/[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 1.221 by wakaba, Sun Jul 26 10:56:23 2009 UTC revision 1.225 by wakaba, Sun Aug 16 05:57:24 2009 UTC
# Line 1872  sub _tree_construction_main ($) { Line 1872  sub _tree_construction_main ($) {
1872          } elsif ({          } elsif ({
1873                    b => 1, big => 1, blockquote => 1, body => 1, br => 1,                    b => 1, big => 1, blockquote => 1, body => 1, br => 1,
1874                    center => 1, code => 1, dd => 1, div => 1, dl => 1, dt => 1,                    center => 1, code => 1, dd => 1, div => 1, dl => 1, dt => 1,
1875                    em => 1, embed => 1, font => 1, h1 => 1, h2 => 1, h3 => 1,                    em => 1, embed => 1, h1 => 1, h2 => 1, h3 => 1,
1876                    h4 => 1, h5 => 1, h6 => 1, head => 1, hr => 1, i => 1,                    h4 => 1, h5 => 1, h6 => 1, head => 1, hr => 1, i => 1,
1877                    img => 1, li => 1, listing => 1, menu => 1, meta => 1,                    img => 1, li => 1, listing => 1, menu => 1, meta => 1,
1878                    nobr => 1, ol => 1, p => 1, pre => 1, ruby => 1, s => 1,                    nobr => 1, ol => 1, p => 1, pre => 1, ruby => 1, s => 1,
1879                    small => 1, span => 1, strong => 1, strike => 1, sub => 1,                    small => 1, span => 1, strong => 1, strike => 1, sub => 1,
1880                    sup => 1, table => 1, tt => 1, u => 1, ul => 1, var => 1,                    sup => 1, table => 1, tt => 1, u => 1, ul => 1, var => 1,
1881                   }->{$token->{tag_name}}) {                   }->{$token->{tag_name}} or
1882                     ($token->{tag_name} eq 'font' and
1883                      ($token->{attributes}->{color} or
1884                       $token->{attributes}->{face} or
1885                       $token->{attributes}->{size}))) {
1886            !!!cp ('t87.2');            !!!cp ('t87.2');
1887            !!!parse-error (type => 'not closed',            !!!parse-error (type => 'not closed',
1888                            text => $self->{open_elements}->[-1]->[0]                            text => $self->{open_elements}->[-1]->[0]
# Line 2261  sub _tree_construction_main ($) { Line 2265  sub _tree_construction_main ($) {
2265    
2266            ## NOTE: There is a "as if in head" code clone.            ## NOTE: There is a "as if in head" code clone.
2267            $parse_rcdata->(RCDATA_CONTENT_MODEL);            $parse_rcdata->(RCDATA_CONTENT_MODEL);
2268            ## ISSUE: A spec bug [Bug 6038]  
2269              ## NOTE: At this point the stack of open elements contain
2270              ## the |head| element (index == -2) and the |script| element
2271              ## (index == -1).  In the "after head" insertion mode the
2272              ## |head| element is inserted only for the purpose of
2273              ## providing the context for the |script| element, and
2274              ## therefore we can now and have to remove the element from
2275              ## the stack.
2276            splice @{$self->{open_elements}}, -2, 1, () # <head>            splice @{$self->{open_elements}}, -2, 1, () # <head>
2277                if ($self->{insertion_mode} & IM_MASK) == AFTER_HEAD_IM;                if ($self->{insertion_mode} & IM_MASK) == AFTER_HEAD_IM;
2278            next B;            next B;
# Line 2496  sub _tree_construction_main ($) { Line 2507  sub _tree_construction_main ($) {
2507                ## Ignore the token                ## Ignore the token
2508                !!!next-token;                !!!next-token;
2509                next B;                next B;
2510              } elsif ($token->{tag_name} eq 'br') {          } elsif ($token->{tag_name} eq 'br') {
2511                if ($self->{insertion_mode} == BEFORE_HEAD_IM) {            if ($self->{insertion_mode} == BEFORE_HEAD_IM) {
2512                  !!!cp ('t142.2');              !!!cp ('t142.2');
2513                  ## (before head) as if <head>, (in head) as if </head>              ## (before head) as if <head>, (in head) as if </head>
2514                  !!!create-element ($self->{head_element}, $HTML_NS, 'head',, $token);              !!!create-element ($self->{head_element}, $HTML_NS, 'head',, $token);
2515                  $self->{open_elements}->[-1]->[0]->append_child ($self->{head_element});              $self->{open_elements}->[-1]->[0]->append_child ($self->{head_element});
2516                  $self->{insertion_mode} = AFTER_HEAD_IM;              $self->{insertion_mode} = AFTER_HEAD_IM;
2517        
2518                  ## Reprocess in the "after head" insertion mode...              ## Reprocess in the "after head" insertion mode...
2519                } elsif ($self->{insertion_mode} == IN_HEAD_IM) {            } elsif ($self->{insertion_mode} == IN_HEAD_IM) {
2520                  !!!cp ('t143.2');              !!!cp ('t143.2');
2521                  ## As if </head>              ## As if </head>
2522                  pop @{$self->{open_elements}};              pop @{$self->{open_elements}};
2523                  $self->{insertion_mode} = AFTER_HEAD_IM;              $self->{insertion_mode} = AFTER_HEAD_IM;
2524        
2525                  ## Reprocess in the "after head" insertion mode...              ## Reprocess in the "after head" insertion mode...
2526                } elsif ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {            } elsif ($self->{insertion_mode} == IN_HEAD_NOSCRIPT_IM) {
2527                  !!!cp ('t143.3');              !!!cp ('t143.3');
2528                  ## NOTE: Two parse errors for <head><noscript></br>              ## NOTE: Two parse errors for <head><noscript></br>
2529                  !!!parse-error (type => 'unmatched end tag',              !!!parse-error (type => 'unmatched end tag',
2530                                  text => 'br', token => $token);                              text => 'br', token => $token);
2531                  ## As if </noscript>              ## As if </noscript>
2532                  pop @{$self->{open_elements}};              pop @{$self->{open_elements}};
2533                  $self->{insertion_mode} = IN_HEAD_IM;              $self->{insertion_mode} = IN_HEAD_IM;
2534    
2535                  ## Reprocess in the "in head" insertion mode...              ## Reprocess in the "in head" insertion mode...
2536                  ## As if </head>              ## As if </head>
2537                  pop @{$self->{open_elements}};              pop @{$self->{open_elements}};
2538                  $self->{insertion_mode} = AFTER_HEAD_IM;              $self->{insertion_mode} = AFTER_HEAD_IM;
2539    
2540                  ## Reprocess in the "after head" insertion mode...              ## Reprocess in the "after head" insertion mode...
2541                } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {            } elsif ($self->{insertion_mode} == AFTER_HEAD_IM) {
2542                  !!!cp ('t143.4');              !!!cp ('t143.4');
2543                  #              #
2544                } else {            } else {
2545                  die "$0: $self->{insertion_mode}: Unknown insertion mode";              die "$0: $self->{insertion_mode}: Unknown insertion mode";
2546                }            }
2547    
2548                ## ISSUE: does not agree with IE7 - it doesn't ignore </br>.            #
2549                !!!parse-error (type => 'unmatched end tag',          } else { ## Other end tags
                               text => 'br', token => $token);  
               ## Ignore the token  
               !!!next-token;  
               next B;  
             } else {  
2550                !!!cp ('t145');                !!!cp ('t145');
2551                !!!parse-error (type => 'unmatched end tag',                !!!parse-error (type => 'unmatched end tag',
2552                                text => $token->{tag_name}, token => $token);                                text => $token->{tag_name}, token => $token);
# Line 2584  sub _tree_construction_main ($) { Line 2590  sub _tree_construction_main ($) {
2590              !!!insert-element ('body',, $token);              !!!insert-element ('body',, $token);
2591              $self->{insertion_mode} = IN_BODY_IM;              $self->{insertion_mode} = IN_BODY_IM;
2592              ## reprocess              ## reprocess
2593              next B;          next B;
2594        } elsif ($token->{type} == END_OF_FILE_TOKEN) {        } elsif ($token->{type} == END_OF_FILE_TOKEN) {
2595          if ($self->{insertion_mode} == BEFORE_HEAD_IM) {          if ($self->{insertion_mode} == BEFORE_HEAD_IM) {
2596            !!!cp ('t149.1');            !!!cp ('t149.1');
# Line 3850  sub _tree_construction_main ($) { Line 3856  sub _tree_construction_main ($) {
3856                     tbody => 1, tfoot => 1, thead => 1,                     tbody => 1, tfoot => 1, thead => 1,
3857                     tr => 1, td => 1, th => 1,                     tr => 1, td => 1, th => 1,
3858                    }->{$token->{tag_name}})) {                    }->{$token->{tag_name}})) {
3859            ## TODO: The type below is not good - <select> is replaced by </select>  
3860            !!!parse-error (type => 'not closed', text => 'select',            ## 1. Parse error.
3861                            token => $token);            if ($token->{tag_name} eq 'select') {
3862            ## NOTE: As if the token were </select> (<select> case) or                !!!parse-error (type => 'select in select', ## XXX: documentation
3863            ## as if there were </select> (otherwise).                                token => $token);
3864            ## have an element in table scope            } else {
3865                !!!parse-error (type => 'not closed', text => 'select',
3866                                token => $token);
3867              }
3868    
3869              ## 2./<select>-1. Unless "have an element in table scope" (select):
3870            my $i;            my $i;
3871            INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {            INSCOPE: for (reverse 0..$#{$self->{open_elements}}) {
3872              my $node = $self->{open_elements}->[$_];              my $node = $self->{open_elements}->[$_];
# Line 3870  sub _tree_construction_main ($) { Line 3881  sub _tree_construction_main ($) {
3881            } # INSCOPE            } # INSCOPE
3882            unless (defined $i) {            unless (defined $i) {
3883              !!!cp ('t280');              !!!cp ('t280');
3884              !!!parse-error (type => 'unmatched end tag',              if ($token->{tag_name} eq 'select') {
3885                              text => 'select', token => $token);                ## NOTE: This error would be raised when
3886              ## Ignore the token                ## |select.innerHTML = '<select>'| is executed; in this
3887                  ## case two errors, "select in select" and "unmatched
3888                  ## end tags" are reported to the user, the latter might
3889                  ## be confusing but this is what the spec requires.
3890                  !!!parse-error (type => 'unmatched end tag',
3891                                  text => 'select',
3892                                  token => $token);
3893                }
3894                ## Ignore the token.
3895              !!!nack ('t280.1');              !!!nack ('t280.1');
3896              !!!next-token;              !!!next-token;
3897              next B;              next B;
3898            }            }
3899    
3900              ## 3. Otherwise, as if there were <select>:
3901                                
3902            !!!cp ('t281');            !!!cp ('t281');
3903            splice @{$self->{open_elements}}, $i;            splice @{$self->{open_elements}}, $i;
# Line 4412  sub _tree_construction_main ($) { Line 4433  sub _tree_construction_main ($) {
4433                  table => 1,                  table => 1,
4434                  hr => 1,                  hr => 1,
4435                 }->{$token->{tag_name}}) {                 }->{$token->{tag_name}}) {
4436    
4437            ## 1. When there is an opening |form| element:
4438          if ($token->{tag_name} eq 'form' and defined $self->{form_element}) {          if ($token->{tag_name} eq 'form' and defined $self->{form_element}) {
4439            !!!cp ('t350');            !!!cp ('t350');
4440            !!!parse-error (type => 'in form:form', token => $token);            !!!parse-error (type => 'in form:form', token => $token);
# Line 4421  sub _tree_construction_main ($) { Line 4444  sub _tree_construction_main ($) {
4444            next B;            next B;
4445          }          }
4446    
4447            ## 2. Close the |p| element, if any.
4448          if ($token->{tag_name} ne 'table' or # The Hixie Quirk          if ($token->{tag_name} ne 'table' or # The Hixie Quirk
4449              $self->{document}->manakai_compat_mode ne 'quirks') {              $self->{document}->manakai_compat_mode ne 'quirks') {
4450            ## has a p element in scope            ## has a p element in scope
# Line 4437  sub _tree_construction_main ($) { Line 4461  sub _tree_construction_main ($) {
4461              }              }
4462            } # INSCOPE            } # INSCOPE
4463          }          }
4464              
4465            ## 3. Close the opening <hn> element, if any.
4466            if ({h1 => 1, h2 => 1, h3 => 1,
4467                 h4 => 1, h5 => 1, h6 => 1}->{$token->{tag_name}}) {
4468              if ($self->{open_elements}->[-1]->[1] == HEADING_EL) {
4469                !!!parse-error (type => 'not closed',
4470                                text => $self->{open_elements}->[-1]->[0]->manakai_local_name,
4471                                token => $token);
4472                pop @{$self->{open_elements}};
4473              }
4474            }
4475    
4476            ## 4. Insertion.
4477          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
4478          if ($token->{tag_name} eq 'pre' or $token->{tag_name} eq 'listing') {          if ($token->{tag_name} eq 'pre' or $token->{tag_name} eq 'listing') {
4479            !!!nack ('t346.1');            !!!nack ('t346.1');
# Line 4481  sub _tree_construction_main ($) { Line 4517  sub _tree_construction_main ($) {
4517        } elsif ($token->{tag_name} eq 'li') {        } elsif ($token->{tag_name} eq 'li') {
4518          ## NOTE: As normal, but imply </li> when there's another <li> ...          ## NOTE: As normal, but imply </li> when there's another <li> ...
4519    
4520          ## NOTE: Special, Scope (<li><foo><li> == <li><foo><li/></foo></li>)          ## NOTE: Special, Scope (<li><foo><li> == <li><foo><li/></foo></li>)::
4521            ## Interpreted as <li><foo/></li><li/> (non-conforming)            ## Interpreted as <li><foo/></li><li/> (non-conforming):
4522            ## blockquote (O9.27), center (O), dd (Fx3, O, S3.1.2, IE7),            ## blockquote (O9.27), center (O), dd (Fx3, O, S3.1.2, IE7),
4523            ## 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),
4524            ## 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),
4525            ## object (Fx)            ## object (Fx)
4526            ## Generate non-tree (non-conforming)            ## Generate non-tree (non-conforming):
4527            ## basefont (IE7 (where basefont is non-void)), center (IE),            ## basefont (IE7 (where basefont is non-void)), center (IE),
4528            ## form (IE), hn (IE)            ## form (IE), hn (IE)
4529          ## address, div, p (<li><foo><li> == <li><foo/></li><li/>)          ## address, div, p (<li><foo><li> == <li><foo/></li><li/>)::
4530            ## Interpreted as <li><foo><li/></foo></li> (non-conforming)            ## Interpreted as <li><foo><li/></foo></li> (non-conforming):
4531            ## div (Fx, S)            ## div (Fx, S)
4532    
4533          my $non_optional;          my $non_optional;
# Line 4838  sub _tree_construction_main ($) { Line 4874  sub _tree_construction_main ($) {
4874            next B;            next B;
4875          }          }
4876        } elsif ($token->{tag_name} eq 'textarea') {        } elsif ($token->{tag_name} eq 'textarea') {
4877          ## Step 1          ## 1. Insert
4878          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);          !!!insert-element-t ($token->{tag_name}, $token->{attributes}, $token);
4879                    
4880          ## Step 2          ## Step 2 # XXX
4881          ## TODO: $self->{form_element} if defined          ## TODO: $self->{form_element} if defined
4882    
4883          ## Step 3          ## 2. Drop U+000A LINE FEED
4884          $self->{ignore_newline} = 1;          $self->{ignore_newline} = 1;
4885    
4886          ## Step 4          ## 3. RCDATA
         ## ISSUE: This step is wrong. (r2302 enbugged)  
   
         ## Step 5  
4887          $self->{content_model} = RCDATA_CONTENT_MODEL;          $self->{content_model} = RCDATA_CONTENT_MODEL;
4888          delete $self->{escape}; # MUST          delete $self->{escape}; # MUST
4889    
4890          ## Step 6-7          ## 4., 6. Insertion mode
4891          $self->{insertion_mode} |= IN_CDATA_RCDATA_IM;          $self->{insertion_mode} |= IN_CDATA_RCDATA_IM;
4892    
4893            ## XXX: 5. frameset-ok flag
4894    
4895          !!!nack ('t392.1');          !!!nack ('t392.1');
4896          !!!next-token;          !!!next-token;
4897          next B;          next B;
# Line 5028  sub _tree_construction_main ($) { Line 5063  sub _tree_construction_main ($) {
5063        }        }
5064      } elsif ($token->{type} == END_TAG_TOKEN) {      } elsif ($token->{type} == END_TAG_TOKEN) {
5065        if ($token->{tag_name} eq 'body') {        if ($token->{tag_name} eq 'body') {
5066          ## has a |body| element in scope  
5067            ## 1. If not "have an element in scope":
5068            ## "has a |body| element in scope"
5069          my $i;          my $i;
5070          INSCOPE: {          INSCOPE: {
5071            for (reverse @{$self->{open_elements}}) {            for (reverse @{$self->{open_elements}}) {
# Line 5051  sub _tree_construction_main ($) { Line 5088  sub _tree_construction_main ($) {
5088            next B;            next B;
5089          } # INSCOPE          } # INSCOPE
5090    
5091            ## 2. If unclosed elements:
5092          for (@{$self->{open_elements}}) {          for (@{$self->{open_elements}}) {
5093            unless ($_->[1] & ALL_END_TAG_OPTIONAL_EL ||            unless ($_->[1] & ALL_END_TAG_OPTIONAL_EL ||
5094                    $_->[1] == OPTGROUP_EL ||                    $_->[1] == OPTGROUP_EL ||
# Line 5066  sub _tree_construction_main ($) { Line 5104  sub _tree_construction_main ($) {
5104            }            }
5105          }          }
5106    
5107            ## 3. Switch the insertion mode.
5108          $self->{insertion_mode} = AFTER_BODY_IM;          $self->{insertion_mode} = AFTER_BODY_IM;
5109          !!!next-token;          !!!next-token;
5110          next B;          next B;
Legend:
Removed from v.1.221  
changed lines
  Added in v.1.225
admin@suikawiki.org
 ViewVC Help
Powered by ViewVC 1.1.24