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

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

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

revision 1.21 by wakaba, Sun Oct 19 09:25:21 2008 UTC revision 1.27 by wakaba, Thu Jul 2 22:24:28 2009 UTC
# Line 1740  sub _get_next_token ($) { Line 1740  sub _get_next_token ($) {
1740    
1741          redo A;          redo A;
1742        } else {        } else {
1743          if ($self->{nc} == 0x003D) { # =          if ($self->{nc} == 0x003D or $self->{nc} == 0x003C) { # =, <
1744                        
1745            ## XML5: Not a parse error.            ## XML5: Not a parse error.
1746            $self->{parse_error}->(level => $self->{level}->{must}, type => 'bad attribute value');            $self->{parse_error}->(level => $self->{level}->{must}, type => 'bad attribute value');
# Line 1816  sub _get_next_token ($) { Line 1816  sub _get_next_token ($) {
1816      }      }
1817        
1818          redo A;          redo A;
1819          } elsif ($self->{is_xml} and
1820                   $is_space->{$self->{nc}}) {
1821            
1822            $self->{ca}->{value} .= ' ';
1823            ## Stay in the state.
1824            
1825        if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
1826          $self->{line_prev} = $self->{line};
1827          $self->{column_prev} = $self->{column};
1828          $self->{column}++;
1829          $self->{nc}
1830              = ord substr ($self->{char_buffer}, $self->{char_buffer_pos}++, 1);
1831        } else {
1832          $self->{set_nc}->($self);
1833        }
1834      
1835            redo A;
1836        } elsif ($self->{nc} == -1) {        } elsif ($self->{nc} == -1) {
1837          $self->{parse_error}->(level => $self->{level}->{must}, type => 'unclosed attribute value');          $self->{parse_error}->(level => $self->{level}->{must}, type => 'unclosed attribute value');
1838          if ($self->{ct}->{type} == START_TAG_TOKEN) {          if ($self->{ct}->{type} == START_TAG_TOKEN) {
# Line 1863  sub _get_next_token ($) { Line 1880  sub _get_next_token ($) {
1880          }          }
1881          $self->{ca}->{value} .= chr ($self->{nc});          $self->{ca}->{value} .= chr ($self->{nc});
1882          $self->{read_until}->($self->{ca}->{value},          $self->{read_until}->($self->{ca}->{value},
1883                                q["&<],                                qq["&<\x09\x0C\x20],
1884                                length $self->{ca}->{value});                                length $self->{ca}->{value});
1885    
1886          ## Stay in the state          ## Stay in the state
# Line 1930  sub _get_next_token ($) { Line 1947  sub _get_next_token ($) {
1947      }      }
1948        
1949          redo A;          redo A;
1950          } elsif ($self->{is_xml} and
1951                   $is_space->{$self->{nc}}) {
1952            
1953            $self->{ca}->{value} .= ' ';
1954            ## Stay in the state.
1955            
1956        if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
1957          $self->{line_prev} = $self->{line};
1958          $self->{column_prev} = $self->{column};
1959          $self->{column}++;
1960          $self->{nc}
1961              = ord substr ($self->{char_buffer}, $self->{char_buffer_pos}++, 1);
1962        } else {
1963          $self->{set_nc}->($self);
1964        }
1965      
1966            redo A;
1967        } elsif ($self->{nc} == -1) {        } elsif ($self->{nc} == -1) {
1968          $self->{parse_error}->(level => $self->{level}->{must}, type => 'unclosed attribute value');          $self->{parse_error}->(level => $self->{level}->{must}, type => 'unclosed attribute value');
1969          if ($self->{ct}->{type} == START_TAG_TOKEN) {          if ($self->{ct}->{type} == START_TAG_TOKEN) {
# Line 1977  sub _get_next_token ($) { Line 2011  sub _get_next_token ($) {
2011          }          }
2012          $self->{ca}->{value} .= chr ($self->{nc});          $self->{ca}->{value} .= chr ($self->{nc});
2013          $self->{read_until}->($self->{ca}->{value},          $self->{read_until}->($self->{ca}->{value},
2014                                q['&<],                                qq['&<\x09\x0C\x20],
2015                                length $self->{ca}->{value});                                length $self->{ca}->{value});
2016    
2017          ## Stay in the state          ## Stay in the state
# Line 2149  sub _get_next_token ($) { Line 2183  sub _get_next_token ($) {
2183               0x0022 => 1, # "               0x0022 => 1, # "
2184               0x0027 => 1, # '               0x0027 => 1, # '
2185               0x003D => 1, # =               0x003D => 1, # =
2186                 0x003C => 1, # <
2187              }->{$self->{nc}}) {              }->{$self->{nc}}) {
2188                        
2189            ## XML5: Not a parse error.            ## XML5: Not a parse error.
# Line 2158  sub _get_next_token ($) { Line 2193  sub _get_next_token ($) {
2193          }          }
2194          $self->{ca}->{value} .= chr ($self->{nc});          $self->{ca}->{value} .= chr ($self->{nc});
2195          $self->{read_until}->($self->{ca}->{value},          $self->{read_until}->($self->{ca}->{value},
2196                                q["'=& >],                                qq["'=& \x09\x0C>],
2197                                length $self->{ca}->{value});                                length $self->{ca}->{value});
2198    
2199          ## Stay in the state          ## Stay in the state
# Line 2964  sub _get_next_token ($) { Line 2999  sub _get_next_token ($) {
2999          redo A;          redo A;
3000        } else {        } else {
3001                    
         ## XML5: Not a parse error.  
         $self->{parse_error}->(level => $self->{level}->{must}, type => 'dash in comment',  
                         line => $self->{line_prev},  
                         column => $self->{column_prev});  
3002          $self->{ct}->{data} .= '--' . chr ($self->{nc}); # comment          $self->{ct}->{data} .= '--' . chr ($self->{nc}); # comment
3003          $self->{state} = COMMENT_STATE;          $self->{state} = COMMENT_STATE;
3004                    
# Line 4628  sub _get_next_token ($) { Line 4659  sub _get_next_token ($) {
4659              0x003C => 1, 0x0026 => 1, -1 => 1, # <, &              0x003C => 1, 0x0026 => 1, -1 => 1, # <, &
4660              $self->{entity_add} => 1,              $self->{entity_add} => 1,
4661            }->{$self->{nc}}) {            }->{$self->{nc}}) {
4662                    if ($self->{is_xml}) {
4663              
4664              $self->{parse_error}->(level => $self->{level}->{must}, type => 'bare ero',
4665                              line => $self->{line_prev},
4666                              column => $self->{column_prev}
4667                                  + ($self->{nc} == -1 ? 1 : 0));
4668            } else {
4669              
4670              ## No error
4671            }
4672          ## Don't consume          ## Don't consume
         ## No error  
4673          ## Return nothing.          ## Return nothing.
4674          #          #
4675        } elsif ($self->{nc} == 0x0023) { # #        } elsif ($self->{nc} == 0x0023) { # #
# Line 4649  sub _get_next_token ($) { Line 4688  sub _get_next_token ($) {
4688      }      }
4689        
4690          redo A;          redo A;
4691        } elsif ((0x0041 <= $self->{nc} and        } elsif ($self->{is_xml} or
4692                   (0x0041 <= $self->{nc} and
4693                  $self->{nc} <= 0x005A) or # A..Z                  $self->{nc} <= 0x005A) or # A..Z
4694                 (0x0061 <= $self->{nc} and                 (0x0061 <= $self->{nc} and
4695                  $self->{nc} <= 0x007A)) { # a..z                  $self->{nc} <= 0x007A)) { # a..z
# Line 4828  sub _get_next_token ($) { Line 4868  sub _get_next_token ($) {
4868        my $code = $self->{kwd};        my $code = $self->{kwd};
4869        my $l = $self->{line_prev};        my $l = $self->{line_prev};
4870        my $c = $self->{column_prev};        my $c = $self->{column_prev};
4871        if ($charref_map->{$code}) {        if ((not $self->{is_xml} and $charref_map->{$code}) or
4872              ($self->{is_xml} and 0xD800 <= $code and $code <= 0xDFFF) or
4873              ($self->{is_xml} and $code == 0x0000)) {
4874                    
4875          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',
4876                          text => (sprintf 'U+%04X', $code),                          text => (sprintf 'U+%04X', $code),
# Line 4981  sub _get_next_token ($) { Line 5023  sub _get_next_token ($) {
5023        my $code = $self->{kwd};        my $code = $self->{kwd};
5024        my $l = $self->{line_prev};        my $l = $self->{line_prev};
5025        my $c = $self->{column_prev};        my $c = $self->{column_prev};
5026        if ($charref_map->{$code}) {        if ((not $self->{is_xml} and $charref_map->{$code}) or
5027              ($self->{is_xml} and 0xD800 <= $code and $code <= 0xDFFF) or
5028              ($self->{is_xml} and $code == 0x0000)) {
5029                    
5030          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',
5031                          text => (sprintf 'U+%04X', $code),                          text => (sprintf 'U+%04X', $code),
# Line 5021  sub _get_next_token ($) { Line 5065  sub _get_next_token ($) {
5065             $self->{nc} <= 0x007A) or # z             $self->{nc} <= 0x007A) or # z
5066            (0x0030 <= $self->{nc} and # 0            (0x0030 <= $self->{nc} and # 0
5067             $self->{nc} <= 0x0039) or # 9             $self->{nc} <= 0x0039) or # 9
5068            $self->{nc} == 0x003B) { # ;            $self->{nc} == 0x003B or # ;
5069              ($self->{is_xml} and
5070               not ($is_space->{$self->{nc}} or
5071                    {
5072                      0x003C => 1, 0x0026 => 1, -1 => 1, # <, &
5073                      $self->{entity_add} => 1,
5074                    }->{$self->{nc}}))) {
5075          our $EntityChar;          our $EntityChar;
5076          $self->{kwd} .= chr $self->{nc};          $self->{kwd} .= chr $self->{nc};
5077          if (defined $EntityChar->{$self->{kwd}} or          if (defined $EntityChar->{$self->{kwd}} or
# Line 5454  sub _get_next_token ($) { Line 5504  sub _get_next_token ($) {
5504          ## XML5: Not defined yet.          ## XML5: Not defined yet.
5505    
5506          ## TODO:          ## TODO:
5507    
5508            if (not $self->{stop_processing} and
5509                not $self->{document}->xml_standalone) {
5510              $self->{parse_error}->(level => $self->{level}->{must}, type => 'stop processing', ## TODO: type
5511                              level => $self->{level}->{info});
5512              $self->{stop_processing} = 1;
5513            }
5514    
5515                    
5516      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
5517        $self->{line_prev} = $self->{line};        $self->{line_prev} = $self->{line};
# Line 5888  sub _get_next_token ($) { Line 5946  sub _get_next_token ($) {
5946          }          }
5947          $self->{ct} = {type => ELEMENT_TOKEN, name => '',          $self->{ct} = {type => ELEMENT_TOKEN, name => '',
5948                         line => $self->{line_prev},                         line => $self->{line_prev},
5949                         column => $self->{column_prev} - 6};                         column => $self->{column_prev} - 7};
5950          $self->{state} = DOCTYPE_MD_STATE;          $self->{state} = DOCTYPE_MD_STATE;
5951                    
5952      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
# Line 5956  sub _get_next_token ($) { Line 6014  sub _get_next_token ($) {
6014          $self->{ct} = {type => ATTLIST_TOKEN, name => '',          $self->{ct} = {type => ATTLIST_TOKEN, name => '',
6015                         attrdefs => [],                         attrdefs => [],
6016                         line => $self->{line_prev},                         line => $self->{line_prev},
6017                         column => $self->{column_prev} - 6};                         column => $self->{column_prev} - 7};
6018          $self->{state} = DOCTYPE_MD_STATE;          $self->{state} = DOCTYPE_MD_STATE;
6019                    
6020      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
# Line 6025  sub _get_next_token ($) { Line 6083  sub _get_next_token ($) {
6083          }          }
6084          $self->{ct} = {type => NOTATION_TOKEN, name => '',          $self->{ct} = {type => NOTATION_TOKEN, name => '',
6085                         line => $self->{line_prev},                         line => $self->{line_prev},
6086                         column => $self->{column_prev} - 6};                         column => $self->{column_prev} - 8};
6087          $self->{state} = DOCTYPE_MD_STATE;          $self->{state} = DOCTYPE_MD_STATE;
6088                    
6089      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
# Line 7887  sub _get_next_token ($) { Line 7945  sub _get_next_token ($) {
7945          redo A;          redo A;
7946        }        }
7947      } elsif ($self->{state} == ENTITY_VALUE_ENTITY_STATE) {      } elsif ($self->{state} == ENTITY_VALUE_ENTITY_STATE) {
       ## TODO: XMLize  
   
7948        if ($is_space->{$self->{nc}} or        if ($is_space->{$self->{nc}} or
7949            {            {
7950              0x003C => 1, 0x0026 => 1, -1 => 1, # <, &              0x003C => 1, 0x0026 => 1, -1 => 1, # <, &
7951              $self->{entity_add} => 1,              $self->{entity_add} => 1,
7952            }->{$self->{nc}}) {            }->{$self->{nc}}) {
7953            $self->{parse_error}->(level => $self->{level}->{must}, type => 'bare ero',
7954                            line => $self->{line_prev},
7955                            column => $self->{column_prev}
7956                                + ($self->{nc} == -1 ? 1 : 0));
7957          ## Don't consume          ## Don't consume
         ## No error  
7958          ## Return nothing.          ## Return nothing.
7959          #          #
7960        } elsif ($self->{nc} == 0x0023) { # #        } elsif ($self->{nc} == 0x0023) { # #
# Line 7914  sub _get_next_token ($) { Line 7973  sub _get_next_token ($) {
7973      }      }
7974        
7975          redo A;          redo A;
       } elsif ((0x0041 <= $self->{nc} and  
                 $self->{nc} <= 0x005A) or # A..Z  
                (0x0061 <= $self->{nc} and  
                 $self->{nc} <= 0x007A)) { # a..z  
         #  
7976        } else {        } else {
         $self->{parse_error}->(level => $self->{level}->{must}, type => 'bare ero');  
         ## Return nothing.  
7977          #          #
7978        }        }
7979    
Legend:
Removed from v.1.21  
changed lines
  Added in v.1.27
admin@suikawiki.org
 ViewVC Help
Powered by ViewVC 1.1.24