/[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 1.223 by wakaba, Sun Aug 16 04:59:53 2009 UTC revision 1.237 by wakaba, Sun Sep 6 08:29:32 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 195  my $el_category = { Line 194  my $el_category = {
194    h6 => HEADING_EL,    h6 => HEADING_EL,
195    head => MISC_SPECIAL_EL,    head => MISC_SPECIAL_EL,
196    header => MISC_SPECIAL_EL,    header => MISC_SPECIAL_EL,
197      hgroup => MISC_SPECIAL_EL,
198    hr => MISC_SPECIAL_EL,    hr => MISC_SPECIAL_EL,
199    html => HTML_EL,    html => HTML_EL,
200    i => FORMATTING_EL,    i => FORMATTING_EL,
# Line 203  my $el_category = { Line 203  my $el_category = {
203    #image => MISC_SPECIAL_EL, ## NOTE: Commented out in the spec.    #image => MISC_SPECIAL_EL, ## NOTE: Commented out in the spec.
204    input => MISC_SPECIAL_EL,    input => MISC_SPECIAL_EL,
205    isindex => MISC_SPECIAL_EL,    isindex => MISC_SPECIAL_EL,
206      ## XXX keygen? (Whether a void element is in Special or not does not
207      ## affect to the processing, however.)
208    li => LI_EL,    li => LI_EL,
209    link => MISC_SPECIAL_EL,    link => MISC_SPECIAL_EL,
210    listing => MISC_SPECIAL_EL,    listing => MISC_SPECIAL_EL,
# Line 247  my $el_category = { Line 249  my $el_category = {
249    u => FORMATTING_EL,    u => FORMATTING_EL,
250    ul => MISC_SPECIAL_EL,    ul => MISC_SPECIAL_EL,
251    wbr => MISC_SPECIAL_EL,    wbr => MISC_SPECIAL_EL,
252      xmp => MISC_SPECIAL_EL,
253  };  };
254    
255  my $el_category_f = {  my $el_category_f = {
# Line 647  sub parse_char_stream ($$$;$$) { Line 650  sub parse_char_stream ($$$;$$) {
650    
651    ## NOTE: |set_inner_html| copies most of this method's code    ## NOTE: |set_inner_html| copies most of this method's code
652    
653      ## Confidence: irrelevant.
654    $self->{confident} = 1 unless exists $self->{confident};    $self->{confident} = 1 unless exists $self->{confident};
655    
656    $self->{document}->input_encoding ($self->{input_encoding})    $self->{document}->input_encoding ($self->{input_encoding})
657        if defined $self->{input_encoding};        if defined $self->{input_encoding};
658  ## TODO: |{input_encoding}| is needless?  ## TODO: |{input_encoding}| is needless?
# Line 911  sub _tree_construction_initial ($) { Line 916  sub _tree_construction_initial ($) {
916    
917    INITIAL: {    INITIAL: {
918      if ($token->{type} == DOCTYPE_TOKEN) {      if ($token->{type} == DOCTYPE_TOKEN) {
919        ## NOTE: Conformance checkers MAY, instead of reporting "not HTML5"        ## NOTE: Conformance checkers MAY, instead of reporting "not
920        ## error, switch to a conformance checking mode for another        ## HTML5" error, switch to a conformance checking mode for
921        ## language.        ## another language.  (We don't support such mode switchings; it
922          ## is nonsense to do anything different from what browsers do.)
923        my $doctype_name = $token->{name};        my $doctype_name = $token->{name};
924        $doctype_name = '' unless defined $doctype_name;        $doctype_name = '' unless defined $doctype_name;
925        $doctype_name =~ tr/a-z/A-Z/; # ASCII case-insensitive        my $doctype = $self->{document}->create_document_type_definition
926        if (not defined $token->{name} or # <!DOCTYPE>            ($doctype_name);
927            defined $token->{sysid}) {  
928          $doctype_name =~ tr/A-Z/a-z/; # ASCII case-insensitive
929          if ($doctype_name ne 'html') {
930          !!!cp ('t1');          !!!cp ('t1');
931          !!!parse-error (type => 'not HTML5', token => $token);          !!!parse-error (type => 'not HTML5', token => $token);
932        } elsif ($doctype_name ne 'HTML') {        } elsif (defined $token->{pubid}) {
933          !!!cp ('t2');          !!!cp ('t2');
934            ## XXX Obsolete permitted DOCTYPEs
935          !!!parse-error (type => 'not HTML5', token => $token);          !!!parse-error (type => 'not HTML5', token => $token);
936        } elsif (defined $token->{pubid}) {        } elsif (defined $token->{sysid}) {
937          if ($token->{pubid} eq 'XSLT-compat') {          if ($token->{sysid} eq 'about:legacy-compat') {
938            !!!cp ('t1.2');            !!!cp ('t1.2'); ## <!DOCTYPE HTML SYSTEM "about:legacy-compat">
939            !!!parse-error (type => 'XSLT-compat', token => $token,            !!!parse-error (type => 'XSLT-compat', token => $token,
940                            level => $self->{level}->{should});                            level => $self->{level}->{should});
941          } else {          } else {
942            !!!parse-error (type => 'not HTML5', token => $token);            !!!parse-error (type => 'not HTML5', token => $token);
943          }          }
944        } else {        } else { ## <!DOCTYPE HTML>
945          !!!cp ('t3');          !!!cp ('t3');
946          #          #
947        }        }
948                
       my $doctype = $self->{document}->create_document_type_definition  
         ($token->{name}); ## ISSUE: If name is missing (e.g. <!DOCTYPE>)?  
949        ## NOTE: Default value for both |public_id| and |system_id| attributes        ## NOTE: Default value for both |public_id| and |system_id| attributes
950        ## 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.
951        $doctype->public_id ($token->{pubid}) if defined $token->{pubid};        $doctype->public_id ($token->{pubid}) if defined $token->{pubid};
952        $doctype->system_id ($token->{sysid}) if defined $token->{sysid};        $doctype->system_id ($token->{sysid}) if defined $token->{sysid};
953    
954        ## NOTE: Other DocumentType attributes are null or empty lists.        ## NOTE: Other DocumentType attributes are null or empty lists.
955        ## In Firefox3, |internalSubset| attribute is set to the empty        ## In Firefox3, |internalSubset| attribute is set to the empty
956        ## string, while |null| is an allowed value for the attribute        ## string, while |null| is an allowed value for the attribute
957        ## according to DOM3 Core.        ## according to DOM3 Core.
958        $self->{document}->append_child ($doctype);        $self->{document}->append_child ($doctype);
959                
960        if ($token->{quirks} or $doctype_name ne 'HTML') {        if ($token->{quirks} or $doctype_name ne 'html') {
961          !!!cp ('t4');          !!!cp ('t4');
962          $self->{document}->manakai_compat_mode ('quirks');          $self->{document}->manakai_compat_mode ('quirks');
963        } elsif (defined $token->{pubid}) {        } elsif (defined $token->{pubid}) {
# Line 1435  sub _tree_construction_main ($) { Line 1443  sub _tree_construction_main ($) {
1443    }; # $script_start_tag    }; # $script_start_tag
1444    
1445    ## NOTE: $open_tables->[-1]->[0] is the "current table" element node.    ## NOTE: $open_tables->[-1]->[0] is the "current table" element node.
1446    ## NOTE: $open_tables->[-1]->[1] is the "tainted" flag.    ## NOTE: $open_tables->[-1]->[1] is the "tainted" flag (OBSOLETE; unused).
1447    ## 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.
1448    my $open_tables = [[$self->{open_elements}->[0]->[0]]];    my $open_tables = [[$self->{open_elements}->[0]->[0]]];
1449    
# Line 1605  sub _tree_construction_main ($) { Line 1613  sub _tree_construction_main ($) {
1613                
1614        ## Step 8        ## Step 8
1615        if ($common_ancestor_node->[1] & TABLE_ROWS_EL) {        if ($common_ancestor_node->[1] & TABLE_ROWS_EL) {
1616            ## Foster parenting.
1617          my $foster_parent_element;          my $foster_parent_element;
1618          my $next_sibling;          my $next_sibling;
1619          OE: for (reverse 0..$#{$self->{open_elements}}) {          OE: for (reverse 0..$#{$self->{open_elements}}) {
1620            if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {            if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {
1621                               my $parent = $self->{open_elements}->[$_]->[0]->parent_node;              !!!cp ('t65.2');
1622                               if (defined $parent and $parent->node_type == 1) {              $foster_parent_element = $self->{open_elements}->[$_ - 1]->[0];
1623                                 !!!cp ('t65.1');              $next_sibling = $self->{open_elements}->[$_]->[0];
1624                                 $foster_parent_element = $parent;              undef $next_sibling
1625                                 $next_sibling = $self->{open_elements}->[$_]->[0];                  unless $next_sibling->parent_node eq $foster_parent_element;
1626                               } else {              last OE;
1627                                 !!!cp ('t65.2');            }
1628                                 $foster_parent_element          } # OE
1629                                   = $self->{open_elements}->[$_ - 1]->[0];          $foster_parent_element ||= $self->{open_elements}->[0]->[0];
1630                               }  
                              last OE;  
                            }  
                          } # OE  
                          $foster_parent_element = $self->{open_elements}->[0]->[0]  
                            unless defined $foster_parent_element;  
1631          $foster_parent_element->insert_before ($last_node->[0], $next_sibling);          $foster_parent_element->insert_before ($last_node->[0], $next_sibling);
1632          $open_tables->[-1]->[1] = 1; # tainted          $open_tables->[-1]->[1] = 1; # tainted
1633        } else {        } else {
# Line 1679  sub _tree_construction_main ($) { Line 1683  sub _tree_construction_main ($) {
1683      $self->{open_elements}->[-1]->[0]->append_child ($_[0]);      $self->{open_elements}->[-1]->[0]->append_child ($_[0]);
1684    }; # $insert_to_current    }; # $insert_to_current
1685    
1686      ## Foster parenting.  Note that there are three "foster parenting"
1687      ## code in the parser: for elements (this one), for texts, and for
1688      ## elements in the AAA code.
1689    my $insert_to_foster = sub {    my $insert_to_foster = sub {
1690      my $child = shift;      my $child = shift;
1691      if ($self->{open_elements}->[-1]->[1] & TABLE_ROWS_EL) {      if ($self->{open_elements}->[-1]->[1] & TABLE_ROWS_EL) {
# Line 1687  sub _tree_construction_main ($) { Line 1694  sub _tree_construction_main ($) {
1694        my $next_sibling;        my $next_sibling;
1695        OE: for (reverse 0..$#{$self->{open_elements}}) {        OE: for (reverse 0..$#{$self->{open_elements}}) {
1696          if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {          if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {
1697                               my $parent = $self->{open_elements}->[$_]->[0]->parent_node;            !!!cp ('t71');
1698                               if (defined $parent and $parent->node_type == 1) {            $foster_parent_element = $self->{open_elements}->[$_ - 1]->[0];
1699                                 !!!cp ('t70');            $next_sibling = $self->{open_elements}->[$_]->[0];
1700                                 $foster_parent_element = $parent;            undef $next_sibling
1701                                 $next_sibling = $self->{open_elements}->[$_]->[0];                unless $next_sibling->parent_node eq $foster_parent_element;
1702                               } else {            last OE;
1703                                 !!!cp ('t71');          }
1704                                 $foster_parent_element        } # OE
1705                                   = $self->{open_elements}->[$_ - 1]->[0];        $foster_parent_element ||= $self->{open_elements}->[0]->[0];
1706                               }  
1707                               last OE;        $foster_parent_element->insert_before ($child, $next_sibling);
                            }  
                          } # OE  
                          $foster_parent_element = $self->{open_elements}->[0]->[0]  
                            unless defined $foster_parent_element;  
                          $foster_parent_element->insert_before  
                            ($child, $next_sibling);  
1708        $open_tables->[-1]->[1] = 1; # tainted        $open_tables->[-1]->[1] = 1; # tainted
1709      } else {      } else {
1710        !!!cp ('t72');        !!!cp ('t72');
# Line 1731  sub _tree_construction_main ($) { Line 1732  sub _tree_construction_main ($) {
1732    ## document.write ("b")</script>|    ## document.write ("b")</script>|
1733    
1734    B: while (1) {    B: while (1) {
1735    
1736        ## The "in table text" insertion mode.
1737        if ($self->{insertion_mode} & TABLE_IMS and
1738            not $self->{insertion_mode} & IN_FOREIGN_CONTENT_IM and
1739            not $self->{insertion_mode} & IN_CDATA_RCDATA_IM) {
1740          C: {
1741            my $s;
1742            if ($token->{type} == CHARACTER_TOKEN) {
1743              !!!cp ('t194');
1744              $self->{pending_chars} ||= [];
1745              push @{$self->{pending_chars}}, $token;
1746              !!!next-token;
1747              next B;
1748            } else {
1749              if ($self->{pending_chars}) {
1750                $s = join '', map { $_->{data} } @{$self->{pending_chars}};
1751                delete $self->{pending_chars};
1752                if ($s =~ /[^\x09\x0A\x0C\x0D\x20]/) {
1753                  !!!cp ('t195');
1754                  #
1755                } else {
1756                  !!!cp ('t195.1');
1757                  #$self->{open_elements}->[-1]->[0]->manakai_append_text ($s);
1758                  $self->{open_elements}->[-1]->[0]->append_child
1759                      ($self->{document}->create_text_node ($s));
1760                  last C;
1761                }
1762              } else {
1763                !!!cp ('t195.2');
1764                last C;
1765              }
1766            }
1767    
1768            ## Foster parenting.
1769            !!!parse-error (type => 'in table:#text', token => $token);
1770    
1771            ## NOTE: As if in body, but insert into the foster parent element.
1772            $reconstruct_active_formatting_elements->($insert_to_foster);
1773                
1774            if ($self->{open_elements}->[-1]->[1] & TABLE_ROWS_EL) {
1775              # MUST
1776              my $foster_parent_element;
1777              my $next_sibling;
1778              OE: for (reverse 0..$#{$self->{open_elements}}) {
1779                if ($self->{open_elements}->[$_]->[1] == TABLE_EL) {
1780                  !!!cp ('t197');
1781                  $foster_parent_element = $self->{open_elements}->[$_ - 1]->[0];
1782                  $next_sibling = $self->{open_elements}->[$_]->[0];
1783                  undef $next_sibling
1784                    unless $next_sibling->parent_node eq $foster_parent_element;
1785                  last OE;
1786                }
1787              } # OE
1788              $foster_parent_element ||= $self->{open_elements}->[0]->[0];
1789    
1790              !!!cp ('t199');
1791              $foster_parent_element->insert_before
1792                  ($self->{document}->create_text_node ($s), $next_sibling);
1793    
1794              $open_tables->[-1]->[1] = 1; # tainted
1795              $open_tables->[-1]->[2] = 1; # ~node inserted
1796            } else {
1797              ## NOTE: Fragment case or in a foster parent'ed element
1798              ## (e.g. |<table><span>a|).  In fragment case, whether the
1799              ## character is appended to existing node or a new node is
1800              ## created is irrelevant, since the foster parent'ed nodes
1801              ## are discarded and fragment parsing does not invoke any
1802              ## script.
1803              !!!cp ('t200');
1804              $self->{open_elements}->[-1]->[0]->manakai_append_text ($s);
1805            }
1806          } # C
1807        } # TABLE_IMS
1808    
1809      if ($token->{type} == DOCTYPE_TOKEN) {      if ($token->{type} == DOCTYPE_TOKEN) {
1810        !!!cp ('t73');        !!!cp ('t73');
1811        !!!parse-error (type => 'in html:#DOCTYPE', token => $token);        !!!parse-error (type => 'in html:#DOCTYPE', token => $token);
# Line 2149  sub _tree_construction_main ($) { Line 2224  sub _tree_construction_main ($) {
2224            !!!ack ('t103.1');            !!!ack ('t103.1');
2225            !!!next-token;            !!!next-token;
2226            next B;            next B;
2227          } elsif ($token->{tag_name} eq 'command' or          } elsif ($token->{tag_name} eq 'command') {
                  $token->{tag_name} eq 'eventsource') {  
2228            if ($self->{insertion_mode} == IN_HEAD_IM) {            if ($self->{insertion_mode} == IN_HEAD_IM) {
2229              ## NOTE: If the insertion mode at the time of the emission              ## NOTE: If the insertion mode at the time of the emission
2230              ## of the token was "before head", $self->{insertion_mode}              ## of the token was "before head", $self->{insertion_mode}
# Line 2265  sub _tree_construction_main ($) { Line 2339  sub _tree_construction_main ($) {
2339    
2340            ## NOTE: There is a "as if in head" code clone.            ## NOTE: There is a "as if in head" code clone.
2341            $parse_rcdata->(RCDATA_CONTENT_MODEL);            $parse_rcdata->(RCDATA_CONTENT_MODEL);
2342            ## ISSUE: A spec bug [Bug 6038]  
2343              ## NOTE: At this point the stack of open elements contain
2344              ## the |head| element (index == -2) and the |script| element
2345              ## (index == -1).  In the "after head" insertion mode the
2346              ## |head| element is inserted only for the purpose of
2347              ## providing the context for the |script| element, and
2348              ## therefore we can now and have to remove the element from
2349              ## the stack.
2350            splice @{$self->{open_elements}}, -2, 1, () # <head>            splice @{$self->{open_elements}}, -2, 1, () # <head>
2351                if ($self->{insertion_mode} & IM_MASK) == AFTER_HEAD_IM;                if ($self->{insertion_mode} & IM_MASK) == AFTER_HEAD_IM;
2352            next B;            next B;
# Line 2500  sub _tree_construction_main ($) { Line 2581  sub _tree_construction_main ($) {
2581                ## Ignore the token                ## Ignore the token
2582                !!!next-token;                !!!next-token;
2583                next B;                next B;
2584              } elsif ($token->{tag_name} eq 'br') {          } elsif ($token->{tag_name} eq 'br') {
2585                if ($self->{insertion_mode} == BEFORE_HEAD_IM) {            if ($self->{insertion_mode} == BEFORE_HEAD_IM) {
2586                  !!!cp ('t142.2');              !!!cp ('t142.2');
2587                  ## (before head) as if <head>, (in head) as if </head>              ## (before head) as if <head>, (in head) as if </head>
2588                  !!!create-element ($self->{head_element}, $HTML_NS, 'head',, $token);              !!!create-element ($self->{head_element}, $HTML_NS, 'head',, $token);
2589                  $self->{open_elements}->[-1]->[0]->append_child ($self->{head_element});              $self->{open_elements}->[-1]->[0]->append_child ($self->{head_element});
2590                  $self->{insertion_mode} = AFTER_HEAD_IM;              $self->{insertion_mode} = AFTER_HEAD_IM;
2591        
2592                  ## Reprocess in the "after head" insertion mode...              ## Reprocess in the "after head" insertion mode...
2593                } elsif ($self->{insertion_mode} == IN_HEAD_IM) {            } elsif ($self->{insertion_mode} == IN_HEAD_IM) {
2594                  !!!cp ('t143.2');              !!!cp ('t143.2');
2595                  ## As if </head>              ## As if </head>
2596                  pop @{$self->{open_elements}};              pop @{$self->{open_elements}};
2597                  $self->{insertion_mode} = AFTER_HEAD_IM;              $self->{insertion_mode} = AFTER_HEAD_IM;
2598        
2599                  ## Reprocess in the "after head" insertion mode...              ## Reprocess in the "after head" insertion mode...
2600                } elsif ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {            } elsif ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
2601                  !!!cp ('t143.3');              !!!cp ('t143.3');
2602                  ## NOTE: Two parse errors for <head><noscript></br>              ## NOTE: Two parse errors for <head><noscript></br>
2603                  !!!parse-error (type => 'unmatched end tag',              !!!parse-error (type => 'unmatched end tag',
2604                                  text => 'br', token => $token);                              text => 'br', token => $token);
2605                  ## As if </noscript>              ## As if </noscript>
2606                  pop @{$self->{open_elements}};              pop @{$self->{open_elements}};
2607                  $self->{insertion_mode} = IN_HEAD_IM;              $self->{insertion_mode} = IN_HEAD_IM;
2608    
2609                  ## Reprocess in the "in head" insertion mode...              ## Reprocess in the "in head" insertion mode...
2610                  ## As if </head>              ## As if </head>
2611                  pop @{$self->{open_elements}};              pop @{$self->{open_elements}};
2612                  $self->{insertion_mode} = AFTER_HEAD_IM;              $self->{insertion_mode} = AFTER_HEAD_IM;
2613    
2614                  ## Reprocess in the "after head" insertion mode...              ## Reprocess in the "after head" insertion mode...
2615                } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {            } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {
2616                  !!!cp ('t143.4');              !!!cp ('t143.4');
2617                  #              #
2618                } else {            } else {
2619                  die "$0: $self->{insertion_mode}: Unknown insertion mode";              die "$0: $self->{insertion_mode}: Unknown insertion mode";
2620                }            }
2621    
2622                ## ISSUE: does not agree with IE7 - it doesn't ignore </br>.            #
2623                !!!parse-error (type => 'unmatched end tag',          } else { ## Other end tags
                               text => 'br', token => $token);  
               ## Ignore the token  
               !!!next-token;  
               next B;  
             } else {  
2624                !!!cp ('t145');                !!!cp ('t145');
2625                !!!parse-error (type => 'unmatched end tag',                !!!parse-error (type => 'unmatched end tag',
2626                                text => $token->{tag_name}, token => $token);                                text => $token->{tag_name}, token => $token);
# Line 2588  sub _tree_construction_main ($) { Line 2664  sub _tree_construction_main ($) {
2664              !!!insert-element ('body',, $token);              !!!insert-element ('body',, $token);
2665              $self->{insertion_mode} = IN_BODY_IM;              $self->{insertion_mode} = IN_BODY_IM;
2666              ## reprocess              ## reprocess
2667              next B;          next B;
2668        } elsif ($token->{type} == END_OF_FILE_TOKEN) {        } elsif ($token->{type} == END_OF_FILE_TOKEN) {
2669          if ($self->{insertion_mode} == BEFORE_HEAD_IM) {          if ($self->{insertion_mode} == BEFORE_HEAD_IM) {
2670            !!!cp ('t149.1');            !!!cp ('t149.1');
# Line 3016  sub _tree_construction_main ($) { Line 3092  sub _tree_construction_main ($) {
3092        $insert = $insert_to_current;        $insert = $insert_to_current;
3093        #        #
3094      } elsif ($self->{insertion_mode} & TABLE_IMS) {      } elsif ($self->{insertion_mode} & TABLE_IMS) {
3095        if ($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 ($token->{type} == START_TAG_TOKEN) {  
3096          if ({          if ({
3097               tr => (($self->{insertion_mode} & IM_MASK) != IN_ROW_IM),               tr => (($self->{insertion_mode} & IM_MASK) != IN_ROW_IM),
3098               th => 1, td => 1,               th => 1, td => 1,
# Line 3356  sub _tree_construction_main ($) { Line 3360  sub _tree_construction_main ($) {
3360            !!!ack-later;            !!!ack-later;
3361            next B;            next B;
3362          } elsif ($token->{tag_name} eq 'style') {          } elsif ($token->{tag_name} eq 'style') {
3363            if (not $open_tables->[-1]->[1]) { # tainted            !!!cp ('t227.8');
3364              !!!cp ('t227.8');            ## NOTE: This is a "as if in head" code clone.
3365              ## NOTE: This is a "as if in head" code clone.            $parse_rcdata->(CDATA_CONTENT_MODEL);
3366              $parse_rcdata->(CDATA_CONTENT_MODEL);            $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
3367              $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted            next B;
             next B;  
           } else {  
             !!!cp ('t227.7');  
             #  
           }  
3368          } elsif ($token->{tag_name} eq 'script') {          } elsif ($token->{tag_name} eq 'script') {
3369            if (not $open_tables->[-1]->[1]) { # tainted            !!!cp ('t227.6');
3370              !!!cp ('t227.6');            ## NOTE: This is a "as if in head" code clone.
3371              ## NOTE: This is a "as if in head" code clone.            $script_start_tag->();
3372              $script_start_tag->();            $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
3373              $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted            next B;
             next B;  
           } else {  
             !!!cp ('t227.5');  
             #  
           }  
3374          } elsif ($token->{tag_name} eq 'input') {          } elsif ($token->{tag_name} eq 'input') {
3375            if (not $open_tables->[-1]->[1]) { # tainted            if ($token->{attributes}->{type}) {
3376              if ($token->{attributes}->{type}) { ## TODO: case              my $type = $token->{attributes}->{type}->{value};
3377                my $type = lc $token->{attributes}->{type}->{value};              $type =~ tr/A-Z/a-z/; ## ASCII case-insensitive.
3378                if ($type eq 'hidden') {              if ($type eq 'hidden') {
3379                  !!!cp ('t227.3');                !!!cp ('t227.3');
3380                  !!!parse-error (type => 'in table',                !!!parse-error (type => 'in table',
3381                                  text => $token->{tag_name}, token => $token);                                text => $token->{tag_name}, token => $token);
3382    
3383                  !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);                !!!insert-element ($token->{tag_name}, $token->{attributes}, $token);
3384                  $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted                $open_tables->[-1]->[2] = 0 if @$open_tables; # ~node inserted
3385    
3386                  ## TODO: form element pointer                ## TODO: form element pointer
3387    
3388                  pop @{$self->{open_elements}};                pop @{$self->{open_elements}};
3389    
3390                  !!!next-token;                !!!next-token;
3391                  !!!ack ('t227.2.1');                !!!ack ('t227.2.1');
3392                  next B;                next B;
               } else {  
                 !!!cp ('t227.2');  
                 #  
               }  
3393              } else {              } else {
3394                !!!cp ('t227.1');                !!!cp ('t227.1');
3395                #                #
# Line 3912  sub _tree_construction_main ($) { Line 3902  sub _tree_construction_main ($) {
3902              ## Reprocess the token.              ## Reprocess the token.
3903              next B;              next B;
3904            }            }
3905            } elsif ($token->{tag_name} eq 'script') {
3906              !!!cp ('t281.3');
3907              ## NOTE: This is an "as if in head" code clone
3908              $script_start_tag->();
3909              next B;
3910          } else {          } else {
3911            !!!cp ('t282');            !!!cp ('t282');
3912            !!!parse-error (type => 'in select',            !!!parse-error (type => 'in select',
# Line 4323  sub _tree_construction_main ($) { Line 4318  sub _tree_construction_main ($) {
4318          $parse_rcdata->(CDATA_CONTENT_MODEL);          $parse_rcdata->(CDATA_CONTENT_MODEL);
4319          next B;          next B;
4320        } elsif ({        } elsif ({
4321                  base => 1, command => 1, eventsource => 1, link => 1,                  base => 1, command => 1, link => 1,
4322                 }->{$token->{tag_name}}) {                 }->{$token->{tag_name}}) {
4323          !!!cp ('t334');          !!!cp ('t334');
4324          ## NOTE: This is an "as if in head" code clone, only "-t" differs          ## NOTE: This is an "as if in head" code clone, only "-t" differs
# Line 4421  sub _tree_construction_main ($) { Line 4416  sub _tree_construction_main ($) {
4416                  center => 1, datagrid => 1, details => 1, dialog => 1,                  center => 1, datagrid => 1, details => 1, dialog => 1,
4417                  dir => 1, div => 1, dl => 1, fieldset => 1, figure => 1,                  dir => 1, div => 1, dl => 1, fieldset => 1, figure => 1,
4418                  footer => 1, h1 => 1, h2 => 1, h3 => 1, h4 => 1, h5 => 1,                  footer => 1, h1 => 1, h2 => 1, h3 => 1, h4 => 1, h5 => 1,
4419                  h6 => 1, header => 1, menu => 1, nav => 1, ol => 1, p => 1,                  h6 => 1, header => 1, hgroup => 1,
4420                    menu => 1, nav => 1, ol => 1, p => 1,
4421                  section => 1, ul => 1,                  section => 1, ul => 1,
4422                  ## NOTE: As normal, but drops leading newline                  ## NOTE: As normal, but drops leading newline
4423                  pre => 1, listing => 1,                  pre => 1, listing => 1,
# Line 4431  sub _tree_construction_main ($) { Line 4427  sub _tree_construction_main ($) {
4427                  table => 1,                  table => 1,
4428                  hr => 1,                  hr => 1,
4429                 }->{$token->{tag_name}}) {                 }->{$token->{tag_name}}) {
4430    
4431            ## 1. When there is an opening |form| element:
4432          if ($token->{tag_name} eq 'form' and defined $self->{form_element}) {          if ($token->{tag_name} eq 'form' and defined $self->{form_element}) {
4433            !!!cp ('t350');            !!!cp ('t350');
4434            !!!parse-error (type => 'in form:form', token => $token);            !!!parse-error (type => 'in form:form', token => $token);
# Line 4440  sub _tree_construction_main ($) { Line 4438  sub _tree_construction_main ($) {
4438            next B;            next B;
4439          }          }
4440    
4441            ## 2. Close the |p| element, if any.
4442          if ($token->{tag_name} ne 'table' or # The Hixie Quirk          if ($token->{tag_name} ne 'table' or # The Hixie Quirk
4443              $self->{document}->manakai_compat_mode ne 'quirks') {              $self->{document}->manakai_compat_mode ne 'quirks') {
4444            ## has a p element in scope            ## has a p element in scope
# Line 4456  sub _tree_construction_main ($) { Line 4455  sub _tree_construction_main ($) {
4455              }              }
4456            } # INSCOPE            } # INSCOPE
4457          }          }
4458              
4459            ## 3. Close the opening <hn> element, if any.
4460            if ({h1 => 1, h2 => 1, h3 => 1,
4461                 h4 => 1, h5 => 1, h6 => 1}->{$token->{tag_name}}) {
4462              if ($self->{open_elements}->[-1]->[1] == HEADING_EL) {
4463                !!!parse-error (type => 'not closed',
4464                                text => $self->{open_elements}->[-1]->[0]->manakai_local_name,
4465                                token => $token);
4466                pop @{$self->{open_elements}};
4467              }
4468            }
4469    
4470            ## 4. Insertion.
4471          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
4472          if ($token->{tag_name} eq 'pre' or $token->{tag_name} eq 'listing') {          if ($token->{tag_name} eq 'pre' or $token->{tag_name} eq 'listing') {
4473            !!!nack ('t346.1');            !!!nack ('t346.1');
# Line 4500  sub _tree_construction_main ($) { Line 4511  sub _tree_construction_main ($) {
4511        } elsif ($token->{tag_name} eq 'li') {        } elsif ($token->{tag_name} eq 'li') {
4512          ## NOTE: As normal, but imply </li> when there's another <li> ...          ## NOTE: As normal, but imply </li> when there's another <li> ...
4513    
4514          ## NOTE: Special, Scope (<li><foo><li> == <li><foo><li/></foo></li>)          ## NOTE: Special, Scope (<li><foo><li> == <li><foo><li/></foo></li>)::
4515            ## Interpreted as <li><foo/></li><li/> (non-conforming)            ## Interpreted as <li><foo/></li><li/> (non-conforming):
4516            ## blockquote (O9.27), center (O), dd (Fx3, O, S3.1.2, IE7),            ## blockquote (O9.27), center (O), dd (Fx3, O, S3.1.2, IE7),
4517            ## dt (Fx, O, S, IE), dl (O), fieldset (O, S, IE), form (Fx, O, S),            ## dt (Fx, O, S, IE), dl (O), fieldset (O, S, IE), form (Fx, O, S),
4518            ## hn (O), pre (O), applet (O, S), button (O, S), marquee (Fx, O, S),            ## hn (O), pre (O), applet (O, S), button (O, S), marquee (Fx, O, S),
4519            ## object (Fx)            ## object (Fx)
4520            ## Generate non-tree (non-conforming)            ## Generate non-tree (non-conforming):
4521            ## basefont (IE7 (where basefont is non-void)), center (IE),            ## basefont (IE7 (where basefont is non-void)), center (IE),
4522            ## form (IE), hn (IE)            ## form (IE), hn (IE)
4523          ## address, div, p (<li><foo><li> == <li><foo/></li><li/>)          ## address, div, p (<li><foo><li> == <li><foo/></li><li/>)::
4524            ## Interpreted as <li><foo><li/></foo></li> (non-conforming)            ## Interpreted as <li><foo><li/></foo></li> (non-conforming):
4525            ## div (Fx, S)            ## div (Fx, S)
4526    
4527          my $non_optional;          my $non_optional;
# Line 4857  sub _tree_construction_main ($) { Line 4868  sub _tree_construction_main ($) {
4868            next B;            next B;
4869          }          }
4870        } elsif ($token->{tag_name} eq 'textarea') {        } elsif ($token->{tag_name} eq 'textarea') {
4871          ## Step 1          ## 1. Insert
4872          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
4873                    
4874          ## Step 2          ## Step 2 # XXX
4875          ## TODO: $self->{form_element} if defined          ## TODO: $self->{form_element} if defined
4876    
4877          ## Step 3          ## 2. Drop U+000A LINE FEED
4878          $self->{ignore_newline} = 1;          $self->{ignore_newline} = 1;
4879    
4880          ## Step 4          ## 3. RCDATA
         ## ISSUE: This step is wrong. (r2302 enbugged)  
   
         ## Step 5  
4881          $self->{content_model} = RCDATA_CONTENT_MODEL;          $self->{content_model} = RCDATA_CONTENT_MODEL;
4882          delete $self->{escape}; # MUST          delete $self->{escape}; # MUST
4883    
4884          ## Step 6-7          ## 4., 6. Insertion mode
4885          $self->{insertion_mode} |= IN_CDATA_RCDATA_IM;          $self->{insertion_mode} |= IN_CDATA_RCDATA_IM;
4886    
4887            ## XXX: 5. frameset-ok flag
4888    
4889          !!!nack ('t392.1');          !!!nack ('t392.1');
4890          !!!next-token;          !!!next-token;
4891          next B;          next B;
# Line 5021  sub _tree_construction_main ($) { Line 5031  sub _tree_construction_main ($) {
5031          } elsif ({          } elsif ({
5032                    area => 1, basefont => 1, bgsound => 1, br => 1,                    area => 1, basefont => 1, bgsound => 1, br => 1,
5033                    embed => 1, img => 1, spacer => 1, wbr => 1,                    embed => 1, img => 1, spacer => 1, wbr => 1,
5034                      keygen => 1,
5035                   }->{$token->{tag_name}}) {                   }->{$token->{tag_name}}) {
5036            !!!cp ('t388.1');            !!!cp ('t388.1');
5037            pop @{$self->{open_elements}};            pop @{$self->{open_elements}};
# Line 5047  sub _tree_construction_main ($) { Line 5058  sub _tree_construction_main ($) {
5058        }        }
5059      } elsif ($token->{type} == END_TAG_TOKEN) {      } elsif ($token->{type} == END_TAG_TOKEN) {
5060        if ($token->{tag_name} eq 'body') {        if ($token->{tag_name} eq 'body') {
5061          ## has a |body| element in scope  
5062            ## 1. If not "have an element in scope":
5063            ## "has a |body| element in scope"
5064          my $i;          my $i;
5065          INSCOPE: {          INSCOPE: {
5066            for (reverse @{$self->{open_elements}}) {            for (reverse @{$self->{open_elements}}) {
# Line 5070  sub _tree_construction_main ($) { Line 5083  sub _tree_construction_main ($) {
5083            next B;            next B;
5084          } # INSCOPE          } # INSCOPE
5085    
5086            ## 2. If unclosed elements:
5087          for (@{$self->{open_elements}}) {          for (@{$self->{open_elements}}) {
5088            unless ($_->[1] & ALL_END_TAG_OPTIONAL_EL ||            unless ($_->[1] & ALL_END_TAG_OPTIONAL_EL ||
5089                    $_->[1] == OPTGROUP_EL ||                    $_->[1] == OPTGROUP_EL ||
# Line 5085  sub _tree_construction_main ($) { Line 5099  sub _tree_construction_main ($) {
5099            }            }
5100          }          }
5101    
5102            ## 3. Switch the insertion mode.
5103          $self->{insertion_mode} = AFTER_BODY_IM;          $self->{insertion_mode} = AFTER_BODY_IM;
5104          !!!next-token;          !!!next-token;
5105          next B;          next B;
# Line 5120  sub _tree_construction_main ($) { Line 5135  sub _tree_construction_main ($) {
5135                  address => 1, article => 1, aside => 1, blockquote => 1,                  address => 1, article => 1, aside => 1, blockquote => 1,
5136                  center => 1, datagrid => 1, details => 1, dialog => 1,                  center => 1, datagrid => 1, details => 1, dialog => 1,
5137                  dir => 1, div => 1, dl => 1, fieldset => 1, figure => 1,                  dir => 1, div => 1, dl => 1, fieldset => 1, figure => 1,
5138                  footer => 1, header => 1, listing => 1, menu => 1, nav => 1,                  footer => 1, header => 1, hgroup => 1,
5139                    listing => 1, menu => 1, nav => 1,
5140                  ol => 1, pre => 1, section => 1, ul => 1,                  ol => 1, pre => 1, section => 1, ul => 1,
5141    
5142                  ## NOTE: As normal, but ... optional tags                  ## NOTE: As normal, but ... optional tags
# Line 5480  sub set_inner_html ($$$$;$) { Line 5496  sub set_inner_html ($$$$;$) {
5496    my $onerror = $_[1];    my $onerror = $_[1];
5497    my $get_wrapper = $_[2] || sub ($) { return $_[0] };    my $get_wrapper = $_[2] || sub ($) { return $_[0] };
5498    
   ## ISSUE: Should {confident} be true?  
   
5499    my $nt = $node->node_type;    my $nt = $node->node_type;
5500    if ($nt == 9) { # Document (invoke the algorithm with no /context/ element)    if ($nt == 9) { # Document (invoke the algorithm with no /context/ element)
5501      # MUST      # MUST
# Line 5696  sub set_inner_html ($$$$;$) { Line 5710  sub set_inner_html ($$$$;$) {
5710        $anode = $anode->parent_node;        $anode = $anode->parent_node;
5711      } # AN      } # AN
5712    
5713        ## F.5. Set the input stream.
5714        $p->{confident} = 1; ## Confident: irrelevant.
5715    
5716      ## F.6. Start the parser.      ## F.6. Start the parser.
5717      {      {
5718        my $self = $p;        my $self = $p;
Legend:
Removed from v.1.223  
changed lines
  Added in v.1.237
[email protected]
 ViewVC Help
Powered by ViewVC 1.1.24