/[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 1.22 by wakaba, Sun Oct 19 10:12:54 2008 UTC revision 1.26 by wakaba, Thu Jul 2 21:42:43 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 $self->{ca}->{value});
2198    
2199          ## Stay in the state          ## Stay in the state
# Line 4837  sub _get_next_token ($) { Line 4872  sub _get_next_token ($) {
4872        my $code = $self->{kwd};        my $code = $self->{kwd};
4873        my $l = $self->{line_prev};        my $l = $self->{line_prev};
4874        my $c = $self->{column_prev};        my $c = $self->{column_prev};
4875        if ($charref_map->{$code}) {        if ((not $self->{is_xml} and $charref_map->{$code}) or
4876              ($self->{is_xml} and 0xD800 <= $code and $code <= 0xDFFF) or
4877              ($self->{is_xml} and $code == 0x0000)) {
4878                    
4879          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',
4880                          text => (sprintf 'U+%04X', $code),                          text => (sprintf 'U+%04X', $code),
# Line 4990  sub _get_next_token ($) { Line 5027  sub _get_next_token ($) {
5027        my $code = $self->{kwd};        my $code = $self->{kwd};
5028        my $l = $self->{line_prev};        my $l = $self->{line_prev};
5029        my $c = $self->{column_prev};        my $c = $self->{column_prev};
5030        if ($charref_map->{$code}) {        if ((not $self->{is_xml} and $charref_map->{$code}) or
5031              ($self->{is_xml} and 0xD800 <= $code and $code <= 0xDFFF) or
5032              ($self->{is_xml} and $code == 0x0000)) {
5033                    
5034          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',          $self->{parse_error}->(level => $self->{level}->{must}, type => 'invalid character reference',
5035                          text => (sprintf 'U+%04X', $code),                          text => (sprintf 'U+%04X', $code),
# Line 5469  sub _get_next_token ($) { Line 5508  sub _get_next_token ($) {
5508          ## XML5: Not defined yet.          ## XML5: Not defined yet.
5509    
5510          ## TODO:          ## TODO:
5511    
5512            if (not $self->{stop_processing} and
5513                not $self->{document}->xml_standalone) {
5514              $self->{parse_error}->(level => $self->{level}->{must}, type => 'stop processing', ## TODO: type
5515                              level => $self->{level}->{info});
5516              $self->{stop_processing} = 1;
5517            }
5518    
5519                    
5520      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
5521        $self->{line_prev} = $self->{line};        $self->{line_prev} = $self->{line};
# Line 5903  sub _get_next_token ($) { Line 5950  sub _get_next_token ($) {
5950          }          }
5951          $self->{ct} = {type => ELEMENT_TOKEN, name => '',          $self->{ct} = {type => ELEMENT_TOKEN, name => '',
5952                         line => $self->{line_prev},                         line => $self->{line_prev},
5953                         column => $self->{column_prev} - 6};                         column => $self->{column_prev} - 7};
5954          $self->{state} = DOCTYPE_MD_STATE;          $self->{state} = DOCTYPE_MD_STATE;
5955                    
5956      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
# Line 5971  sub _get_next_token ($) { Line 6018  sub _get_next_token ($) {
6018          $self->{ct} = {type => ATTLIST_TOKEN, name => '',          $self->{ct} = {type => ATTLIST_TOKEN, name => '',
6019                         attrdefs => [],                         attrdefs => [],
6020                         line => $self->{line_prev},                         line => $self->{line_prev},
6021                         column => $self->{column_prev} - 6};                         column => $self->{column_prev} - 7};
6022          $self->{state} = DOCTYPE_MD_STATE;          $self->{state} = DOCTYPE_MD_STATE;
6023                    
6024      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {
# Line 6040  sub _get_next_token ($) { Line 6087  sub _get_next_token ($) {
6087          }          }
6088          $self->{ct} = {type => NOTATION_TOKEN, name => '',          $self->{ct} = {type => NOTATION_TOKEN, name => '',
6089                         line => $self->{line_prev},                         line => $self->{line_prev},
6090                         column => $self->{column_prev} - 6};                         column => $self->{column_prev} - 8};
6091          $self->{state} = DOCTYPE_MD_STATE;          $self->{state} = DOCTYPE_MD_STATE;
6092                    
6093      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {      if ($self->{char_buffer_pos} < length $self->{char_buffer}) {

Legend:
Removed from v.1.22  
changed lines
  Added in v.1.26

admin@suikawiki.org
ViewVC Help
Powered by ViewVC 1.1.24