->{column_prev}};
-           $self->{state} = TAG_NAME_STATE;
-           !!!next-input-character;
-           redo A;
-         } elsif ($self->{next_char} == 0x003E) { # >
-           !!!cp (21);
-           !!!parse-error (type => 'empty start tag',
-                           line => $self->{line_prev},
-                           column => $self->{column_prev});
-           $self->{state} = DATA_STATE;
-           !!!next-input-character;
-           !!!emit ({type => CHARACTER_TOKEN, data => '<>',
-                     line => $self->{line_prev},
-                     column => $self->{column_prev},
-                    });
-           redo A;
-         } elsif ($self->{next_char} == 0x003F) { # ?
-           !!!cp (22);
-           !!!parse-error (type => 'pio',
-                           line => $self->{line_prev},
-                           column => $self->{column_prev});
-           $self->{state} = BOGUS_COMMENT_STATE;
-           $self->{current_token} = {type => COMMENT_TOKEN, data => '',
-                                     line => $self->{line_prev},
-                                     column => $self->{column_prev},
-                                    };
-           ## $self->{next_char} is intentionally left as is
-           redo A;
-         } else {
-           !!!cp (23);
-           !!!parse-error (type => 'bare stago',
-                           line => $self->{line_prev},
-                           column => $self->{column_prev});
-           $self->{state} = DATA_STATE;
-           ## reconsume
-           !!!emit ({type => CHARACTER_TOKEN, data => '<',
-                     line => $self->{line_prev},
-                     column => $self->{column_prev},
-                    });
-           redo A;
-         }
-       } else {
-         die "$0: $self->{content_model} in tag open";
-       }
-     } elsif ($self->{state} == CLOSE_TAG_OPEN_STATE) {
-       my ($l, $c) = ($self->{line_prev}, $self->{column_prev} - 1); # "<"of"</"
-       if ($self->{content_model} & CM_LIMITED_MARKUP) { # RCDATA | CDATA
-         if (defined $self->{last_emitted_start_tag_name}) {
-           ## NOTE: <http://krijnhoetmer.nl/irc-logs/whatwg/20070626#l-564>
-           my @next_char;
-           TAGNAME: for (my $i = 0; $i < length $self->{last_emitted_start_tag_name}; $i++) {
-             push @next_char, $self->{next_char};
-             my $c = ord substr ($self->{last_emitted_start_tag_name}, $i, 1);
-             my $C = 0x0061 <= $c && $c <= 0x007A ? $c - 0x0020 : $c;
-             if ($self->{next_char} == $c or $self->{next_char} == $C) {
-               !!!cp (24);
-               !!!next-input-character;
-               next TAGNAME;
-             } else {
-               !!!cp (25);
-               $self->{next_char} = shift @next_char; # reconsume
-               !!!back-next-input-character (@next_char);
-               $self->{state} = DATA_STATE;
-               !!!emit ({type => CHARACTER_TOKEN, data => '</',
-                         line => $l, column => $c,
-                        });
-               redo A;
-             }
-           }
-           push @next_char, $self->{next_char};
-           unless ($self->{next_char} == 0x0009 or # HT
-                   $self->{next_char} == 0x000A or # LF
-                   $self->{next_char} == 0x000B or # VT
-                   $self->{next_char} == 0x000C or # FF
-                   $self->{next_char} == 0x0020 or # SP
-                   $self->{next_char} == 0x003E or # >
-                   $self->{next_char} == 0x002F or # /
-                   $self->{next_char} == -1) {
-             !!!cp (26);
-             $self->{next_char} = shift @next_char; # reconsume
-             !!!back-next-input-character (@next_char);
-             $self->{state} = DATA_STATE;
-             !!!emit ({type => CHARACTER_TOKEN, data => '</',
-                       line => $l, column => $c,
-                      });
-             redo A;
-           } else {
-             !!!cp (27);
-             $self->{next_char} = shift @next_char;
-             !!!back-next-input-character (@next_char);
-             # and consume...
-           }
-         } else {
-           ## No start tag token has ever been emitted
-           !!!cp (28);
-           # next-input-character is already done
-           $self->{state} = DATA_STATE;
-           !!!emit ({type => CHARACTER_TOKEN, data => '</',
-                     line => $l, column => $c,
-                    });
-           redo A;
-         }
-       }
-       if (0x0041 <= $self->{next_char} and
-           $self->{next_char} <= 0x005A) { # A..Z
-         !!!cp (29);
-         $self->{current_token}
-             = {type => END_TAG_TOKEN,
-                tag_name => chr ($self->{next_char} + 0x0020),
-                line => $l, column => $c};
-         $self->{state} = TAG_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif (0x0061 <= $self->{next_char} and
-                $self->{next_char} <= 0x007A) { # a..z
-         !!!cp (30);
-         $self->{current_token} = {type => END_TAG_TOKEN,
-                                   tag_name => chr ($self->{next_char}),
-                                   line => $l, column => $c};
-         $self->{state} = TAG_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (31);
-         !!!parse-error (type => 'empty end tag',
-                         line => $self->{line_prev}, ## "<" in "</>"
-                         column => $self->{column_prev} - 1);
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (32);
-         !!!parse-error (type => 'bare etago');
-         $self->{state} = DATA_STATE;
-         # reconsume
-         !!!emit ({type => CHARACTER_TOKEN, data => '</',
-                   line => $l, column => $c,
-                  });
-         redo A;
-       } else {
-         !!!cp (33);
-         !!!parse-error (type => 'bogus end tag');
-         $self->{state} = BOGUS_COMMENT_STATE;
-         $self->{current_token} = {type => COMMENT_TOKEN, data => '',
-                                   line => $self->{line_prev}, # "<" of "</"
-                                   column => $self->{column_prev} - 1,
-                                  };
-         ## $self->{next_char} is intentionally left as is
-         redo A;
-       }
-     } elsif ($self->{state} == TAG_NAME_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (34);
-         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (35);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           #if ($self->{current_token}->{attributes}) {
-           #  ## NOTE: This should never be reached.
-           #  !!! cp (36);
-           #  !!! parse-error (type => 'end tag attribute');
-           #} else {
-             !!!cp (37);
-           #}
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif (0x0041 <= $self->{next_char} and
-                $self->{next_char} <= 0x005A) { # A..Z
-         !!!cp (38);
-         $self->{current_token}->{tag_name} .= chr ($self->{next_char} + 0x0020);
-           # start tag or end tag
-         ## Stay in this state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (39);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           #if ($self->{current_token}->{attributes}) {
-           #  ## NOTE: This state should never be reached.
-           #  !!! cp (40);
-           #  !!! parse-error (type => 'end tag attribute');
-           #} else {
-             !!!cp (41);
-           #}
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         # reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif ($self->{next_char} == 0x002F) { # /
-         !!!cp (42);
-         $self->{state} = SELF_CLOSING_START_TAG_STATE;
-         !!!next-input-character;
-         redo A;
-       } else {
-         !!!cp (44);
-         $self->{current_token}->{tag_name} .= chr $self->{next_char};
-           # start tag or end tag
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == BEFORE_ATTRIBUTE_NAME_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (45);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (46);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (47);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             !!!cp (48);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif (0x0041 <= $self->{next_char} and
-                $self->{next_char} <= 0x005A) { # A..Z
-         !!!cp (49);
-         $self->{current_attribute}
-             = {name => chr ($self->{next_char} + 0x0020),
-                value => '',
-                line => $self->{line}, column => $self->{column}};
-         $self->{state} = ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x002F) { # /
-         !!!cp (50);
-         $self->{state} = SELF_CLOSING_START_TAG_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (52);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (53);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             !!!cp (54);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         # reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         if ({
-x0022 => 1, # "
-x0027 => 1, # '
-x003D => 1, # =
-             }->{$self->{next_char}}) {
-           !!!cp (55);
-           !!!parse-error (type => 'bad attribute name');
-         } else {
-           !!!cp (56);
-         }
-         $self->{current_attribute}
-             = {name => chr ($self->{next_char}),
-                value => '',
-                line => $self->{line}, column => $self->{column}};
-         $self->{state} = ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == ATTRIBUTE_NAME_STATE) {
-       my $before_leave = sub {
-         if (exists $self->{current_token}->{attributes} # start tag or end tag
-             ->{$self->{current_attribute}->{name}}) { # MUST
-           !!!cp (57);
-           !!!parse-error (type => 'duplicate attribute', text => $self->{current_attribute}->{name}, line => $self->{current_attribute}->{line}, column => $self->{current_attribute}->{column});
-           ## Discard $self->{current_attribute} # MUST
-         } else {
-           !!!cp (58);
-           $self->{current_token}->{attributes}->{$self->{current_attribute}->{name}}
-             = $self->{current_attribute};
-         }
-       }; # $before_leave
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (59);
-         $before_leave->();
-         $self->{state} = AFTER_ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003D) { # =
-         !!!cp (60);
-         $before_leave->();
-         $self->{state} = BEFORE_ATTRIBUTE_VALUE_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         $before_leave->();
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (61);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           !!!cp (62);
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!parse-error (type => 'end tag attribute');
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif (0x0041 <= $self->{next_char} and
-                $self->{next_char} <= 0x005A) { # A..Z
-         !!!cp (63);
-         $self->{current_attribute}->{name} .= chr ($self->{next_char} + 0x0020);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x002F) { # /
-         !!!cp (64);
-         $before_leave->();
-         $self->{state} = SELF_CLOSING_START_TAG_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         $before_leave->();
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (66);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (67);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (68);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         # reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         if ($self->{next_char} == 0x0022 or # "
-             $self->{next_char} == 0x0027) { # '
-           !!!cp (69);
-           !!!parse-error (type => 'bad attribute name');
-         } else {
-           !!!cp (70);
-         }
-         $self->{current_attribute}->{name} .= chr ($self->{next_char});
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == AFTER_ATTRIBUTE_NAME_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (71);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003D) { # =
-         !!!cp (72);
-         $self->{state} = BEFORE_ATTRIBUTE_VALUE_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (73);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (74);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (75);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif (0x0041 <= $self->{next_char} and
-                $self->{next_char} <= 0x005A) { # A..Z
-         !!!cp (76);
-         $self->{current_attribute}
-             = {name => chr ($self->{next_char} + 0x0020),
-                value => '',
-                line => $self->{line}, column => $self->{column}};
-         $self->{state} = ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x002F) { # /
-         !!!cp (77);
-         $self->{state} = SELF_CLOSING_START_TAG_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (79);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (80);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (81);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         # reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         if ($self->{next_char} == 0x0022 or # "
-             $self->{next_char} == 0x0027) { # '
-           !!!cp (78);
-           !!!parse-error (type => 'bad attribute name');
-         } else {
-           !!!cp (82);
-         }
-         $self->{current_attribute}
-             = {name => chr ($self->{next_char}),
-                value => '',
-                line => $self->{line}, column => $self->{column}};
-         $self->{state} = ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == BEFORE_ATTRIBUTE_VALUE_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (83);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0022) { # "
-         !!!cp (84);
-         $self->{state} = ATTRIBUTE_VALUE_DOUBLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0026) { # &
-         !!!cp (85);
-         $self->{state} = ATTRIBUTE_VALUE_UNQUOTED_STATE;
-         ## reconsume
-         redo A;
-       } elsif ($self->{next_char} == 0x0027) { # '
-         !!!cp (86);
-         $self->{state} = ATTRIBUTE_VALUE_SINGLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!parse-error (type => 'empty unquoted attribute value');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (87);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (88);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (89);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (90);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (91);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (92);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         if ($self->{next_char} == 0x003D) { # =
-           !!!cp (93);
-           !!!parse-error (type => 'bad attribute value');
-         } else {
-           !!!cp (94);
-         }
-         $self->{current_attribute}->{value} .= chr ($self->{next_char});
-         $self->{state} = ATTRIBUTE_VALUE_UNQUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == ATTRIBUTE_VALUE_DOUBLE_QUOTED_STATE) {
-       if ($self->{next_char} == 0x0022) { # "
-         !!!cp (95);
-         $self->{state} = AFTER_ATTRIBUTE_VALUE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0026) { # &
-         !!!cp (96);
-         $self->{last_attribute_value_state} = $self->{state};
-         $self->{state} = ENTITY_IN_ATTRIBUTE_VALUE_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed attribute value');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (97);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (98);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (99);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         !!!cp (100);
-         $self->{current_attribute}->{value} .= chr ($self->{next_char});
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == ATTRIBUTE_VALUE_SINGLE_QUOTED_STATE) {
-       if ($self->{next_char} == 0x0027) { # '
-         !!!cp (101);
-         $self->{state} = AFTER_ATTRIBUTE_VALUE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0026) { # &
-         !!!cp (102);
-         $self->{last_attribute_value_state} = $self->{state};
-         $self->{state} = ENTITY_IN_ATTRIBUTE_VALUE_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed attribute value');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (103);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (104);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (105);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         !!!cp (106);
-         $self->{current_attribute}->{value} .= chr ($self->{next_char});
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == ATTRIBUTE_VALUE_UNQUOTED_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # HT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (107);
-         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0026) { # &
-         !!!cp (108);
-         $self->{last_attribute_value_state} = $self->{state};
-         $self->{state} = ENTITY_IN_ATTRIBUTE_VALUE_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (109);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (110);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (111);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (112);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (113);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (114);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         if ({
-x0022 => 1, # "
-x0027 => 1, # '
-x003D => 1, # =
-             }->{$self->{next_char}}) {
-           !!!cp (115);
-           !!!parse-error (type => 'bad attribute value');
-         } else {
-           !!!cp (116);
-         }
-         $self->{current_attribute}->{value} .= chr ($self->{next_char});
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == ENTITY_IN_ATTRIBUTE_VALUE_STATE) {
-       my $token = $self->_tokenize_attempt_to_consume_an_entity
-           (1,
-            $self->{last_attribute_value_state}
-              == ATTRIBUTE_VALUE_DOUBLE_QUOTED_STATE ? 0x0022 : # "
-            $self->{last_attribute_value_state}
-              == ATTRIBUTE_VALUE_SINGLE_QUOTED_STATE ? 0x0027 : # '
-            -1);
-       unless (defined $token) {
-         !!!cp (117);
-         $self->{current_attribute}->{value} .= '&';
-       } else {
-         !!!cp (118);
-         $self->{current_attribute}->{value} .= $token->{data};
-         $self->{current_attribute}->{has_reference} = $token->{has_reference};
-         ## ISSUE: spec says "append the returned character token to the current attribute's value"
-       }
-       $self->{state} = $self->{last_attribute_value_state};
-       # next-input-character is already done
-       redo A;
-     } elsif ($self->{state} == AFTER_ATTRIBUTE_VALUE_QUOTED_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (118);
-         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (119);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (120);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (121);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif ($self->{next_char} == 0x002F) { # /
-         !!!cp (122);
-         $self->{state} = SELF_CLOSING_START_TAG_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (122.3);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (122.1);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (122.2);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         ## Reconsume.
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         !!!cp ('124.1');
-         !!!parse-error (type => 'no space between attributes');
-         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
-         ## reconsume
-         redo A;
-       }
-     } elsif ($self->{state} == SELF_CLOSING_START_TAG_STATE) {
-       if ($self->{next_char} == 0x003E) { # >
-         if ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           !!!cp ('124.2');
-           !!!parse-error (type => 'nestc', token => $self->{current_token});
-           ## TODO: Different type than slash in start tag
-           $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
-           if ($self->{current_token}->{attributes}) {
-             !!!cp ('124.4');
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             !!!cp ('124.5');
-           }
-           ## TODO: Test |<title></title/>|
-         } else {
-           !!!cp ('124.3');
-           $self->{self_closing} = 1;
-         }
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
-           !!!cp (124.7);
-           $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
-           if ($self->{current_token}->{attributes}) {
-             !!!cp (124.5);
-             !!!parse-error (type => 'end tag attribute');
-           } else {
-             ## NOTE: This state should never be reached.
-             !!!cp (124.6);
-           }
-         } else {
-           die "$0: $self->{current_token}->{type}: Unknown token type";
-         }
-         $self->{state} = DATA_STATE;
-         ## Reconsume.
-         !!!emit ($self->{current_token}); # start tag or end tag
-         redo A;
-       } else {
-         !!!cp ('124.4');
-         !!!parse-error (type => 'nestc');
-         ## TODO: This error type is wrong.
-         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
-         ## Reconsume.
-         redo A;
-       }
-     } elsif ($self->{state} == BOGUS_COMMENT_STATE) {
-       ## (only happen if PCDATA state)
-       ## NOTE: Set by the previous state
-       #my $token = {type => COMMENT_TOKEN, data => ''};
-       BC: {
-         if ($self->{next_char} == 0x003E) { # >
-           !!!cp (124);
-           $self->{state} = DATA_STATE;
-           !!!next-input-character;
-           !!!emit ($self->{current_token}); # comment
-           redo A;
-         } elsif ($self->{next_char} == -1) {
-           !!!cp (125);
-           $self->{state} = DATA_STATE;
-           ## reconsume
-           !!!emit ($self->{current_token}); # comment
-           redo A;
-         } else {
-           !!!cp (126);
-           $self->{current_token}->{data} .= chr ($self->{next_char}); # comment
-           !!!next-input-character;
-           redo BC;
-         }
-       } # BC
-       die "$0: _get_next_token: unexpected case [BC]";
-     } elsif ($self->{state} == MARKUP_DECLARATION_OPEN_STATE) {
-       ## (only happen if PCDATA state)
-       my ($l, $c) = ($self->{line_prev}, $self->{column_prev} - 1);
-       my @next_char;
-       push @next_char, $self->{next_char};
-       if ($self->{next_char} == 0x002D) { # -
-         !!!next-input-character;
-         push @next_char, $self->{next_char};
-         if ($self->{next_char} == 0x002D) { # -
-           !!!cp (127);
-           $self->{current_token} = {type => COMMENT_TOKEN, data => '',
-                                     line => $l, column => $c,
-                                    };
-           $self->{state} = COMMENT_START_STATE;
-           !!!next-input-character;
-           redo A;
-         } else {
-           !!!cp (128);
-         }
-       } elsif ($self->{next_char} == 0x0044 or # D
-                $self->{next_char} == 0x0064) { # d
-         !!!next-input-character;
-         push @next_char, $self->{next_char};
-         if ($self->{next_char} == 0x004F or # O
-             $self->{next_char} == 0x006F) { # o
-           !!!next-input-character;
-           push @next_char, $self->{next_char};
-           if ($self->{next_char} == 0x0043 or # C
-               $self->{next_char} == 0x0063) { # c
-             !!!next-input-character;
-             push @next_char, $self->{next_char};
-             if ($self->{next_char} == 0x0054 or # T
-                 $self->{next_char} == 0x0074) { # t
-               !!!next-input-character;
-               push @next_char, $self->{next_char};
-               if ($self->{next_char} == 0x0059 or # Y
-                   $self->{next_char} == 0x0079) { # y
-                 !!!next-input-character;
-                 push @next_char, $self->{next_char};
-                 if ($self->{next_char} == 0x0050 or # P
-                     $self->{next_char} == 0x0070) { # p
-                   !!!next-input-character;
-                   push @next_char, $self->{next_char};
-                   if ($self->{next_char} == 0x0045 or # E
-                       $self->{next_char} == 0x0065) { # e
-                     !!!cp (129);
-                     ## TODO: What a stupid code this is!
-                     $self->{state} = DOCTYPE_STATE;
-                     $self->{current_token} = {type => DOCTYPE_TOKEN,
-                                               quirks => 1,
-                                               line => $l, column => $c,
-                                              };
-                     !!!next-input-character;
-                     redo A;
-                   } else {
-                     !!!cp (130);
-                   }
-                 } else {
-                   !!!cp (131);
-                 }
-               } else {
-                 !!!cp (132);
-               }
-             } else {
-               !!!cp (133);
-             }
-           } else {
-             !!!cp (134);
-           }
-         } else {
-           !!!cp (135);
-         }
-       } elsif ($self->{insertion_mode} & IN_FOREIGN_CONTENT_IM and
-                $self->{open_elements}->[-1]->[1] & FOREIGN_EL and
-                $self->{next_char} == 0x005B) { # [
-         !!!next-input-character;
-         push @next_char, $self->{next_char};
-         if ($self->{next_char} == 0x0043) { # C
-           !!!next-input-character;
-           push @next_char, $self->{next_char};
-           if ($self->{next_char} == 0x0044) { # D
-             !!!next-input-character;
-             push @next_char, $self->{next_char};
-             if ($self->{next_char} == 0x0041) { # A
-               !!!next-input-character;
-               push @next_char, $self->{next_char};
-               if ($self->{next_char} == 0x0054) { # T
-                 !!!next-input-character;
-                 push @next_char, $self->{next_char};
-                 if ($self->{next_char} == 0x0041) { # A
-                   !!!next-input-character;
-                   push @next_char, $self->{next_char};
-                   if ($self->{next_char} == 0x005B) { # [
-                     !!!cp (135.1);
-                     $self->{state} = CDATA_BLOCK_STATE;
-                     !!!next-input-character;
-                     redo A;
-                   } else {
-                     !!!cp (135.2);
-                   }
-                 } else {
-                   !!!cp (135.3);
-                 }
-               } else {
-                 !!!cp (135.4);
-               }
-             } else {
-               !!!cp (135.5);
-             }
-           } else {
-             !!!cp (135.6);
-           }
-         } else {
-           !!!cp (135.7);
-         }
-       } else {
-         !!!cp (136);
-       }
-       !!!parse-error (type => 'bogus comment');
-       $self->{next_char} = shift @next_char;
-       !!!back-next-input-character (@next_char);
-       $self->{state} = BOGUS_COMMENT_STATE;
-       $self->{current_token} = {type => COMMENT_TOKEN, data => '',
-                                 line => $l, column => $c,
-                                };
-       redo A;
-       ## ISSUE: typos in spec: chacacters, is is a parse error
-       ## ISSUE: spec is somewhat unclear on "is the first character that will be in the comment"; what is "that will be in the comment" is what the algorithm defines, isn't it?
-     } elsif ($self->{state} == COMMENT_START_STATE) {
-       if ($self->{next_char} == 0x002D) { # -
-         !!!cp (137);
-         $self->{state} = COMMENT_START_DASH_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (138);
-         !!!parse-error (type => 'bogus comment');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (139);
-         !!!parse-error (type => 'unclosed comment');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } else {
-         !!!cp (140);
-         $self->{current_token}->{data} # comment
-             .= chr ($self->{next_char});
-         $self->{state} = COMMENT_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == COMMENT_START_DASH_STATE) {
-       if ($self->{next_char} == 0x002D) { # -
-         !!!cp (141);
-         $self->{state} = COMMENT_END_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (142);
-         !!!parse-error (type => 'bogus comment');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (143);
-         !!!parse-error (type => 'unclosed comment');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } else {
-         !!!cp (144);
-         $self->{current_token}->{data} # comment
-             .= '-' . chr ($self->{next_char});
-         $self->{state} = COMMENT_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == COMMENT_STATE) {
-       if ($self->{next_char} == 0x002D) { # -
-         !!!cp (145);
-         $self->{state} = COMMENT_END_DASH_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (146);
-         !!!parse-error (type => 'unclosed comment');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } else {
-         !!!cp (147);
-         $self->{current_token}->{data} .= chr ($self->{next_char}); # comment
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == COMMENT_END_DASH_STATE) {
-       if ($self->{next_char} == 0x002D) { # -
-         !!!cp (148);
-         $self->{state} = COMMENT_END_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (149);
-         !!!parse-error (type => 'unclosed comment');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } else {
-         !!!cp (150);
-         $self->{current_token}->{data} .= '-' . chr ($self->{next_char}); # comment
-         $self->{state} = COMMENT_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == COMMENT_END_STATE) {
-       if ($self->{next_char} == 0x003E) { # >
-         !!!cp (151);
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } elsif ($self->{next_char} == 0x002D) { # -
-         !!!cp (152);
-         !!!parse-error (type => 'dash in comment',
-                         line => $self->{line_prev},
-                         column => $self->{column_prev});
-         $self->{current_token}->{data} .= '-'; # comment
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (153);
-         !!!parse-error (type => 'unclosed comment');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # comment
-         redo A;
-       } else {
-         !!!cp (154);
-         !!!parse-error (type => 'dash in comment',
-                         line => $self->{line_prev},
-                         column => $self->{column_prev});
-         $self->{current_token}->{data} .= '--' . chr ($self->{next_char}); # comment
-         $self->{state} = COMMENT_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == DOCTYPE_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (155);
-         $self->{state} = BEFORE_DOCTYPE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } else {
-         !!!cp (156);
-         !!!parse-error (type => 'no space before DOCTYPE name');
-         $self->{state} = BEFORE_DOCTYPE_NAME_STATE;
-         ## reconsume
-         redo A;
-       }
-     } elsif ($self->{state} == BEFORE_DOCTYPE_NAME_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (157);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (158);
-         !!!parse-error (type => 'no DOCTYPE name');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # DOCTYPE (quirks)
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (159);
-         !!!parse-error (type => 'no DOCTYPE name');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # DOCTYPE (quirks)
-         redo A;
-       } else {
-         !!!cp (160);
-         $self->{current_token}->{name} = chr $self->{next_char};
-         delete $self->{current_token}->{quirks};
- ## ISSUE: "Set the token's name name to the" in the spec
-         $self->{state} = DOCTYPE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == DOCTYPE_NAME_STATE) {
- ## ISSUE: Redundant "First," in the spec.
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (161);
-         $self->{state} = AFTER_DOCTYPE_NAME_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (162);
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (163);
-         !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (164);
-         $self->{current_token}->{name}
-           .= chr ($self->{next_char}); # DOCTYPE
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == AFTER_DOCTYPE_NAME_STATE) {
-       if ($self->{next_char} == 0x0009 or # HT
-           $self->{next_char} == 0x000A or # LF
-           $self->{next_char} == 0x000B or # VT
-           $self->{next_char} == 0x000C or # FF
-           $self->{next_char} == 0x0020) { # SP
-         !!!cp (165);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (166);
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (167);
-         !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == 0x0050 or # P
-                $self->{next_char} == 0x0070) { # p
-         !!!next-input-character;
-         if ($self->{next_char} == 0x0055 or # U
-             $self->{next_char} == 0x0075) { # u
-           !!!next-input-character;
-           if ($self->{next_char} == 0x0042 or # B
-               $self->{next_char} == 0x0062) { # b
-             !!!next-input-character;
-             if ($self->{next_char} == 0x004C or # L
-                 $self->{next_char} == 0x006C) { # l
-               !!!next-input-character;
-               if ($self->{next_char} == 0x0049 or # I
-                   $self->{next_char} == 0x0069) { # i
-                 !!!next-input-character;
-                 if ($self->{next_char} == 0x0043 or # C
-                     $self->{next_char} == 0x0063) { # c
-                   !!!cp (168);
-                   $self->{state} = BEFORE_DOCTYPE_PUBLIC_IDENTIFIER_STATE;
-                   !!!next-input-character;
-                   redo A;
-                 } else {
-                   !!!cp (169);
-                 }
-               } else {
-                 !!!cp (170);
-               }
-             } else {
-               !!!cp (171);
-             }
-           } else {
-             !!!cp (172);
-           }
-         } else {
-           !!!cp (173);
-         }
-         #
-       } elsif ($self->{next_char} == 0x0053 or # S
-                $self->{next_char} == 0x0073) { # s
-         !!!next-input-character;
-         if ($self->{next_char} == 0x0059 or # Y
-             $self->{next_char} == 0x0079) { # y
-           !!!next-input-character;
-           if ($self->{next_char} == 0x0053 or # S
-               $self->{next_char} == 0x0073) { # s
-             !!!next-input-character;
-             if ($self->{next_char} == 0x0054 or # T
-                 $self->{next_char} == 0x0074) { # t
-               !!!next-input-character;
-               if ($self->{next_char} == 0x0045 or # E
-                   $self->{next_char} == 0x0065) { # e
-                 !!!next-input-character;
-                 if ($self->{next_char} == 0x004D or # M
-                     $self->{next_char} == 0x006D) { # m
-                   !!!cp (174);
-                   $self->{state} = BEFORE_DOCTYPE_SYSTEM_IDENTIFIER_STATE;
-                   !!!next-input-character;
-                   redo A;
-                 } else {
-                   !!!cp (175);
-                 }
-               } else {
-                 !!!cp (176);
-               }
-             } else {
-               !!!cp (177);
-             }
-           } else {
-             !!!cp (178);
-           }
-         } else {
-           !!!cp (179);
-         }
-         #
-       } else {
-         !!!cp (180);
-         !!!next-input-character;
-         #
-       }
-       !!!parse-error (type => 'string after DOCTYPE name');
-       $self->{current_token}->{quirks} = 1;
-       $self->{state} = BOGUS_DOCTYPE_STATE;
-       # next-input-character is already done
-       redo A;
-     } elsif ($self->{state} == BEFORE_DOCTYPE_PUBLIC_IDENTIFIER_STATE) {
-       if ({
-x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
-             #0x000D => 1, # HT, LF, VT, FF, SP, CR
-           }->{$self->{next_char}}) {
-         !!!cp (181);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} eq 0x0022) { # "
-         !!!cp (182);
-         $self->{current_token}->{public_identifier} = ''; # DOCTYPE
-         $self->{state} = DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} eq 0x0027) { # '
-         !!!cp (183);
-         $self->{current_token}->{public_identifier} = ''; # DOCTYPE
-         $self->{state} = DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} eq 0x003E) { # >
-         !!!cp (184);
-         !!!parse-error (type => 'no PUBLIC literal');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (185);
-         !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (186);
-         !!!parse-error (type => 'string after PUBLIC');
-         $self->{current_token}->{quirks} = 1;
-         $self->{state} = BOGUS_DOCTYPE_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED_STATE) {
-       if ($self->{next_char} == 0x0022) { # "
-         !!!cp (187);
-         $self->{state} = AFTER_DOCTYPE_PUBLIC_IDENTIFIER_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (188);
-         !!!parse-error (type => 'unclosed PUBLIC literal');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (189);
-         !!!parse-error (type => 'unclosed PUBLIC literal');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (190);
-         $self->{current_token}->{public_identifier} # DOCTYPE
-             .= chr $self->{next_char};
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED_STATE) {
-       if ($self->{next_char} == 0x0027) { # '
-         !!!cp (191);
-         $self->{state} = AFTER_DOCTYPE_PUBLIC_IDENTIFIER_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (192);
-         !!!parse-error (type => 'unclosed PUBLIC literal');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (193);
-         !!!parse-error (type => 'unclosed PUBLIC literal');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (194);
-         $self->{current_token}->{public_identifier} # DOCTYPE
-             .= chr $self->{next_char};
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == AFTER_DOCTYPE_PUBLIC_IDENTIFIER_STATE) {
-       if ({
-x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
-             #0x000D => 1, # HT, LF, VT, FF, SP, CR
-           }->{$self->{next_char}}) {
-         !!!cp (195);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0022) { # "
-         !!!cp (196);
-         $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0027) { # '
-         !!!cp (197);
-         $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (198);
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (199);
-         !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (200);
-         !!!parse-error (type => 'string after PUBLIC literal');
-         $self->{current_token}->{quirks} = 1;
-         $self->{state} = BOGUS_DOCTYPE_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == BEFORE_DOCTYPE_SYSTEM_IDENTIFIER_STATE) {
-       if ({
-x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
-             #0x000D => 1, # HT, LF, VT, FF, SP, CR
-           }->{$self->{next_char}}) {
-         !!!cp (201);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0022) { # "
-         !!!cp (202);
-         $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x0027) { # '
-         !!!cp (203);
-         $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (204);
-         !!!parse-error (type => 'no SYSTEM literal');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (205);
-         !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (206);
-         !!!parse-error (type => 'string after SYSTEM');
-         $self->{current_token}->{quirks} = 1;
-         $self->{state} = BOGUS_DOCTYPE_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED_STATE) {
-       if ($self->{next_char} == 0x0022) { # "
-         !!!cp (207);
-         $self->{state} = AFTER_DOCTYPE_SYSTEM_IDENTIFIER_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (208);
-         !!!parse-error (type => 'unclosed SYSTEM literal');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (209);
-         !!!parse-error (type => 'unclosed SYSTEM literal');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (210);
-         $self->{current_token}->{system_identifier} # DOCTYPE
-             .= chr $self->{next_char};
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED_STATE) {
-       if ($self->{next_char} == 0x0027) { # '
-         !!!cp (211);
-         $self->{state} = AFTER_DOCTYPE_SYSTEM_IDENTIFIER_STATE;
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (212);
-         !!!parse-error (type => 'unclosed SYSTEM literal');
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (213);
-         !!!parse-error (type => 'unclosed SYSTEM literal');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (214);
-         $self->{current_token}->{system_identifier} # DOCTYPE
-             .= chr $self->{next_char};
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == AFTER_DOCTYPE_SYSTEM_IDENTIFIER_STATE) {
-       if ({
-x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
-             #0x000D => 1, # HT, LF, VT, FF, SP, CR
-           }->{$self->{next_char}}) {
-         !!!cp (215);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       } elsif ($self->{next_char} == 0x003E) { # >
-         !!!cp (216);
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (217);
-         !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         $self->{current_token}->{quirks} = 1;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (218);
-         !!!parse-error (type => 'string after SYSTEM literal');
-         #$self->{current_token}->{quirks} = 1;
-         $self->{state} = BOGUS_DOCTYPE_STATE;
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == BOGUS_DOCTYPE_STATE) {
-       if ($self->{next_char} == 0x003E) { # >
-         !!!cp (219);
-         $self->{state} = DATA_STATE;
-         !!!next-input-character;
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } elsif ($self->{next_char} == -1) {
-         !!!cp (220);
-         !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = DATA_STATE;
-         ## reconsume
-         !!!emit ($self->{current_token}); # DOCTYPE
-         redo A;
-       } else {
-         !!!cp (221);
-         ## Stay in the state
-         !!!next-input-character;
-         redo A;
-       }
-     } elsif ($self->{state} == CDATA_BLOCK_STATE) {
-       my $s = '';
-       my ($l, $c) = ($self->{line}, $self->{column});
-       CS: while ($self->{next_char} != -1) {
-         if ($self->{next_char} == 0x005D) { # ]
-           !!!next-input-character;
-           if ($self->{next_char} == 0x005D) { # ]
-             !!!next-input-character;
-             MDC: {
-               if ($self->{next_char} == 0x003E) { # >
-                 !!!cp (221.1);
-                 !!!next-input-character;
-                 last CS;
-               } elsif ($self->{next_char} == 0x005D) { # ]
-                 !!!cp (221.2);
-                 $s .= ']';
-                 !!!next-input-character;
-                 redo MDC;
-               } else {
-                 !!!cp (221.3);
-                 $s .= ']]';
-                 #
-               }
-             } # MDC
-           } else {
-             !!!cp (221.4);
-             $s .= ']';
-             #
-           }
-         } else {
-           !!!cp (221.5);
-           #
-         }
-         $s .= chr $self->{next_char};
-         !!!next-input-character;
-       } # CS
-       $self->{state} = DATA_STATE;
-       ## next-input-character done or EOF, which is reconsumed.
-       if (length $s) {
-         !!!cp (221.6);
-         !!!emit ({type => CHARACTER_TOKEN, data => $s,
-                   line => $l, column => $c});
-       } else {
-         !!!cp (221.7);
-       }
-       redo A;
-       ## ISSUE: "text tokens" in spec.
-       ## TODO: Streaming support
-     } else {
-       die "$0: $self->{state}: Unknown state";
-     }
-   } # A
-   die "$0: _get_next_token: unexpected case";
- } # _get_next_token
- sub _tokenize_attempt_to_consume_an_entity ($$$) {
-   my ($self, $in_attr, $additional) = @_;
-   my ($l, $c) = ($self->{line_prev}, $self->{column_prev});
-   if ({
-x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, # HT, LF, VT, FF,
-x0020 => 1, 0x003C => 1, 0x0026 => 1, -1 => 1, # SP, <, & # 0x000D # CR
-        $additional => 1,
-       }->{$self->{next_char}}) {
-     !!!cp (1001);
-     ## Don't consume
-     ## No error
-     return undef;
-   } elsif ($self->{next_char} == 0x0023) { # #
-     !!!next-input-character;
-     if ($self->{next_char} == 0x0078 or # x
-         $self->{next_char} == 0x0058) { # X
-       my $code;
-       X: {
-         my $x_char = $self->{next_char};
-         !!!next-input-character;
-         if (0x0030 <= $self->{next_char} and
-             $self->{next_char} <= 0x0039) { # 0..9
-           !!!cp (1002);
-           $code ||= 0;
-           $code *= 0x10;
-           $code += $self->{next_char} - 0x0030;
-           redo X;
-         } elsif (0x0061 <= $self->{next_char} and
-                  $self->{next_char} <= 0x0066) { # a..f
-           !!!cp (1003);
-           $code ||= 0;
-           $code *= 0x10;
-           $code += $self->{next_char} - 0x0060 + 9;
-           redo X;
-         } elsif (0x0041 <= $self->{next_char} and
-                  $self->{next_char} <= 0x0046) { # A..F
-           !!!cp (1004);
-           $code ||= 0;
-           $code *= 0x10;
-           $code += $self->{next_char} - 0x0040 + 9;
-           redo X;
-         } elsif (not defined $code) { # no hexadecimal digit
-           !!!cp (1005);
-           !!!parse-error (type => 'bare hcro', line => $l, column => $c);
-           !!!back-next-input-character ($x_char, $self->{next_char});
-           $self->{next_char} = 0x0023; # #
-           return undef;
-         } elsif ($self->{next_char} == 0x003B) { # ;
-           !!!cp (1006);
-           !!!next-input-character;
-         } else {
-           !!!cp (1007);
-           !!!parse-error (type => 'no refc', line => $l, column => $c);
-         }
-         if ($code == 0 or (0xD800 <= $code and $code <= 0xDFFF)) {
-           !!!cp (1008);
-           !!!parse-error (type => 'invalid character reference',
-                           text => (sprintf 'U+%04X', $code),
-                           line => $l, column => $c);
-           $code = 0xFFFD;
-         } elsif ($code > 0x10FFFF) {
-           !!!cp (1009);
-           !!!parse-error (type => 'invalid character reference',
-                           text => (sprintf 'U-%08X', $code),
-                           line => $l, column => $c);
-           $code = 0xFFFD;
-         } elsif ($code == 0x000D) {
-           !!!cp (1010);
-           !!!parse-error (type => 'CR character reference', line => $l, column => $c);
-           $code = 0x000A;
-         } elsif (0x80 <= $code and $code <= 0x9F) {
-           !!!cp (1011);
-           !!!parse-error (type => 'C1 character reference', text => (sprintf 'U+%04X', $code), line => $l, column => $c);
-           $code = $c1_entity_char->{$code};
-         }
-         return {type => CHARACTER_TOKEN, data => chr $code,
-                 has_reference => 1,
-                 line => $l, column => $c,
-                };
-       } # X
-     } elsif (0x0030 <= $self->{next_char} and
-              $self->{next_char} <= 0x0039) { # 0..9
-       my $code = $self->{next_char} - 0x0030;
-       !!!next-input-character;
-       while (0x0030 <= $self->{next_char} and
-                 $self->{next_char} <= 0x0039) { # 0..9
-         !!!cp (1012);
-         $code *= 10;
-         $code += $self->{next_char} - 0x0030;
-         !!!next-input-character;
-       }
-       if ($self->{next_char} == 0x003B) { # ;
-         !!!cp (1013);
-         !!!next-input-character;
-       } else {
-         !!!cp (1014);
-         !!!parse-error (type => 'no refc', line => $l, column => $c);
-       }
-       if ($code == 0 or (0xD800 <= $code and $code <= 0xDFFF)) {
-         !!!cp (1015);
-         !!!parse-error (type => 'invalid character reference',
-                         text => (sprintf 'U+%04X', $code),
-                         line => $l, column => $c);
-         $code = 0xFFFD;
-       } elsif ($code > 0x10FFFF) {
-         !!!cp (1016);
-         !!!parse-error (type => 'invalid character reference',
-                         text => (sprintf 'U-%08X', $code),
-                         line => $l, column => $c);
-         $code = 0xFFFD;
-       } elsif ($code == 0x000D) {
-         !!!cp (1017);
-         !!!parse-error (type => 'CR character reference',
-                         line => $l, column => $c);
-         $code = 0x000A;
-       } elsif (0x80 <= $code and $code <= 0x9F) {
-         !!!cp (1018);
-         !!!parse-error (type => 'C1 character reference',
-                         text => (sprintf 'U+%04X', $code),
-                         line => $l, column => $c);
-         $code = $c1_entity_char->{$code};
-       }
-       return {type => CHARACTER_TOKEN, data => chr $code, has_reference => 1,
-               line => $l, column => $c,
-              };
-     } else {
-       !!!cp (1019);
-       !!!parse-error (type => 'bare nero', line => $l, column => $c);
-       !!!back-next-input-character ($self->{next_char});
-       $self->{next_char} = 0x0023; # #
-       return undef;
-     }
-   } elsif ((0x0041 <= $self->{next_char} and
-             $self->{next_char} <= 0x005A) or
-            (0x0061 <= $self->{next_char} and
-             $self->{next_char} <= 0x007A)) {
-     my $entity_name = chr $self->{next_char};
-     !!!next-input-character;
-     my $value = $entity_name;
-     my $match = 0;
-     require Whatpm::_NamedEntityList;
-     our $EntityChar;
-     while (length $entity_name < 30 and
-            ## NOTE: Some number greater than the maximum length of entity name
-            ((0x0041 <= $self->{next_char} and # a
-              $self->{next_char} <= 0x005A) or # x
-             (0x0061 <= $self->{next_char} and # a
-              $self->{next_char} <= 0x007A) or # z
-             (0x0030 <= $self->{next_char} and # 0
-              $self->{next_char} <= 0x0039) or # 9
-             $self->{next_char} == 0x003B)) { # ;
-       $entity_name .= chr $self->{next_char};
-       if (defined $EntityChar->{$entity_name}) {
-         if ($self->{next_char} == 0x003B) { # ;
-           !!!cp (1020);
-           $value = $EntityChar->{$entity_name};
-           $match = 1;
-           !!!next-input-character;
-           last;
-         } else {
-           !!!cp (1021);
-           $value = $EntityChar->{$entity_name};
-           $match = -1;
-           !!!next-input-character;
-         }
-       } else {
-         !!!cp (1022);
-         $value .= chr $self->{next_char};
-         $match *= 2;
-         !!!next-input-character;
-       }
-     }
-     if ($match > 0) {
-       !!!cp (1023);
-       return {type => CHARACTER_TOKEN, data => $value, has_reference => 1,
-               line => $l, column => $c,
-              };
-     } elsif ($match < 0) {
-       !!!parse-error (type => 'no refc', line => $l, column => $c);
-       if ($in_attr and $match < -1) {
-         !!!cp (1024);
-         return {type => CHARACTER_TOKEN, data => '&'.$entity_name,
-                 line => $l, column => $c,
-                };
-       } else {
-         !!!cp (1025);
-         return {type => CHARACTER_TOKEN, data => $value, has_reference => 1,
-                 line => $l, column => $c,
-                };
-       }
-     } else {
-       !!!cp (1026);
-       !!!parse-error (type => 'bare ero', line => $l, column => $c);
-       ## NOTE: "No characters are consumed" in the spec.
-       return {type => CHARACTER_TOKEN, data => '&'.$value,
-               line => $l, column => $c,
-              };
-     }
-   } else {
-     !!!cp (1027);
-     ## no characters are consumed
-     !!!parse-error (type => 'bare ero', line => $l, column => $c);
-     return undef;
-   }
- } # _tokenize_attempt_to_consume_an_entity
  sub _initialize_tree_constructor ($) {
    my $self = shift;
    ## NOTE: $self->{document} MUST be specified before this method is called
-Line 3119 
 sub _construct_tree ($) {
+Line 873 
 sub _construct_tree ($) {
    ## When an interactive UA render the $self->{document} available
    ## to the user, or when it begin accepting user input, are
    ## not defined.
-   ## Append a character: collect it and all subsequent consecutive
-   ## characters and insert one Text node whose data is concatenation
-   ## of all those characters. # MUST
    !!!next-token;
    undef $self->{form_element};
    undef $self->{head_element};
+   undef $self->{head_element_inserted};
    $self->{open_elements} = [];
    undef $self->{inner_html_node};
+   undef $self->{ignore_newline};
    ## NOTE: The "initial" insertion mode.
    $self->_tree_construction_initial; # MUST
-Line 3154 
 sub _tree_construction_initial ($) {
+Line 906 
 sub _tree_construction_initial ($) {
        ## language.
        my $doctype_name = $token->{name};
        $doctype_name = '' unless defined $doctype_name;
-       $doctype_name =~ tr/a-z/A-Z/;
+       $doctype_name =~ tr/a-z/A-Z/; # ASCII case-insensitive
        if (not defined $token->{name} or # <!DOCTYPE>
-           defined $token->{public_identifier} or
+           defined $token->{sysid}) {
-           defined $token->{system_identifier}) {
          !!!cp ('t1');
          !!!parse-error (type => 'not HTML5', token => $token);
        } elsif ($doctype_name ne 'HTML') {
          !!!cp ('t2');
-         ## ISSUE: ASCII case-insensitive? (in fact it does not matter)
          !!!parse-error (type => 'not HTML5', token => $token);
+       } elsif (defined $token->{pubid}) {
+         if ($token->{pubid} eq 'XSLT-compat') {
+           !!!cp ('t1.2');
+           !!!parse-error (type => 'XSLT-compat', token => $token,
+                           level => $self->{level}->{should});
+         } else {
+           !!!parse-error (type => 'not HTML5', token => $token);
+         }
        } else {
          !!!cp ('t3');
+         #
        }
        my $doctype = $self->{document}->create_document_type_definition
          ($token->{name}); ## ISSUE: If name is missing (e.g. <!DOCTYPE>)?
        ## NOTE: Default value for both |public_id| and |system_id| attributes
        ## are empty strings, so that we don't set any value in missing cases.
-       $doctype->public_id ($token->{public_identifier})
+       $doctype->public_id ($token->{pubid}) if defined $token->{pubid};
-           if defined $token->{public_identifier};
+       $doctype->system_id ($token->{sysid}) if defined $token->{sysid};
-       $doctype->system_id ($token->{system_identifier})
-           if defined $token->{system_identifier};
        ## NOTE: Other DocumentType attributes are null or empty lists.
        ## ISSUE: internalSubset = null??
        $self->{document}->append_child ($doctype);
-Line 3183 
 sub _tree_construction_initial ($) {
+Line 940 
 sub _tree_construction_initial ($) {
        if ($token->{quirks} or $doctype_name ne 'HTML') {
          !!!cp ('t4');
          $self->{document}->manakai_compat_mode ('quirks');
-       } elsif (defined $token->{public_identifier}) {
+       } elsif (defined $token->{pubid}) {
-         my $pubid = $token->{public_identifier};
+         my $pubid = $token->{pubid};
          $pubid =~ tr/a-z/A-z/;
          my $prefix = [
            "+//SILMARIL//DTD HTML PRO V0R11 19970101//",
-Line 3258 
 sub _tree_construction_initial ($) {
+Line 1015 
 sub _tree_construction_initial ($) {
            $self->{document}->manakai_compat_mode ('quirks');
          } elsif ($pubid =~ m[^-//W3C//DTD HTML 4.01 FRAMESET//] or
                   $pubid =~ m[^-//W3C//DTD HTML 4.01 TRANSITIONAL//]) {
-           if (defined $token->{system_identifier}) {
+           if (defined $token->{sysid}) {
              !!!cp ('t6');
              $self->{document}->manakai_compat_mode ('quirks');
            } else {
-Line 3275 
 sub _tree_construction_initial ($) {
+Line 1032 
 sub _tree_construction_initial ($) {
        } else {
          !!!cp ('t10');
        }
-       if (defined $token->{system_identifier}) {
+       if (defined $token->{sysid}) {
-         my $sysid = $token->{system_identifier};
+         my $sysid = $token->{sysid};
          $sysid =~ tr/A-Z/a-z/;
          if ($sysid eq "http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd") {
            ## NOTE: Ensure that |PUBLIC "(limited quirks)" "(quirks)"| is
-Line 3306 
 sub _tree_construction_initial ($) {
+Line 1063 
 sub _tree_construction_initial ($) {
        !!!ack-later;
        return;
      } elsif ($token->{type} == CHARACTER_TOKEN) {
-       if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) { # \x0D
+       if ($token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {
          ## Ignore the token
          unless (length $token->{data}) {
-Line 3363 
 sub _tree_construction_root_element ($)
+Line 1120 
 sub _tree_construction_root_element ($)
          !!!next-token;
          redo B;
        } elsif ($token->{type} == CHARACTER_TOKEN) {
-         if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) { # \x0D
+         if ($token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {
            ## Ignore the token.
            unless (length $token->{data}) {
-Line 3430 
 sub _tree_construction_root_element ($)
+Line 1187 
 sub _tree_construction_root_element ($)
      ## NOTE: Reprocess the token.
      !!!ack-later;
      return; ## Go to the "before head" insertion mode.
-     ## ISSUE: There is an issue in the spec
    } # B
    die "$0: _tree_construction_root_element: This should never be reached";
-Line 3467 
 sub _reset_insertion_mode ($) {
+Line 1222 
 sub _reset_insertion_mode ($) {
          ## SVG elements.  Currently the HTML syntax supports only MathML and
          ## SVG elements as foreigners.
          $new_mode = IN_BODY_IM | IN_FOREIGN_CONTENT_IM;
-       } elsif ($node->[1] & TABLE_CELL_EL) {
+       } elsif ($node->[1] == TABLE_CELL_EL) {
          if ($last) {
            !!!cp ('t28.2');
            #
-Line 3496 
 sub _reset_insertion_mode ($) {
+Line 1251 
 sub _reset_insertion_mode ($) {
        $self->{insertion_mode} = $new_mode and return if defined $new_mode;
        ## Step 15
-       if ($node->[1] & HTML_EL) {
+       if ($node->[1] == HTML_EL) {
          unless (defined $self->{head_element}) {
            !!!cp ('t29');
            $self->{insertion_mode} = BEFORE_HEAD_IM;
-Line 3628 
 sub _tree_construction_main ($) {
+Line 1383 
 sub _tree_construction_main ($) {
      ## Step 1
      my $start_tag_name = $token->{tag_name};
-     my $el;
+     !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
-     !!!create-element ($el, $HTML_NS, $start_tag_name, $token->{attributes}, $token);
      ## Step 2
-     $insert->($el);
-     ## Step 3
      $self->{content_model} = $content_model_flag; # CDATA or RCDATA
      delete $self->{escape}; # MUST
-     ## Step 4
+     ## Step 3, 4
-     my $text = '';
+     $self->{insertion_mode} |= IN_CDATA_RCDATA_IM;
-     !!!nack ('t40.1');
-     !!!next-token;
-     while ($token->{type} == CHARACTER_TOKEN) { # or until stop tokenizing
-       !!!cp ('t40');
-       $text .= $token->{data};
-       !!!next-token;
-     }
-     ## Step 5
-     if (length $text) {
-       !!!cp ('t41');
-       my $text = $self->{document}->create_text_node ($text);
-       $el->append_child ($text);
-     }
-     ## Step 6
+     !!!nack ('t40.1');
-     $self->{content_model} = PCDATA_CONTENT_MODEL;
-     ## Step 7
-     if ($token->{type} == END_TAG_TOKEN and
-         $token->{tag_name} eq $start_tag_name) {
-       !!!cp ('t42');
-       ## Ignore the token
-     } else {
-       ## NOTE: An end-of-file token.
-       if ($content_model_flag == CDATA_CONTENT_MODEL) {
-         !!!cp ('t43');
-         !!!parse-error (type => 'in CDATA:#eof', token => $token);
-       } elsif ($content_model_flag == RCDATA_CONTENT_MODEL) {
-         !!!cp ('t44');
-         !!!parse-error (type => 'in RCDATA:#eof', token => $token);
-       } else {
-         die "$0: $content_model_flag in parse_rcdata";
-       }
-     }
      !!!next-token;
    }; # $parse_rcdata
    my $script_start_tag = sub () {
+     ## Step 1
      my $script_el;
      !!!create-element ($script_el, $HTML_NS, 'script', $token->{attributes}, $token);
+     ## Step 2
      ## TODO: mark as "parser-inserted"
+     ## Step 3
+     ## TODO: Mark as "already executed", if ...
+     ## Step 4
+     $insert->($script_el);
+     ## ISSUE: $script_el is not put into the stack
+     push @{$self->{open_elements}}, [$script_el, $el_category->{script}];
+     ## Step 5
      $self->{content_model} = CDATA_CONTENT_MODEL;
      delete $self->{escape}; # MUST
-     my $text = '';
-     !!!nack ('t45.1');
-     !!!next-token;
-     while ($token->{type} == CHARACTER_TOKEN) {
-       !!!cp ('t45');
-       $text .= $token->{data};
-       !!!next-token;
-     } # stop if non-character token or tokenizer stops tokenising
-     if (length $text) {
-       !!!cp ('t46');
-       $script_el->manakai_append_text ($text);
-     }
-     $self->{content_model} = PCDATA_CONTENT_MODEL;
-     if ($token->{type} == END_TAG_TOKEN and
+     ## Step 6-7
-         $token->{tag_name} eq 'script') {
+     $self->{insertion_mode} |= IN_CDATA_RCDATA_IM;
-       !!!cp ('t47');
-       ## Ignore the token
-     } else {
-       !!!cp ('t48');
-       !!!parse-error (type => 'in CDATA:#eof', token => $token);
-       ## ISSUE: And ignore?
-       ## TODO: mark as "already executed"
-     }
-     if (defined $self->{inner_html_node}) {
-       !!!cp ('t49');
-       ## TODO: mark as "already executed"
-     } else {
-       !!!cp ('t50');
-       ## TODO: $old_insertion_point = current insertion point
-       ## TODO: insertion point = just before the next input character
-       $insert->($script_el);
+     !!!nack ('t40.2');
-       ## TODO: insertion point = $old_insertion_point (might be "undefined")
-       ## TODO: if there is a script that will execute as soon as the parser resume, then...
-     }
      !!!next-token;
    }; # $script_start_tag
    ## NOTE: $open_tables->[-1]->[0] is the "current table" element node.
    ## NOTE: $open_tables->[-1]->[1] is the "tainted" flag.
+   ## NOTE: $open_tables->[-1]->[2] is set false when non-Text node inserted.
    my $open_tables = [[$self->{open_elements}->[0]->[0]]];
    my $formatting_end_tag = sub {
-Line 3816 
 sub _tree_construction_main ($) {
+Line 1511 
 sub _tree_construction_main ($) {
            !!!cp ('t59');
            $furthest_block = $node;
            $furthest_block_i_in_open = $_;
+           ## NOTE: The topmost (eldest) node.
          } elsif ($node->[0] eq $formatting_element->[0]) {
            !!!cp ('t60');
            last OE;
-Line 3902 
 sub _tree_construction_main ($) {
+Line 1598 
 sub _tree_construction_main ($) {
          my $foster_parent_element;
          my $next_sibling;
          OE: for (reverse 0..$#{$self->{open_elements}}) {
-           if ($self->{open_elements}->[$_]->[1] & TABLE_EL) {
+           if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {
                               my $parent = $self->{open_elements}->[$_]->[0]->parent_node;
                               if (defined $parent and $parent->node_type == 1) {
                                 !!!cp ('t65.1');
-Line 3962 
 sub _tree_construction_main ($) {
+Line 1658 
 sub _tree_construction_main ($) {
            $i = $_;
          }
        } # OE
-       splice @{$self->{open_elements}}, $i + 1, 1, $clone;
+       splice @{$self->{open_elements}}, $i + 1, 0, $clone;
        ## Step 14
        redo FET;
-Line 3980 
 sub _tree_construction_main ($) {
+Line 1676 
 sub _tree_construction_main ($) {
        my $foster_parent_element;
        my $next_sibling;
        OE: for (reverse 0..$#{$self->{open_elements}}) {
-         if ($self->{open_elements}->[$_]->[1] & TABLE_EL) {
+         if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {
                               my $parent = $self->{open_elements}->[$_]->[0]->parent_node;
                               if (defined $parent and $parent->node_type == 1) {
                                 !!!cp ('t70');
-Line 4005 
 sub _tree_construction_main ($) {
+Line 1701 
 sub _tree_construction_main ($) {
      }
    }; # $insert_to_foster
+   ## NOTE: Insert a character (MUST): When a character is inserted, if
+   ## the last node that was inserted by the parser is a Text node and
+   ## the character has to be inserted after that node, then the
+   ## character is appended to the Text node.  However, if any other
+   ## node is inserted by the parser, then a new Text node is created
+   ## and the character is appended as that Text node.  If I'm not
+   ## wrong, for a parser with scripting disabled, there are only two
+   ## cases where this occurs.  One is the case where an element node
+   ## is inserted to the |head| element.  This is covered by using the
+   ## |$self->{head_element_inserted}| flag.  Another is the case where
+   ## an element or comment is inserted into the |table| subtree while
+   ## foster parenting happens.  This is covered by using the [2] flag
+   ## of the |$open_tables| structure.  All other cases are handled
+   ## simply by calling |manakai_append_text| method.
+   ## TODO: |<body><script>document.write("a<br>");
+   ## document.body.removeChild (document.body.lastChild);
+   ## document.write ("b")</script>|
    B: while (1) {
      if ($token->{type} == DOCTYPE_TOKEN) {
        !!!cp ('t73');
-Line 4052 
 sub _tree_construction_main ($) {
+Line 1767 
 sub _tree_construction_main ($) {
        } else {
          !!!cp ('t87');
          $self->{open_elements}->[-1]->[0]->append_child ($comment);
+         $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
        }
        !!!next-token;
        next B;
+     } elsif ($self->{insertion_mode} & IN_CDATA_RCDATA_IM) {
+       if ($token->{type} == CHARACTER_TOKEN) {
+         $token->{data} =~ s/^\x0A// if $self->{ignore_newline};
+         delete $self->{ignore_newline};
+         if (length $token->{data}) {
+           !!!cp ('t43');
+           $self->{open_elements}->[-1]->[0]->manakai_append_text
+               ($token->{data});
+         } else {
+           !!!cp ('t43.1');
+         }
+         !!!next-token;
+         next B;
+       } elsif ($token->{type} == END_TAG_TOKEN) {
+         delete $self->{ignore_newline};
+         if ($token->{tag_name} eq 'script') {
+           !!!cp ('t50');
+           ## Para 1-2
+           my $script = pop @{$self->{open_elements}};
+           ## Para 3
+           $self->{insertion_mode} &= ~ IN_CDATA_RCDATA_IM;
+           ## Para 4
+           ## TODO: $old_insertion_point = $current_insertion_point;
+           ## TODO: $current_insertion_point = just before $self->{nc};
+           ## Para 5
+           ## TODO: Run the $script->[0].
+           ## Para 6
+           ## TODO: $current_insertion_point = $old_insertion_point;
+           ## Para 7
+           ## TODO: if ($pending_external_script) {
+             ## TODO: ...
+           ## TODO: }
+           !!!next-token;
+           next B;
+         } else {
+           !!!cp ('t42');
+           pop @{$self->{open_elements}};
+           $self->{insertion_mode} &= ~ IN_CDATA_RCDATA_IM;
+           !!!next-token;
+           next B;
+         }
+       } elsif ($token->{type} == END_OF_FILE_TOKEN) {
+         delete $self->{ignore_newline};
+         !!!cp ('t44');
+         !!!parse-error (type => 'not closed',
+                         text => $self->{open_elements}->[-1]->[0]
+                             ->manakai_local_name,
+                         token => $token);
+         #if ($self->{open_elements}->[-1]->[1] == SCRIPT_EL) {
+         #  ## TODO: Mark as "already executed"
+         #}
+         pop @{$self->{open_elements}};
+         $self->{insertion_mode} &= ~ IN_CDATA_RCDATA_IM;
+         ## Reprocess.
+         next B;
+       } else {
+         die "$0: $token->{type}: In CDATA/RCDATA: Unknown token type";
+       }
      } elsif ($self->{insertion_mode} & IN_FOREIGN_CONTENT_IM) {
        if ($token->{type} == CHARACTER_TOKEN) {
          !!!cp ('t87.1');
-Line 4066 
 sub _tree_construction_main ($) {
+Line 1855 
 sub _tree_construction_main ($) {
               $self->{open_elements}->[-1]->[1] & FOREIGN_FLOW_CONTENT_EL) or
              not ($self->{open_elements}->[-1]->[1] & FOREIGN_EL) or
              ($token->{tag_name} eq 'svg' and
-              $self->{open_elements}->[-1]->[1] & MML_AXML_EL)) {
+              $self->{open_elements}->[-1]->[1] == MML_AXML_EL)) {
            ## NOTE: "using the rules for secondary insertion mode"then"continue"
            !!!cp ('t87.2');
            #
-Line 4167 
 sub _tree_construction_main ($) {
+Line 1956 
 sub _tree_construction_main ($) {
          pop @{$self->{open_elements}}
              while $self->{open_elements}->[-1]->[1] & FOREIGN_EL;
+         ## NOTE: |<span><svg>| ... two parse errors, |<svg>| ... a parse error.
          $self->{insertion_mode} &= ~ IN_FOREIGN_CONTENT_IM;
          ## Reprocess.
          next B;
-Line 4177 
 sub _tree_construction_main ($) {
+Line 1968 
 sub _tree_construction_main ($) {
      if ($self->{insertion_mode} & HEAD_IMS) {
        if ($token->{type} == CHARACTER_TOKEN) {
-         if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
+         if ($token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {
            unless ($self->{insertion_mode} == BEFORE_HEAD_IM) {
-             !!!cp ('t88.2');
+             if ($self->{head_element_inserted}) {
-             $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
+               !!!cp ('t88.3');
+               $self->{open_elements}->[-1]->[0]->append_child
+                 ($self->{document}->create_text_node ($1));
+               delete $self->{head_element_inserted};
+               ## NOTE: |</head> <link> |
+               #
+             } else {
+               !!!cp ('t88.2');
+               $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
+               ## NOTE: |</head> &#x20;|
+               #
+             }
            } else {
              !!!cp ('t88.1');
              ## Ignore the token.
-             !!!next-token;
+             #
-             next B;
            }
            unless (length $token->{data}) {
              !!!cp ('t88');
              !!!next-token;
              next B;
            }
+ ## TODO: set $token->{column} appropriately
          }
          if ($self->{insertion_mode} == BEFORE_HEAD_IM) {
-Line 4276 
 sub _tree_construction_main ($) {
+Line 2078 
 sub _tree_construction_main ($) {
            !!!cp ('t97');
          }
-             if ($token->{tag_name} eq 'base') {
+         if ($token->{tag_name} eq 'base') {
-               if ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
+           if ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
-                 !!!cp ('t98');
+             !!!cp ('t98');
-                 ## As if </noscript>
+             ## As if </noscript>
-                 pop @{$self->{open_elements}};
+             pop @{$self->{open_elements}};
-                 !!!parse-error (type => 'in noscript', text => 'base',
+             !!!parse-error (type => 'in noscript', text => 'base',
-                                 token => $token);
+                             token => $token);
-                 $self->{insertion_mode} = IN_HEAD_IM;
+             $self->{insertion_mode} = IN_HEAD_IM;
-                 ## Reprocess in the "in head" insertion mode...
+             ## Reprocess in the "in head" insertion mode...
-               } else {
+           } else {
-                 !!!cp ('t99');
+             !!!cp ('t99');
-               }
+           }
-               ## NOTE: There is a "as if in head" code clone.
+           ## NOTE: There is a "as if in head" code clone.
-               if ($self->{insertion_mode} == AFTER_HEAD_IM) {
+           if ($self->{insertion_mode} == AFTER_HEAD_IM) {
-                 !!!cp ('t100');
+             !!!cp ('t100');
-                 !!!parse-error (type => 'after head',
+             !!!parse-error (type => 'after head',
-                                 text => $token->{tag_name}, token => $token);
+                             text => $token->{tag_name}, token => $token);
-                 push @{$self->{open_elements}},
+             push @{$self->{open_elements}},
-                     [$self->{head_element}, $el_category->{head}];
+                 [$self->{head_element}, $el_category->{head}];
-               } else {
+             $self->{head_element_inserted} = 1;
-                 !!!cp ('t101');
+           } else {
-               }
+             !!!cp ('t101');
-               !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+           }
-               pop @{$self->{open_elements}}; ## ISSUE: This step is missing in the spec.
+           !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
-               pop @{$self->{open_elements}} # <head>
+           pop @{$self->{open_elements}};
-                   if $self->{insertion_mode} == AFTER_HEAD_IM;
+           pop @{$self->{open_elements}} # <head>
-               !!!nack ('t101.1');
+               if $self->{insertion_mode} == AFTER_HEAD_IM;
-               !!!next-token;
+           !!!nack ('t101.1');
-               next B;
+           !!!next-token;
-             } elsif ($token->{tag_name} eq 'link') {
+           next B;
-               ## NOTE: There is a "as if in head" code clone.
+         } elsif ($token->{tag_name} eq 'link') {
-               if ($self->{insertion_mode} == AFTER_HEAD_IM) {
+           ## NOTE: There is a "as if in head" code clone.
-                 !!!cp ('t102');
+           if ($self->{insertion_mode} == AFTER_HEAD_IM) {
-                 !!!parse-error (type => 'after head',
+             !!!cp ('t102');
-                                 text => $token->{tag_name}, token => $token);
+             !!!parse-error (type => 'after head',
-                 push @{$self->{open_elements}},
+                             text => $token->{tag_name}, token => $token);
-                     [$self->{head_element}, $el_category->{head}];
+             push @{$self->{open_elements}},
-               } else {
+                 [$self->{head_element}, $el_category->{head}];
-                 !!!cp ('t103');
+             $self->{head_element_inserted} = 1;
-               }
+           } else {
-               !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+             !!!cp ('t103');
-               pop @{$self->{open_elements}}; ## ISSUE: This step is missing in the spec.
+           }
-               pop @{$self->{open_elements}} # <head>
+           !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
-                   if $self->{insertion_mode} == AFTER_HEAD_IM;
+           pop @{$self->{open_elements}};
-               !!!ack ('t103.1');
+           pop @{$self->{open_elements}} # <head>
-               !!!next-token;
+               if $self->{insertion_mode} == AFTER_HEAD_IM;
-               next B;
+           !!!ack ('t103.1');
-             } elsif ($token->{tag_name} eq 'meta') {
+           !!!next-token;
-               ## NOTE: There is a "as if in head" code clone.
+           next B;
-               if ($self->{insertion_mode} == AFTER_HEAD_IM) {
+         } elsif ($token->{tag_name} eq 'command' or
-                 !!!cp ('t104');
+                  $token->{tag_name} eq 'eventsource') {
-                 !!!parse-error (type => 'after head',
+           if ($self->{insertion_mode} == IN_HEAD_IM) {
-                                 text => $token->{tag_name}, token => $token);
+             ## NOTE: If the insertion mode at the time of the emission
-                 push @{$self->{open_elements}},
+             ## of the token was "before head", $self->{insertion_mode}
-                     [$self->{head_element}, $el_category->{head}];
+             ## is already changed to |IN_HEAD_IM|.
-               } else {
-                 !!!cp ('t105');
+             ## NOTE: There is a "as if in head" code clone.
-               }
+             !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
-               !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+             pop @{$self->{open_elements}};
-               my $meta_el = pop @{$self->{open_elements}}; ## ISSUE: This step is missing in the spec.
+             pop @{$self->{open_elements}} # <head>
+                 if $self->{insertion_mode} == AFTER_HEAD_IM;
+             !!!ack ('t103.2');
+             !!!next-token;
+             next B;
+           } else {
+             ## NOTE: "in head noscript" or "after head" insertion mode
+             ## - in these cases, these tags are treated as same as
+             ## normal in-body tags.
+             !!!cp ('t103.3');
+             #
+           }
+         } elsif ($token->{tag_name} eq 'meta') {
+           ## NOTE: There is a "as if in head" code clone.
+           if ($self->{insertion_mode} == AFTER_HEAD_IM) {
+             !!!cp ('t104');
+             !!!parse-error (type => 'after head',
+                             text => $token->{tag_name}, token => $token);
+             push @{$self->{open_elements}},
+                 [$self->{head_element}, $el_category->{head}];
+             $self->{head_element_inserted} = 1;
+           } else {
+             !!!cp ('t105');
+           }
+           !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+           my $meta_el = pop @{$self->{open_elements}};
                unless ($self->{confident}) {
                  if ($token->{attributes}->{charset}) {
-Line 4355 
 sub _tree_construction_main ($) {
+Line 2182 
 sub _tree_construction_main ($) {
                  } elsif ($token->{attributes}->{content}) {
                    if ($token->{attributes}->{content}->{value}
                        =~ /[Cc][Hh][Aa][Rr][Ss][Ee][Tt]
-                           [\x09-\x0D\x20]*=
+                           [\x09\x0A\x0C\x0D\x20]*=
-                           [\x09-\x0D\x20]*(?>"([^"]*)"|'([^']*)'|
+                           [\x09\x0A\x0C\x0D\x20]*(?>"([^"]*)"|'([^']*)'|
-                           ([^"'\x09-\x0D\x20][^\x09-\x0D\x20\x3B]*))/x) {
+                           ([^"'\x09\x0A\x0C\x0D\x20]
+                            [^\x09\x0A\x0C\x0D\x20\x3B]*))/x) {
                      !!!cp ('t107');
                      ## NOTE: Whether the encoding is supported or not is handled
                      ## in the {change_encoding} callback.
-Line 4394 
 sub _tree_construction_main ($) {
+Line 2222 
 sub _tree_construction_main ($) {
                !!!ack ('t110.1');
                !!!next-token;
                next B;
-             } elsif ($token->{tag_name} eq 'title') {
+         } elsif ($token->{tag_name} eq 'title') {
-               if ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
+           if ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
-                 !!!cp ('t111');
+             !!!cp ('t111');
-                 ## As if </noscript>
+             ## As if </noscript>
-                 pop @{$self->{open_elements}};
+             pop @{$self->{open_elements}};
-                 !!!parse-error (type => 'in noscript', text => 'title',
+             !!!parse-error (type => 'in noscript', text => 'title',
-                                 token => $token);
+                             token => $token);
-                 $self->{insertion_mode} = IN_HEAD_IM;
+             $self->{insertion_mode} = IN_HEAD_IM;
-                 ## Reprocess in the "in head" insertion mode...
+             ## Reprocess in the "in head" insertion mode...
-               } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {
+           } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {
-                 !!!cp ('t112');
+             !!!cp ('t112');
-                 !!!parse-error (type => 'after head',
+             !!!parse-error (type => 'after head',
-                                 text => $token->{tag_name}, token => $token);
+                             text => $token->{tag_name}, token => $token);
-                 push @{$self->{open_elements}},
+             push @{$self->{open_elements}},
-                     [$self->{head_element}, $el_category->{head}];
+                 [$self->{head_element}, $el_category->{head}];
-               } else {
+             $self->{head_element_inserted} = 1;
-                 !!!cp ('t113');
+           } else {
-               }
+             !!!cp ('t113');
+           }
-               ## NOTE: There is a "as if in head" code clone.
+           ## NOTE: There is a "as if in head" code clone.
-               my $parent = defined $self->{head_element} ? $self->{head_element}
+           $parse_rcdata->(RCDATA_CONTENT_MODEL);
-                   : $self->{open_elements}->[-1]->[0];
+           ## ISSUE: A spec bug [Bug 6038]
-               $parse_rcdata->(RCDATA_CONTENT_MODEL);
+           splice @{$self->{open_elements}}, -2, 1, () # <head>
-               pop @{$self->{open_elements}} # <head>
+               if ($self->{insertion_mode} & AFTER_HEAD_IM) == AFTER_HEAD_IM;
-                   if $self->{insertion_mode} == AFTER_HEAD_IM;
+           next B;
-               next B;
+         } elsif ($token->{tag_name} eq 'style' or
-             } elsif ($token->{tag_name} eq 'style' or
+                  $token->{tag_name} eq 'noframes') {
-                      $token->{tag_name} eq 'noframes') {
+           ## NOTE: Or (scripting is enabled and tag_name eq 'noscript' and
-               ## NOTE: Or (scripting is enabled and tag_name eq 'noscript' and
+           ## insertion mode IN_HEAD_IM)
-               ## insertion mode IN_HEAD_IM)
+           ## NOTE: There is a "as if in head" code clone.
-               ## NOTE: There is a "as if in head" code clone.
+           if ($self->{insertion_mode} == AFTER_HEAD_IM) {
-               if ($self->{insertion_mode} == AFTER_HEAD_IM) {
+             !!!cp ('t114');
-                 !!!cp ('t114');
+             !!!parse-error (type => 'after head',
-                 !!!parse-error (type => 'after head',
+                             text => $token->{tag_name}, token => $token);
-                                 text => $token->{tag_name}, token => $token);
+             push @{$self->{open_elements}},
-                 push @{$self->{open_elements}},
+                 [$self->{head_element}, $el_category->{head}];
-                     [$self->{head_element}, $el_category->{head}];
+             $self->{head_element_inserted} = 1;
-               } else {
+           } else {
-                 !!!cp ('t115');
+             !!!cp ('t115');
-               }
+           }
-               $parse_rcdata->(CDATA_CONTENT_MODEL);
+           $parse_rcdata->(CDATA_CONTENT_MODEL);
-               pop @{$self->{open_elements}} # <head>
+           ## ISSUE: A spec bug [Bug 6038]
-                   if $self->{insertion_mode} == AFTER_HEAD_IM;
+           splice @{$self->{open_elements}}, -2, 1, () # <head>
-               next B;
+               if ($self->{insertion_mode} & AFTER_HEAD_IM) == AFTER_HEAD_IM;
-             } elsif ($token->{tag_name} eq 'noscript') {
+           next B;
+         } elsif ($token->{tag_name} eq 'noscript') {
                if ($self->{insertion_mode} == IN_HEAD_IM) {
                  !!!cp ('t116');
                  ## NOTE: and scripting is disalbed
-Line 4460 
 sub _tree_construction_main ($) {
+Line 2290 
 sub _tree_construction_main ($) {
                  !!!cp ('t118');
                  #
                }
-             } elsif ($token->{tag_name} eq 'script') {
+         } elsif ($token->{tag_name} eq 'script') {
-               if ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
+           if ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
-                 !!!cp ('t119');
+             !!!cp ('t119');
-                 ## As if </noscript>
+             ## As if </noscript>
-                 pop @{$self->{open_elements}};
+             pop @{$self->{open_elements}};
-                 !!!parse-error (type => 'in noscript', text => 'script',
+             !!!parse-error (type => 'in noscript', text => 'script',
-                                 token => $token);
+                             token => $token);
-                 $self->{insertion_mode} = IN_HEAD_IM;
+             $self->{insertion_mode} = IN_HEAD_IM;
-                 ## Reprocess in the "in head" insertion mode...
+             ## Reprocess in the "in head" insertion mode...
-               } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {
+           } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {
-                 !!!cp ('t120');
+             !!!cp ('t120');
-                 !!!parse-error (type => 'after head',
+             !!!parse-error (type => 'after head',
-                                 text => $token->{tag_name}, token => $token);
+                             text => $token->{tag_name}, token => $token);
-                 push @{$self->{open_elements}},
+             push @{$self->{open_elements}},
-                     [$self->{head_element}, $el_category->{head}];
+                 [$self->{head_element}, $el_category->{head}];
-               } else {
+             $self->{head_element_inserted} = 1;
-                 !!!cp ('t121');
+           } else {
-               }
+             !!!cp ('t121');
+           }
-               ## NOTE: There is a "as if in head" code clone.
+           ## NOTE: There is a "as if in head" code clone.
-               $script_start_tag->();
+           $script_start_tag->();
-               pop @{$self->{open_elements}} # <head>
+           ## ISSUE: A spec bug  [Bug 6038]
-                   if $self->{insertion_mode} == AFTER_HEAD_IM;
+           splice @{$self->{open_elements}}, -2, 1 # <head>
-               next B;
+               if ($self->{insertion_mode} & AFTER_HEAD_IM) == AFTER_HEAD_IM;
-             } elsif ($token->{tag_name} eq 'body' or
+           next B;
-                      $token->{tag_name} eq 'frameset') {
+         } elsif ($token->{tag_name} eq 'body' or
+                  $token->{tag_name} eq 'frameset') {
                if ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
                  !!!cp ('t122');
                  ## As if </noscript>
-Line 4621 
 sub _tree_construction_main ($) {
+Line 2453 
 sub _tree_construction_main ($) {
              } elsif ({
                        body => 1, html => 1,
                       }->{$token->{tag_name}}) {
-               if ($self->{insertion_mode} == BEFORE_HEAD_IM or
+               ## TODO: This branch is entirely redundant.
+               if ($self->{insertion_mode} == BEFORE_HEAD_IM or
                    $self->{insertion_mode} == IN_HEAD_IM or
                    $self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
                  !!!cp ('t140');
-Line 4793 
 sub _tree_construction_main ($) {
+Line 2626 
 sub _tree_construction_main ($) {
        } else {
          die "$0: $token->{type}: Unknown token type";
        }
-           ## ISSUE: An issue in the spec.
      } elsif ($self->{insertion_mode} & BODY_IMS) {
            if ($token->{type} == CHARACTER_TOKEN) {
              !!!cp ('t150');
-Line 4814 
 sub _tree_construction_main ($) {
+Line 2645 
 sub _tree_construction_main ($) {
                  ## have an element in table scope
                  for (reverse 0..$#{$self->{open_elements}}) {
                    my $node = $self->{open_elements}->[$_];
-                   if ($node->[1] & TABLE_CELL_EL) {
+                   if ($node->[1] == TABLE_CELL_EL) {
                      !!!cp ('t151');
                      ## Close the cell
-Line 4848 
 sub _tree_construction_main ($) {
+Line 2679 
 sub _tree_construction_main ($) {
                  INSCOPE: {
                    for (reverse 0..$#{$self->{open_elements}}) {
                      my $node = $self->{open_elements}->[$_];
-                     if ($node->[1] & CAPTION_EL) {
+                     if ($node->[1] == CAPTION_EL) {
                        !!!cp ('t155');
                        $i = $_;
                        last INSCOPE;
-Line 4874 
 sub _tree_construction_main ($) {
+Line 2705 
 sub _tree_construction_main ($) {
                    pop @{$self->{open_elements}};
                  }
-                 unless ($self->{open_elements}->[-1]->[1] & CAPTION_EL) {
+                 unless ($self->{open_elements}->[-1]->[1] == CAPTION_EL) {
                    !!!cp ('t159');
                    !!!parse-error (type => 'not closed',
                                    text => $self->{open_elements}->[-1]->[0]
-Line 4971 
 sub _tree_construction_main ($) {
+Line 2802 
 sub _tree_construction_main ($) {
                  INSCOPE: {
                    for (reverse 0..$#{$self->{open_elements}}) {
                      my $node = $self->{open_elements}->[$_];
-                     if ($node->[1] & CAPTION_EL) {
+                     if ($node->[1] == CAPTION_EL) {
                        !!!cp ('t171');
                        $i = $_;
                        last INSCOPE;
-Line 4996 
 sub _tree_construction_main ($) {
+Line 2827 
 sub _tree_construction_main ($) {
                    pop @{$self->{open_elements}};
                  }
-                 unless ($self->{open_elements}->[-1]->[1] & CAPTION_EL) {
+                 unless ($self->{open_elements}->[-1]->[1] == CAPTION_EL) {
                    !!!cp ('t175');
                    !!!parse-error (type => 'not closed',
                                    text => $self->{open_elements}->[-1]->[0]
-Line 5046 
 sub _tree_construction_main ($) {
+Line 2877 
 sub _tree_construction_main ($) {
                                line => $token->{line},
                                column => $token->{column}};
                      next B;
-                   } elsif ($node->[1] & TABLE_CELL_EL) {
+                   } elsif ($node->[1] == TABLE_CELL_EL) {
                      !!!cp ('t180');
                      $tn = $node->[0]->manakai_local_name;
                      ## NOTE: There is exactly one |td| or |th| element
-Line 5075 
 sub _tree_construction_main ($) {
+Line 2906 
 sub _tree_construction_main ($) {
                my $i;
                INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                  my $node = $self->{open_elements}->[$_];
-                 if ($node->[1] & CAPTION_EL) {
+                 if ($node->[1] == CAPTION_EL) {
                    !!!cp ('t184');
                    $i = $_;
                    last INSCOPE;
-Line 5099 
 sub _tree_construction_main ($) {
+Line 2930 
 sub _tree_construction_main ($) {
                  pop @{$self->{open_elements}};
                }
-               unless ($self->{open_elements}->[-1]->[1] & CAPTION_EL) {
+               unless ($self->{open_elements}->[-1]->[1] == CAPTION_EL) {
                  !!!cp ('t188');
                  !!!parse-error (type => 'not closed',
                                  text => $self->{open_elements}->[-1]->[0]
-Line 5166 
 sub _tree_construction_main ($) {
+Line 2997 
 sub _tree_construction_main ($) {
      } elsif ($self->{insertion_mode} & TABLE_IMS) {
        if ($token->{type} == CHARACTER_TOKEN) {
          if (not $open_tables->[-1]->[1] and # tainted
-             $token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
+             $token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {
            $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
            unless (length $token->{data}) {
-Line 5180 
 sub _tree_construction_main ($) {
+Line 3011 
 sub _tree_construction_main ($) {
          !!!parse-error (type => 'in table:#text', token => $token);
-             ## As if in body, but insert into foster parent element
+         ## NOTE: As if in body, but insert into the foster parent element.
-             ## ISSUE: Spec says that "whenever a node would be inserted
+         $reconstruct_active_formatting_elements->($insert_to_foster);
-             ## into the current node" while characters might not be
-             ## result in a new Text node.
-             $reconstruct_active_formatting_elements->($insert_to_foster);
-             if ($self->{open_elements}->[-1]->[1] & TABLE_ROWS_EL) {
+         if ($self->{open_elements}->[-1]->[1] & TABLE_ROWS_EL) {
-               # MUST
+           # MUST
-               my $foster_parent_element;
+           my $foster_parent_element;
-               my $next_sibling;
+           my $next_sibling;
-               my $prev_sibling;
+           my $prev_sibling;
-               OE: for (reverse 0..$#{$self->{open_elements}}) {
+           OE: for (reverse 0..$#{$self->{open_elements}}) {
-                 if ($self->{open_elements}->[$_]->[1] & TABLE_EL) {
+             if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {
-                   my $parent = $self->{open_elements}->[$_]->[0]->parent_node;
+               my $parent = $self->{open_elements}->[$_]->[0]->parent_node;
-                   if (defined $parent and $parent->node_type == 1) {
+               if (defined $parent and $parent->node_type == 1) {
-                     !!!cp ('t196');
+                 $foster_parent_element = $parent;
-                     $foster_parent_element = $parent;
+                 !!!cp ('t196');
-                     $next_sibling = $self->{open_elements}->[$_]->[0];
+                 $next_sibling = $self->{open_elements}->[$_]->[0];
-                     $prev_sibling = $next_sibling->previous_sibling;
+                 $prev_sibling = $next_sibling->previous_sibling;
-                   } else {
+                 #
-                     !!!cp ('t197');
-                     $foster_parent_element = $self->{open_elements}->[$_ - 1]->[0];
-                     $prev_sibling = $foster_parent_element->last_child;
-                   }
-                   last OE;
-                 }
-               } # OE
-               $foster_parent_element = $self->{open_elements}->[0]->[0] and
-               $prev_sibling = $foster_parent_element->last_child
-                 unless defined $foster_parent_element;
-               if (defined $prev_sibling and
-                   $prev_sibling->node_type == 3) {
-                 !!!cp ('t198');
-                 $prev_sibling->manakai_append_text ($token->{data});
                } else {
-                 !!!cp ('t199');
+                 !!!cp ('t197');
-                 $foster_parent_element->insert_before
+                 $foster_parent_element = $self->{open_elements}->[$_ - 1]->[0];
-                   ($self->{document}->create_text_node ($token->{data}),
+                 $prev_sibling = $foster_parent_element->last_child;
-                    $next_sibling);
+                 #
                }
+               last OE;
+             }
+           } # OE
+           $foster_parent_element = $self->{open_elements}->[0]->[0] and
+           $prev_sibling = $foster_parent_element->last_child
+               unless defined $foster_parent_element;
+           undef $prev_sibling unless $open_tables->[-1]->[2]; # ~node inserted
+           if (defined $prev_sibling and
+               $prev_sibling->node_type == 3) {
+             !!!cp ('t198');
+             $prev_sibling->manakai_append_text ($token->{data});
+           } else {
+             !!!cp ('t199');
+             $foster_parent_element->insert_before
+                 ($self->{document}->create_text_node ($token->{data}),
+                  $next_sibling);
+           }
            $open_tables->[-1]->[1] = 1; # tainted
+           $open_tables->[-1]->[2] = 1; # ~node inserted
          } else {
+           ## NOTE: Fragment case or in a foster parent'ed element
+           ## (e.g. |<table><span>a|).  In fragment case, whether the
+           ## character is appended to existing node or a new node is
+           ## created is irrelevant, since the foster parent'ed nodes
+           ## are discarded and fragment parsing does not invoke any
+           ## script.
            !!!cp ('t200');
-           $self->{open_elements}->[-1]->[0]->manakai_append_text ($token->{data});
+           $self->{open_elements}->[-1]->[0]->manakai_append_text
+               ($token->{data});
          }
          !!!next-token;
-Line 5260 
 sub _tree_construction_main ($) {
+Line 3099 
 sub _tree_construction_main ($) {
                pop @{$self->{open_elements}};
              }
-                 $self->{insertion_mode} = IN_ROW_IM;
+             $self->{insertion_mode} = IN_ROW_IM;
-                 if ($token->{tag_name} eq 'tr') {
+             if ($token->{tag_name} eq 'tr') {
-                   !!!cp ('t204');
+               !!!cp ('t204');
-                   !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+               !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
-                   !!!nack ('t204');
+               $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
-                   !!!next-token;
+               !!!nack ('t204');
-                   next B;
+               !!!next-token;
-                 } else {
+               next B;
-                   !!!cp ('t205');
+             } else {
-                   !!!insert-element ('tr',, $token);
+               !!!cp ('t205');
-                   ## reprocess in the "in row" insertion mode
+               !!!insert-element ('tr',, $token);
-                 }
+               ## reprocess in the "in row" insertion mode
-               } else {
+             }
-                 !!!cp ('t206');
+           } else {
-               }
+             !!!cp ('t206');
+           }
                ## Clear back to table row context
                while (not ($self->{open_elements}->[-1]->[1]
-Line 5283 
 sub _tree_construction_main ($) {
+Line 3123 
 sub _tree_construction_main ($) {
                  pop @{$self->{open_elements}};
                }
-               !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+           !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
-               $self->{insertion_mode} = IN_CELL_IM;
+           $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
+           $self->{insertion_mode} = IN_CELL_IM;
-               push @$active_formatting_elements, ['#marker', ''];
+           push @$active_formatting_elements, ['#marker', ''];
-               !!!nack ('t207.1');
+           !!!nack ('t207.1');
+           !!!next-token;
+           next B;
+         } elsif ({
+                   caption => 1, col => 1, colgroup => 1,
+                   tbody => 1, tfoot => 1, thead => 1,
+                   tr => 1, # $self->{insertion_mode} == IN_ROW_IM
+                  }->{$token->{tag_name}}) {
+           if ($self->{insertion_mode} == IN_ROW_IM) {
+             ## As if </tr>
+             ## have an element in table scope
+             my $i;
+             INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
+               my $node = $self->{open_elements}->[$_];
+               if ($node->[1] == TABLE_ROW_EL) {
+                 !!!cp ('t208');
+                 $i = $_;
+                 last INSCOPE;
+               } elsif ($node->[1] & TABLE_SCOPING_EL) {
+                 !!!cp ('t209');
+                 last INSCOPE;
+               }
+             } # INSCOPE
+             unless (defined $i) {
+               !!!cp ('t210');
+               ## TODO: This type is wrong.
+               !!!parse-error (type => 'unmacthed end tag',
+                               text => $token->{tag_name}, token => $token);
+               ## Ignore the token
+               !!!nack ('t210.1');
                !!!next-token;
                next B;
-             } elsif ({
+             }
-                       caption => 1, col => 1, colgroup => 1,
-                       tbody => 1, tfoot => 1, thead => 1,
-                       tr => 1, # $self->{insertion_mode} == IN_ROW_IM
-                      }->{$token->{tag_name}}) {
-               if ($self->{insertion_mode} == IN_ROW_IM) {
-                 ## As if </tr>
-                 ## have an element in table scope
-                 my $i;
-                 INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
-                   my $node = $self->{open_elements}->[$_];
-                   if ($node->[1] & TABLE_ROW_EL) {
-                     !!!cp ('t208');
-                     $i = $_;
-                     last INSCOPE;
-                   } elsif ($node->[1] & TABLE_SCOPING_EL) {
-                     !!!cp ('t209');
-                     last INSCOPE;
-                   }
-                 } # INSCOPE
-                 unless (defined $i) {
-                   !!!cp ('t210');
- ## TODO: This type is wrong.
-                   !!!parse-error (type => 'unmacthed end tag',
-                                   text => $token->{tag_name}, token => $token);
-                   ## Ignore the token
-                   !!!nack ('t210.1');
-                   !!!next-token;
-                   next B;
-                 }
                  ## Clear back to table row context
                  while (not ($self->{open_elements}->[-1]->[1]
-Line 5348 
 sub _tree_construction_main ($) {
+Line 3189 
 sub _tree_construction_main ($) {
                  my $i;
                  INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                    my $node = $self->{open_elements}->[$_];
-                   if ($node->[1] & TABLE_ROW_GROUP_EL) {
+                   if ($node->[1] == TABLE_ROW_GROUP_EL) {
                      !!!cp ('t214');
                      $i = $_;
                      last INSCOPE;
-Line 5390 
 sub _tree_construction_main ($) {
+Line 3231 
 sub _tree_construction_main ($) {
                  !!!cp ('t218');
                }
-               if ($token->{tag_name} eq 'col') {
+           if ($token->{tag_name} eq 'col') {
-                 ## Clear back to table context
+             ## Clear back to table context
-                 while (not ($self->{open_elements}->[-1]->[1]
+             while (not ($self->{open_elements}->[-1]->[1]
-                                 & TABLE_SCOPING_EL)) {
+                             & TABLE_SCOPING_EL)) {
-                   !!!cp ('t219');
+               !!!cp ('t219');
-                   ## ISSUE: Can this state be reached?
+               ## ISSUE: Can this state be reached?
-                   pop @{$self->{open_elements}};
+               pop @{$self->{open_elements}};
-                 }
+             }
-                 !!!insert-element ('colgroup',, $token);
+             !!!insert-element ('colgroup',, $token);
-                 $self->{insertion_mode} = IN_COLUMN_GROUP_IM;
+             $self->{insertion_mode} = IN_COLUMN_GROUP_IM;
-                 ## reprocess
+             ## reprocess
-                 !!!ack-later;
+             $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
-                 next B;
+             !!!ack-later;
-               } elsif ({
+             next B;
-                         caption => 1,
+           } elsif ({
-                         colgroup => 1,
+                     caption => 1,
-                         tbody => 1, tfoot => 1, thead => 1,
+                     colgroup => 1,
-                        }->{$token->{tag_name}}) {
+                     tbody => 1, tfoot => 1, thead => 1,
-                 ## Clear back to table context
+                    }->{$token->{tag_name}}) {
+             ## Clear back to table context
                  while (not ($self->{open_elements}->[-1]->[1]
                                  & TABLE_SCOPING_EL)) {
                    !!!cp ('t220');
-Line 5417 
 sub _tree_construction_main ($) {
+Line 3259 
 sub _tree_construction_main ($) {
                    pop @{$self->{open_elements}};
                  }
-                 push @$active_formatting_elements, ['#marker', '']
+             push @$active_formatting_elements, ['#marker', '']
-                     if $token->{tag_name} eq 'caption';
+                 if $token->{tag_name} eq 'caption';
-                 !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+             !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
-                 $self->{insertion_mode} = {
+             $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
-                                            caption => IN_CAPTION_IM,
+             $self->{insertion_mode} = {
-                                            colgroup => IN_COLUMN_GROUP_IM,
+                                        caption => IN_CAPTION_IM,
-                                            tbody => IN_TABLE_BODY_IM,
+                                        colgroup => IN_COLUMN_GROUP_IM,
-                                            tfoot => IN_TABLE_BODY_IM,
+                                        tbody => IN_TABLE_BODY_IM,
-                                            thead => IN_TABLE_BODY_IM,
+                                        tfoot => IN_TABLE_BODY_IM,
-                                           }->{$token->{tag_name}};
+                                        thead => IN_TABLE_BODY_IM,
-                 !!!next-token;
+                                       }->{$token->{tag_name}};
-                 !!!nack ('t220.1');
+             !!!next-token;
-                 next B;
+             !!!nack ('t220.1');
-               } else {
+             next B;
-                 die "$0: in table: <>: $token->{tag_name}";
+           } else {
-               }
+             die "$0: in table: <>: $token->{tag_name}";
+           }
              } elsif ($token->{tag_name} eq 'table') {
                !!!parse-error (type => 'not closed',
                                text => $self->{open_elements}->[-1]->[0]
-Line 5445 
 sub _tree_construction_main ($) {
+Line 3288 
 sub _tree_construction_main ($) {
                my $i;
                INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                  my $node = $self->{open_elements}->[$_];
-                 if ($node->[1] & TABLE_EL) {
+                 if ($node->[1] == TABLE_EL) {
                    !!!cp ('t221');
                    $i = $_;
                    last INSCOPE;
-Line 5472 
 sub _tree_construction_main ($) {
+Line 3315 
 sub _tree_construction_main ($) {
                  pop @{$self->{open_elements}};
                }
-               unless ($self->{open_elements}->[-1]->[1] & TABLE_EL) {
+               unless ($self->{open_elements}->[-1]->[1] == TABLE_EL) {
                  !!!cp ('t225');
                  ## NOTE: |<table><tr><table>|
                  !!!parse-error (type => 'not closed',
-Line 5496 
 sub _tree_construction_main ($) {
+Line 3339 
 sub _tree_construction_main ($) {
              !!!cp ('t227.8');
              ## NOTE: This is a "as if in head" code clone.
              $parse_rcdata->(CDATA_CONTENT_MODEL);
+             $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
              next B;
            } else {
              !!!cp ('t227.7');
-Line 5506 
 sub _tree_construction_main ($) {
+Line 3350 
 sub _tree_construction_main ($) {
              !!!cp ('t227.6');
              ## NOTE: This is a "as if in head" code clone.
              $script_start_tag->();
+             $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
              next B;
            } else {
              !!!cp ('t227.5');
-Line 5521 
 sub _tree_construction_main ($) {
+Line 3366 
 sub _tree_construction_main ($) {
                                  text => $token->{tag_name}, token => $token);
                  !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
+                 $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
                  ## TODO: form element pointer
-Line 5558 
 sub _tree_construction_main ($) {
+Line 3404 
 sub _tree_construction_main ($) {
                my $i;
                INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                  my $node = $self->{open_elements}->[$_];
-                 if ($node->[1] & TABLE_ROW_EL) {
+                 if ($node->[1] == TABLE_ROW_EL) {
                    !!!cp ('t228');
                    $i = $_;
                    last INSCOPE;
-Line 5599 
 sub _tree_construction_main ($) {
+Line 3445 
 sub _tree_construction_main ($) {
                  my $i;
                  INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                    my $node = $self->{open_elements}->[$_];
-                   if ($node->[1] & TABLE_ROW_EL) {
+                   if ($node->[1] == TABLE_ROW_EL) {
                      !!!cp ('t233');
                      $i = $_;
                      last INSCOPE;
-Line 5637 
 sub _tree_construction_main ($) {
+Line 3483 
 sub _tree_construction_main ($) {
                  my $i;
                  INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                    my $node = $self->{open_elements}->[$_];
-                   if ($node->[1] & TABLE_ROW_GROUP_EL) {
+                   if ($node->[1] == TABLE_ROW_GROUP_EL) {
                      !!!cp ('t237');
                      $i = $_;
                      last INSCOPE;
-Line 5684 
 sub _tree_construction_main ($) {
+Line 3530 
 sub _tree_construction_main ($) {
                my $i;
                INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                  my $node = $self->{open_elements}->[$_];
-                 if ($node->[1] & TABLE_EL) {
+                 if ($node->[1] == TABLE_EL) {
                    !!!cp ('t241');
                    $i = $_;
                    last INSCOPE;
-Line 5743 
 sub _tree_construction_main ($) {
+Line 3589 
 sub _tree_construction_main ($) {
                  my $i;
                  INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
                    my $node = $self->{open_elements}->[$_];
-                   if ($node->[1] & TABLE_ROW_EL) {
+                   if ($node->[1] == TABLE_ROW_EL) {
                      !!!cp ('t250');
                      $i = $_;
                      last INSCOPE;
-Line 5833 
 sub _tree_construction_main ($) {
+Line 3679 
 sub _tree_construction_main ($) {
            #
          }
        } elsif ($token->{type} == END_OF_FILE_TOKEN) {
-         unless ($self->{open_elements}->[-1]->[1] & HTML_EL and
+         unless ($self->{open_elements}->[-1]->[1] == HTML_EL and
                  @{$self->{open_elements}} == 1) { # redundant, maybe
            !!!parse-error (type => 'in body:#eof', token => $token);
            !!!cp ('t259.1');
-Line 5850 
 sub _tree_construction_main ($) {
+Line 3696 
 sub _tree_construction_main ($) {
        }
      } elsif ($self->{insertion_mode} == IN_COLUMN_GROUP_IM) {
            if ($token->{type} == CHARACTER_TOKEN) {
-             if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
+             if ($token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {
                $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
                unless (length $token->{data}) {
                  !!!cp ('t260');
-Line 5875 
 sub _tree_construction_main ($) {
+Line 3721 
 sub _tree_construction_main ($) {
              }
            } elsif ($token->{type} == END_TAG_TOKEN) {
              if ($token->{tag_name} eq 'colgroup') {
-               if ($self->{open_elements}->[-1]->[1] & HTML_EL) {
+               if ($self->{open_elements}->[-1]->[1] == HTML_EL) {
                  !!!cp ('t264');
                  !!!parse-error (type => 'unmatched end tag',
                                  text => 'colgroup', token => $token);
-Line 5901 
 sub _tree_construction_main ($) {
+Line 3747 
 sub _tree_construction_main ($) {
                #
              }
        } elsif ($token->{type} == END_OF_FILE_TOKEN) {
-         if ($self->{open_elements}->[-1]->[1] & HTML_EL and
+         if ($self->{open_elements}->[-1]->[1] == HTML_EL and
              @{$self->{open_elements}} == 1) { # redundant, maybe
            !!!cp ('t270.2');
            ## Stop parsing.
-Line 5919 
 sub _tree_construction_main ($) {
+Line 3765 
 sub _tree_construction_main ($) {
        }
            ## As if </colgroup>
-           if ($self->{open_elements}->[-1]->[1] & HTML_EL) {
+           if ($self->{open_elements}->[-1]->[1] == HTML_EL) {
              !!!cp ('t269');
  ## TODO: Wrong error type?
              !!!parse-error (type => 'unmatched end tag',
-Line 5944 
 sub _tree_construction_main ($) {
+Line 3790 
 sub _tree_construction_main ($) {
          next B;
        } elsif ($token->{type} == START_TAG_TOKEN) {
          if ($token->{tag_name} eq 'option') {
-           if ($self->{open_elements}->[-1]->[1] & OPTION_EL) {
+           if ($self->{open_elements}->[-1]->[1] == OPTION_EL) {
              !!!cp ('t272');
              ## As if </option>
              pop @{$self->{open_elements}};
-Line 5957 
 sub _tree_construction_main ($) {
+Line 3803 
 sub _tree_construction_main ($) {
            !!!next-token;
            next B;
          } elsif ($token->{tag_name} eq 'optgroup') {
-           if ($self->{open_elements}->[-1]->[1] & OPTION_EL) {
+           if ($self->{open_elements}->[-1]->[1] == OPTION_EL) {
              !!!cp ('t274');
              ## As if </option>
              pop @{$self->{open_elements}};
-Line 5965 
 sub _tree_construction_main ($) {
+Line 3811 
 sub _tree_construction_main ($) {
              !!!cp ('t275');
            }
-           if ($self->{open_elements}->[-1]->[1] & OPTGROUP_EL) {
+           if ($self->{open_elements}->[-1]->[1] == OPTGROUP_EL) {
              !!!cp ('t276');
              ## As if </optgroup>
              pop @{$self->{open_elements}};
-Line 5995 
 sub _tree_construction_main ($) {
+Line 3841 
 sub _tree_construction_main ($) {
            my $i;
            INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
              my $node = $self->{open_elements}->[$_];
-             if ($node->[1] & SELECT_EL) {
+             if ($node->[1] == SELECT_EL) {
                !!!cp ('t278');
                $i = $_;
                last INSCOPE;
-Line 6040 
 sub _tree_construction_main ($) {
+Line 3886 
 sub _tree_construction_main ($) {
          }
        } elsif ($token->{type} == END_TAG_TOKEN) {
          if ($token->{tag_name} eq 'optgroup') {
-           if ($self->{open_elements}->[-1]->[1] & OPTION_EL and
+           if ($self->{open_elements}->[-1]->[1] == OPTION_EL and
-               $self->{open_elements}->[-2]->[1] & OPTGROUP_EL) {
+               $self->{open_elements}->[-2]->[1] == OPTGROUP_EL) {
              !!!cp ('t283');
              ## As if </option>
              splice @{$self->{open_elements}}, -2;
-           } elsif ($self->{open_elements}->[-1]->[1] & OPTGROUP_EL) {
+           } elsif ($self->{open_elements}->[-1]->[1] == OPTGROUP_EL) {
              !!!cp ('t284');
              pop @{$self->{open_elements}};
            } else {
-Line 6058 
 sub _tree_construction_main ($) {
+Line 3904 
 sub _tree_construction_main ($) {
            !!!next-token;
            next B;
          } elsif ($token->{tag_name} eq 'option') {
-           if ($self->{open_elements}->[-1]->[1] & OPTION_EL) {
+           if ($self->{open_elements}->[-1]->[1] == OPTION_EL) {
              !!!cp ('t286');
              pop @{$self->{open_elements}};
            } else {
-Line 6075 
 sub _tree_construction_main ($) {
+Line 3921 
 sub _tree_construction_main ($) {
            my $i;
            INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
              my $node = $self->{open_elements}->[$_];
-             if ($node->[1] & SELECT_EL) {
+             if ($node->[1] == SELECT_EL) {
                !!!cp ('t288');
                $i = $_;
                last INSCOPE;
-Line 6137 
 sub _tree_construction_main ($) {
+Line 3983 
 sub _tree_construction_main ($) {
            undef $i;
            INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
              my $node = $self->{open_elements}->[$_];
-             if ($node->[1] & SELECT_EL) {
+             if ($node->[1] == SELECT_EL) {
                !!!cp ('t295');
                $i = $_;
                last INSCOPE;
-Line 6176 
 sub _tree_construction_main ($) {
+Line 4022 
 sub _tree_construction_main ($) {
            next B;
          }
        } elsif ($token->{type} == END_OF_FILE_TOKEN) {
-         unless ($self->{open_elements}->[-1]->[1] & HTML_EL and
+         unless ($self->{open_elements}->[-1]->[1] == HTML_EL and
                  @{$self->{open_elements}} == 1) { # redundant, maybe
            !!!cp ('t299.1');
            !!!parse-error (type => 'in body:#eof', token => $token);
-Line 6191 
 sub _tree_construction_main ($) {
+Line 4037 
 sub _tree_construction_main ($) {
        }
      } elsif ($self->{insertion_mode} & BODY_AFTER_IMS) {
        if ($token->{type} == CHARACTER_TOKEN) {
-         if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
+         if ($token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {
            my $data = $1;
            ## As if in body
            $reconstruct_active_formatting_elements->($insert_to_current);
-Line 6208 
 sub _tree_construction_main ($) {
+Line 4054 
 sub _tree_construction_main ($) {
          if ($self->{insertion_mode} == AFTER_HTML_BODY_IM) {
            !!!cp ('t301');
            !!!parse-error (type => 'after html:#text', token => $token);
+           #
-           ## Reprocess in the "after body" insertion mode.
          } else {
            !!!cp ('t302');
+           ## "after body" insertion mode
+           !!!parse-error (type => 'after body:#text', token => $token);
+           #
          }
-         ## "after body" insertion mode
-         !!!parse-error (type => 'after body:#text', token => $token);
          $self->{insertion_mode} = IN_BODY_IM;
          ## reprocess
-Line 6225 
 sub _tree_construction_main ($) {
+Line 4070 
 sub _tree_construction_main ($) {
            !!!cp ('t303');
            !!!parse-error (type => 'after html',
                            text => $token->{tag_name}, token => $token);
+           #
-           ## Reprocess in the "after body" insertion mode.
          } else {
            !!!cp ('t304');
+           ## "after body" insertion mode
+           !!!parse-error (type => 'after body',
+                           text => $token->{tag_name}, token => $token);
+           #
          }
-         ## "after body" insertion mode
-         !!!parse-error (type => 'after body',
-                         text => $token->{tag_name}, token => $token);
          $self->{insertion_mode} = IN_BODY_IM;
          !!!ack-later;
          ## reprocess
-Line 6245 
 sub _tree_construction_main ($) {
+Line 4089 
 sub _tree_construction_main ($) {
            !!!parse-error (type => 'after html:/',
                            text => $token->{tag_name}, token => $token);
-           $self->{insertion_mode} = AFTER_BODY_IM;
+           $self->{insertion_mode} = IN_BODY_IM;
-           ## Reprocess in the "after body" insertion mode.
+           ## Reprocess.
+           next B;
          } else {
            !!!cp ('t306');
          }
-Line 6284 
 sub _tree_construction_main ($) {
+Line 4129 
 sub _tree_construction_main ($) {
        }
      } elsif ($self->{insertion_mode} & FRAME_IMS) {
        if ($token->{type} == CHARACTER_TOKEN) {
-         if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
+         if ($token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {
            $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
            unless (length $token->{data}) {
-Line 6294 
 sub _tree_construction_main ($) {
+Line 4139 
 sub _tree_construction_main ($) {
            }
          }
-         if ($token->{data} =~ s/^[^\x09\x0A\x0B\x0C\x20]+//) {
+         if ($token->{data} =~ s/^[^\x09\x0A\x0C\x20]+//) {
            if ($self->{insertion_mode} == IN_FRAMESET_IM) {
              !!!cp ('t311');
              !!!parse-error (type => 'in frameset:#text', token => $token);
            } elsif ($self->{insertion_mode} == AFTER_FRAMESET_IM) {
              !!!cp ('t312');
              !!!parse-error (type => 'after frameset:#text', token => $token);
-           } else { # "after html frameset"
+           } else { # "after after frameset"
              !!!cp ('t313');
              !!!parse-error (type => 'after html:#text', token => $token);
-             $self->{insertion_mode} = AFTER_FRAMESET_IM;
-             ## Reprocess in the "after frameset" insertion mode.
-             !!!parse-error (type => 'after frameset:#text', token => $token);
            }
            ## Ignore the token.
-Line 6323 
 sub _tree_construction_main ($) {
+Line 4164 
 sub _tree_construction_main ($) {
          die qq[$0: Character "$token->{data}"];
        } elsif ($token->{type} == START_TAG_TOKEN) {
-         if ($self->{insertion_mode} == AFTER_HTML_FRAMESET_IM) {
-           !!!cp ('t316');
-           !!!parse-error (type => 'after html',
-                           text => $token->{tag_name}, token => $token);
-           $self->{insertion_mode} = AFTER_FRAMESET_IM;
-           ## Process in the "after frameset" insertion mode.
-         } else {
-           !!!cp ('t317');
-         }
          if ($token->{tag_name} eq 'frameset' and
              $self->{insertion_mode} == IN_FRAMESET_IM) {
            !!!cp ('t318');
-Line 6354 
 sub _tree_construction_main ($) {
+Line 4184 
 sub _tree_construction_main ($) {
            ## NOTE: As if in head.
            $parse_rcdata->(CDATA_CONTENT_MODEL);
            next B;
+           ## NOTE: |<!DOCTYPE HTML><frameset></frameset></html><noframes></noframes>|
+           ## has no parse error.
          } else {
            if ($self->{insertion_mode} == IN_FRAMESET_IM) {
              !!!cp ('t321');
              !!!parse-error (type => 'in frameset',
                              text => $token->{tag_name}, token => $token);
-           } else {
+           } elsif ($self->{insertion_mode} == AFTER_FRAMESET_IM) {
              !!!cp ('t322');
              !!!parse-error (type => 'after frameset',
                              text => $token->{tag_name}, token => $token);
+           } else { # "after after frameset"
+             !!!cp ('t322.2');
+             !!!parse-error (type => 'after after frameset',
+                             text => $token->{tag_name}, token => $token);
            }
            ## Ignore the token
            !!!nack ('t322.1');
-Line 6370 
 sub _tree_construction_main ($) {
+Line 4207 
 sub _tree_construction_main ($) {
            next B;
          }
        } elsif ($token->{type} == END_TAG_TOKEN) {
-         if ($self->{insertion_mode} == AFTER_HTML_FRAMESET_IM) {
-           !!!cp ('t323');
-           !!!parse-error (type => 'after html:/',
-                           text => $token->{tag_name}, token => $token);
-           $self->{insertion_mode} = AFTER_FRAMESET_IM;
-           ## Process in the "after frameset" insertion mode.
-         } else {
-           !!!cp ('t324');
-         }
          if ($token->{tag_name} eq 'frameset' and
              $self->{insertion_mode} == IN_FRAMESET_IM) {
-           if ($self->{open_elements}->[-1]->[1] & HTML_EL and
+           if ($self->{open_elements}->[-1]->[1] == HTML_EL and
                @{$self->{open_elements}} == 1) {
              !!!cp ('t325');
              !!!parse-error (type => 'unmatched end tag',
-Line 6397 
 sub _tree_construction_main ($) {
+Line 4223 
 sub _tree_construction_main ($) {
            }
            if (not defined $self->{inner_html_node} and
-               not ($self->{open_elements}->[-1]->[1] & FRAMESET_EL)) {
+               not ($self->{open_elements}->[-1]->[1] == FRAMESET_EL)) {
              !!!cp ('t327');
              $self->{insertion_mode} = AFTER_FRAMESET_IM;
            } else {
-Line 6415 
 sub _tree_construction_main ($) {
+Line 4241 
 sub _tree_construction_main ($) {
              !!!cp ('t330');
              !!!parse-error (type => 'in frameset:/',
                              text => $token->{tag_name}, token => $token);
-           } else {
+           } elsif ($self->{insertion_mode} == AFTER_FRAMESET_IM) {
-             !!!cp ('t331');
+             !!!cp ('t330.1');
              !!!parse-error (type => 'after frameset:/',
                              text => $token->{tag_name}, token => $token);
+           } else { # "after after html"
+             !!!cp ('t331');
+             !!!parse-error (type => 'after after frameset:/',
+                             text => $token->{tag_name}, token => $token);
            }
            ## Ignore the token
            !!!next-token;
            next B;
          }
        } elsif ($token->{type} == END_OF_FILE_TOKEN) {
-         unless ($self->{open_elements}->[-1]->[1] & HTML_EL and
+         unless ($self->{open_elements}->[-1]->[1] == HTML_EL and
                  @{$self->{open_elements}} == 1) { # redundant, maybe
            !!!cp ('t331.1');
            !!!parse-error (type => 'in body:#eof', token => $token);
-Line 6438 
 sub _tree_construction_main ($) {
+Line 4268 
 sub _tree_construction_main ($) {
        } else {
          die "$0: $token->{type}: Unknown token type";
        }
-       ## ISSUE: An issue in spec here
      } else {
        die "$0: $self->{insertion_mode}: Unknown insertion mode";
      }
-Line 6457 
 sub _tree_construction_main ($) {
+Line 4285 
 sub _tree_construction_main ($) {
          $parse_rcdata->(CDATA_CONTENT_MODEL);
          next B;
        } elsif ({
-                 base => 1, link => 1,
+                 base => 1, command => 1, eventsource => 1, link => 1,
                 }->{$token->{tag_name}}) {
          !!!cp ('t334');
          ## NOTE: This is an "as if in head" code clone, only "-t" differs
          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
-         pop @{$self->{open_elements}}; ## ISSUE: This step is missing in the spec.
+         pop @{$self->{open_elements}};
          !!!ack ('t334.1');
          !!!next-token;
          next B;
        } elsif ($token->{tag_name} eq 'meta') {
          ## NOTE: This is an "as if in head" code clone, only "-t" differs
          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
-         my $meta_el = pop @{$self->{open_elements}}; ## ISSUE: This step is missing in the spec.
+         my $meta_el = pop @{$self->{open_elements}};
          unless ($self->{confident}) {
            if ($token->{attributes}->{charset}) {
-Line 6486 
 sub _tree_construction_main ($) {
+Line 4314 
 sub _tree_construction_main ($) {
            } elsif ($token->{attributes}->{content}) {
              if ($token->{attributes}->{content}->{value}
                  =~ /[Cc][Hh][Aa][Rr][Ss][Ee][Tt]
-                     [\x09-\x0D\x20]*=
+                     [\x09\x0A\x0C\x0D\x20]*=
-                     [\x09-\x0D\x20]*(?>"([^"]*)"|'([^']*)'|
+                     [\x09\x0A\x0C\x0D\x20]*(?>"([^"]*)"|'([^']*)'|
-                     ([^"'\x09-\x0D\x20][^\x09-\x0D\x20\x3B]*))/x) {
+                     ([^"'\x09\x0A\x0C\x0D\x20][^\x09\x0A\x0C\x0D\x20\x3B]*))
+                    /x) {
                !!!cp ('t336');
                ## NOTE: Whether the encoding is supported or not is handled
                ## in the {change_encoding} callback.
-Line 6529 
 sub _tree_construction_main ($) {
+Line 4358 
 sub _tree_construction_main ($) {
          !!!parse-error (type => 'in body', text => 'body', token => $token);
          if (@{$self->{open_elements}} == 1 or
-             not ($self->{open_elements}->[1]->[1] & BODY_EL)) {
+             not ($self->{open_elements}->[1]->[1] == BODY_EL)) {
            !!!cp ('t342');
            ## Ignore the token
          } else {
-Line 6547 
 sub _tree_construction_main ($) {
+Line 4376 
 sub _tree_construction_main ($) {
          !!!next-token;
          next B;
        } elsif ({
-                 address => 1, blockquote => 1, center => 1, dir => 1,
+                 ## NOTE: Start tags for non-phrasing flow content elements
-                 div => 1, dl => 1, fieldset => 1,
-                 h1 => 1, h2 => 1, h3 => 1, h4 => 1, h5 => 1, h6 => 1,
+                 ## NOTE: The normal one
-                 menu => 1, ol => 1, p => 1, ul => 1,
+                 address => 1, article => 1, aside => 1, blockquote => 1,
+                 center => 1, datagrid => 1, details => 1, dialog => 1,
+                 dir => 1, div => 1, dl => 1, fieldset => 1, figure => 1,
+                 footer => 1, h1 => 1, h2 => 1, h3 => 1, h4 => 1, h5 => 1,
+                 h6 => 1, header => 1, menu => 1, nav => 1, ol => 1, p => 1,
+                 section => 1, ul => 1,
+                 ## NOTE: As normal, but drops leading newline
                  pre => 1, listing => 1,
+                 ## NOTE: As normal, but interacts with the form element pointer
                  form => 1,
                  table => 1,
                  hr => 1,
                 }->{$token->{tag_name}}) {
-Line 6567 
 sub _tree_construction_main ($) {
+Line 4404 
 sub _tree_construction_main ($) {
          ## has a p element in scope
          INSCOPE: for (reverse @{$self->{open_elements}}) {
-           if ($_->[1] & P_EL) {
+           if ($_->[1] == P_EL) {
              !!!cp ('t344');
              !!!back-token; # <form>
              $token = {type => END_TAG_TOKEN, tag_name => 'p',
-Line 6619 
 sub _tree_construction_main ($) {
+Line 4456 
 sub _tree_construction_main ($) {
            !!!next-token;
          }
          next B;
-       } elsif ({li => 1, dt => 1, dd => 1}->{$token->{tag_name}}) {
+       } elsif ($token->{tag_name} eq 'li') {
-         ## has a p element in scope
+         ## NOTE: As normal, but imply </li> when there's another <li> ...
+         ## NOTE: Special, Scope (<li><foo><li> == <li><foo><li/></foo></li>)
+           ## Interpreted as <li><foo/></li><li/> (non-conforming)
+           ## blockquote (O9.27), center (O), dd (Fx3, O, S3.1.2, IE7),
+           ## dt (Fx, O, S, IE), dl (O), fieldset (O, S, IE), form (Fx, O, S),
+           ## hn (O), pre (O), applet (O, S), button (O, S), marquee (Fx, O, S),
+           ## object (Fx)
+           ## Generate non-tree (non-conforming)
+           ## basefont (IE7 (where basefont is non-void)), center (IE),
+           ## form (IE), hn (IE)
+         ## address, div, p (<li><foo><li> == <li><foo/></li><li/>)
+           ## Interpreted as <li><foo><li/></foo></li> (non-conforming)
+           ## div (Fx, S)
+         my $non_optional;
+         my $i = -1;
+         ## 1.
+         for my $node (reverse @{$self->{open_elements}}) {
+           if ($node->[1] == LI_EL) {
+             ## 2. (a) As if </li>
+             {
+               ## If no </li> - not applied
+               #
+               ## Otherwise
+               ## 1. generate implied end tags, except for </li>
+               #
+               ## 2. If current node != "li", parse error
+               if ($non_optional) {
+                 !!!parse-error (type => 'not closed',
+                                 text => $non_optional->[0]->manakai_local_name,
+                                 token => $token);
+                 !!!cp ('t355');
+               } else {
+                 !!!cp ('t356');
+               }
+               ## 3. Pop
+               splice @{$self->{open_elements}}, $i;
+             }
+             last; ## 2. (b) goto 5.
+           } elsif (
+                    ## NOTE: not "formatting" and not "phrasing"
+                    ($node->[1] & SPECIAL_EL or
+                     $node->[1] & SCOPING_EL) and
+                    ## NOTE: "li", "dt", and "dd" are in |SPECIAL_EL|.
+                    (not $node->[1] & ADDRESS_DIV_P_EL)
+                   ) {
+             ## 3.
+             !!!cp ('t357');
+             last; ## goto 5.
+           } elsif ($node->[1] & END_TAG_OPTIONAL_EL) {
+             !!!cp ('t358');
+             #
+           } else {
+             !!!cp ('t359');
+             $non_optional ||= $node;
+             #
+           }
+           ## 4.
+           ## goto 2.
+           $i--;
+         }
+         ## 5. (a) has a |p| element in scope
          INSCOPE: for (reverse @{$self->{open_elements}}) {
-           if ($_->[1] & P_EL) {
+           if ($_->[1] == P_EL) {
              !!!cp ('t353');
+             ## NOTE: |<p><li>|, for example.
              !!!back-token; # <x>
              $token = {type => END_TAG_TOKEN, tag_name => 'p',
                        line => $token->{line}, column => $token->{column}};
-Line 6633 
 sub _tree_construction_main ($) {
+Line 4542 
 sub _tree_construction_main ($) {
              last INSCOPE;
            }
          } # INSCOPE
-         ## Step 1
+         ## 5. (b) insert
+         !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
+         !!!nack ('t359.1');
+         !!!next-token;
+         next B;
+       } elsif ($token->{tag_name} eq 'dt' or
+                $token->{tag_name} eq 'dd') {
+         ## NOTE: As normal, but imply </dt> or </dd> when ...
+         my $non_optional;
          my $i = -1;
-         my $node = $self->{open_elements}->[$i];
-         my $li_or_dtdd = {li => {li => 1},
+         ## 1.
-                           dt => {dt => 1, dd => 1},
+         for my $node (reverse @{$self->{open_elements}}) {
-                           dd => {dt => 1, dd => 1}}->{$token->{tag_name}};
+           if ($node->[1] == DTDD_EL) {
-         LI: {
+             ## 2. (a) As if </li>
-           ## Step 2
+             {
-           if ($li_or_dtdd->{$node->[0]->manakai_local_name}) {
+               ## If no </li> - not applied
-             if ($i != -1) {
+               #
-               !!!cp ('t355');
-               !!!parse-error (type => 'not closed',
+               ## Otherwise
-                               text => $self->{open_elements}->[-1]->[0]
-                                   ->manakai_local_name,
+               ## 1. generate implied end tags, except for </dt> or </dd>
-                               token => $token);
+               #
-             } else {
-               !!!cp ('t356');
+               ## 2. If current node != "dt"|"dd", parse error
+               if ($non_optional) {
+                 !!!parse-error (type => 'not closed',
+                                 text => $non_optional->[0]->manakai_local_name,
+                                 token => $token);
+                 !!!cp ('t355.1');
+               } else {
+                 !!!cp ('t356.1');
+               }
+               ## 3. Pop
+               splice @{$self->{open_elements}}, $i;
              }
-             splice @{$self->{open_elements}}, $i;
-             last LI;
+             last; ## 2. (b) goto 5.
+           } elsif (
+                    ## NOTE: not "formatting" and not "phrasing"
+                    ($node->[1] & SPECIAL_EL or
+                     $node->[1] & SCOPING_EL) and
+                    ## NOTE: "li", "dt", and "dd" are in |SPECIAL_EL|.
+                    (not $node->[1] & ADDRESS_DIV_P_EL)
+                   ) {
+             ## 3.
+             !!!cp ('t357.1');
+             last; ## goto 5.
+           } elsif ($node->[1] & END_TAG_OPTIONAL_EL) {
+             !!!cp ('t358.1');
+             #
            } else {
-             !!!cp ('t357');
+             !!!cp ('t359.1');
-           }
+             $non_optional ||= $node;
+             #
-           ## Step 3
-           if (not ($node->[1] & FORMATTING_EL) and
-               #not $phrasing_category->{$node->[1]} and
-               ($node->[1] & SPECIAL_EL or
-                $node->[1] & SCOPING_EL) and
-               not ($node->[1] & ADDRESS_EL) and
-               not ($node->[1] & DIV_EL)) {
-             !!!cp ('t358');
-             last LI;
            }
+           ## 4.
-           !!!cp ('t359');
+           ## goto 2.
-           ## Step 4
            $i--;
-           $node = $self->{open_elements}->[$i];
+         }
-           redo LI;
-         } # LI
+         ## 5. (a) has a |p| element in scope
+         INSCOPE: for (reverse @{$self->{open_elements}}) {
+           if ($_->[1] == P_EL) {
+             !!!cp ('t353.1');
+             !!!back-token; # <x>
+             $token = {type => END_TAG_TOKEN, tag_name => 'p',
+                       line => $token->{line}, column => $token->{column}};
+             next B;
+           } elsif ($_->[1] & SCOPING_EL) {
+             !!!cp ('t354.1');
+             last INSCOPE;
+           }
+         } # INSCOPE
+         ## 5. (b) insert
          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
-         !!!nack ('t359.1');
+         !!!nack ('t359.2');
          !!!next-token;
          next B;
        } elsif ($token->{tag_name} eq 'plaintext') {
+         ## NOTE: As normal, but effectively ends parsing
          ## has a p element in scope
          INSCOPE: for (reverse @{$self->{open_elements}}) {
-           if ($_->[1] & P_EL) {
+           if ($_->[1] == P_EL) {
              !!!cp ('t367');
              !!!back-token; # <plaintext>
              $token = {type => END_TAG_TOKEN, tag_name => 'p',
-Line 6705 
 sub _tree_construction_main ($) {
+Line 4653 
 sub _tree_construction_main ($) {
        } elsif ($token->{tag_name} eq 'a') {
          AFE: for my $i (reverse 0..$#$active_formatting_elements) {
            my $node = $active_formatting_elements->[$i];
-           if ($node->[1] & A_EL) {
+           if ($node->[1] == A_EL) {
              !!!cp ('t371');
              !!!parse-error (type => 'in a:a', token => $token);
-Line 6749 
 sub _tree_construction_main ($) {
+Line 4697 
 sub _tree_construction_main ($) {
          ## has a |nobr| element in scope
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
            my $node = $self->{open_elements}->[$_];
-           if ($node->[1] & NOBR_EL) {
+           if ($node->[1] == NOBR_EL) {
              !!!cp ('t376');
              !!!parse-error (type => 'in nobr:nobr', token => $token);
              !!!back-token; # <nobr>
-Line 6772 
 sub _tree_construction_main ($) {
+Line 4720 
 sub _tree_construction_main ($) {
          ## has a button element in scope
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
            my $node = $self->{open_elements}->[$_];
-           if ($node->[1] & BUTTON_EL) {
+           if ($node->[1] == BUTTON_EL) {
              !!!cp ('t378');
              !!!parse-error (type => 'in button:button', token => $token);
              !!!back-token; # <button>
-Line 6872 
 sub _tree_construction_main ($) {
+Line 4820 
 sub _tree_construction_main ($) {
            next B;
          }
        } elsif ($token->{tag_name} eq 'textarea') {
-         my $tag_name = $token->{tag_name};
+         ## Step 1
-         my $el;
+         !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
-         !!!create-element ($el, $HTML_NS, $token->{tag_name}, $token->{attributes}, $token);
+         ## Step 2
          ## TODO: $self->{form_element} if defined
+         ## Step 3
+         $self->{ignore_newline} = 1;
+         ## Step 4
+         ## ISSUE: This step is wrong. (r2302 enbugged)
+         ## Step 5
          $self->{content_model} = RCDATA_CONTENT_MODEL;
          delete $self->{escape}; # MUST
-         $insert->($el);
+         ## Step 6-7
+         $self->{insertion_mode} |= IN_CDATA_RCDATA_IM;
-         my $text = '';
          !!!nack ('t392.1');
          !!!next-token;
-         if ($token->{type} == CHARACTER_TOKEN) {
+         next B;
-           $token->{data} =~ s/^\x0A//;
+       } elsif ($token->{tag_name} eq 'optgroup' or
-           unless (length $token->{data}) {
+                $token->{tag_name} eq 'option') {
-             !!!cp ('t392');
+         ## has an |option| element in scope
-             !!!next-token;
+         INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
-           } else {
+           my $node = $self->{open_elements}->[$_];
-             !!!cp ('t393');
+           if ($node->[1] == OPTION_EL) {
+             !!!cp ('t397.1');
+             ## NOTE: As if </option>
+             !!!back-token; # <option> or <optgroup>
+             $token = {type => END_TAG_TOKEN, tag_name => 'option',
+                       line => $token->{line}, column => $token->{column}};
+             next B;
+           } elsif ($node->[1] & SCOPING_EL) {
+             !!!cp ('t397.2');
+             last INSCOPE;
            }
-         } else {
+         } # INSCOPE
-           !!!cp ('t394');
-         }
+         $reconstruct_active_formatting_elements->($insert_to_current);
-         while ($token->{type} == CHARACTER_TOKEN) {
-           !!!cp ('t395');
+         !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
-           $text .= $token->{data};
-           !!!next-token;
+         !!!nack ('t397.3');
-         }
-         if (length $text) {
-           !!!cp ('t396');
-           $el->manakai_append_text ($text);
-         }
-         $self->{content_model} = PCDATA_CONTENT_MODEL;
-         if ($token->{type} == END_TAG_TOKEN and
-             $token->{tag_name} eq $tag_name) {
-           !!!cp ('t397');
-           ## Ignore the token
-         } else {
-           !!!cp ('t398');
-           !!!parse-error (type => 'in RCDATA:#eof', token => $token);
-         }
          !!!next-token;
-         next B;
+         redo B;
        } elsif ($token->{tag_name} eq 'rt' or
                 $token->{tag_name} eq 'rp') {
          ## has a |ruby| element in scope
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
            my $node = $self->{open_elements}->[$_];
-           if ($node->[1] & RUBY_EL) {
+           if ($node->[1] == RUBY_EL) {
              !!!cp ('t398.1');
              ## generate implied end tags
              while ($self->{open_elements}->[-1]->[1] & END_TAG_OPTIONAL_EL) {
                !!!cp ('t398.2');
                pop @{$self->{open_elements}};
              }
-             unless ($self->{open_elements}->[-1]->[1] & RUBY_EL) {
+             unless ($self->{open_elements}->[-1]->[1] == RUBY_EL) {
                !!!cp ('t398.3');
                !!!parse-error (type => 'not closed',
                                text => $self->{open_elements}->[-1]->[0]
                                    ->manakai_local_name,
                                token => $token);
                pop @{$self->{open_elements}}
-                   while not $self->{open_elements}->[-1]->[1] & RUBY_EL;
+                   while not $self->{open_elements}->[-1]->[1] == RUBY_EL;
              }
              last INSCOPE;
            } elsif ($node->[1] & SCOPING_EL) {
-Line 6965 
 sub _tree_construction_main ($) {
+Line 4914 
 sub _tree_construction_main ($) {
          if ($self->{self_closing}) {
            pop @{$self->{open_elements}};
-           !!!ack ('t398.1');
+           !!!ack ('t398.6');
          } else {
-           !!!cp ('t398.2');
+           !!!cp ('t398.7');
            $self->{insertion_mode} |= IN_FOREIGN_CONTENT_IM;
            ## NOTE: |<body><math><mi><svg>| -> "in foreign content" insertion
            ## mode, "in body" (not "in foreign content") secondary insertion
-Line 6978 
 sub _tree_construction_main ($) {
+Line 4927 
 sub _tree_construction_main ($) {
          next B;
        } elsif ({
                  caption => 1, col => 1, colgroup => 1, frame => 1,
-                 frameset => 1, head => 1, option => 1, optgroup => 1,
+                 frameset => 1, head => 1,
                  tbody => 1, td => 1, tfoot => 1, th => 1,
                  thead => 1, tr => 1,
                 }->{$token->{tag_name}}) {
-Line 6989 
 sub _tree_construction_main ($) {
+Line 4938 
 sub _tree_construction_main ($) {
          !!!nack ('t401.1'); ## NOTE: |<col/>| or |<frame/>| here is an error.
          !!!next-token;
          next B;
+       } elsif ($token->{tag_name} eq 'param' or
-         ## ISSUE: An issue on HTML5 new elements in the spec.
+                $token->{tag_name} eq 'source') {
+         !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
+         pop @{$self->{open_elements}};
+         !!!ack ('t398.5');
+         !!!next-token;
+         redo B;
        } else {
          if ($token->{tag_name} eq 'image') {
            !!!cp ('t384');
-Line 7013 
 sub _tree_construction_main ($) {
+Line 4968 
 sub _tree_construction_main ($) {
            !!!nack ('t380.1');
          } elsif ({
                    b => 1, big => 1, em => 1, font => 1, i => 1,
-                   s => 1, small => 1, strile => 1,
+                   s => 1, small => 1, strike => 1,
                    strong => 1, tt => 1, u => 1,
                   }->{$token->{tag_name}}) {
            !!!cp ('t375');
-Line 7026 
 sub _tree_construction_main ($) {
+Line 4981 
 sub _tree_construction_main ($) {
            !!!ack ('t388.2');
          } elsif ({
                    area => 1, basefont => 1, bgsound => 1, br => 1,
-                   embed => 1, img => 1, param => 1, spacer => 1, wbr => 1,
+                   embed => 1, img => 1, spacer => 1, wbr => 1,
-                   #image => 1,
                   }->{$token->{tag_name}}) {
            !!!cp ('t388.1');
            pop @{$self->{open_elements}};
-Line 7058 
 sub _tree_construction_main ($) {
+Line 5012 
 sub _tree_construction_main ($) {
          my $i;
          INSCOPE: {
            for (reverse @{$self->{open_elements}}) {
-             if ($_->[1] & BODY_EL) {
+             if ($_->[1] == BODY_EL) {
                !!!cp ('t405');
                $i = $_;
                last INSCOPE;
-Line 7068 
 sub _tree_construction_main ($) {
+Line 5022 
 sub _tree_construction_main ($) {
              }
            }
-           !!!parse-error (type => 'start tag not allowed',
+           ## NOTE: |<marquee></body>|, |<svg><foreignobject></body>|
+           !!!parse-error (type => 'unmatched end tag',
                            text => $token->{tag_name}, token => $token);
            ## NOTE: Ignore the token.
            !!!next-token;
-Line 7094 
 sub _tree_construction_main ($) {
+Line 5050 
 sub _tree_construction_main ($) {
          ## TODO: Update this code.  It seems that the code below is not
          ## up-to-date, though it has same effect as speced.
          if (@{$self->{open_elements}} > 1 and
-             $self->{open_elements}->[1]->[1] & BODY_EL) {
+             $self->{open_elements}->[1]->[1] == BODY_EL) {
-           ## ISSUE: There is an issue in the spec.
+           unless ($self->{open_elements}->[-1]->[1] == BODY_EL) {
-           unless ($self->{open_elements}->[-1]->[1] & BODY_EL) {
              !!!cp ('t406');
              !!!parse-error (type => 'not closed',
                              text => $self->{open_elements}->[1]->[0]
-Line 7117 
 sub _tree_construction_main ($) {
+Line 5072 
 sub _tree_construction_main ($) {
            next B;
          }
        } elsif ({
-                 address => 1, blockquote => 1, center => 1, dir => 1,
+                 ## NOTE: End tags for non-phrasing flow content elements
-                 div => 1, dl => 1, fieldset => 1, listing => 1,
-                 menu => 1, ol => 1, pre => 1, ul => 1,
+                 ## NOTE: The normal ones
+                 address => 1, article => 1, aside => 1, blockquote => 1,
+                 center => 1, datagrid => 1, details => 1, dialog => 1,
+                 dir => 1, div => 1, dl => 1, fieldset => 1, figure => 1,
+                 footer => 1, header => 1, listing => 1, menu => 1, nav => 1,
+                 ol => 1, pre => 1, section => 1, ul => 1,
+                 ## NOTE: As normal, but ... optional tags
                  dd => 1, dt => 1, li => 1,
                  applet => 1, button => 1, marquee => 1, object => 1,
                 }->{$token->{tag_name}}) {
+         ## NOTE: Code for <li> start tags includes "as if </li>" code.
+         ## Code for <dt> or <dd> start tags includes "as if </dt> or
+         ## </dd>" code.
          ## has an element in scope
          my $i;
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
-Line 7141 
 sub _tree_construction_main ($) {
+Line 5108 
 sub _tree_construction_main ($) {
            !!!cp ('t413');
            !!!parse-error (type => 'unmatched end tag',
                            text => $token->{tag_name}, token => $token);
+           ## NOTE: Ignore the token.
          } else {
            ## Step 1. generate implied end tags
            while ({
-Line 7148 
 sub _tree_construction_main ($) {
+Line 5116 
 sub _tree_construction_main ($) {
                    dd => ($token->{tag_name} ne 'dd'),
                    dt => ($token->{tag_name} ne 'dt'),
                    li => ($token->{tag_name} ne 'li'),
+                   option => 1,
+                   optgroup => 1,
                    p => 1,
                    rt => 1,
                    rp => 1,
-Line 7180 
 sub _tree_construction_main ($) {
+Line 5150 
 sub _tree_construction_main ($) {
          !!!next-token;
          next B;
        } elsif ($token->{tag_name} eq 'form') {
+         ## NOTE: As normal, but interacts with the form element pointer
          undef $self->{form_element};
          ## has an element in scope
          my $i;
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
            my $node = $self->{open_elements}->[$_];
-           if ($node->[1] & FORM_EL) {
+           if ($node->[1] == FORM_EL) {
              !!!cp ('t418');
              $i = $_;
              last INSCOPE;
-Line 7200 
 sub _tree_construction_main ($) {
+Line 5172 
 sub _tree_construction_main ($) {
            !!!cp ('t421');
            !!!parse-error (type => 'unmatched end tag',
                            text => $token->{tag_name}, token => $token);
+           ## NOTE: Ignore the token.
          } else {
            ## Step 1. generate implied end tags
            while ($self->{open_elements}->[-1]->[1] & END_TAG_OPTIONAL_EL) {
-Line 7226 
 sub _tree_construction_main ($) {
+Line 5199 
 sub _tree_construction_main ($) {
          !!!next-token;
          next B;
        } elsif ({
+                 ## NOTE: As normal, except acts as a closer for any ...
                  h1 => 1, h2 => 1, h3 => 1, h4 => 1, h5 => 1, h6 => 1,
                 }->{$token->{tag_name}}) {
          ## has an element in scope
          my $i;
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
            my $node = $self->{open_elements}->[$_];
-           if ($node->[1] & HEADING_EL) {
+           if ($node->[1] == HEADING_EL) {
              !!!cp ('t423');
              $i = $_;
              last INSCOPE;
-Line 7246 
 sub _tree_construction_main ($) {
+Line 5220 
 sub _tree_construction_main ($) {
            !!!cp ('t425.1');
            !!!parse-error (type => 'unmatched end tag',
                            text => $token->{tag_name}, token => $token);
+           ## NOTE: Ignore the token.
          } else {
            ## Step 1. generate implied end tags
            while ($self->{open_elements}->[-1]->[1] & END_TAG_OPTIONAL_EL) {
-Line 7270 
 sub _tree_construction_main ($) {
+Line 5245 
 sub _tree_construction_main ($) {
          !!!next-token;
          next B;
        } elsif ($token->{tag_name} eq 'p') {
+         ## NOTE: As normal, except </p> implies <p> and ...
          ## has an element in scope
+         my $non_optional;
          my $i;
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
            my $node = $self->{open_elements}->[$_];
-           if ($node->[1] & P_EL) {
+           if ($node->[1] == P_EL) {
              !!!cp ('t410.1');
              $i = $_;
              last INSCOPE;
            } elsif ($node->[1] & SCOPING_EL) {
              !!!cp ('t411.1');
              last INSCOPE;
+           } elsif ($node->[1] & END_TAG_OPTIONAL_EL) {
+             ## NOTE: |END_TAG_OPTIONAL_EL| includes "p"
+             !!!cp ('t411.2');
+             #
+           } else {
+             !!!cp ('t411.3');
+             $non_optional ||= $node;
+             #
            }
          } # INSCOPE
          if (defined $i) {
-           if ($self->{open_elements}->[-1]->[0]->manakai_local_name
+           ## 1. Generate implied end tags
-                   ne $token->{tag_name}) {
+           #
+           ## 2. If current node != "p", parse error
+           if ($non_optional) {
              !!!cp ('t412.1');
              !!!parse-error (type => 'not closed',
-                             text => $self->{open_elements}->[-1]->[0]
+                             text => $non_optional->[0]->manakai_local_name,
-                                 ->manakai_local_name,
                              token => $token);
            } else {
              !!!cp ('t414.1');
            }
+           ## 3. Pop
            splice @{$self->{open_elements}}, $i;
          } else {
            !!!cp ('t413.1');
-Line 7315 
 sub _tree_construction_main ($) {
+Line 5304 
 sub _tree_construction_main ($) {
        } elsif ({
                  a => 1,
                  b => 1, big => 1, em => 1, font => 1, i => 1,
-                 nobr => 1, s => 1, small => 1, strile => 1,
+                 nobr => 1, s => 1, small => 1, strike => 1,
                  strong => 1, tt => 1, u => 1,
                 }->{$token->{tag_name}}) {
          !!!cp ('t427');
-Line 7336 
 sub _tree_construction_main ($) {
+Line 5325 
 sub _tree_construction_main ($) {
          ## Ignore the token.
          !!!next-token;
          next B;
-       } elsif ({
-                 caption => 1, col => 1, colgroup => 1, frame => 1,
-                 frameset => 1, head => 1, option => 1, optgroup => 1,
-                 tbody => 1, td => 1, tfoot => 1, th => 1,
-                 thead => 1, tr => 1,
-                 area => 1, basefont => 1, bgsound => 1,
-                 embed => 1, hr => 1, iframe => 1, image => 1,
-                 img => 1, input => 1, isindex => 1, noembed => 1,
-                 noframes => 1, param => 1, select => 1, spacer => 1,
-                 table => 1, textarea => 1, wbr => 1,
-                 noscript => 0, ## TODO: if scripting is enabled
-                }->{$token->{tag_name}}) {
-         !!!cp ('t429');
-         !!!parse-error (type => 'unmatched end tag',
-                         text => $token->{tag_name}, token => $token);
-         ## Ignore the token
-         !!!next-token;
-         next B;
-         ## ISSUE: Issue on HTML5 new elements in spec
        } else {
+         if ($token->{tag_name} eq 'sarcasm') {
+           sleep 0.001; # take a deep breath
+         }
          ## Step 1
          my $node_i = -1;
          my $node = $self->{open_elements}->[$node_i];
          ## Step 2
          S2: {
-           if ($node->[0]->manakai_local_name eq $token->{tag_name}) {
+           my $node_tag_name = $node->[0]->manakai_local_name;
+           $node_tag_name =~ tr/A-Z/a-z/; # for SVG camelCase tag names
+           if ($node_tag_name eq $token->{tag_name}) {
              ## Step 1
              ## generate implied end tags
              while ($self->{open_elements}->[-1]->[1] & END_TAG_OPTIONAL_EL) {
-Line 7377 
 sub _tree_construction_main ($) {
+Line 5351 
 sub _tree_construction_main ($) {
              }
              ## Step 2
-             if ($self->{open_elements}->[-1]->[0]->manakai_local_name
+             my $current_tag_name
-                     ne $token->{tag_name}) {
+                 = $self->{open_elements}->[-1]->[0]->manakai_local_name;
+             $current_tag_name =~ tr/A-Z/a-z/;
+             if ($current_tag_name ne $token->{tag_name}) {
                !!!cp ('t431');
                ## NOTE: <x><y></x>
                !!!parse-error (type => 'not closed',
-Line 7406 
 sub _tree_construction_main ($) {
+Line 5382 
 sub _tree_construction_main ($) {
                ## Ignore the token
                !!!next-token;
                last S2;
-             }
+               ## NOTE: |<span><dd></span>a|: In Safari 3.1.2 and Opera
+               ## 9.27, "a" is a child of <dd> (conforming).  In
+               ## Firefox 3.0.2, "a" is a child of <body>.  In WinIE 7,
+               ## "a" is a child of both <body> and <dd>.
+             }
              !!!cp ('t434');
            }
-Line 7448 
 sub _tree_construction_main ($) {
+Line 5429 
 sub _tree_construction_main ($) {
    ## TODO: script stuffs
  } # _tree_construct_main
- sub set_inner_html ($$$) {
+ sub set_inner_html ($$$$;$) {
    my $class = shift;
    my $node = shift;
-   my $s = \$_[0];
+   #my $s = \$_[0];
    my $onerror = $_[1];
+   my $get_wrapper = $_[2] || sub ($) { return $_[0] };
    ## ISSUE: Should {confident} be true?
-Line 7471 
 sub set_inner_html ($$$) {
+Line 5453 
 sub set_inner_html ($$$) {
      }
      ## Step 3, 4, 5 # MUST
-     $class->parse_string ($$s => $node, $onerror);
+     $class->parse_char_string ($_[0] => $node, $onerror, $get_wrapper);
    } elsif ($nt == 1) {
      ## TODO: If non-html element
      ## NOTE: Most of this code is copied from |parse_string|
+ ## TODO: Support for $get_wrapper
      ## Step 1 # MUST
      my $this_doc = $node->owner_document;
      my $doc = $this_doc->implementation->create_document;
-Line 7488 
 sub set_inner_html ($$$) {
+Line 5472 
 sub set_inner_html ($$$) {
      my $i = 0;
      $p->{line_prev} = $p->{line} = 1;
      $p->{column_prev} = $p->{column} = 0;
-     $p->{set_next_char} = sub {
+     require Whatpm::Charset::DecodeHandle;
+     my $input = Whatpm::Charset::DecodeHandle::CharString->new (\($_[0]));
+     $input = $get_wrapper->($input);
+     $p->{set_nc} = sub {
        my $self = shift;
-       pop @{$self->{prev_char}};
+       my $char = '';
-       unshift @{$self->{prev_char}}, $self->{next_char};
+       if (defined $self->{next_nc}) {
+         $char = $self->{next_nc};
-       $self->{next_char} = -1 and return if $i >= length $$s;
+         delete $self->{next_nc};
-       $self->{next_char} = ord substr $$s, $i++, 1;
+         $self->{nc} = ord $char;
+       } else {
+         $self->{char_buffer} = '';
+         $self->{char_buffer_pos} = 0;
+         my $count = $input->manakai_read_until
+             ($self->{char_buffer}, qr/[^\x00\x0A\x0D]/,
+              $self->{char_buffer_pos});
+         if ($count) {
+           $self->{line_prev} = $self->{line};
+           $self->{column_prev} = $self->{column};
+           $self->{column}++;
+           $self->{nc}
+               = ord substr ($self->{char_buffer},
+                             $self->{char_buffer_pos}++, 1);
+           return;
+         }
+         if ($input->read ($char, 1)) {
+           $self->{nc} = ord $char;
+         } else {
+           $self->{nc} = -1;
+           return;
+         }
+       }
        ($p->{line_prev}, $p->{column_prev}) = ($p->{line}, $p->{column});
        $p->{column}++;
-       if ($self->{next_char} == 0x000A) { # LF
+       if ($self->{nc} == 0x000A) { # LF
          $p->{line}++;
          $p->{column} = 0;
          !!!cp ('i1');
-       } elsif ($self->{next_char} == 0x000D) { # CR
+       } elsif ($self->{nc} == 0x000D) { # CR
-         $i++ if substr ($$s, $i, 1) eq "\x0A";
+ ## TODO: support for abort/streaming
-         $self->{next_char} = 0x000A; # LF # MUST
+         my $next = '';
+         if ($input->read ($next, 1) and $next ne "\x0A") {
+           $self->{next_nc} = $next;
+         }
+         $self->{nc} = 0x000A; # LF # MUST
          $p->{line}++;
          $p->{column} = 0;
          !!!cp ('i2');
-       } elsif ($self->{next_char} > 0x10FFFF) {
+       } elsif ($self->{nc} == 0x0000) { # NULL
-         $self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
-         !!!cp ('i3');
-       } elsif ($self->{next_char} == 0x0000) { # NULL
          !!!cp ('i4');
          !!!parse-error (type => 'NULL');
-         $self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
+         $self->{nc} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
-       } elsif ($self->{next_char} <= 0x0008 or
-                (0x000E <= $self->{next_char} and
-                 $self->{next_char} <= 0x001F) or
-                (0x007F <= $self->{next_char} and
-                 $self->{next_char} <= 0x009F) or
-                (0xD800 <= $self->{next_char} and
-                 $self->{next_char} <= 0xDFFF) or
-                (0xFDD0 <= $self->{next_char} and
-                 $self->{next_char} <= 0xFDDF) or
-                {
-xFFFE => 1, 0xFFFF => 1, 0x1FFFE => 1, 0x1FFFF => 1,
-x2FFFE => 1, 0x2FFFF => 1, 0x3FFFE => 1, 0x3FFFF => 1,
-x4FFFE => 1, 0x4FFFF => 1, 0x5FFFE => 1, 0x5FFFF => 1,
-x6FFFE => 1, 0x6FFFF => 1, 0x7FFFE => 1, 0x7FFFF => 1,
-x8FFFE => 1, 0x8FFFF => 1, 0x9FFFE => 1, 0x9FFFF => 1,
-xAFFFE => 1, 0xAFFFF => 1, 0xBFFFE => 1, 0xBFFFF => 1,
-xCFFFE => 1, 0xCFFFF => 1, 0xDFFFE => 1, 0xDFFFF => 1,
-xEFFFE => 1, 0xEFFFF => 1, 0xFFFFE => 1, 0xFFFFF => 1,
-x10FFFE => 1, 0x10FFFF => 1,
-                }->{$self->{next_char}}) {
-         !!!cp ('i4.1');
-         if ($self->{next_char} < 0x10000) {
-           !!!parse-error (type => 'control char',
-                           text => (sprintf 'U+%04X', $self->{next_char}));
-         } else {
-           !!!parse-error (type => 'control char',
-                           text => (sprintf 'U-%08X', $self->{next_char}));
-         }
        }
      };
-     $p->{prev_char} = [-1, -1, -1];
-     $p->{next_char} = -1;
+     $p->{read_until} = sub {
+       #my ($scalar, $specials_range, $offset) = @_;
+       return 0 if defined $p->{next_nc};
+       my $pattern = qr/[^$_[1]\x00\x0A\x0D]/;
+       my $offset = $_[2] || 0;
+       if ($p->{char_buffer_pos} < length $p->{char_buffer}) {
+         pos ($p->{char_buffer}) = $p->{char_buffer_pos};
+         if ($p->{char_buffer} =~ /\G(?>$pattern)+/) {
+           substr ($_[0], $offset)
+               = substr ($p->{char_buffer}, $-[0], $+[0] - $-[0]);
+           my $count = $+[0] - $-[0];
+           if ($count) {
+             $p->{column} += $count;
+             $p->{char_buffer_pos} += $count;
+             $p->{line_prev} = $p->{line};
+             $p->{column_prev} = $p->{column} - 1;
+             $p->{nc} = -1;
+           }
+           return $count;
+         } else {
+           return 0;
+         }
+       } else {
+         my $count = $input->manakai_read_until ($_[0], $pattern, $_[2]);
+         if ($count) {
+           $p->{column} += $count;
+           $p->{column_prev} += $count;
+           $p->{nc} = -1;
+         }
+         return $count;
+       }
+     }; # $p->{read_until}
      my $ponerror = $onerror || sub {
        my (%opt) = @_;
        my $line = $opt{line};
-Line 7564 
 sub set_inner_html ($$$) {
+Line 5581 
 sub set_inner_html ($$$) {
        $ponerror->(line => $p->{line}, column => $p->{column}, @_);
      };
+     my $char_onerror = sub {
+       my (undef, $type, %opt) = @_;
+       $ponerror->(layer => 'encode',
+                   line => $p->{line}, column => $p->{column} + 1,
+                   %opt, type => $type);
+     }; # $char_onerror
+     $input->onerror ($char_onerror);
      $p->_initialize_tokenizer;
      $p->_initialize_tree_constructor;
-Line 7599 
 sub set_inner_html ($$$) {
+Line 5624 
 sub set_inner_html ($$$) {
      push @{$p->{open_elements}}, [$root, $el_category->{html}];
      undef $p->{head_element};
+     undef $p->{head_element_inserted};
      ## Step 6 # MUST
      $p->_reset_insertion_mode;

 Legend:



Removed from v.1.156
 


changed lines


 
Added in v.1.208
 Legend:



Removed from v.1.156
 


changed lines


 
Added in v.1.208
-Removed from v.1.156
+Added in v.1.208

[email protected]	ViewVC Help
Powered by ViewVC 1.1.24