/[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 1.54 by wakaba,
Sat Aug 11 06:37:12 2007 UTC
+revision 1.62 by wakaba,
Sun Nov 11 04:59:35 2007 UTC
 Line 147 
 sub new ($) {
    $self->{parse_error} = sub {
      #
    };
+   $self->{application_cache_selection} = sub {
+     #
+   };
    return $self;
  } # new
-Line 159 
 sub CDATA_CONTENT_MODEL () { CM_LIMITED_
+Line 162 
 sub CDATA_CONTENT_MODEL () { CM_LIMITED_
  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 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 }
  sub BODY_IMS ()       { 0b10000 }
- sub BODY_TABLE_IMS () { 0b100000 | BODY_IMS }
+ sub BODY_TABLE_IMS () { 0b100000 }
  sub TABLE_IMS ()      { 0b1000000 }
- sub ROW_IMS ()        { 0b10000000 | TABLE_IMS }
+ sub ROW_IMS ()        { 0b10000000 }
  sub BODY_AFTER_IMS () { 0b100000000 }
  sub FRAME_IMS ()      { 0b1000000000 }
-Line 175 
 sub IN_HEAD_NOSCRIPT_IM () { HEAD_IMS |
+Line 219 
 sub IN_HEAD_NOSCRIPT_IM () { HEAD_IMS |
  sub AFTER_HEAD_IM () { HEAD_IMS | 0b10 }
  sub BEFORE_HEAD_IM () { HEAD_IMS | 0b11 }
  sub IN_BODY_IM () { BODY_IMS }
- sub IN_CELL_IM () { BODY_TABLE_IMS | 0b01 }
+ sub IN_CELL_IM () { BODY_IMS | BODY_TABLE_IMS | 0b01 }
- sub IN_CAPTION_IM () { BODY_TABLE_IMS | 0b10 }
+ sub IN_CAPTION_IM () { BODY_IMS | BODY_TABLE_IMS | 0b10 }
- sub IN_ROW_IM () { ROW_IMS | 0b01 }
+ sub IN_ROW_IM () { TABLE_IMS | ROW_IMS | 0b01 }
- sub IN_TABLE_BODY_IM () { ROW_IMS | 0b10 }
+ sub IN_TABLE_BODY_IM () { TABLE_IMS | ROW_IMS | 0b10 }
  sub IN_TABLE_IM () { TABLE_IMS }
  sub AFTER_BODY_IM () { BODY_AFTER_IMS }
  sub IN_FRAMESET_IM () { FRAME_IMS | 0b01 }
-Line 190 
 sub IN_COLUMN_GROUP_IM () { 0b10 }
+Line 234 
 sub IN_COLUMN_GROUP_IM () { 0b10 }
  sub _initialize_tokenizer ($) {
    my $self = shift;
-   $self->{state} = 'data'; # MUST
+   $self->{state} = DATA_STATE; # MUST
    $self->{content_model} = PCDATA_CONTENT_MODEL; # be
    undef $self->{current_token}; # start tag, end tag, comment, or DOCTYPE
    undef $self->{current_attribute};
-Line 204 
 sub _initialize_tokenizer ($) {
+Line 248 
 sub _initialize_tokenizer ($) {
  } # _initialize_tokenizer
  ## A token has:
- ##   ->{type} eq 'DOCTYPE', 'start tag', 'end tag', 'comment',
+ ##   ->{type} == DOCTYPE_TOKEN, START_TAG_TOKEN, END_TAG_TOKEN, COMMENT_TOKEN,
- ##       'character', or 'end-of-file'
+ ##       CHARACTER_TOKEN, or END_OF_FILE_TOKEN
- ##   ->{name} (DOCTYPE, start tag (tag name), end tag (tag name))
+ ##   ->{name} (DOCTYPE_TOKEN)
- ##   ->{public_identifier} (DOCTYPE)
+ ##   ->{tag_name} (START_TAG_TOKEN, END_TAG_TOKEN)
- ##   ->{system_identifier} (DOCTYPE)
+ ##   ->{public_identifier} (DOCTYPE_TOKEN)
- ##   ->{correct} == 1 or 0 (DOCTYPE)
+ ##   ->{system_identifier} (DOCTYPE_TOKEN)
- ##   ->{attributes} isa HASH (start tag, end tag)
+ ##   ->{correct} == 1 or 0 (DOCTYPE_TOKEN)
- ##   ->{data} (comment, character)
+ ##   ->{attributes} isa HASH (START_TAG_TOKEN, END_TAG_TOKEN)
+ ##   ->{data} (COMMENT_TOKEN, CHARACTER_TOKEN)
  ## Emitted token MUST immediately be handled by the tree construction state.
-Line 221 
 sub _initialize_tokenizer ($) {
+Line 266 
 sub _initialize_tokenizer ($) {
  ## has completed loading.  If one has, then it MUST be executed
  ## and removed from the list.
+ ## NOTE: HTML5 "Writing HTML documents" section, applied to
+ ## documents and not to user agents and conformance checkers,
+ ## contains some requirements that are not detected by the
+ ## parsing algorithm:
+ ## - Some requirements on character encoding declarations. ## TODO
+ ## - "Elements MUST NOT contain content that their content model disallows."
+ ##   ... Some are parse error, some are not (will be reported by c.c.).
+ ## - Polytheistic slash SHOULD NOT be used. (Applied only to atheists.) ## TODO
+ ## - Text (in elements, attributes, and comments) SHOULD NOT contain
+ ##   control characters other than space characters. ## TODO: (what is control character? C0, C1 and DEL?  Unicode control character?)
+ ## TODO: HTML5 poses authors two SHOULD-level requirements that cannot
+ ## be detected by the HTML5 parsing algorithm:
+ ## - Text,
  sub _get_next_token ($) {
    my $self = shift;
    if (@{$self->{token}}) {
-Line 228 
 sub _get_next_token ($) {
+Line 288 
 sub _get_next_token ($) {
    }
    A: {
-     if ($self->{state} eq 'data') {
+     if ($self->{state} == DATA_STATE) {
        if ($self->{next_input_character} == 0x0026) { # &
          if ($self->{content_model} & CM_ENTITY) { # PCDATA | RCDATA
-           $self->{state} = 'entity data';
+           $self->{state} = ENTITY_DATA_STATE;
            !!!next-input-character;
            redo A;
          } else {
-Line 253 
 sub _get_next_token ($) {
+Line 313 
 sub _get_next_token ($) {
          if ($self->{content_model} & CM_FULL_MARKUP or # PCDATA
              (($self->{content_model} & CM_LIMITED_MARKUP) and # CDATA | RCDATA
               not $self->{escape})) {
-           $self->{state} = 'tag open';
+           $self->{state} = TAG_OPEN_STATE;
            !!!next-input-character;
            redo A;
          } else {
-Line 270 
 sub _get_next_token ($) {
+Line 330 
 sub _get_next_token ($) {
          #
        } elsif ($self->{next_input_character} == -1) {
-         !!!emit ({type => 'end-of-file'});
+         !!!emit ({type => END_OF_FILE_TOKEN});
          last A; ## TODO: ok?
        }
        # Anything else
-       my $token = {type => 'character',
+       my $token = {type => CHARACTER_TOKEN,
                     data => chr $self->{next_input_character}};
        ## Stay in the data state
        !!!next-input-character;
-Line 282 
 sub _get_next_token ($) {
+Line 342 
 sub _get_next_token ($) {
        !!!emit ($token);
        redo A;
-     } elsif ($self->{state} eq 'entity data') {
+     } elsif ($self->{state} == ENTITY_DATA_STATE) {
        ## (cannot happen in CDATA state)
        my $token = $self->_tokenize_attempt_to_consume_an_entity (0);
-       $self->{state} = 'data';
+       $self->{state} = DATA_STATE;
        # next-input-character is already done
        unless (defined $token) {
-         !!!emit ({type => 'character', data => '&'});
+         !!!emit ({type => CHARACTER_TOKEN, data => '&'});
        } else {
          !!!emit ($token);
        }
        redo A;
-     } elsif ($self->{state} eq 'tag open') {
+     } elsif ($self->{state} == TAG_OPEN_STATE) {
        if ($self->{content_model} & CM_LIMITED_MARKUP) { # RCDATA | CDATA
          if ($self->{next_input_character} == 0x002F) { # /
            !!!next-input-character;
-           $self->{state} = 'close tag open';
+           $self->{state} = CLOSE_TAG_OPEN_STATE;
            redo A;
          } else {
            ## reconsume
-           $self->{state} = 'data';
+           $self->{state} = DATA_STATE;
-           !!!emit ({type => 'character', data => '<'});
+           !!!emit ({type => CHARACTER_TOKEN, data => '<'});
            redo A;
          }
        } elsif ($self->{content_model} & CM_FULL_MARKUP) { # PCDATA
          if ($self->{next_input_character} == 0x0021) { # !
-           $self->{state} = 'markup declaration open';
+           $self->{state} = MARKUP_DECLARATION_OPEN_STATE;
            !!!next-input-character;
            redo A;
          } elsif ($self->{next_input_character} == 0x002F) { # /
-           $self->{state} = 'close tag open';
+           $self->{state} = CLOSE_TAG_OPEN_STATE;
            !!!next-input-character;
            redo A;
          } elsif (0x0041 <= $self->{next_input_character} and
                   $self->{next_input_character} <= 0x005A) { # A..Z
            $self->{current_token}
-             = {type => 'start tag',
+             = {type => START_TAG_TOKEN,
                 tag_name => chr ($self->{next_input_character} + 0x0020)};
-           $self->{state} = 'tag name';
+           $self->{state} = TAG_NAME_STATE;
            !!!next-input-character;
            redo A;
          } elsif (0x0061 <= $self->{next_input_character} and
                   $self->{next_input_character} <= 0x007A) { # a..z
-           $self->{current_token} = {type => 'start tag',
+           $self->{current_token} = {type => START_TAG_TOKEN,
                              tag_name => chr ($self->{next_input_character})};
-           $self->{state} = 'tag name';
+           $self->{state} = TAG_NAME_STATE;
            !!!next-input-character;
            redo A;
          } elsif ($self->{next_input_character} == 0x003E) { # >
            !!!parse-error (type => 'empty start tag');
-           $self->{state} = 'data';
+           $self->{state} = DATA_STATE;
            !!!next-input-character;
-           !!!emit ({type => 'character', data => '<>'});
+           !!!emit ({type => CHARACTER_TOKEN, data => '<>'});
            redo A;
          } elsif ($self->{next_input_character} == 0x003F) { # ?
            !!!parse-error (type => 'pio');
-           $self->{state} = 'bogus comment';
+           $self->{state} = BOGUS_COMMENT_STATE;
            ## $self->{next_input_character} is intentionally left as is
            redo A;
          } else {
            !!!parse-error (type => 'bare stago');
-           $self->{state} = 'data';
+           $self->{state} = DATA_STATE;
            ## reconsume
-           !!!emit ({type => 'character', data => '<'});
+           !!!emit ({type => CHARACTER_TOKEN, data => '<'});
            redo A;
          }
        } else {
          die "$0: $self->{content_model} in tag open";
        }
-     } elsif ($self->{state} eq 'close tag open') {
+     } elsif ($self->{state} == CLOSE_TAG_OPEN_STATE) {
        if ($self->{content_model} & CM_LIMITED_MARKUP) { # RCDATA | CDATA
          if (defined $self->{last_emitted_start_tag_name}) {
            ## NOTE: <http://krijnhoetmer.nl/irc-logs/whatwg/20070626#l-564>
-Line 375 
 sub _get_next_token ($) {
+Line 435 
 sub _get_next_token ($) {
              } else {
                $self->{next_input_character} = shift @next_char; # reconsume
                !!!back-next-input-character (@next_char);
-               $self->{state} = 'data';
+               $self->{state} = DATA_STATE;
-               !!!emit ({type => 'character', data => '</'});
+               !!!emit ({type => CHARACTER_TOKEN, data => '</'});
                redo A;
              }
-Line 394 
 sub _get_next_token ($) {
+Line 454 
 sub _get_next_token ($) {
                    $self->{next_input_character} == -1) {
              $self->{next_input_character} = shift @next_char; # reconsume
              !!!back-next-input-character (@next_char);
-             $self->{state} = 'data';
+             $self->{state} = DATA_STATE;
-             !!!emit ({type => 'character', data => '</'});
+             !!!emit ({type => CHARACTER_TOKEN, data => '</'});
              redo A;
            } else {
              $self->{next_input_character} = shift @next_char;
-Line 405 
 sub _get_next_token ($) {
+Line 465 
 sub _get_next_token ($) {
          } else {
            ## No start tag token has ever been emitted
            # next-input-character is already done
-           $self->{state} = 'data';
+           $self->{state} = DATA_STATE;
-           !!!emit ({type => 'character', data => '</'});
+           !!!emit ({type => CHARACTER_TOKEN, data => '</'});
            redo A;
          }
        }
        if (0x0041 <= $self->{next_input_character} and
            $self->{next_input_character} <= 0x005A) { # A..Z
-         $self->{current_token} = {type => 'end tag',
+         $self->{current_token} = {type => END_TAG_TOKEN,
                            tag_name => chr ($self->{next_input_character} + 0x0020)};
-         $self->{state} = 'tag name';
+         $self->{state} = TAG_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif (0x0061 <= $self->{next_input_character} and
                 $self->{next_input_character} <= 0x007A) { # a..z
-         $self->{current_token} = {type => 'end tag',
+         $self->{current_token} = {type => END_TAG_TOKEN,
                            tag_name => chr ($self->{next_input_character})};
-         $self->{state} = 'tag name';
+         $self->{state} = TAG_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
          !!!parse-error (type => 'empty end tag');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'bare etago');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          # reconsume
-         !!!emit ({type => 'character', data => '</'});
+         !!!emit ({type => CHARACTER_TOKEN, data => '</'});
          redo A;
        } else {
          !!!parse-error (type => 'bogus end tag');
-         $self->{state} = 'bogus comment';
+         $self->{state} = BOGUS_COMMENT_STATE;
          ## $self->{next_input_character} is intentionally left as is
          redo A;
        }
-     } elsif ($self->{state} eq 'tag name') {
+     } elsif ($self->{state} == TAG_NAME_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
            $self->{next_input_character} == 0x000C or # FF
            $self->{next_input_character} == 0x0020) { # SP
-         $self->{state} = 'before attribute name';
+         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 466 
 sub _get_next_token ($) {
+Line 526 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 481 
 sub _get_next_token ($) {
+Line 541 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 493 
 sub _get_next_token ($) {
+Line 553 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          # reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 502 
 sub _get_next_token ($) {
+Line 562 
 sub _get_next_token ($) {
        } elsif ($self->{next_input_character} == 0x002F) { # /
          !!!next-input-character;
          if ($self->{next_input_character} == 0x003E and # >
-             $self->{current_token}->{type} eq 'start tag' and
+             $self->{current_token}->{type} == START_TAG_TOKEN and
              $permitted_slash_tag_name->{$self->{current_token}->{tag_name}}) {
            # permitted slash
            #
          } else {
            !!!parse-error (type => 'nestc');
          }
-         $self->{state} = 'before attribute name';
+         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
          # next-input-character is already done
          redo A;
        } else {
-Line 519 
 sub _get_next_token ($) {
+Line 579 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'before attribute name') {
+     } elsif ($self->{state} == BEFORE_ATTRIBUTE_NAME_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
-Line 529 
 sub _get_next_token ($) {
+Line 589 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 541 
 sub _get_next_token ($) {
+Line 601 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 551 
 sub _get_next_token ($) {
+Line 611 
 sub _get_next_token ($) {
                 $self->{next_input_character} <= 0x005A) { # A..Z
          $self->{current_attribute} = {name => chr ($self->{next_input_character} + 0x0020),
                                value => ''};
-         $self->{state} = 'attribute name';
+         $self->{state} = ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x002F) { # /
          !!!next-input-character;
          if ($self->{next_input_character} == 0x003E and # >
-             $self->{current_token}->{type} eq 'start tag' and
+             $self->{current_token}->{type} == START_TAG_TOKEN and
              $permitted_slash_tag_name->{$self->{current_token}->{tag_name}}) {
            # permitted slash
            #
-Line 569 
 sub _get_next_token ($) {
+Line 629 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 581 
 sub _get_next_token ($) {
+Line 641 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          # reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 590 
 sub _get_next_token ($) {
+Line 650 
 sub _get_next_token ($) {
        } else {
          $self->{current_attribute} = {name => chr ($self->{next_input_character}),
                                value => ''};
-         $self->{state} = 'attribute name';
+         $self->{state} = ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'attribute name') {
+     } 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
-Line 612 
 sub _get_next_token ($) {
+Line 672 
 sub _get_next_token ($) {
            $self->{next_input_character} == 0x000C or # FF
            $self->{next_input_character} == 0x0020) { # SP
          $before_leave->();
-         $self->{state} = 'after attribute name';
+         $self->{state} = AFTER_ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003D) { # =
          $before_leave->();
-         $self->{state} = 'before attribute value';
+         $self->{state} = BEFORE_ATTRIBUTE_VALUE_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
          $before_leave->();
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 634 
 sub _get_next_token ($) {
+Line 694 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 650 
 sub _get_next_token ($) {
+Line 710 
 sub _get_next_token ($) {
          $before_leave->();
          !!!next-input-character;
          if ($self->{next_input_character} == 0x003E and # >
-             $self->{current_token}->{type} eq 'start tag' and
+             $self->{current_token}->{type} == START_TAG_TOKEN and
              $permitted_slash_tag_name->{$self->{current_token}->{tag_name}}) {
            # permitted slash
            #
          } else {
            !!!parse-error (type => 'nestc');
          }
-         $self->{state} = 'before attribute name';
+         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
          # next-input-character is already done
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed tag');
          $before_leave->();
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 675 
 sub _get_next_token ($) {
+Line 735 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          # reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 687 
 sub _get_next_token ($) {
+Line 747 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'after attribute name') {
+     } elsif ($self->{state} == AFTER_ATTRIBUTE_NAME_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
-Line 697 
 sub _get_next_token ($) {
+Line 757 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003D) { # =
-         $self->{state} = 'before attribute value';
+         $self->{state} = BEFORE_ATTRIBUTE_VALUE_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 713 
 sub _get_next_token ($) {
+Line 773 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 723 
 sub _get_next_token ($) {
+Line 783 
 sub _get_next_token ($) {
                 $self->{next_input_character} <= 0x005A) { # A..Z
          $self->{current_attribute} = {name => chr ($self->{next_input_character} + 0x0020),
                                value => ''};
-         $self->{state} = 'attribute name';
+         $self->{state} = ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x002F) { # /
          !!!next-input-character;
          if ($self->{next_input_character} == 0x003E and # >
-             $self->{current_token}->{type} eq 'start tag' and
+             $self->{current_token}->{type} == START_TAG_TOKEN and
              $permitted_slash_tag_name->{$self->{current_token}->{tag_name}}) {
            # permitted slash
            #
-Line 737 
 sub _get_next_token ($) {
+Line 797 
 sub _get_next_token ($) {
            !!!parse-error (type => 'nestc');
            ## TODO: Different error type for <aa / bb> than <aa/>
          }
-         $self->{state} = 'before attribute name';
+         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
          # next-input-character is already done
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 754 
 sub _get_next_token ($) {
+Line 814 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          # reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 763 
 sub _get_next_token ($) {
+Line 823 
 sub _get_next_token ($) {
        } else {
          $self->{current_attribute} = {name => chr ($self->{next_input_character}),
                                value => ''};
-         $self->{state} = 'attribute name';
+         $self->{state} = ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'before attribute value') {
+     } elsif ($self->{state} == BEFORE_ATTRIBUTE_VALUE_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
-Line 777 
 sub _get_next_token ($) {
+Line 837 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x0022) { # "
-         $self->{state} = 'attribute value (double-quoted)';
+         $self->{state} = ATTRIBUTE_VALUE_DOUBLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x0026) { # &
-         $self->{state} = 'attribute value (unquoted)';
+         $self->{state} = ATTRIBUTE_VALUE_UNQUOTED_STATE;
          ## reconsume
          redo A;
        } elsif ($self->{next_input_character} == 0x0027) { # '
-         $self->{state} = 'attribute value (single-quoted)';
+         $self->{state} = ATTRIBUTE_VALUE_SINGLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 801 
 sub _get_next_token ($) {
+Line 861 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 809 
 sub _get_next_token ($) {
+Line 869 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 821 
 sub _get_next_token ($) {
+Line 881 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 829 
 sub _get_next_token ($) {
+Line 889 
 sub _get_next_token ($) {
          redo A;
        } else {
          $self->{current_attribute}->{value} .= chr ($self->{next_input_character});
-         $self->{state} = 'attribute value (unquoted)';
+         $self->{state} = ATTRIBUTE_VALUE_UNQUOTED_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'attribute value (double-quoted)') {
+     } elsif ($self->{state} == ATTRIBUTE_VALUE_DOUBLE_QUOTED_STATE) {
        if ($self->{next_input_character} == 0x0022) { # "
-         $self->{state} = 'before attribute name';
+         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x0026) { # &
-         $self->{last_attribute_value_state} = 'attribute value (double-quoted)';
+         $self->{last_attribute_value_state} = $self->{state};
-         $self->{state} = 'entity in attribute value';
+         $self->{state} = ENTITY_IN_ATTRIBUTE_VALUE_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed attribute value');
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 857 
 sub _get_next_token ($) {
+Line 917 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 869 
 sub _get_next_token ($) {
+Line 929 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'attribute value (single-quoted)') {
+     } elsif ($self->{state} == ATTRIBUTE_VALUE_SINGLE_QUOTED_STATE) {
        if ($self->{next_input_character} == 0x0027) { # '
-         $self->{state} = 'before attribute name';
+         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x0026) { # &
-         $self->{last_attribute_value_state} = 'attribute value (single-quoted)';
+         $self->{last_attribute_value_state} = $self->{state};
-         $self->{state} = 'entity in attribute value';
+         $self->{state} = ENTITY_IN_ATTRIBUTE_VALUE_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed attribute value');
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 893 
 sub _get_next_token ($) {
+Line 953 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 905 
 sub _get_next_token ($) {
+Line 965 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'attribute value (unquoted)') {
+     } elsif ($self->{state} == ATTRIBUTE_VALUE_UNQUOTED_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # HT
            $self->{next_input_character} == 0x000C or # FF
            $self->{next_input_character} == 0x0020) { # SP
-         $self->{state} = 'before attribute name';
+         $self->{state} = BEFORE_ATTRIBUTE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x0026) { # &
-         $self->{last_attribute_value_state} = 'attribute value (unquoted)';
+         $self->{last_attribute_value_state} = $self->{state};
-         $self->{state} = 'entity in attribute value';
+         $self->{state} = ENTITY_IN_ATTRIBUTE_VALUE_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 932 
 sub _get_next_token ($) {
+Line 992 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 940 
 sub _get_next_token ($) {
+Line 1000 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed tag');
-         if ($self->{current_token}->{type} eq 'start tag') {
+         if ($self->{current_token}->{type} == START_TAG_TOKEN) {
            $self->{current_token}->{first_start_tag}
                = not defined $self->{last_emitted_start_tag_name};
            $self->{last_emitted_start_tag_name} = $self->{current_token}->{tag_name};
-         } elsif ($self->{current_token}->{type} eq 'end tag') {
+         } elsif ($self->{current_token}->{type} == END_TAG_TOKEN) {
            $self->{content_model} = PCDATA_CONTENT_MODEL; # MUST
            if ($self->{current_token}->{attributes}) {
              !!!parse-error (type => 'end tag attribute');
-Line 952 
 sub _get_next_token ($) {
+Line 1012 
 sub _get_next_token ($) {
          } else {
            die "$0: $self->{current_token}->{type}: Unknown token type";
          }
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # start tag or end tag
-Line 964 
 sub _get_next_token ($) {
+Line 1024 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'entity in attribute value') {
+     } elsif ($self->{state} == ENTITY_IN_ATTRIBUTE_VALUE_STATE) {
        my $token = $self->_tokenize_attempt_to_consume_an_entity (1);
        unless (defined $token) {
-Line 977 
 sub _get_next_token ($) {
+Line 1037 
 sub _get_next_token ($) {
        $self->{state} = $self->{last_attribute_value_state};
        # next-input-character is already done
        redo A;
-     } elsif ($self->{state} eq 'bogus comment') {
+     } elsif ($self->{state} == BOGUS_COMMENT_STATE) {
        ## (only happen if PCDATA state)
-       my $token = {type => 'comment', data => ''};
+       my $token = {type => COMMENT_TOKEN, data => ''};
        BC: {
          if ($self->{next_input_character} == 0x003E) { # >
-           $self->{state} = 'data';
+           $self->{state} = DATA_STATE;
            !!!next-input-character;
            !!!emit ($token);
            redo A;
          } elsif ($self->{next_input_character} == -1) {
-           $self->{state} = 'data';
+           $self->{state} = DATA_STATE;
            ## reconsume
            !!!emit ($token);
-Line 1003 
 sub _get_next_token ($) {
+Line 1063 
 sub _get_next_token ($) {
            redo BC;
          }
        } # BC
-     } elsif ($self->{state} eq 'markup declaration open') {
+     } elsif ($self->{state} == MARKUP_DECLARATION_OPEN_STATE) {
        ## (only happen if PCDATA state)
        my @next_char;
-Line 1013 
 sub _get_next_token ($) {
+Line 1073 
 sub _get_next_token ($) {
          !!!next-input-character;
          push @next_char, $self->{next_input_character};
          if ($self->{next_input_character} == 0x002D) { # -
-           $self->{current_token} = {type => 'comment', data => ''};
+           $self->{current_token} = {type => COMMENT_TOKEN, data => ''};
-           $self->{state} = 'comment start';
+           $self->{state} = COMMENT_START_STATE;
            !!!next-input-character;
            redo A;
          }
-Line 1045 
 sub _get_next_token ($) {
+Line 1105 
 sub _get_next_token ($) {
                    if ($self->{next_input_character} == 0x0045 or # E
                        $self->{next_input_character} == 0x0065) { # e
                      ## ISSUE: What a stupid code this is!
-                     $self->{state} = 'DOCTYPE';
+                     $self->{state} = DOCTYPE_STATE;
                      !!!next-input-character;
                      redo A;
                    }
-Line 1059 
 sub _get_next_token ($) {
+Line 1119 
 sub _get_next_token ($) {
        !!!parse-error (type => 'bogus comment');
        $self->{next_input_character} = shift @next_char;
        !!!back-next-input-character (@next_char);
-       $self->{state} = 'bogus comment';
+       $self->{state} = BOGUS_COMMENT_STATE;
        redo A;
        ## ISSUE: typos in spec: chacacters, is is a parse error
        ## ISSUE: spec is somewhat unclear on "is the first character that will be in the comment"; what is "that will be in the comment" is what the algorithm defines, isn't it?
-     } elsif ($self->{state} eq 'comment start') {
+     } elsif ($self->{state} == COMMENT_START_STATE) {
        if ($self->{next_input_character} == 0x002D) { # -
-         $self->{state} = 'comment start dash';
+         $self->{state} = COMMENT_START_DASH_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
          !!!parse-error (type => 'bogus comment');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # comment
-Line 1079 
 sub _get_next_token ($) {
+Line 1139 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed comment');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # comment
-Line 1088 
 sub _get_next_token ($) {
+Line 1148 
 sub _get_next_token ($) {
        } else {
          $self->{current_token}->{data} # comment
              .= chr ($self->{next_input_character});
-         $self->{state} = 'comment';
+         $self->{state} = COMMENT_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'comment start dash') {
+     } elsif ($self->{state} == COMMENT_START_DASH_STATE) {
        if ($self->{next_input_character} == 0x002D) { # -
-         $self->{state} = 'comment end';
+         $self->{state} = COMMENT_END_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
          !!!parse-error (type => 'bogus comment');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # comment
-Line 1107 
 sub _get_next_token ($) {
+Line 1167 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed comment');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # comment
-Line 1116 
 sub _get_next_token ($) {
+Line 1176 
 sub _get_next_token ($) {
        } else {
          $self->{current_token}->{data} # comment
              .= '-' . chr ($self->{next_input_character});
-         $self->{state} = 'comment';
+         $self->{state} = COMMENT_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'comment') {
+     } elsif ($self->{state} == COMMENT_STATE) {
        if ($self->{next_input_character} == 0x002D) { # -
-         $self->{state} = 'comment end dash';
+         $self->{state} = COMMENT_END_DASH_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed comment');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # comment
-Line 1139 
 sub _get_next_token ($) {
+Line 1199 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'comment end dash') {
+     } elsif ($self->{state} == COMMENT_END_DASH_STATE) {
        if ($self->{next_input_character} == 0x002D) { # -
-         $self->{state} = 'comment end';
+         $self->{state} = COMMENT_END_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed comment');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # comment
-Line 1154 
 sub _get_next_token ($) {
+Line 1214 
 sub _get_next_token ($) {
          redo A;
        } else {
          $self->{current_token}->{data} .= '-' . chr ($self->{next_input_character}); # comment
-         $self->{state} = 'comment';
+         $self->{state} = COMMENT_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'comment end') {
+     } elsif ($self->{state} == COMMENT_END_STATE) {
        if ($self->{next_input_character} == 0x003E) { # >
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # comment
-Line 1174 
 sub _get_next_token ($) {
+Line 1234 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed comment');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          !!!emit ($self->{current_token}); # comment
-Line 1183 
 sub _get_next_token ($) {
+Line 1243 
 sub _get_next_token ($) {
        } else {
          !!!parse-error (type => 'dash in comment');
          $self->{current_token}->{data} .= '--' . chr ($self->{next_input_character}); # comment
-         $self->{state} = 'comment';
+         $self->{state} = COMMENT_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'DOCTYPE') {
+     } elsif ($self->{state} == DOCTYPE_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
            $self->{next_input_character} == 0x000C or # FF
            $self->{next_input_character} == 0x0020) { # SP
-         $self->{state} = 'before DOCTYPE name';
+         $self->{state} = BEFORE_DOCTYPE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } else {
          !!!parse-error (type => 'no space before DOCTYPE name');
-         $self->{state} = 'before DOCTYPE name';
+         $self->{state} = BEFORE_DOCTYPE_NAME_STATE;
          ## reconsume
          redo A;
        }
-     } elsif ($self->{state} eq 'before DOCTYPE name') {
+     } elsif ($self->{state} == BEFORE_DOCTYPE_NAME_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
-Line 1213 
 sub _get_next_token ($) {
+Line 1273 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
          !!!parse-error (type => 'no DOCTYPE name');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
-         !!!emit ({type => 'DOCTYPE'}); # incorrect
+         !!!emit ({type => DOCTYPE_TOKEN}); # incorrect
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'no DOCTYPE name');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
-         !!!emit ({type => 'DOCTYPE'}); # incorrect
+         !!!emit ({type => DOCTYPE_TOKEN}); # incorrect
          redo A;
        } else {
          $self->{current_token}
-             = {type => 'DOCTYPE',
+             = {type => DOCTYPE_TOKEN,
                 name => chr ($self->{next_input_character}),
                 correct => 1};
  ## ISSUE: "Set the token's name name to the" in the spec
-         $self->{state} = 'DOCTYPE name';
+         $self->{state} = DOCTYPE_NAME_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'DOCTYPE name') {
+     } elsif ($self->{state} == DOCTYPE_NAME_STATE) {
  ## ISSUE: Redundant "First," in the spec.
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
            $self->{next_input_character} == 0x000C or # FF
            $self->{next_input_character} == 0x0020) { # SP
-         $self->{state} = 'after DOCTYPE name';
+         $self->{state} = AFTER_DOCTYPE_NAME_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # DOCTYPE
-Line 1256 
 sub _get_next_token ($) {
+Line 1316 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1270 
 sub _get_next_token ($) {
+Line 1330 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'after DOCTYPE name') {
+     } elsif ($self->{state} == AFTER_DOCTYPE_NAME_STATE) {
        if ($self->{next_input_character} == 0x0009 or # HT
            $self->{next_input_character} == 0x000A or # LF
            $self->{next_input_character} == 0x000B or # VT
-Line 1280 
 sub _get_next_token ($) {
+Line 1340 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # DOCTYPE
-Line 1288 
 sub _get_next_token ($) {
+Line 1348 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1312 
 sub _get_next_token ($) {
+Line 1372 
 sub _get_next_token ($) {
                  !!!next-input-character;
                  if ($self->{next_input_character} == 0x0043 or # C
                      $self->{next_input_character} == 0x0063) { # c
-                   $self->{state} = 'before DOCTYPE public identifier';
+                   $self->{state} = BEFORE_DOCTYPE_PUBLIC_IDENTIFIER_STATE;
                    !!!next-input-character;
                    redo A;
                  }
-Line 1339 
 sub _get_next_token ($) {
+Line 1399 
 sub _get_next_token ($) {
                  !!!next-input-character;
                  if ($self->{next_input_character} == 0x004D or # M
                      $self->{next_input_character} == 0x006D) { # m
-                   $self->{state} = 'before DOCTYPE system identifier';
+                   $self->{state} = BEFORE_DOCTYPE_SYSTEM_IDENTIFIER_STATE;
                    !!!next-input-character;
                    redo A;
                  }
-Line 1355 
 sub _get_next_token ($) {
+Line 1415 
 sub _get_next_token ($) {
        }
        !!!parse-error (type => 'string after DOCTYPE name');
-       $self->{state} = 'bogus DOCTYPE';
+       $self->{state} = BOGUS_DOCTYPE_STATE;
        # next-input-character is already done
        redo A;
-     } elsif ($self->{state} eq 'before DOCTYPE public identifier') {
+     } elsif ($self->{state} == BEFORE_DOCTYPE_PUBLIC_IDENTIFIER_STATE) {
        if ({
 x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
              #0x000D => 1, # HT, LF, VT, FF, SP, CR
-Line 1368 
 sub _get_next_token ($) {
+Line 1428 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} eq 0x0022) { # "
          $self->{current_token}->{public_identifier} = ''; # DOCTYPE
-         $self->{state} = 'DOCTYPE public identifier (double-quoted)';
+         $self->{state} = DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} eq 0x0027) { # '
          $self->{current_token}->{public_identifier} = ''; # DOCTYPE
-         $self->{state} = 'DOCTYPE public identifier (single-quoted)';
+         $self->{state} = DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} eq 0x003E) { # >
          !!!parse-error (type => 'no PUBLIC literal');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          delete $self->{current_token}->{correct};
-Line 1389 
 sub _get_next_token ($) {
+Line 1449 
 sub _get_next_token ($) {
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1398 
 sub _get_next_token ($) {
+Line 1458 
 sub _get_next_token ($) {
          redo A;
        } else {
          !!!parse-error (type => 'string after PUBLIC');
-         $self->{state} = 'bogus DOCTYPE';
+         $self->{state} = BOGUS_DOCTYPE_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'DOCTYPE public identifier (double-quoted)') {
+     } elsif ($self->{state} == DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED_STATE) {
        if ($self->{next_input_character} == 0x0022) { # "
-         $self->{state} = 'after DOCTYPE public identifier';
+         $self->{state} = AFTER_DOCTYPE_PUBLIC_IDENTIFIER_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed PUBLIC literal');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1424 
 sub _get_next_token ($) {
+Line 1484 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'DOCTYPE public identifier (single-quoted)') {
+     } elsif ($self->{state} == DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED_STATE) {
        if ($self->{next_input_character} == 0x0027) { # '
-         $self->{state} = 'after DOCTYPE public identifier';
+         $self->{state} = AFTER_DOCTYPE_PUBLIC_IDENTIFIER_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed PUBLIC literal');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1446 
 sub _get_next_token ($) {
+Line 1506 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'after DOCTYPE public identifier') {
+     } elsif ($self->{state} == AFTER_DOCTYPE_PUBLIC_IDENTIFIER_STATE) {
        if ({
 x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
              #0x000D => 1, # HT, LF, VT, FF, SP, CR
-Line 1456 
 sub _get_next_token ($) {
+Line 1516 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == 0x0022) { # "
          $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = 'DOCTYPE system identifier (double-quoted)';
+         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x0027) { # '
          $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = 'DOCTYPE system identifier (single-quoted)';
+         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # DOCTYPE
-Line 1474 
 sub _get_next_token ($) {
+Line 1534 
 sub _get_next_token ($) {
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1483 
 sub _get_next_token ($) {
+Line 1543 
 sub _get_next_token ($) {
          redo A;
        } else {
          !!!parse-error (type => 'string after PUBLIC literal');
-         $self->{state} = 'bogus DOCTYPE';
+         $self->{state} = BOGUS_DOCTYPE_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'before DOCTYPE system identifier') {
+     } elsif ($self->{state} == BEFORE_DOCTYPE_SYSTEM_IDENTIFIER_STATE) {
        if ({
 x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
              #0x000D => 1, # HT, LF, VT, FF, SP, CR
-Line 1497 
 sub _get_next_token ($) {
+Line 1557 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == 0x0022) { # "
          $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = 'DOCTYPE system identifier (double-quoted)';
+         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x0027) { # '
          $self->{current_token}->{system_identifier} = ''; # DOCTYPE
-         $self->{state} = 'DOCTYPE system identifier (single-quoted)';
+         $self->{state} = DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
          !!!parse-error (type => 'no SYSTEM literal');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          delete $self->{current_token}->{correct};
-Line 1517 
 sub _get_next_token ($) {
+Line 1577 
 sub _get_next_token ($) {
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1526 
 sub _get_next_token ($) {
+Line 1586 
 sub _get_next_token ($) {
          redo A;
        } else {
          !!!parse-error (type => 'string after SYSTEM');
-         $self->{state} = 'bogus DOCTYPE';
+         $self->{state} = BOGUS_DOCTYPE_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'DOCTYPE system identifier (double-quoted)') {
+     } elsif ($self->{state} == DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED_STATE) {
        if ($self->{next_input_character} == 0x0022) { # "
-         $self->{state} = 'after DOCTYPE system identifier';
+         $self->{state} = AFTER_DOCTYPE_SYSTEM_IDENTIFIER_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed SYSTEM literal');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1552 
 sub _get_next_token ($) {
+Line 1612 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'DOCTYPE system identifier (single-quoted)') {
+     } elsif ($self->{state} == DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED_STATE) {
        if ($self->{next_input_character} == 0x0027) { # '
-         $self->{state} = 'after DOCTYPE system identifier';
+         $self->{state} = AFTER_DOCTYPE_SYSTEM_IDENTIFIER_STATE;
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed SYSTEM literal');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1574 
 sub _get_next_token ($) {
+Line 1634 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'after DOCTYPE system identifier') {
+     } elsif ($self->{state} == AFTER_DOCTYPE_SYSTEM_IDENTIFIER_STATE) {
        if ({
 x0009 => 1, 0x000A => 1, 0x000B => 1, 0x000C => 1, 0x0020 => 1,
              #0x000D => 1, # HT, LF, VT, FF, SP, CR
-Line 1583 
 sub _get_next_token ($) {
+Line 1643 
 sub _get_next_token ($) {
          !!!next-input-character;
          redo A;
        } elsif ($self->{next_input_character} == 0x003E) { # >
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          !!!emit ($self->{current_token}); # DOCTYPE
-Line 1592 
 sub _get_next_token ($) {
+Line 1652 
 sub _get_next_token ($) {
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1601 
 sub _get_next_token ($) {
+Line 1661 
 sub _get_next_token ($) {
          redo A;
        } else {
          !!!parse-error (type => 'string after SYSTEM literal');
-         $self->{state} = 'bogus DOCTYPE';
+         $self->{state} = BOGUS_DOCTYPE_STATE;
          !!!next-input-character;
          redo A;
        }
-     } elsif ($self->{state} eq 'bogus DOCTYPE') {
+     } elsif ($self->{state} == BOGUS_DOCTYPE_STATE) {
        if ($self->{next_input_character} == 0x003E) { # >
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          !!!next-input-character;
          delete $self->{current_token}->{correct};
-Line 1616 
 sub _get_next_token ($) {
+Line 1676 
 sub _get_next_token ($) {
          redo A;
        } elsif ($self->{next_input_character} == -1) {
          !!!parse-error (type => 'unclosed DOCTYPE');
-         $self->{state} = 'data';
+         $self->{state} = DATA_STATE;
          ## reconsume
          delete $self->{current_token}->{correct};
-Line 1697 
 sub _tokenize_attempt_to_consume_an_enti
+Line 1757 
 sub _tokenize_attempt_to_consume_an_enti
            $code = $c1_entity_char->{$code};
          }
-         return {type => 'character', data => chr $code};
+         return {type => CHARACTER_TOKEN, data => chr $code};
        } # X
      } elsif (0x0030 <= $self->{next_input_character} and
               $self->{next_input_character} <= 0x0039) { # 0..9
-Line 1732 
 sub _tokenize_attempt_to_consume_an_enti
+Line 1792 
 sub _tokenize_attempt_to_consume_an_enti
          $code = $c1_entity_char->{$code};
        }
-       return {type => 'character', data => chr $code};
+       return {type => CHARACTER_TOKEN, data => chr $code};
      } else {
        !!!parse-error (type => 'bare nero');
        !!!back-next-input-character ($self->{next_input_character});
-Line 1780 
 sub _tokenize_attempt_to_consume_an_enti
+Line 1840 
 sub _tokenize_attempt_to_consume_an_enti
      }
      if ($match > 0) {
-       return {type => 'character', data => $value};
+       return {type => CHARACTER_TOKEN, data => $value};
      } elsif ($match < 0) {
        !!!parse-error (type => 'no refc');
        if ($in_attr and $match < -1) {
-         return {type => 'character', data => '&'.$entity_name};
+         return {type => CHARACTER_TOKEN, data => '&'.$entity_name};
        } else {
-         return {type => 'character', data => $value};
+         return {type => CHARACTER_TOKEN, data => $value};
        }
      } else {
        !!!parse-error (type => 'bare ero');
        ## NOTE: No characters are consumed in the spec.
-       return {type => 'character', data => '&'.$value};
+       return {type => CHARACTER_TOKEN, data => '&'.$value};
      }
    } else {
      ## no characters are consumed
-Line 1847 
 sub _construct_tree ($) {
+Line 1907 
 sub _construct_tree ($) {
  sub _tree_construction_initial ($) {
    my $self = shift;
    INITIAL: {
-     if ($token->{type} eq 'DOCTYPE') {
+     if ($token->{type} == DOCTYPE_TOKEN) {
        ## NOTE: Conformance checkers MAY, instead of reporting "not HTML5"
        ## error, switch to a conformance checking mode for another
        ## language.
-Line 1974 
 sub _tree_construction_initial ($) {
+Line 2034 
 sub _tree_construction_initial ($) {
        !!!next-token;
        return;
      } elsif ({
-               'start tag' => 1,
+               START_TAG_TOKEN, 1,
-               'end tag' => 1,
+               END_TAG_TOKEN, 1,
-               'end-of-file' => 1,
+               END_OF_FILE_TOKEN, 1,
               }->{$token->{type}}) {
        !!!parse-error (type => 'no DOCTYPE');
        $self->{document}->manakai_compat_mode ('quirks');
        ## Go to the root element phase
        ## reprocess
        return;
-     } elsif ($token->{type} eq 'character') {
+     } elsif ($token->{type} == CHARACTER_TOKEN) {
        if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) { # \x0D
          ## Ignore the token
-Line 1999 
 sub _tree_construction_initial ($) {
+Line 2059 
 sub _tree_construction_initial ($) {
        ## Go to the root element phase
        ## reprocess
        return;
-     } elsif ($token->{type} eq 'comment') {
+     } elsif ($token->{type} == COMMENT_TOKEN) {
        my $comment = $self->{document}->create_comment ($token->{data});
        $self->{document}->append_child ($comment);
-Line 2007 
 sub _tree_construction_initial ($) {
+Line 2067 
 sub _tree_construction_initial ($) {
        !!!next-token;
        redo INITIAL;
      } else {
-       die "$0: $token->{type}: Unknown token";
+       die "$0: $token->{type}: Unknown token type";
      }
    } # INITIAL
  } # _tree_construction_initial
-Line 2016 
 sub _tree_construction_root_element ($)
+Line 2076 
 sub _tree_construction_root_element ($)
    my $self = shift;
    B: {
-       if ($token->{type} eq 'DOCTYPE') {
+       if ($token->{type} == DOCTYPE_TOKEN) {
          !!!parse-error (type => 'in html:#DOCTYPE');
          ## Ignore the token
          ## Stay in the phase
          !!!next-token;
          redo B;
-       } elsif ($token->{type} eq 'comment') {
+       } elsif ($token->{type} == COMMENT_TOKEN) {
          my $comment = $self->{document}->create_comment ($token->{data});
          $self->{document}->append_child ($comment);
          ## Stay in the phase
          !!!next-token;
          redo B;
-       } elsif ($token->{type} eq 'character') {
+       } elsif ($token->{type} == CHARACTER_TOKEN) {
          if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) { # \x0D
            ## Ignore the token.
-Line 2038 
 sub _tree_construction_root_element ($)
+Line 2098 
 sub _tree_construction_root_element ($)
              redo B;
            }
          }
+         $self->{application_cache_selection}->(undef);
+         #
+       } elsif ($token->{type} == START_TAG_TOKEN) {
+         if ($token->{tag_name} eq 'html' and
+             $token->{attributes}->{manifest}) { ## ISSUE: Spec spells as "application"
+           $self->{application_cache_selection}
+                ->($token->{attributes}->{manifest}->{value});
+           ## ISSUE: No relative reference resolution?
+         } else {
+           $self->{application_cache_selection}->(undef);
+         }
+         ## ISSUE: There is an issue in the spec
          #
        } elsif ({
-                 'start tag' => 1,
+                 END_TAG_TOKEN, 1,
-                 'end tag' => 1,
+                 END_OF_FILE_TOKEN, 1,
-                 'end-of-file' => 1,
                 }->{$token->{type}}) {
+         $self->{application_cache_selection}->(undef);
          ## ISSUE: There is an issue in the spec
          #
        } else {
-         die "$0: $token->{type}: Unknown token";
+         die "$0: $token->{type}: Unknown token type";
        }
        my $root_element; !!!create-element ($root_element, 'html');
        $self->{document}->append_child ($root_element);
        push @{$self->{open_elements}}, [$root_element, 'html'];
-Line 2230 
 sub _tree_construction_main ($) {
+Line 2307 
 sub _tree_construction_main ($) {
      ## Step 4
      my $text = '';
      !!!next-token;
-     while ($token->{type} eq 'character') { # or until stop tokenizing
+     while ($token->{type} == CHARACTER_TOKEN) { # or until stop tokenizing
        $text .= $token->{data};
        !!!next-token;
      }
-Line 2245 
 sub _tree_construction_main ($) {
+Line 2322 
 sub _tree_construction_main ($) {
      $self->{content_model} = PCDATA_CONTENT_MODEL;
      ## Step 7
-     if ($token->{type} eq 'end tag' and $token->{tag_name} eq $start_tag_name) {
+     if ($token->{type} == END_TAG_TOKEN and $token->{tag_name} eq $start_tag_name) {
        ## Ignore the token
      } elsif ($content_model_flag == CDATA_CONTENT_MODEL) {
        !!!parse-error (type => 'in CDATA:#'.$token->{type});
-Line 2268 
 sub _tree_construction_main ($) {
+Line 2345 
 sub _tree_construction_main ($) {
      my $text = '';
      !!!next-token;
-     while ($token->{type} eq 'character') {
+     while ($token->{type} == CHARACTER_TOKEN) {
        $text .= $token->{data};
        !!!next-token;
      } # stop if non-character token or tokenizer stops tokenising
-Line 2278 
 sub _tree_construction_main ($) {
+Line 2355 
 sub _tree_construction_main ($) {
      $self->{content_model} = PCDATA_CONTENT_MODEL;
-     if ($token->{type} eq 'end tag' and
+     if ($token->{type} == END_TAG_TOKEN and
          $token->{tag_name} eq 'script') {
        ## Ignore the token
      } else {
-Line 2524 
 sub _tree_construction_main ($) {
+Line 2601 
 sub _tree_construction_main ($) {
    my $insert;
    B: {
-     if ($token->{type} eq 'DOCTYPE') {
+     if ($token->{type} == DOCTYPE_TOKEN) {
        !!!parse-error (type => 'DOCTYPE in the middle');
        ## Ignore the token
        ## Stay in the phase
        !!!next-token;
        redo B;
-     } elsif ($token->{type} eq 'end-of-file') {
+     } elsif ($token->{type} == END_OF_FILE_TOKEN) {
-       if ($self->{insertion_mode} == AFTER_HTML_BODY_IM or
+       if ($self->{insertion_mode} & AFTER_HTML_IMS) {
-           $self->{insertion_mode} == AFTER_HTML_FRAMESET_IM) {
          #
        } else {
          ## Generate implied end tags
-Line 2541 
 sub _tree_construction_main ($) {
+Line 2617 
 sub _tree_construction_main ($) {
               tbody => 1, tfoot=> 1, thead => 1,
              }->{$self->{open_elements}->[-1]->[1]}) {
            !!!back-token;
-           $token = {type => 'end tag', tag_name => $self->{open_elements}->[-1]->[1]};
+           $token = {type => END_TAG_TOKEN, tag_name => $self->{open_elements}->[-1]->[1]};
            redo B;
          }
-Line 2559 
 sub _tree_construction_main ($) {
+Line 2635 
 sub _tree_construction_main ($) {
        ## Stop parsing
        last B;
-     } elsif ($token->{type} eq 'start tag' and
+     } elsif ($token->{type} == START_TAG_TOKEN and
               $token->{tag_name} eq 'html') {
        if ($self->{insertion_mode} == AFTER_HTML_BODY_IM) {
          ## Turn into the main phase
-Line 2586 
 sub _tree_construction_main ($) {
+Line 2662 
 sub _tree_construction_main ($) {
        }
        !!!next-token;
        redo B;
-     } elsif ($token->{type} eq 'comment') {
+     } elsif ($token->{type} == COMMENT_TOKEN) {
        my $comment = $self->{document}->create_comment ($token->{data});
-       if ($self->{insertion_mode} == AFTER_HTML_BODY_IM or
+       if ($self->{insertion_mode} & AFTER_HTML_IMS) {
-           $self->{insertion_mode} == AFTER_HTML_FRAMESET_IM) {
          $self->{document}->append_child ($comment);
        } elsif ($self->{insertion_mode} == AFTER_BODY_IM) {
          $self->{open_elements}->[0]->[0]->append_child ($comment);
-Line 2598 
 sub _tree_construction_main ($) {
+Line 2673 
 sub _tree_construction_main ($) {
        }
        !!!next-token;
        redo B;
-     } elsif ($self->{insertion_mode} == IN_HEAD_IM or
+     } elsif ($self->{insertion_mode} & HEAD_IMS) {
-              $self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM or
+       if ($token->{type} == CHARACTER_TOKEN) {
-              $self->{insertion_mode} == AFTER_HEAD_IM or
-              $self->{insertion_mode} == BEFORE_HEAD_IM) {
-       if ($token->{type} eq 'character') {
          if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
            $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
            unless (length $token->{data}) {
-Line 2643 
 sub _tree_construction_main ($) {
+Line 2715 
 sub _tree_construction_main ($) {
              $self->{insertion_mode} = IN_BODY_IM;
              ## reprocess
              redo B;
-           } elsif ($token->{type} eq 'start tag') {
+           } elsif ($token->{type} == START_TAG_TOKEN) {
              if ($token->{tag_name} eq 'head') {
                if ($self->{insertion_mode} == BEFORE_HEAD_IM) {
                  !!!create-element ($self->{head_element}, $token->{tag_name}, $token->{attributes});
-Line 2856 
 sub _tree_construction_main ($) {
+Line 2928 
 sub _tree_construction_main ($) {
              $self->{insertion_mode} = IN_BODY_IM;
              ## reprocess
              redo B;
-           } elsif ($token->{type} eq 'end tag') {
+           } elsif ($token->{type} == END_TAG_TOKEN) {
              if ($token->{tag_name} eq 'head') {
                if ($self->{insertion_mode} == BEFORE_HEAD_IM) {
                  ## As if <head>
-Line 2978 
 sub _tree_construction_main ($) {
+Line 3050 
 sub _tree_construction_main ($) {
            }
            ## ISSUE: An issue in the spec.
-     } elsif ($self->{insertion_mode} == IN_BODY_IM or
+     } elsif ($self->{insertion_mode} & BODY_IMS) {
-              $self->{insertion_mode} == IN_CELL_IM or
+           if ($token->{type} == CHARACTER_TOKEN) {
-              $self->{insertion_mode} == IN_CAPTION_IM) {
-           if ($token->{type} eq 'character') {
              ## NOTE: There is a code clone of "character in body".
              $reconstruct_active_formatting_elements->($insert_to_current);
-Line 2989 
 sub _tree_construction_main ($) {
+Line 3059 
 sub _tree_construction_main ($) {
              !!!next-token;
              redo B;
-           } elsif ($token->{type} eq 'start tag') {
+           } elsif ($token->{type} == START_TAG_TOKEN) {
              if ({
                   caption => 1, col => 1, colgroup => 1, tbody => 1,
                   td => 1, tfoot => 1, th => 1, thead => 1, tr => 1,
-Line 3017 
 sub _tree_construction_main ($) {
+Line 3087 
 sub _tree_construction_main ($) {
                  ## Close the cell
                  !!!back-token; # <?>
-                 $token = {type => 'end tag', tag_name => $tn};
+                 $token = {type => END_TAG_TOKEN, tag_name => $tn};
                  redo B;
                } elsif ($self->{insertion_mode} == IN_CAPTION_IM) {
                  !!!parse-error (type => 'not closed:caption');
-Line 3050 
 sub _tree_construction_main ($) {
+Line 3120 
 sub _tree_construction_main ($) {
                       tbody => 1, tfoot=> 1, thead => 1,
                      }->{$self->{open_elements}->[-1]->[1]}) {
                    !!!back-token; # <?>
-                   $token = {type => 'end tag', tag_name => 'caption'};
+                   $token = {type => END_TAG_TOKEN, tag_name => 'caption'};
                    !!!back-token;
-                   $token = {type => 'end tag',
+                   $token = {type => END_TAG_TOKEN,
                              tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                    redo B;
                  }
-Line 3075 
 sub _tree_construction_main ($) {
+Line 3145 
 sub _tree_construction_main ($) {
              } else {
                #
              }
-           } elsif ($token->{type} eq 'end tag') {
+           } elsif ($token->{type} == END_TAG_TOKEN) {
              if ($token->{tag_name} eq 'td' or $token->{tag_name} eq 'th') {
                if ($self->{insertion_mode} == IN_CELL_IM) {
                  ## have an element in table scope
-Line 3107 
 sub _tree_construction_main ($) {
+Line 3177 
 sub _tree_construction_main ($) {
                       tbody => 1, tfoot=> 1, thead => 1,
                      }->{$self->{open_elements}->[-1]->[1]}) {
                    !!!back-token;
-                   $token = {type => 'end tag',
+                   $token = {type => END_TAG_TOKEN,
                              tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                    redo B;
                  }
-Line 3161 
 sub _tree_construction_main ($) {
+Line 3231 
 sub _tree_construction_main ($) {
                       tbody => 1, tfoot=> 1, thead => 1,
                      }->{$self->{open_elements}->[-1]->[1]}) {
                    !!!back-token;
-                   $token = {type => 'end tag',
+                   $token = {type => END_TAG_TOKEN,
                              tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                    redo B;
                  }
-Line 3218 
 sub _tree_construction_main ($) {
+Line 3288 
 sub _tree_construction_main ($) {
                ## Close the cell
                !!!back-token; # </?>
-               $token = {type => 'end tag', tag_name => $tn};
+               $token = {type => END_TAG_TOKEN, tag_name => $tn};
                redo B;
              } elsif ($token->{tag_name} eq 'table' and
                       $self->{insertion_mode} == IN_CAPTION_IM) {
-Line 3252 
 sub _tree_construction_main ($) {
+Line 3322 
 sub _tree_construction_main ($) {
                     tbody => 1, tfoot=> 1, thead => 1,
                    }->{$self->{open_elements}->[-1]->[1]}) {
                  !!!back-token; # </table>
-                 $token = {type => 'end tag', tag_name => 'caption'};
+                 $token = {type => END_TAG_TOKEN, tag_name => 'caption'};
                  !!!back-token;
-                 $token = {type => 'end tag',
+                 $token = {type => END_TAG_TOKEN,
                            tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                  redo B;
                }
-Line 3274 
 sub _tree_construction_main ($) {
+Line 3344 
 sub _tree_construction_main ($) {
              } elsif ({
                        body => 1, col => 1, colgroup => 1, html => 1,
                       }->{$token->{tag_name}}) {
-               if ($self->{insertion_mode} == IN_CELL_IM or
+               if ($self->{insertion_mode} & BODY_TABLE_IMS) {
-                   $self->{insertion_mode} == IN_CAPTION_IM) {
                  !!!parse-error (type => 'unmatched end tag:'.$token->{tag_name});
                  ## Ignore the token
                  !!!next-token;
-Line 3301 
 sub _tree_construction_main ($) {
+Line 3370 
 sub _tree_construction_main ($) {
        $insert = $insert_to_current;
        #
-     } elsif ($self->{insertion_mode} == IN_ROW_IM or
+     } elsif ($self->{insertion_mode} & TABLE_IMS) {
-              $self->{insertion_mode} == IN_TABLE_BODY_IM or
+       if ($token->{type} == CHARACTER_TOKEN) {
-              $self->{insertion_mode} == IN_TABLE_IM) {
-           if ($token->{type} eq 'character') {
-             ## NOTE: There are "character in table" code clones.
              if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
                $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
-Line 3362 
 sub _tree_construction_main ($) {
+Line 3428 
 sub _tree_construction_main ($) {
              !!!next-token;
              redo B;
-           } elsif ($token->{type} eq 'start tag') {
+       } elsif ($token->{type} == START_TAG_TOKEN) {
              if ({
                   tr => ($self->{insertion_mode} != IN_ROW_IM),
                   th => 1, td => 1,
-Line 3548 
 sub _tree_construction_main ($) {
+Line 3614 
 sub _tree_construction_main ($) {
                  die "$0: in table: <>: $token->{tag_name}";
                }
              } elsif ($token->{tag_name} eq 'table') {
-               ## NOTE: There are code clones for this "table in table"
                !!!parse-error (type => 'not closed:'.$self->{open_elements}->[-1]->[1]);
                ## As if </table>
-Line 3579 
 sub _tree_construction_main ($) {
+Line 3644 
 sub _tree_construction_main ($) {
                     tbody => 1, tfoot=> 1, thead => 1,
                    }->{$self->{open_elements}->[-1]->[1]}) {
                  !!!back-token; # <table>
-                 $token = {type => 'end tag', tag_name => 'table'};
+                 $token = {type => END_TAG_TOKEN, tag_name => 'table'};
                  !!!back-token;
-                 $token = {type => 'end tag',
+                 $token = {type => END_TAG_TOKEN,
                            tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                  redo B;
                }
-Line 3596 
 sub _tree_construction_main ($) {
+Line 3661 
 sub _tree_construction_main ($) {
                ## reprocess
                redo B;
-             } else {
+         } else {
-               #
+           !!!parse-error (type => 'in table:'.$token->{tag_name});
-             }
-           } elsif ($token->{type} eq 'end tag') {
+           $insert = $insert_to_foster;
+           #
+         }
+       } elsif ($token->{type} == END_TAG_TOKEN) {
              if ($token->{tag_name} eq 'tr' and
                  $self->{insertion_mode} == IN_ROW_IM) {
                ## have an element in table scope
-Line 3740 
 sub _tree_construction_main ($) {
+Line 3808 
 sub _tree_construction_main ($) {
                     tbody => 1, tfoot=> 1, thead => 1,
                    }->{$self->{open_elements}->[-1]->[1]}) {
                  !!!back-token;
-                 $token = {type => 'end tag',
+                 $token = {type => END_TAG_TOKEN,
                            tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                  redo B;
                }
-Line 3758 
 sub _tree_construction_main ($) {
+Line 3826 
 sub _tree_construction_main ($) {
              } elsif ({
                        tbody => 1, tfoot => 1, thead => 1,
                       }->{$token->{tag_name}} and
-                      ($self->{insertion_mode} == IN_ROW_IM or
+                      $self->{insertion_mode} & ROW_IMS) {
-                       $self->{insertion_mode} == IN_TABLE_BODY_IM)) {
                if ($self->{insertion_mode} == IN_ROW_IM) {
                  ## have an element in table scope
                  my $i;
-Line 3857 
 sub _tree_construction_main ($) {
+Line 3924 
 sub _tree_construction_main ($) {
                ## Ignore the token
                !!!next-token;
                redo B;
-             } else {
+         } else {
-               #
+           !!!parse-error (type => 'in table:/'.$token->{tag_name});
-             }
-           } else {
-             die "$0: $token->{type}: Unknown token type";
-           }
-       !!!parse-error (type => 'in table:'.$token->{tag_name});
-       $insert = $insert_to_foster;
+           $insert = $insert_to_foster;
-       #
+           #
+         }
+       } else {
+         die "$0: $token->{type}: Unknown token type";
+       }
      } elsif ($self->{insertion_mode} == IN_COLUMN_GROUP_IM) {
-           if ($token->{type} eq 'character') {
+           if ($token->{type} == CHARACTER_TOKEN) {
              if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
                $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
                unless (length $token->{data}) {
-Line 3879 
 sub _tree_construction_main ($) {
+Line 3944 
 sub _tree_construction_main ($) {
              }
              #
-           } elsif ($token->{type} eq 'start tag') {
+           } elsif ($token->{type} == START_TAG_TOKEN) {
              if ($token->{tag_name} eq 'col') {
                !!!insert-element ($token->{tag_name}, $token->{attributes});
                pop @{$self->{open_elements}};
-Line 3888 
 sub _tree_construction_main ($) {
+Line 3953 
 sub _tree_construction_main ($) {
              } else {
                #
              }
-           } elsif ($token->{type} eq 'end tag') {
+           } elsif ($token->{type} == END_TAG_TOKEN) {
              if ($token->{tag_name} eq 'colgroup') {
                if ($self->{open_elements}->[-1]->[1] eq 'html') {
                  !!!parse-error (type => 'unmatched end tag:colgroup');
-Line 3926 
 sub _tree_construction_main ($) {
+Line 3991 
 sub _tree_construction_main ($) {
              redo B;
            }
      } elsif ($self->{insertion_mode} == IN_SELECT_IM) {
-           if ($token->{type} eq 'character') {
+       if ($token->{type} == CHARACTER_TOKEN) {
-             $self->{open_elements}->[-1]->[0]->manakai_append_text ($token->{data});
+         $self->{open_elements}->[-1]->[0]->manakai_append_text ($token->{data});
-             !!!next-token;
+         !!!next-token;
-             redo B;
+         redo B;
-           } elsif ($token->{type} eq 'start tag') {
+       } elsif ($token->{type} == START_TAG_TOKEN) {
              if ($token->{tag_name} eq 'option') {
                if ($self->{open_elements}->[-1]->[1] eq 'option') {
                  ## As if </option>
-Line 3983 
 sub _tree_construction_main ($) {
+Line 4048 
 sub _tree_construction_main ($) {
                !!!next-token;
                redo B;
-             } else {
+         } else {
-               #
+           !!!parse-error (type => 'in select:'.$token->{tag_name});
-             }
+           ## Ignore the token
-           } elsif ($token->{type} eq 'end tag') {
+           !!!next-token;
+           redo B;
+         }
+       } elsif ($token->{type} == END_TAG_TOKEN) {
              if ($token->{tag_name} eq 'optgroup') {
                if ($self->{open_elements}->[-1]->[1] eq 'option' and
                    $self->{open_elements}->[-2]->[1] eq 'optgroup') {
-Line 4088 
 sub _tree_construction_main ($) {
+Line 4156 
 sub _tree_construction_main ($) {
                ## reprocess
                redo B;
-             } else {
+         } else {
-               #
+           !!!parse-error (type => 'in select:/'.$token->{tag_name});
-             }
-           } else {
-             #
-           }
-           !!!parse-error (type => 'in select:'.$token->{tag_name});
            ## Ignore the token
            !!!next-token;
            redo B;
-     } elsif ($self->{insertion_mode} == AFTER_BODY_IM or
+         }
-              $self->{insertion_mode} == AFTER_HTML_BODY_IM) {
+       } else {
-       if ($token->{type} eq 'character') {
+         die "$0: $token->{type}: Unknown token type";
+       }
+     } elsif ($self->{insertion_mode} & BODY_AFTER_IMS) {
+       if ($token->{type} == CHARACTER_TOKEN) {
          if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
            my $data = $1;
            ## As if in body
-Line 4127 
 sub _tree_construction_main ($) {
+Line 4192 
 sub _tree_construction_main ($) {
          $self->{insertion_mode} = IN_BODY_IM;
          ## reprocess
          redo B;
-       } elsif ($token->{type} eq 'start tag') {
+       } elsif ($token->{type} == START_TAG_TOKEN) {
          if ($self->{insertion_mode} == AFTER_HTML_BODY_IM) {
            !!!parse-error (type => 'after html:'.$token->{tag_name});
-Line 4140 
 sub _tree_construction_main ($) {
+Line 4205 
 sub _tree_construction_main ($) {
          $self->{insertion_mode} = IN_BODY_IM;
          ## reprocess
          redo B;
-       } elsif ($token->{type} eq 'end tag') {
+       } elsif ($token->{type} == END_TAG_TOKEN) {
          if ($self->{insertion_mode} == AFTER_HTML_BODY_IM) {
            !!!parse-error (type => 'after html:/'.$token->{tag_name});
-Line 4170 
 sub _tree_construction_main ($) {
+Line 4235 
 sub _tree_construction_main ($) {
        } else {
          die "$0: $token->{type}: Unknown token type";
        }
-     } elsif ($self->{insertion_mode} == IN_FRAMESET_IM or
+     } elsif ($self->{insertion_mode} & FRAME_IMS) {
-              $self->{insertion_mode} == AFTER_FRAMESET_IM or
+       if ($token->{type} == CHARACTER_TOKEN) {
-              $self->{insertion_mode} == AFTER_HTML_FRAMESET_IM) {
-       if ($token->{type} eq 'character') {
          if ($token->{data} =~ s/^([\x09\x0A\x0B\x0C\x20]+)//) {
            $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);
-Line 4206 
 sub _tree_construction_main ($) {
+Line 4269 
 sub _tree_construction_main ($) {
          }
          die qq[$0: Character "$token->{data}"];
-       } elsif ($token->{type} eq 'start tag') {
+       } elsif ($token->{type} == START_TAG_TOKEN) {
          if ($self->{insertion_mode} == AFTER_HTML_FRAMESET_IM) {
            !!!parse-error (type => 'after html:'.$token->{tag_name});
-Line 4239 
 sub _tree_construction_main ($) {
+Line 4302 
 sub _tree_construction_main ($) {
            !!!next-token;
            redo B;
          }
-       } elsif ($token->{type} eq 'end tag') {
+       } elsif ($token->{type} == END_TAG_TOKEN) {
          if ($self->{insertion_mode} == AFTER_HTML_FRAMESET_IM) {
            !!!parse-error (type => 'after html:/'.$token->{tag_name});
-Line 4289 
 sub _tree_construction_main ($) {
+Line 4352 
 sub _tree_construction_main ($) {
      }
      ## "in body" insertion mode
-     if ($token->{type} eq 'start tag') {
+     if ($token->{type} == START_TAG_TOKEN) {
        if ($token->{tag_name} eq 'script') {
          ## NOTE: This is an "as if in head" code clone
          $script_start_tag->($insert);
-Line 4369 
 sub _tree_construction_main ($) {
+Line 4432 
 sub _tree_construction_main ($) {
          INSCOPE: for (reverse @{$self->{open_elements}}) {
            if ($_->[1] eq 'p') {
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'p'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'p'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4382 
 sub _tree_construction_main ($) {
+Line 4445 
 sub _tree_construction_main ($) {
          !!!insert-element-t ($token->{tag_name}, $token->{attributes});
          if ($token->{tag_name} eq 'pre') {
            !!!next-token;
-           if ($token->{type} eq 'character') {
+           if ($token->{type} == CHARACTER_TOKEN) {
              $token->{data} =~ s/^\x0A//;
              unless (length $token->{data}) {
                !!!next-token;
-Line 4403 
 sub _tree_construction_main ($) {
+Line 4466 
 sub _tree_construction_main ($) {
            INSCOPE: for (reverse @{$self->{open_elements}}) {
              if ($_->[1] eq 'p') {
                !!!back-token;
-               $token = {type => 'end tag', tag_name => 'p'};
+               $token = {type => END_TAG_TOKEN, tag_name => 'p'};
                redo B;
              } elsif ({
                        table => 1, caption => 1, td => 1, th => 1,
-Line 4423 
 sub _tree_construction_main ($) {
+Line 4486 
 sub _tree_construction_main ($) {
          INSCOPE: for (reverse @{$self->{open_elements}}) {
            if ($_->[1] eq 'p') {
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'p'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'p'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4470 
 sub _tree_construction_main ($) {
+Line 4533 
 sub _tree_construction_main ($) {
          INSCOPE: for (reverse @{$self->{open_elements}}) {
            if ($_->[1] eq 'p') {
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'p'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'p'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4517 
 sub _tree_construction_main ($) {
+Line 4580 
 sub _tree_construction_main ($) {
          INSCOPE: for (reverse @{$self->{open_elements}}) {
            if ($_->[1] eq 'p') {
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'p'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'p'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4541 
 sub _tree_construction_main ($) {
+Line 4604 
 sub _tree_construction_main ($) {
            my $node = $self->{open_elements}->[$_];
            if ($node->[1] eq 'p') {
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'p'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'p'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4585 
 sub _tree_construction_main ($) {
+Line 4648 
 sub _tree_construction_main ($) {
              !!!parse-error (type => 'in a:a');
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'a'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'a'};
              $formatting_end_tag->($token->{tag_name});
              AFE2: for (reverse 0..$#$active_formatting_elements) {
-Line 4632 
 sub _tree_construction_main ($) {
+Line 4695 
 sub _tree_construction_main ($) {
          INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
            my $node = $self->{open_elements}->[$_];
            if ($node->[1] eq 'nobr') {
-             !!!parse-error (type => 'not closed:nobr');
+             !!!parse-error (type => 'in nobr:nobr');
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'nobr'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'nobr'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4656 
 sub _tree_construction_main ($) {
+Line 4719 
 sub _tree_construction_main ($) {
            if ($node->[1] eq 'button') {
              !!!parse-error (type => 'in button:button');
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'button'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'button'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4691 
 sub _tree_construction_main ($) {
+Line 4754 
 sub _tree_construction_main ($) {
          INSCOPE: for (reverse @{$self->{open_elements}}) {
            if ($_->[1] eq 'p') {
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'p'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'p'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4730 
 sub _tree_construction_main ($) {
+Line 4793 
 sub _tree_construction_main ($) {
          INSCOPE: for (reverse @{$self->{open_elements}}) {
            if ($_->[1] eq 'p') {
              !!!back-token;
-             $token = {type => 'end tag', tag_name => 'p'};
+             $token = {type => END_TAG_TOKEN, tag_name => 'p'};
              redo B;
            } elsif ({
                      table => 1, caption => 1, td => 1, th => 1,
-Line 4770 
 sub _tree_construction_main ($) {
+Line 4833 
 sub _tree_construction_main ($) {
            delete $at->{action};
            delete $at->{prompt};
            my @tokens = (
-                         {type => 'start tag', tag_name => 'form',
+                         {type => START_TAG_TOKEN, tag_name => 'form',
                           attributes => $form_attrs},
-                         {type => 'start tag', tag_name => 'hr'},
+                         {type => START_TAG_TOKEN, tag_name => 'hr'},
-                         {type => 'start tag', tag_name => 'p'},
+                         {type => START_TAG_TOKEN, tag_name => 'p'},
-                         {type => 'start tag', tag_name => 'label'},
+                         {type => START_TAG_TOKEN, tag_name => 'label'},
                         );
            if ($prompt_attr) {
-             push @tokens, {type => 'character', data => $prompt_attr->{value}};
+             push @tokens, {type => CHARACTER_TOKEN, data => $prompt_attr->{value}};
            } else {
-             push @tokens, {type => 'character',
+             push @tokens, {type => CHARACTER_TOKEN,
                             data => 'This is a searchable index. Insert your search keywords here: '}; # SHOULD
              ## TODO: make this configurable
            }
            push @tokens,
-                         {type => 'start tag', tag_name => 'input', attributes => $at},
+                         {type => START_TAG_TOKEN, tag_name => 'input', attributes => $at},
-                         #{type => 'character', data => ''}, # SHOULD
+                         #{type => CHARACTER_TOKEN, data => ''}, # SHOULD
-                         {type => 'end tag', tag_name => 'label'},
+                         {type => END_TAG_TOKEN, tag_name => 'label'},
-                         {type => 'end tag', tag_name => 'p'},
+                         {type => END_TAG_TOKEN, tag_name => 'p'},
-                         {type => 'start tag', tag_name => 'hr'},
+                         {type => START_TAG_TOKEN, tag_name => 'hr'},
-                         {type => 'end tag', tag_name => 'form'};
+                         {type => END_TAG_TOKEN, tag_name => 'form'};
            $token = shift @tokens;
            !!!back-token (@tokens);
            redo B;
-Line 4807 
 sub _tree_construction_main ($) {
+Line 4870 
 sub _tree_construction_main ($) {
          my $text = '';
          !!!next-token;
-         if ($token->{type} eq 'character') {
+         if ($token->{type} == CHARACTER_TOKEN) {
            $token->{data} =~ s/^\x0A//;
            unless (length $token->{data}) {
              !!!next-token;
            }
          }
-         while ($token->{type} eq 'character') {
+         while ($token->{type} == CHARACTER_TOKEN) {
            $text .= $token->{data};
            !!!next-token;
          }
-Line 4823 
 sub _tree_construction_main ($) {
+Line 4886 
 sub _tree_construction_main ($) {
          $self->{content_model} = PCDATA_CONTENT_MODEL;
-         if ($token->{type} eq 'end tag' and
+         if ($token->{type} == END_TAG_TOKEN and
              $token->{tag_name} eq $tag_name) {
            ## Ignore the token
          } else {
-Line 4837 
 sub _tree_construction_main ($) {
+Line 4900 
 sub _tree_construction_main ($) {
                  noframes => 1,
                  noscript => 0, ## TODO: 1 if scripting is enabled
                 }->{$token->{tag_name}}) {
-         ## NOTE: There are two "as if in body" code clones.
+         ## NOTE: There is an "as if in body" code clone.
          $parse_rcdata->(CDATA_CONTENT_MODEL, $insert);
          redo B;
        } elsif ($token->{tag_name} eq 'select') {
-Line 4868 
 sub _tree_construction_main ($) {
+Line 4931 
 sub _tree_construction_main ($) {
          !!!next-token;
          redo B;
        }
-     } elsif ($token->{type} eq 'end tag') {
+     } elsif ($token->{type} == END_TAG_TOKEN) {
        if ($token->{tag_name} eq 'body') {
          if (@{$self->{open_elements}} > 1 and
              $self->{open_elements}->[1]->[1] eq 'body') {
-Line 4929 
 sub _tree_construction_main ($) {
+Line 4992 
 sub _tree_construction_main ($) {
                   tbody => 1, tfoot=> 1, thead => 1,
                  }->{$self->{open_elements}->[-1]->[1]}) {
                !!!back-token;
-               $token = {type => 'end tag',
+               $token = {type => END_TAG_TOKEN,
                          tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                redo B;
              }
-Line 4977 
 sub _tree_construction_main ($) {
+Line 5040 
 sub _tree_construction_main ($) {
                   tbody => 1, tfoot=> 1, thead => 1,
                  }->{$self->{open_elements}->[-1]->[1]}) {
                !!!back-token;
-               $token = {type => 'end tag',
+               $token = {type => END_TAG_TOKEN,
                          tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                redo B;
              }
-Line 4993 
 sub _tree_construction_main ($) {
+Line 5056 
 sub _tree_construction_main ($) {
          if ($self->{open_elements}->[-1]->[1] eq $token->{tag_name}) {
            pop @{$self->{open_elements}};
          } else {
-           !!!parse-error (type => 'not closed:'.$self->{open_elements}->[-1]->[1]);
+           !!!parse-error (type => 'unmatched end tag:'.$token->{tag_name});
          }
          undef $self->{form_element};
-Line 5016 
 sub _tree_construction_main ($) {
+Line 5079 
 sub _tree_construction_main ($) {
                   tbody => 1, tfoot=> 1, thead => 1,
                  }->{$self->{open_elements}->[-1]->[1]}) {
                !!!back-token;
-               $token = {type => 'end tag',
+               $token = {type => END_TAG_TOKEN,
                          tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                redo B;
              }
-Line 5031 
 sub _tree_construction_main ($) {
+Line 5094 
 sub _tree_construction_main ($) {
          } # INSCOPE
          if ($self->{open_elements}->[-1]->[1] ne $token->{tag_name}) {
-           !!!parse-error (type => 'not closed:'.$self->{open_elements}->[-1]->[1]);
+           !!!parse-error (type => 'unmatched end tag:'.$token->{tag_name});
          }
          splice @{$self->{open_elements}}, $i if defined $i;
-Line 5090 
 sub _tree_construction_main ($) {
+Line 5153 
 sub _tree_construction_main ($) {
              if ({
                   dd => 1, dt => 1, li => 1, p => 1,
                   td => 1, th => 1, tr => 1,
-                  tbody => 1, tfoot=> 1, thead => 1,
+                  tbody => 1, tfoot => 1, thead => 1,
                  }->{$self->{open_elements}->[-1]->[1]}) {
                !!!back-token;
-               $token = {type => 'end tag',
+               $token = {type => END_TAG_TOKEN,
                          tag_name => $self->{open_elements}->[-1]->[1]}; # MUST
                redo B;
              }
              ## Step 2
              if ($token->{tag_name} ne $self->{open_elements}->[-1]->[1]) {
+               ## NOTE: <x><y></x>
                !!!parse-error (type => 'not closed:'.$self->{open_elements}->[-1]->[1]);
              }
-Line 5302 
 sub set_inner_html ($$$) {
+Line 5366 
 sub set_inner_html ($$$) {
  } # tree construction stage
- sub get_inner_html ($$$) {
-   my (undef, $node, $on_error) = @_;
-   ## Step 1
-   my $s = '';
-   my $in_cdata;
-   my $parent = $node;
-   while (defined $parent) {
-     if ($parent->node_type == 1 and
-         $parent->namespace_uri eq 'http://www.w3.org/1999/xhtml' and
-         {
-           style => 1, script => 1, xmp => 1, iframe => 1,
-           noembed => 1, noframes => 1, noscript => 1,
-         }->{$parent->local_name}) { ## TODO: case thingy
-       $in_cdata = 1;
-     }
-     $parent = $parent->parent_node;
-   }
-   ## Step 2
-   my @node = @{$node->child_nodes};
-   C: while (@node) {
-     my $child = shift @node;
-     unless (ref $child) {
-       if ($child eq 'cdata-out') {
-         $in_cdata = 0;
-       } else {
-         $s .= $child; # end tag
-       }
-       next C;
-     }
-     my $nt = $child->node_type;
-     if ($nt == 1) { # Element
-       my $tag_name = $child->tag_name; ## TODO: manakai_tag_name
-       $s .= '<' . $tag_name;
-       ## NOTE: Non-HTML case:
-       ## <http://permalink.gmane.org/gmane.org.w3c.whatwg.discuss/11191>
-       my @attrs = @{$child->attributes}; # sort order MUST be stable
-       for my $attr (@attrs) { # order is implementation dependent
-         my $attr_name = $attr->name; ## TODO: manakai_name
-         $s .= ' ' . $attr_name . '="';
-         my $attr_value = $attr->value;
-         ## escape
-         $attr_value =~ s/&/&amp;/g;
-         $attr_value =~ s/</&lt;/g;
-         $attr_value =~ s/>/&gt;/g;
-         $attr_value =~ s/"/&quot;/g;
-         $s .= $attr_value . '"';
-       }
-       $s .= '>';
-       next C if {
-         area => 1, base => 1, basefont => 1, bgsound => 1,
-         br => 1, col => 1, embed => 1, frame => 1, hr => 1,
-         img => 1, input => 1, link => 1, meta => 1, param => 1,
-         spacer => 1, wbr => 1,
-       }->{$tag_name};
-       $s .= "\x0A" if $tag_name eq 'pre' or $tag_name eq 'textarea';
-       if (not $in_cdata and {
-         style => 1, script => 1, xmp => 1, iframe => 1,
-         noembed => 1, noframes => 1, noscript => 1,
-         plaintext => 1,
-       }->{$tag_name}) {
-         unshift @node, 'cdata-out';
-         $in_cdata = 1;
-       }
-       unshift @node, @{$child->child_nodes}, '</' . $tag_name . '>';
-     } elsif ($nt == 3 or $nt == 4) {
-       if ($in_cdata) {
-         $s .= $child->data;
-       } else {
-         my $value = $child->data;
-         $value =~ s/&/&amp;/g;
-         $value =~ s/</&lt;/g;
-         $value =~ s/>/&gt;/g;
-         $value =~ s/"/&quot;/g;
-         $s .= $value;
-       }
-     } elsif ($nt == 8) {
-       $s .= '<!--' . $child->data . '-->';
-     } elsif ($nt == 10) {
-       $s .= '<!DOCTYPE ' . $child->name . '>';
-     } elsif ($nt == 5) { # entrefs
-       push @node, @{$child->child_nodes};
-     } else {
-       $on_error->($child) if defined $on_error;
-     }
-     ## ISSUE: This code does not support PIs.
-   } # C
-   ## Step 3
-   return \$s;
- } # get_inner_html
 ;
  # $Date$

 Legend:



Removed from v.1.54
 


changed lines


 
Added in v.1.62
 Legend:



Removed from v.1.54
 


changed lines


 
Added in v.1.62
-Removed from v.1.54
+Added in v.1.62

[email protected]	ViewVC Help
Powered by ViewVC 1.1.24