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

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

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

revision 1.225 by wakaba, Sun Aug 16 05:57:24 2009 UTC revision 1.233 by wakaba, Sun Sep 6 01:45:58 2009 UTC
# Line 179  my $el_category = { Line 179  my $el_category = {
179    dt => DTDD_EL,    dt => DTDD_EL,
180    em => FORMATTING_EL,    em => FORMATTING_EL,
181    embed => MISC_SPECIAL_EL,    embed => MISC_SPECIAL_EL,
   eventsource => MISC_SPECIAL_EL,  
182    fieldset => MISC_SPECIAL_EL,    fieldset => MISC_SPECIAL_EL,
183    figure => MISC_SPECIAL_EL,    figure => MISC_SPECIAL_EL,
184    font => FORMATTING_EL,    font => FORMATTING_EL,
# Line 203  my $el_category = { Line 202  my $el_category = {
202    #image => MISC_SPECIAL_EL, ## NOTE: Commented out in the spec.    #image => MISC_SPECIAL_EL, ## NOTE: Commented out in the spec.
203    input => MISC_SPECIAL_EL,    input => MISC_SPECIAL_EL,
204    isindex => MISC_SPECIAL_EL,    isindex => MISC_SPECIAL_EL,
205      ## XXX keygen? (Whether a void element is in Special or not does not
206      ## affect to the processing, however.)
207    li => LI_EL,    li => LI_EL,
208    link => MISC_SPECIAL_EL,    link => MISC_SPECIAL_EL,
209    listing => MISC_SPECIAL_EL,    listing => MISC_SPECIAL_EL,
# Line 911  sub _tree_construction_initial ($) { Line 912  sub _tree_construction_initial ($) {
912    
913    INITIAL: {    INITIAL: {
914      if ($token->{type} == DOCTYPE_TOKEN) {      if ($token->{type} == DOCTYPE_TOKEN) {
915        ## NOTE: Conformance checkers MAY, instead of reporting "not HTML5"        ## NOTE: Conformance checkers MAY, instead of reporting "not
916        ## error, switch to a conformance checking mode for another        ## HTML5" error, switch to a conformance checking mode for
917        ## language.        ## another language.  (We don't support such mode switchings; it
918          ## is nonsense to do anything different from what browsers do.)
919        my $doctype_name = $token->{name};        my $doctype_name = $token->{name};
920        $doctype_name = '' unless defined $doctype_name;        $doctype_name = '' unless defined $doctype_name;
921        $doctype_name =~ tr/a-z/A-Z/; # ASCII case-insensitive        my $doctype = $self->{document}->create_document_type_definition
922        if (not defined $token->{name} or # <!DOCTYPE>            ($doctype_name);
923            defined $token->{sysid}) {  
924          $doctype_name =~ tr/A-Z/a-z/; # ASCII case-insensitive
925          if ($doctype_name ne 'html') {
926          !!!cp ('t1');          !!!cp ('t1');
927          !!!parse-error (type => 'not HTML5', token => $token);          !!!parse-error (type => 'not HTML5', token => $token);
928        } elsif ($doctype_name ne 'HTML') {        } elsif (defined $token->{pubid}) {
929          !!!cp ('t2');          !!!cp ('t2');
930            ## XXX Obsolete permitted DOCTYPEs
931          !!!parse-error (type => 'not HTML5', token => $token);          !!!parse-error (type => 'not HTML5', token => $token);
932        } elsif (defined $token->{pubid}) {        } elsif (defined $token->{sysid}) {
933          if ($token->{pubid} eq 'XSLT-compat') {          if ($token->{sysid} eq 'about:legacy-compat') {
934            !!!cp ('t1.2');            !!!cp ('t1.2'); ## <!DOCTYPE HTML SYSTEM "about:legacy-compat">
935            !!!parse-error (type => 'XSLT-compat', token => $token,            !!!parse-error (type => 'XSLT-compat', token => $token,
936                            level => $self->{level}->{should});                            level => $self->{level}->{should});
937          } else {          } else {
938            !!!parse-error (type => 'not HTML5', token => $token);            !!!parse-error (type => 'not HTML5', token => $token);
939          }          }
940        } else {        } else { ## <!DOCTYPE HTML>
941          !!!cp ('t3');          !!!cp ('t3');
942          #          #
943        }        }
944                
       my $doctype = $self->{document}->create_document_type_definition  
         ($token->{name}); ## ISSUE: If name is missing (e.g. <!DOCTYPE>)?  
945        ## NOTE: Default value for both |public_id| and |system_id| attributes        ## NOTE: Default value for both |public_id| and |system_id| attributes
946        ## are empty strings, so that we don't set any value in missing cases.        ## are empty strings, so that we don't set any value in missing cases.
947        $doctype->public_id ($token->{pubid}) if defined $token->{pubid};        $doctype->public_id ($token->{pubid}) if defined $token->{pubid};
948        $doctype->system_id ($token->{sysid}) if defined $token->{sysid};        $doctype->system_id ($token->{sysid}) if defined $token->{sysid};
949    
950        ## NOTE: Other DocumentType attributes are null or empty lists.        ## NOTE: Other DocumentType attributes are null or empty lists.
951        ## In Firefox3, |internalSubset| attribute is set to the empty        ## In Firefox3, |internalSubset| attribute is set to the empty
952        ## string, while |null| is an allowed value for the attribute        ## string, while |null| is an allowed value for the attribute
953        ## according to DOM3 Core.        ## according to DOM3 Core.
954        $self->{document}->append_child ($doctype);        $self->{document}->append_child ($doctype);
955                
956        if ($token->{quirks} or $doctype_name ne 'HTML') {        if ($token->{quirks} or $doctype_name ne 'html') {
957          !!!cp ('t4');          !!!cp ('t4');
958          $self->{document}->manakai_compat_mode ('quirks');          $self->{document}->manakai_compat_mode ('quirks');
959        } elsif (defined $token->{pubid}) {        } elsif (defined $token->{pubid}) {
# Line 1435  sub _tree_construction_main ($) { Line 1439  sub _tree_construction_main ($) {
1439    }; # $script_start_tag    }; # $script_start_tag
1440    
1441    ## NOTE: $open_tables->[-1]->[0] is the "current table" element node.    ## NOTE: $open_tables->[-1]->[0] is the "current table" element node.
1442    ## NOTE: $open_tables->[-1]->[1] is the "tainted" flag.    ## NOTE: $open_tables->[-1]->[1] is the "tainted" flag (OBSOLETE; unused).
1443    ## NOTE: $open_tables->[-1]->[2] is set false when non-Text node inserted.    ## NOTE: $open_tables->[-1]->[2] is set false when non-Text node inserted.
1444    my $open_tables = [[$self->{open_elements}->[0]->[0]]];    my $open_tables = [[$self->{open_elements}->[0]->[0]]];
1445    
# Line 1731  sub _tree_construction_main ($) { Line 1735  sub _tree_construction_main ($) {
1735    ## document.write ("b")</script>|    ## document.write ("b")</script>|
1736    
1737    B: while (1) {    B: while (1) {
1738    
1739        ## The "in table text" insertion mode.
1740        if ($self->{insertion_mode} & TABLE_IMS and
1741            not $self->{insertion_mode} & IN_FOREIGN_CONTENT_IM and
1742            not $self->{insertion_mode} & IN_CDATA_RCDATA_IM) {
1743          C: {
1744            my $s;
1745            if ($token->{type} == CHARACTER_TOKEN) {
1746              !!!cp ('t194');
1747              $self->{pending_chars} ||= [];
1748              push @{$self->{pending_chars}}, $token;
1749              !!!next-token;
1750              next B;
1751            } else {
1752              if ($self->{pending_chars}) {
1753                $s = join '', map { $_->{data} } @{$self->{pending_chars}};
1754                delete $self->{pending_chars};
1755                if ($s =~ /[^\x09\x0A\x0C\x0D\x20]/) {
1756                  !!!cp ('t195');
1757                  #
1758                } else {
1759                  !!!cp ('t195.1');
1760                  #$self->{open_elements}->[-1]->[0]->manakai_append_text ($s);
1761                  $self->{open_elements}->[-1]->[0]->append_child
1762                      ($self->{document}->create_text_node ($s));
1763                  last C;
1764                }
1765              } else {
1766                !!!cp ('t195.2');
1767                last C;
1768              }
1769            }
1770    
1771            ## Foster parenting
1772            !!!parse-error (type => 'in table:#text', token => $token);
1773    
1774            ## NOTE: As if in body, but insert into the foster parent element.
1775            $reconstruct_active_formatting_elements->($insert_to_foster);
1776                
1777            if ($self->{open_elements}->[-1]->[1] & TABLE_ROWS_EL) {
1778              # MUST
1779              my $foster_parent_element;
1780              my $next_sibling;
1781              #my $prev_sibling;
1782              OE: for (reverse 0..$#{$self->{open_elements}}) {
1783                if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {
1784                  my $parent = $self->{open_elements}->[$_]->[0]->parent_node;
1785                  if (defined $parent and $parent->node_type == 1) {
1786                    $foster_parent_element = $parent;
1787                    !!!cp ('t196');
1788                    $next_sibling = $self->{open_elements}->[$_]->[0];
1789              #      $prev_sibling = $next_sibling->previous_sibling;
1790                    #
1791                  } else {
1792                    !!!cp ('t197');
1793                    $foster_parent_element = $self->{open_elements}->[$_ - 1]->[0];
1794              #      $prev_sibling = $foster_parent_element->last_child;
1795                    #
1796                  }
1797                  last OE;
1798                }
1799              } # OE
1800              $foster_parent_element = $self->{open_elements}->[0]->[0] #and
1801              #$prev_sibling = $foster_parent_element->last_child
1802                  unless defined $foster_parent_element;
1803              #undef $prev_sibling unless $open_tables->[-1]->[2]; # ~node inserted
1804              #if (defined $prev_sibling and
1805              #    $prev_sibling->node_type == 3) {
1806              #  !!! cp ('t198');
1807              #  $prev_sibling->manakai_append_text ($s);
1808              #} else {
1809                !!!cp ('t199');
1810                $foster_parent_element->insert_before
1811                    ($self->{document}->create_text_node ($s), $next_sibling);
1812              #}
1813              $open_tables->[-1]->[1] = 1; # tainted
1814              $open_tables->[-1]->[2] = 1; # ~node inserted
1815            } else {
1816              ## NOTE: Fragment case or in a foster parent'ed element
1817              ## (e.g. |<table><span>a|).  In fragment case, whether the
1818              ## character is appended to existing node or a new node is
1819              ## created is irrelevant, since the foster parent'ed nodes
1820              ## are discarded and fragment parsing does not invoke any
1821              ## script.
1822              !!!cp ('t200');
1823              $self->{open_elements}->[-1]->[0]->manakai_append_text ($s);
1824            }
1825          } # C
1826        } # TABLE_IMS
1827    
1828      if ($token->{type} == DOCTYPE_TOKEN) {      if ($token->{type} == DOCTYPE_TOKEN) {
1829        !!!cp ('t73');        !!!cp ('t73');
1830        !!!parse-error (type => 'in html:#DOCTYPE', token => $token);        !!!parse-error (type => 'in html:#DOCTYPE', token => $token);
# Line 2149  sub _tree_construction_main ($) { Line 2243  sub _tree_construction_main ($) {
2243            !!!ack ('t103.1');            !!!ack ('t103.1');
2244            !!!next-token;            !!!next-token;
2245            next B;            next B;
2246          } elsif ($token->{tag_name} eq 'command' or          } elsif ($token->{tag_name} eq 'command') {
                  $token->{tag_name} eq 'eventsource') {  
2247            if ($self->{insertion_mode} == IN_HEAD_IM) {            if ($self->{insertion_mode} == IN_HEAD_IM) {
2248              ## NOTE: If the insertion mode at the time of the emission              ## NOTE: If the insertion mode at the time of the emission
2249              ## of the token was "before head", $self->{insertion_mode}              ## of the token was "before head", $self->{insertion_mode}
# Line 3018  sub _tree_construction_main ($) { Line 3111  sub _tree_construction_main ($) {
3111        $insert = $insert_to_current;        $insert = $insert_to_current;
3112        #        #
3113      } elsif ($self->{insertion_mode} & TABLE_IMS) {      } elsif ($self->{insertion_mode} & TABLE_IMS) {
3114        if ($token->{type} == CHARACTER_TOKEN) {        if ($token->{type} == START_TAG_TOKEN) {
         if (not $open_tables->[-1]->[1] and # tainted  
             $token->{data} =~ s/^([\x09\x0A\x0C\x20]+)//) {  
           $self->{open_elements}->[-1]->[0]->manakai_append_text ($1);  
                 
           unless (length $token->{data}) {  
             !!!cp ('t194');  
             !!!next-token;  
             next B;  
           } else {  
             !!!cp ('t195');  
           }  
         }  
   
         !!!parse-error (type => 'in table:#text', token => $token);  
   
         ## NOTE: As if in body, but insert into the foster parent element.  
         $reconstruct_active_formatting_elements->($insert_to_foster);  
               
         if ($self->{open_elements}->[-1]->[1] & TABLE_ROWS_EL) {  
           # MUST  
           my $foster_parent_element;  
           my $next_sibling;  
           my $prev_sibling;  
           OE: for (reverse 0..$#{$self->{open_elements}}) {  
             if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {  
               my $parent = $self->{open_elements}->[$_]->[0]->parent_node;  
               if (defined $parent and $parent->node_type == 1) {  
                 $foster_parent_element = $parent;  
                 !!!cp ('t196');  
                 $next_sibling = $self->{open_elements}->[$_]->[0];  
                 $prev_sibling = $next_sibling->previous_sibling;  
                 #  
               } else {  
                 !!!cp ('t197');  
                 $foster_parent_element = $self->{open_elements}->[$_ - 1]->[0];  
                 $prev_sibling = $foster_parent_element->last_child;  
                 #  
               }  
               last OE;  
             }  
           } # OE  
           $foster_parent_element = $self->{open_elements}->[0]->[0] and  
           $prev_sibling = $foster_parent_element->last_child  
               unless defined $foster_parent_element;  
           undef $prev_sibling unless $open_tables->[-1]->[2]; # ~node inserted  
           if (defined $prev_sibling and  
               $prev_sibling->node_type == 3) {  
             !!!cp ('t198');  
             $prev_sibling->manakai_append_text ($token->{data});  
           } else {  
             !!!cp ('t199');  
             $foster_parent_element->insert_before  
                 ($self->{document}->create_text_node ($token->{data}),  
                  $next_sibling);  
           }  
           $open_tables->[-1]->[1] = 1; # tainted  
           $open_tables->[-1]->[2] = 1; # ~node inserted  
         } else {  
           ## NOTE: Fragment case or in a foster parent'ed element  
           ## (e.g. |<table><span>a|).  In fragment case, whether the  
           ## character is appended to existing node or a new node is  
           ## created is irrelevant, since the foster parent'ed nodes  
           ## are discarded and fragment parsing does not invoke any  
           ## script.  
           !!!cp ('t200');  
           $self->{open_elements}->[-1]->[0]->manakai_append_text  
               ($token->{data});  
         }  
               
         !!!next-token;  
         next B;  
       } elsif ($token->{type} == START_TAG_TOKEN) {  
3115          if ({          if ({
3116               tr => (($self->{insertion_mode} & IM_MASK) != IN_ROW_IM),               tr => (($self->{insertion_mode} & IM_MASK) != IN_ROW_IM),
3117               th => 1, td => 1,               th => 1, td => 1,
# Line 3358  sub _tree_construction_main ($) { Line 3379  sub _tree_construction_main ($) {
3379            !!!ack-later;            !!!ack-later;
3380            next B;            next B;
3381          } elsif ($token->{tag_name} eq 'style') {          } elsif ($token->{tag_name} eq 'style') {
3382            if (not $open_tables->[-1]->[1]) { # tainted            !!!cp ('t227.8');
3383              !!!cp ('t227.8');            ## NOTE: This is a "as if in head" code clone.
3384              ## NOTE: This is a "as if in head" code clone.            $parse_rcdata->(CDATA_CONTENT_MODEL);
3385              $parse_rcdata->(CDATA_CONTENT_MODEL);            $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
3386              $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted            next B;
             next B;  
           } else {  
             !!!cp ('t227.7');  
             #  
           }  
3387          } elsif ($token->{tag_name} eq 'script') {          } elsif ($token->{tag_name} eq 'script') {
3388            if (not $open_tables->[-1]->[1]) { # tainted            !!!cp ('t227.6');
3389              !!!cp ('t227.6');            ## NOTE: This is a "as if in head" code clone.
3390              ## NOTE: This is a "as if in head" code clone.            $script_start_tag->();
3391              $script_start_tag->();            $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
3392              $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted            next B;
             next B;  
           } else {  
             !!!cp ('t227.5');  
             #  
           }  
3393          } elsif ($token->{tag_name} eq 'input') {          } elsif ($token->{tag_name} eq 'input') {
3394            if (not $open_tables->[-1]->[1]) { # tainted            if ($token->{attributes}->{type}) {
3395              if ($token->{attributes}->{type}) { ## TODO: case              my $type = $token->{attributes}->{type}->{value};
3396                my $type = lc $token->{attributes}->{type}->{value};              $type =~ tr/A-Z/a-z/; ## ASCII case-insensitive.
3397                if ($type eq 'hidden') {              if ($type eq 'hidden') {
3398                  !!!cp ('t227.3');                !!!cp ('t227.3');
3399                  !!!parse-error (type => 'in table',                !!!parse-error (type => 'in table',
3400                                  text => $token->{tag_name}, token => $token);                                text => $token->{tag_name}, token => $token);
3401    
3402                  !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);                !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
3403                  $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted                $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
3404    
3405                  ## TODO: form element pointer                ## TODO: form element pointer
3406    
3407                  pop @{$self->{open_elements}};                pop @{$self->{open_elements}};
3408    
3409                  !!!next-token;                !!!next-token;
3410                  !!!ack ('t227.2.1');                !!!ack ('t227.2.1');
3411                  next B;                next B;
               } else {  
                 !!!cp ('t227.2');  
                 #  
               }  
3412              } else {              } else {
3413                !!!cp ('t227.1');                !!!cp ('t227.1');
3414                #                #
# Line 3914  sub _tree_construction_main ($) { Line 3921  sub _tree_construction_main ($) {
3921              ## Reprocess the token.              ## Reprocess the token.
3922              next B;              next B;
3923            }            }
3924            } elsif ($token->{tag_name} eq 'script') {
3925              !!!cp ('t281.3');
3926              ## NOTE: This is an "as if in head" code clone
3927              $script_start_tag->();
3928              next B;
3929          } else {          } else {
3930            !!!cp ('t282');            !!!cp ('t282');
3931            !!!parse-error (type => 'in select',            !!!parse-error (type => 'in select',
# Line 4325  sub _tree_construction_main ($) { Line 4337  sub _tree_construction_main ($) {
4337          $parse_rcdata->(CDATA_CONTENT_MODEL);          $parse_rcdata->(CDATA_CONTENT_MODEL);
4338          next B;          next B;
4339        } elsif ({        } elsif ({
4340                  base => 1, command => 1, eventsource => 1, link => 1,                  base => 1, command => 1, link => 1,
4341                 }->{$token->{tag_name}}) {                 }->{$token->{tag_name}}) {
4342          !!!cp ('t334');          !!!cp ('t334');
4343          ## NOTE: This is an "as if in head" code clone, only "-t" differs          ## NOTE: This is an "as if in head" code clone, only "-t" differs
# Line 5037  sub _tree_construction_main ($) { Line 5049  sub _tree_construction_main ($) {
5049          } elsif ({          } elsif ({
5050                    area => 1, basefont => 1, bgsound => 1, br => 1,                    area => 1, basefont => 1, bgsound => 1, br => 1,
5051                    embed => 1, img => 1, spacer => 1, wbr => 1,                    embed => 1, img => 1, spacer => 1, wbr => 1,
5052                      keygen => 1,
5053                   }->{$token->{tag_name}}) {                   }->{$token->{tag_name}}) {
5054            !!!cp ('t388.1');            !!!cp ('t388.1');
5055            pop @{$self->{open_elements}};            pop @{$self->{open_elements}};
Legend:
Removed from v.1.225  
changed lines
  Added in v.1.233
admin@suikawiki.org
 ViewVC Help
Powered by ViewVC 1.1.24