PageRenderTime 71ms CodeModel.GetById 20ms RepoModel.GetById 0ms app.codeStats 1ms

/lib/htmlawed/htmLawed.php

https://gitlab.com/OnBlox/OnBlox-Template
PHP | 734 lines | 636 code | 29 blank | 69 comment | 327 complexity | d8dc619502c99283c77c363bd76cd495 MD5 | raw file
    

  1. <?php
    2. 

  2. 

  3. /*
    4. 

  4. htmLawed 1.1.21, 27 February 2016
    5. 

  5. Copyright Santosh Patnaik
    6. 

  6. Dual licensed with LGPL 3 and GPL 2+
    7. 

  7. A PHP Labware internal utility; www.bioinformatics.org/phplabware/internal_utilities/htmLawed
    8. 

    9. 

  9. See htmLawed_README.txt/htm
    10. 

  10. */
    11. 

  11. 

  12. function htmLawed($t, $C=1, $S=array()){
    13. 

  13. $C = is_array($C) ? $C : array();
    14. 

  14. if(!empty($C['valid_xhtml'])){
    15. 

  15.  $C['elements'] = empty($C['elements']) ? '*-center-dir-font-isindex-menu-s-strike-u' : $C['elements'];
    16. 

  16.  $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 2;
    17. 

  17.  $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 2;
    18. 

  18. }
    19. 

  19. // config eles
    20. 

  20. $e = array('a'=>1, 'abbr'=>1, 'acronym'=>1, 'address'=>1, 'applet'=>1, 'area'=>1, 'b'=>1, 'bdo'=>1, 'big'=>1, 'blockquote'=>1, 'br'=>1, 'button'=>1, 'caption'=>1, 'center'=>1, 'cite'=>1, 'code'=>1, 'col'=>1, 'colgroup'=>1, 'dd'=>1, 'del'=>1, 'dfn'=>1, 'dir'=>1, 'div'=>1, 'dl'=>1, 'dt'=>1, 'em'=>1, 'embed'=>1, 'fieldset'=>1, 'font'=>1, 'form'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'hr'=>1, 'i'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'ins'=>1, 'isindex'=>1, 'kbd'=>1, 'label'=>1, 'legend'=>1, 'li'=>1, 'map'=>1, 'menu'=>1, 'noscript'=>1, 'object'=>1, 'ol'=>1, 'optgroup'=>1, 'option'=>1, 'p'=>1, 'param'=>1, 'pre'=>1, 'q'=>1, 'rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1, 'ruby'=>1, 's'=>1, 'samp'=>1, 'script'=>1, 'select'=>1, 'small'=>1, 'span'=>1, 'strike'=>1, 'strong'=>1, 'sub'=>1, 'sup'=>1, 'table'=>1, 'tbody'=>1, 'td'=>1, 'textarea'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1, 'tt'=>1, 'u'=>1, 'ul'=>1, 'var'=>1); // 86/deprecated+embed+ruby
    21. 

  21. if(!empty($C['safe'])){
    22. 

  22.  unset($e['applet'], $e['embed'], $e['iframe'], $e['object'], $e['script']);
    23. 

  23. }
    24. 

  24. $x = !empty($C['elements']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['elements']) : '*';
    25. 

  25. if($x == '-*'){$e = array();}
    26. 

  26. elseif(strpos($x, '*') === false){$e = array_flip(explode(',', $x));}
    27. 

  27. else{
    28. 

  28.  if(isset($x[1])){
    29. 

  29.   preg_match_all('`(?:^|-|\+)[^\-+]+?(?=-|\+|$)`', $x, $m, PREG_SET_ORDER);
    30. 

  30.   for($i=count($m); --$i>=0;){$m[$i] = $m[$i][0];}
    31. 

  31.   foreach($m as $v){
    32. 

  32.    if($v[0] == '+'){$e[substr($v, 1)] = 1;}
    33. 

  33.    if($v[0] == '-' && isset($e[($v = substr($v, 1))]) && !in_array('+'. $v, $m)){unset($e[$v]);}
    34. 

  34.   }
    35. 

  35.  }
    36. 

  36. }
    37. 

  37. $C['elements'] =& $e;
    38. 

  38. // config attrs
    39. 

  39. $x = !empty($C['deny_attribute']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['deny_attribute']) : '';
    40. 

  40. $x = array_flip((isset($x[0]) && $x[0] == '*') ? explode('-', $x) : explode(',', $x. (!empty($C['safe']) ? ',on*' : '')));
    41. 

  41. if(isset($x['on*'])){
    42. 

  42.  unset($x['on*']);
    43. 

  43.  $x += array('onblur'=>1, 'onchange'=>1, 'onclick'=>1, 'ondblclick'=>1, 'onfocus'=>1, 'onkeydown'=>1, 'onkeypress'=>1, 'onkeyup'=>1, 'onmousedown'=>1, 'onmousemove'=>1, 'onmouseout'=>1, 'onmouseover'=>1, 'onmouseup'=>1, 'onreset'=>1, 'onselect'=>1, 'onsubmit'=>1);
    44. 

  44. }
    45. 

  45. $C['deny_attribute'] = $x;
    46. 

  46. // config URL
    47. 

  47. $x = (isset($C['schemes'][2]) && strpos($C['schemes'], ':')) ? strtolower($C['schemes']) : 'href: aim, feed, file, ftp, gopher, http, https, irc, mailto, news, nntp, sftp, ssh, telnet; *:file, http, https';
    48. 

  48. $C['schemes'] = array();
    49. 

  49. foreach(explode(';', str_replace(array(' ', "\t", "\r", "\n"), '', $x)) as $v){
    50. 

  50.  $x = $x2 = null; list($x, $x2) = explode(':', $v, 2);
    51. 

  51.  if($x2){$C['schemes'][$x] = array_flip(explode(',', $x2));}
    52. 

  52. }
    53. 

  53. if(!isset($C['schemes']['*'])){$C['schemes']['*'] = array('file'=>1, 'http'=>1, 'https'=>1,);}
    54. 

  54. if(!empty($C['safe']) && empty($C['schemes']['style'])){$C['schemes']['style'] = array('!'=>1);}
    55. 

  55. $C['abs_url'] = isset($C['abs_url']) ? $C['abs_url'] : 0;
    56. 

  56. if(!isset($C['base_url']) or !preg_match('`^[a-zA-Z\d.+\-]+://[^/]+/(.+?/)?$`', $C['base_url'])){
    57. 

  57.  $C['base_url'] = $C['abs_url'] = 0;
    58. 

  58. }
    59. 

  59. // config rest
    60. 

  60. $C['and_mark'] = empty($C['and_mark']) ? 0 : 1;
    61. 

  61. $C['anti_link_spam'] = (isset($C['anti_link_spam']) && is_array($C['anti_link_spam']) && count($C['anti_link_spam']) == 2 && (empty($C['anti_link_spam'][0]) or hl_regex($C['anti_link_spam'][0])) && (empty($C['anti_link_spam'][1]) or hl_regex($C['anti_link_spam'][1]))) ? $C['anti_link_spam'] : 0;
    62. 

  62. $C['anti_mail_spam'] = isset($C['anti_mail_spam']) ? $C['anti_mail_spam'] : 0;
    63. 

  63. $C['balance'] = isset($C['balance']) ? (bool)$C['balance'] : 1;
    64. 

  64. $C['cdata'] = isset($C['cdata']) ? $C['cdata'] : (empty($C['safe']) ? 3 : 0);
    65. 

  65. $C['clean_ms_char'] = empty($C['clean_ms_char']) ? 0 : $C['clean_ms_char'];
    66. 

  66. $C['comment'] = isset($C['comment']) ? $C['comment'] : (empty($C['safe']) ? 3 : 0);
    67. 

  67. $C['css_expression'] = empty($C['css_expression']) ? 0 : 1;
    68. 

  68. $C['direct_list_nest'] = empty($C['direct_list_nest']) ? 0 : 1;
    69. 

  69. $C['hexdec_entity'] = isset($C['hexdec_entity']) ? $C['hexdec_entity'] : 1;
    70. 

  70. $C['hook'] = (!empty($C['hook']) && function_exists($C['hook'])) ? $C['hook'] : 0;
    71. 

  71. $C['hook_tag'] = (!empty($C['hook_tag']) && function_exists($C['hook_tag'])) ? $C['hook_tag'] : 0;
    72. 

  72. $C['keep_bad'] = isset($C['keep_bad']) ? $C['keep_bad'] : 6;
    73. 

  73. $C['lc_std_val'] = isset($C['lc_std_val']) ? (bool)$C['lc_std_val'] : 1;
    74. 

  74. $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 1;
    75. 

  75. $C['named_entity'] = isset($C['named_entity']) ? (bool)$C['named_entity'] : 1;
    76. 

  76. $C['no_deprecated_attr'] = isset($C['no_deprecated_attr']) ? $C['no_deprecated_attr'] : 1;
    77. 

  77. $C['parent'] = isset($C['parent'][0]) ? strtolower($C['parent']) : 'body';
    78. 

  78. $C['show_setting'] = !empty($C['show_setting']) ? $C['show_setting'] : 0;
    79. 

  79. $C['style_pass'] = empty($C['style_pass']) ? 0 : 1;
    80. 

  80. $C['tidy'] = empty($C['tidy']) ? 0 : $C['tidy'];
    81. 

  81. $C['unique_ids'] = isset($C['unique_ids']) ? $C['unique_ids'] : 1;
    82. 

  82. $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 0;
    83. 

  83. 

  84. if(isset($GLOBALS['C'])){$reC = $GLOBALS['C'];}
    85. 

  85. $GLOBALS['C'] = $C;
    86. 

  86. $S = is_array($S) ? $S : hl_spec($S);
    87. 

  87. if(isset($GLOBALS['S'])){$reS = $GLOBALS['S'];}
    88. 

  88. $GLOBALS['S'] = $S;
    89. 

  89. 

  90. $t = preg_replace('`[\x00-\x08\x0b-\x0c\x0e-\x1f]`', '', $t);
    91. 

  91. if($C['clean_ms_char']){
    92. 

  92.  $x = array("\x7f"=>'', "\x80"=>'&#8364;', "\x81"=>'', "\x83"=>'&#402;', "\x85"=>'&#8230;', "\x86"=>'&#8224;', "\x87"=>'&#8225;', "\x88"=>'&#710;', "\x89"=>'&#8240;', "\x8a"=>'&#352;', "\x8b"=>'&#8249;', "\x8c"=>'&#338;', "\x8d"=>'', "\x8e"=>'&#381;', "\x8f"=>'', "\x90"=>'', "\x95"=>'&#8226;', "\x96"=>'&#8211;', "\x97"=>'&#8212;', "\x98"=>'&#732;', "\x99"=>'&#8482;', "\x9a"=>'&#353;', "\x9b"=>'&#8250;', "\x9c"=>'&#339;', "\x9d"=>'', "\x9e"=>'&#382;', "\x9f"=>'&#376;');
    93. 

  93.  $x = $x + ($C['clean_ms_char'] == 1 ? array("\x82"=>'&#8218;', "\x84"=>'&#8222;', "\x91"=>'&#8216;', "\x92"=>'&#8217;', "\x93"=>'&#8220;', "\x94"=>'&#8221;') : array("\x82"=>'\'', "\x84"=>'"', "\x91"=>'\'', "\x92"=>'\'', "\x93"=>'"', "\x94"=>'"'));
    94. 

  94.  $t = strtr($t, $x);
    95. 

  95. }
    96. 

  96. if($C['cdata'] or $C['comment']){$t = preg_replace_callback('`<!(?:(?:--.*?--)|(?:\[CDATA\[.*?\]\]))>`sm', 'hl_cmtcd', $t);}
    97. 

  97. $t = preg_replace_callback('`&amp;([A-Za-z][A-Za-z0-9]{1,30}|#(?:[0-9]{1,8}|[Xx][0-9A-Fa-f]{1,7}));`', 'hl_ent', str_replace('&', '&amp;', $t));
    98. 

  98. if($C['unique_ids'] && !isset($GLOBALS['hl_Ids'])){$GLOBALS['hl_Ids'] = array();}
    99. 

  99. if($C['hook']){$t = $C['hook']($t, $C, $S);}
    100. 

  100. if($C['show_setting'] && preg_match('`^[a-z][a-z0-9_]*$`i', $C['show_setting'])){
    101. 

  101.  $GLOBALS[$C['show_setting']] = array('config'=>$C, 'spec'=>$S, 'time'=>microtime());
    102. 

  102. }
    103. 

  103. 

  104. // Neutralize not well formatted html tags
    105. 

  105. if ($C['keep_bad'] == 1) {
    106. 

  106.    if (preg_match_all('/<[^\s><\/]+/', $t, $matches)) {
    107. 

  107.       foreach ($matches[0] as $match) {
    108. 

  108.          $search = str_replace('<', '', $match);
    109. 

  109.          if (!in_array($search, array_keys($C['elements']))) {
    110. 

  110.             $t = str_replace($match, '&lt;'.$search, $t);
    111. 

  111.          }
    112. 

  112.       }
    113. 

  113.    }
    114. 

  114. }
    115. 

  115. 

  116. // main
    117. 

  117. $t = preg_replace_callback('`<(?:(?:\s|$)|(?:[^>]*(?:>|$)))|>`m', 'hl_tag', $t);
    118. 

  118. $t = $C['balance'] ? hl_bal($t, $C['keep_bad'], $C['parent']) : $t;
    119. 

  119. $t = (($C['cdata'] or $C['comment']) && strpos($t, "\x01") !== false) ? str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05"), array('', '', '&', '<', '>'), $t) : $t;
    120. 

  120. $t = $C['tidy'] ? hl_tidy($t, $C['tidy'], $C['parent']) : $t;
    121. 

  121. unset($C, $e);
    122. 

  122. if(isset($reC)){$GLOBALS['C'] = $reC;}
    123. 

  123. if(isset($reS)){$GLOBALS['S'] = $reS;}
    124. 

  124. return $t;
    125. 

  125. // eof
    126. 

  126. }
    127. 

  127. 

  128. function hl_attrval($t, $p){
    129. 

  129. // check attr val against $S
    130. 

  130. $o = 1; $l = strlen($t);
    131. 

  131. foreach($p as $k=>$v){
    132. 

  132.  switch($k){
    133. 

  133.   case 'maxlen':if($l > $v){$o = 0;}
    134. 

  134.   break; case 'minlen': if($l < $v){$o = 0;}
    135. 

  135.   break; case 'maxval': if((float)($t) > $v){$o = 0;}
    136. 

  136.   break; case 'minval': if((float)($t) < $v){$o = 0;}
    137. 

  137.   break; case 'match': if(!preg_match($v, $t)){$o = 0;}
    138. 

  138.   break; case 'nomatch': if(preg_match($v, $t)){$o = 0;}
    139. 

  139.   break; case 'oneof':
    140. 

  140.    $m = 0;
    141. 

  141.    foreach(explode('|', $v) as $n){if($t == $n){$m = 1; break;}}
    142. 

  142.    $o = $m;
    143. 

  143.   break; case 'noneof':
    144. 

  144.    $m = 1;
    145. 

  145.    foreach(explode('|', $v) as $n){if($t == $n){$m = 0; break;}}
    146. 

  146.    $o = $m;
    147. 

  147.   break; default:
    148. 

  148.   break;
    149. 

  149.  }
    150. 

  150.  if(!$o){break;}
    151. 

  151. }
    152. 

  152. return ($o ? $t : (isset($p['default']) ? $p['default'] : 0));
    153. 

  153. // eof
    154. 

  154. }
    155. 

  155. 

  156. function hl_bal($t, $do=1, $in='div'){
    157. 

  157. // balance tags
    158. 

  158. // by content
    159. 

  159. $cB = array('blockquote'=>1, 'form'=>1, 'map'=>1, 'noscript'=>1); // Block
    160. 

  160. $cE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty
    161. 

  161. $cF = array('button'=>1, 'del'=>1, 'div'=>1, 'dd'=>1, 'fieldset'=>1, 'iframe'=>1, 'ins'=>1, 'li'=>1, 'noscript'=>1, 'object'=>1, 'td'=>1, 'th'=>1); // Flow; later context-wise dynamic move of ins & del to $cI
    162. 

  162. $cI = array('a'=>1, 'abbr'=>1, 'acronym'=>1, 'address'=>1, 'b'=>1, 'bdo'=>1, 'big'=>1, 'caption'=>1, 'cite'=>1, 'code'=>1, 'dfn'=>1, 'dt'=>1, 'em'=>1, 'font'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'i'=>1, 'kbd'=>1, 'label'=>1, 'legend'=>1, 'p'=>1, 'pre'=>1, 'q'=>1, 'rb'=>1, 'rt'=>1, 's'=>1, 'samp'=>1, 'small'=>1, 'span'=>1, 'strike'=>1, 'strong'=>1, 'sub'=>1, 'sup'=>1, 'tt'=>1, 'u'=>1, 'var'=>1); // Inline
    163. 

  163. $cN = array('a'=>array('a'=>1), 'button'=>array('a'=>1, 'button'=>1, 'fieldset'=>1, 'form'=>1, 'iframe'=>1, 'input'=>1, 'label'=>1, 'select'=>1, 'textarea'=>1), 'fieldset'=>array('fieldset'=>1), 'form'=>array('form'=>1), 'label'=>array('label'=>1), 'noscript'=>array('script'=>1), 'pre'=>array('big'=>1, 'font'=>1, 'img'=>1, 'object'=>1, 'script'=>1, 'small'=>1, 'sub'=>1, 'sup'=>1), 'rb'=>array('ruby'=>1), 'rt'=>array('ruby'=>1)); // Illegal
    164. 

  164. $cN2 = array_keys($cN);
    165. 

  165. $cR = array('blockquote'=>1, 'dir'=>1, 'dl'=>1, 'form'=>1, 'map'=>1, 'menu'=>1, 'noscript'=>1, 'ol'=>1, 'optgroup'=>1, 'rbc'=>1, 'rtc'=>1, 'ruby'=>1, 'select'=>1, 'table'=>1, 'tbody'=>1, 'tfoot'=>1, 'thead'=>1, 'tr'=>1, 'ul'=>1);
    166. 

  166. $cS = array('colgroup'=>array('col'=>1), 'dir'=>array('li'=>1), 'dl'=>array('dd'=>1, 'dt'=>1), 'menu'=>array('li'=>1), 'ol'=>array('li'=>1), 'optgroup'=>array('option'=>1), 'option'=>array('#pcdata'=>1), 'rbc'=>array('rb'=>1), 'rp'=>array('#pcdata'=>1), 'rtc'=>array('rt'=>1), 'ruby'=>array('rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1), 'select'=>array('optgroup'=>1, 'option'=>1), 'script'=>array('#pcdata'=>1), 'table'=>array('caption'=>1, 'col'=>1, 'colgroup'=>1, 'tfoot'=>1, 'tbody'=>1, 'tr'=>1, 'thead'=>1), 'tbody'=>array('tr'=>1), 'tfoot'=>array('tr'=>1), 'textarea'=>array('#pcdata'=>1), 'thead'=>array('tr'=>1), 'tr'=>array('td'=>1, 'th'=>1), 'ul'=>array('li'=>1)); // Specific - immediate parent-child
    167. 

  167. if($GLOBALS['C']['direct_list_nest']){$cS['ol'] = $cS['ul'] += array('ol'=>1, 'ul'=>1);}
    168. 

  168. $cO = array('address'=>array('p'=>1), 'applet'=>array('param'=>1), 'blockquote'=>array('script'=>1), 'fieldset'=>array('legend'=>1, '#pcdata'=>1), 'form'=>array('script'=>1), 'map'=>array('area'=>1), 'object'=>array('param'=>1, 'embed'=>1)); // Other
    169. 

  169. $cT = array('colgroup'=>1, 'dd'=>1, 'dt'=>1, 'li'=>1, 'option'=>1, 'p'=>1, 'td'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1); // Omitable closing
    170. 

  170. // block/inline type; ins & del both type; #pcdata: text
    171. 

  171. $eB = array('address'=>1, 'blockquote'=>1, 'center'=>1, 'del'=>1, 'dir'=>1, 'dl'=>1, 'div'=>1, 'fieldset'=>1, 'form'=>1, 'ins'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'hr'=>1, 'isindex'=>1, 'menu'=>1, 'noscript'=>1, 'ol'=>1, 'p'=>1, 'pre'=>1, 'table'=>1, 'ul'=>1);
    172. 

  172. $eI = array('#pcdata'=>1, 'a'=>1, 'abbr'=>1, 'acronym'=>1, 'applet'=>1, 'b'=>1, 'bdo'=>1, 'big'=>1, 'br'=>1, 'button'=>1, 'cite'=>1, 'code'=>1, 'del'=>1, 'dfn'=>1, 'em'=>1, 'embed'=>1, 'font'=>1, 'i'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'ins'=>1, 'kbd'=>1, 'label'=>1, 'map'=>1, 'object'=>1, 'q'=>1, 'ruby'=>1, 's'=>1, 'samp'=>1, 'select'=>1, 'script'=>1, 'small'=>1, 'span'=>1, 'strike'=>1, 'strong'=>1, 'sub'=>1, 'sup'=>1, 'textarea'=>1, 'tt'=>1, 'u'=>1, 'var'=>1);
    173. 

  173. $eN = array('a'=>1, 'big'=>1, 'button'=>1, 'fieldset'=>1, 'font'=>1, 'form'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'label'=>1, 'object'=>1, 'ruby'=>1, 'script'=>1, 'select'=>1, 'small'=>1, 'sub'=>1, 'sup'=>1, 'textarea'=>1); // Exclude from specific ele; $cN values
    174. 

  174. $eO = array('area'=>1, 'caption'=>1, 'col'=>1, 'colgroup'=>1, 'dd'=>1, 'dt'=>1, 'legend'=>1, 'li'=>1, 'optgroup'=>1, 'option'=>1, 'param'=>1, 'rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1, 'script'=>1, 'tbody'=>1, 'td'=>1, 'tfoot'=>1, 'thead'=>1, 'th'=>1, 'tr'=>1); // Missing in $eB & $eI
    175. 

  175. $eF = $eB + $eI;
    176. 

  176. 

  177. // $in sets allowed child
    178. 

  178. $in = ((isset($eF[$in]) && $in != '#pcdata') or isset($eO[$in])) ? $in : 'div';
    179. 

  179. if(isset($cE[$in])){
    180. 

  180.  return (!$do ? '' : str_replace(array('<', '>'), array('&lt;', '&gt;'), $t));
    181. 

  181. }
    182. 

  182. if(isset($cS[$in])){$inOk = $cS[$in];}
    183. 

  183. elseif(isset($cI[$in])){$inOk = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
    184. 

  184. elseif(isset($cF[$in])){$inOk = $eF; unset($cI['del'], $cI['ins']);}
    185. 

  185. elseif(isset($cB[$in])){$inOk = $eB; unset($cI['del'], $cI['ins']);}
    186. 

  186. if(isset($cO[$in])){$inOk = $inOk + $cO[$in];}
    187. 

  187. if(isset($cN[$in])){$inOk = array_diff_assoc($inOk, $cN[$in]);}
    188. 

  188. 

  189. $t = explode('<', $t);
    190. 

  190. $ok = $q = array(); // $q seq list of open non-empty ele
    191. 

  191. ob_start();
    192. 

  192. 

  193. for($i=-1, $ci=count($t); ++$i<$ci;){
    194. 

  194.  // allowed $ok in parent $p
    195. 

  195.  if($ql = count($q)){
    196. 

  196.   $p = array_pop($q);
    197. 

  197.   $q[] = $p;
    198. 

  198.   if(isset($cS[$p])){$ok = $cS[$p];}
    199. 

  199.   elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
    200. 

  200.   elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);}
    201. 

  201.   elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);}
    202. 

  202.   if(isset($cO[$p])){$ok = $ok + $cO[$p];}
    203. 

  203.   if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);}
    204. 

  204.  }else{$ok = $inOk; unset($cI['del'], $cI['ins']);}
    205. 

  205.  // bad tags, & ele content
    206. 

  206.  if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){
    207. 

  207.   echo '&lt;', $s, $e, $a, '&gt;';
    208. 

  208.  }
    209. 

  209.  if(isset($x[0])){
    210. 

  210.   if(strlen(trim($x)) && (($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql))){
    211. 

  211.    echo '<div>', $x, '</div>';
    212. 

  212.   }
    213. 

  213.   elseif($do < 3 or isset($ok['#pcdata'])){echo $x;}
    214. 

  214.   elseif(strpos($x, "\x02\x04")){
    215. 

  215.    foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){
    216. 

  216.     echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : ''));
    217. 

  217.    }
    218. 

  218.   }elseif($do > 4){echo preg_replace('`\S`', '', $x);}
    219. 

  219.  }
    220. 

  220.  // get markup
    221. 

  221.  if(!preg_match('`^(/?)([a-z1-6]+)([^>]*)>(.*)`sm', $t[$i], $r)){$x = $t[$i]; continue;}
    222. 

  222.  $s = null; $e = null; $a = null; $x = null; list($all, $s, $e, $a, $x) = $r;
    223. 

  223.  // close tag
    224. 

  224.  if($s){
    225. 

  225.   if(isset($cE[$e]) or !in_array($e, $q)){continue;} // Empty/unopen
    226. 

  226.   if($p == $e){array_pop($q); echo '</', $e, '>'; unset($e); continue;} // Last open
    227. 

  227.   $add = ''; // Nesting - close open tags that need to be
    228. 

  228.   for($j=-1, $cj=count($q); ++$j<$cj;){  
    229. 

  229.    if(($d = array_pop($q)) == $e){break;}
    230. 

  230.    else{$add .= "</{$d}>";}
    231. 

  231.   }
    232. 

  232.   echo $add, '</', $e, '>'; unset($e); continue;
    233. 

  233.  }
    234. 

  234.  // open tag
    235. 

  235.  // $cB ele needs $eB ele as child
    236. 

  236.  if(isset($cB[$e]) && strlen(trim($x))){
    237. 

  237.   $t[$i] = "{$e}{$a}>";
    238. 

  238.   array_splice($t, $i+1, 0, 'div>'. $x); unset($e, $x); ++$ci; --$i; continue;
    239. 

  239.  }
    240. 

  240.  if((($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql)) && !isset($eB[$e]) && !isset($ok[$e])){
    241. 

  241.   array_splice($t, $i, 0, 'div>'); unset($e, $x); ++$ci; --$i; continue;
    242. 

  242.  }
    243. 

  243.  // if no open ele, $in = parent; mostly immediate parent-child relation should hold
    244. 

  244.  if(!$ql or !isset($eN[$e]) or !array_intersect($q, $cN2)){
    245. 

  245.   if(!isset($ok[$e])){
    246. 

  246.    if($ql && isset($cT[$p])){echo '</', array_pop($q), '>'; unset($e, $x); --$i;}
    247. 

  247.    continue;
    248. 

  248.   }
    249. 

  249.   if(!isset($cE[$e])){$q[] = $e;}
    250. 

  250.   echo '<', $e, $a, '>'; unset($e); continue;
    251. 

  251.  }
    252. 

  252.  // specific parent-child
    253. 

  253.  if(isset($cS[$p][$e])){
    254. 

  254.   if(!isset($cE[$e])){$q[] = $e;}
    255. 

  255.   echo '<', $e, $a, '>'; unset($e); continue;
    256. 

  256.  }
    257. 

  257.  // nesting
    258. 

  258.  $add = '';
    259. 

  259.  $q2 = array();
    260. 

  260.  for($k=-1, $kc=count($q); ++$k<$kc;){
    261. 

  261.   $d = $q[$k];
    262. 

  262.   $ok2 = array();
    263. 

  263.   if(isset($cS[$d])){$q2[] = $d; continue;}
    264. 

  264.   $ok2 = isset($cI[$d]) ? $eI : $eF;
    265. 

  265.   if(isset($cO[$d])){$ok2 = $ok2 + $cO[$d];}
    266. 

  266.   if(isset($cN[$d])){$ok2 = array_diff_assoc($ok2, $cN[$d]);}
    267. 

  267.   if(!isset($ok2[$e])){
    268. 

  268.    if(!$k && !isset($inOk[$e])){continue 2;}
    269. 

  269.    $add = "</{$d}>";
    270. 

  270.    for(;++$k<$kc;){$add = "</{$q[$k]}>{$add}";}
    271. 

  271.    break;
    272. 

  272.   }
    273. 

  273.   else{$q2[] = $d;}
    274. 

  274.  }
    275. 

  275.  $q = $q2;
    276. 

  276.  if(!isset($cE[$e])){$q[] = $e;}
    277. 

  277.  echo $add, '<', $e, $a, '>'; unset($e); continue;
    278. 

  278. }
    279. 

  279. 

  280. // end
    281. 

  281. if($ql = count($q)){
    282. 

  282.  $p = array_pop($q);
    283. 

  283.  $q[] = $p;
    284. 

  284.  if(isset($cS[$p])){$ok = $cS[$p];}
    285. 

  285.  elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
    286. 

  286.  elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);}
    287. 

  287.  elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);}
    288. 

  288.  if(isset($cO[$p])){$ok = $ok + $cO[$p];}
    289. 

  289.  if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);}
    290. 

  290. }else{$ok = $inOk; unset($cI['del'], $cI['ins']);}
    291. 

  291. if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){
    292. 

  292.  echo '&lt;', $s, $e, $a, '&gt;';
    293. 

  293. }
    294. 

  294. if(isset($x[0])){
    295. 

  295.  if(strlen(trim($x)) && (($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql))){
    296. 

  296.   echo '<div>', $x, '</div>';
    297. 

  297.  }
    298. 

  298.  elseif($do < 3 or isset($ok['#pcdata'])){echo $x;}
    299. 

  299.  elseif(strpos($x, "\x02\x04")){
    300. 

  300.   foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){
    301. 

  301.    echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : ''));
    302. 

  302.   }
    303. 

  303.  }elseif($do > 4){echo preg_replace('`\S`', '', $x);}
    304. 

  304. }
    305. 

  305. while(!empty($q) && ($e = array_pop($q))){echo '</', $e, '>';}
    306. 

  306. $o = ob_get_contents();
    307. 

  307. ob_end_clean();
    308. 

  308. return $o;
    309. 

  309. // eof
    310. 

  310. }
    311. 

  311. 

  312. function hl_cmtcd($t){
    313. 

  313. // comment/CDATA sec handler
    314. 

  314. $t = $t[0];
    315. 

  315. global $C;
    316. 

  316. if(!($v = $C[$n = $t[3] == '-' ? 'comment' : 'cdata'])){return $t;}
    317. 

  317. if($v == 1){return '';}
    318. 

  318. if($n == 'comment'){
    319. 

  319.  if(substr(($t = preg_replace('`--+`', '-', substr($t, 4, -3))), -1) != ' '){$t .= ' ';}
    320. 

  320. }
    321. 

  321. else{$t = substr($t, 1, -1);}
    322. 

  322. $t = $v == 2 ? str_replace(array('&', '<', '>'), array('&amp;', '&lt;', '&gt;'), $t) : $t;
    323. 

  323. return str_replace(array('&', '<', '>'), array("\x03", "\x04", "\x05"), ($n == 'comment' ? "\x01\x02\x04!--$t--\x05\x02\x01" : "\x01\x01\x04$t\x05\x01\x01"));
    324. 

  324. // eof
    325. 

  325. }
    326. 

  326. 

  327. function hl_ent($t){
    328. 

  328. // entitity handler
    329. 

  329. global $C;
    330. 

  330. $t = $t[1];
    331. 

  331. static $U = array('quot'=>1,'amp'=>1,'lt'=>1,'gt'=>1);
    332. 

  332. static $N = array('fnof'=>'402', 'Alpha'=>'913', 'Beta'=>'914', 'Gamma'=>'915', 'Delta'=>'916', 'Epsilon'=>'917', 'Zeta'=>'918', 'Eta'=>'919', 'Theta'=>'920', 'Iota'=>'921', 'Kappa'=>'922', 'Lambda'=>'923', 'Mu'=>'924', 'Nu'=>'925', 'Xi'=>'926', 'Omicron'=>'927', 'Pi'=>'928', 'Rho'=>'929', 'Sigma'=>'931', 'Tau'=>'932', 'Upsilon'=>'933', 'Phi'=>'934', 'Chi'=>'935', 'Psi'=>'936', 'Omega'=>'937', 'alpha'=>'945', 'beta'=>'946', 'gamma'=>'947', 'delta'=>'948', 'epsilon'=>'949', 'zeta'=>'950', 'eta'=>'951', 'theta'=>'952', 'iota'=>'953', 'kappa'=>'954', 'lambda'=>'955', 'mu'=>'956', 'nu'=>'957', 'xi'=>'958', 'omicron'=>'959', 'pi'=>'960', 'rho'=>'961', 'sigmaf'=>'962', 'sigma'=>'963', 'tau'=>'964', 'upsilon'=>'965', 'phi'=>'966', 'chi'=>'967', 'psi'=>'968', 'omega'=>'969', 'thetasym'=>'977', 'upsih'=>'978', 'piv'=>'982', 'bull'=>'8226', 'hellip'=>'8230', 'prime'=>'8242', 'Prime'=>'8243', 'oline'=>'8254', 'frasl'=>'8260', 'weierp'=>'8472', 'image'=>'8465', 'real'=>'8476', 'trade'=>'8482', 'alefsym'=>'8501', 'larr'=>'8592', 'uarr'=>'8593', 'rarr'=>'8594', 'darr'=>'8595', 'harr'=>'8596', 'crarr'=>'8629', 'lArr'=>'8656', 'uArr'=>'8657', 'rArr'=>'8658', 'dArr'=>'8659', 'hArr'=>'8660', 'forall'=>'8704', 'part'=>'8706', 'exist'=>'8707', 'empty'=>'8709', 'nabla'=>'8711', 'isin'=>'8712', 'notin'=>'8713', 'ni'=>'8715', 'prod'=>'8719', 'sum'=>'8721', 'minus'=>'8722', 'lowast'=>'8727', 'radic'=>'8730', 'prop'=>'8733', 'infin'=>'8734', 'ang'=>'8736', 'and'=>'8743', 'or'=>'8744', 'cap'=>'8745', 'cup'=>'8746', 'int'=>'8747', 'there4'=>'8756', 'sim'=>'8764', 'cong'=>'8773', 'asymp'=>'8776', 'ne'=>'8800', 'equiv'=>'8801', 'le'=>'8804', 'ge'=>'8805', 'sub'=>'8834', 'sup'=>'8835', 'nsub'=>'8836', 'sube'=>'8838', 'supe'=>'8839', 'oplus'=>'8853', 'otimes'=>'8855', 'perp'=>'8869', 'sdot'=>'8901', 'lceil'=>'8968', 'rceil'=>'8969', 'lfloor'=>'8970', 'rfloor'=>'8971', 'lang'=>'9001', 'rang'=>'9002', 'loz'=>'9674', 'spades'=>'9824', 'clubs'=>'9827', 'hearts'=>'9829', 'diams'=>'9830', 'apos'=>'39',  'OElig'=>'338', 'oelig'=>'339', 'Scaron'=>'352', 'scaron'=>'353', 'Yuml'=>'376', 'circ'=>'710', 'tilde'=>'732', 'ensp'=>'8194', 'emsp'=>'8195', 'thinsp'=>'8201', 'zwnj'=>'8204', 'zwj'=>'8205', 'lrm'=>'8206', 'rlm'=>'8207', 'ndash'=>'8211', 'mdash'=>'8212', 'lsquo'=>'8216', 'rsquo'=>'8217', 'sbquo'=>'8218', 'ldquo'=>'8220', 'rdquo'=>'8221', 'bdquo'=>'8222', 'dagger'=>'8224', 'Dagger'=>'8225', 'permil'=>'8240', 'lsaquo'=>'8249', 'rsaquo'=>'8250', 'euro'=>'8364', 'nbsp'=>'160', 'iexcl'=>'161', 'cent'=>'162', 'pound'=>'163', 'curren'=>'164', 'yen'=>'165', 'brvbar'=>'166', 'sect'=>'167', 'uml'=>'168', 'copy'=>'169', 'ordf'=>'170', 'laquo'=>'171', 'not'=>'172', 'shy'=>'173', 'reg'=>'174', 'macr'=>'175', 'deg'=>'176', 'plusmn'=>'177', 'sup2'=>'178', 'sup3'=>'179', 'acute'=>'180', 'micro'=>'181', 'para'=>'182', 'middot'=>'183', 'cedil'=>'184', 'sup1'=>'185', 'ordm'=>'186', 'raquo'=>'187', 'frac14'=>'188', 'frac12'=>'189', 'frac34'=>'190', 'iquest'=>'191', 'Agrave'=>'192', 'Aacute'=>'193', 'Acirc'=>'194', 'Atilde'=>'195', 'Auml'=>'196', 'Aring'=>'197', 'AElig'=>'198', 'Ccedil'=>'199', 'Egrave'=>'200', 'Eacute'=>'201', 'Ecirc'=>'202', 'Euml'=>'203', 'Igrave'=>'204', 'Iacute'=>'205', 'Icirc'=>'206', 'Iuml'=>'207', 'ETH'=>'208', 'Ntilde'=>'209', 'Ograve'=>'210', 'Oacute'=>'211', 'Ocirc'=>'212', 'Otilde'=>'213', 'Ouml'=>'214', 'times'=>'215', 'Oslash'=>'216', 'Ugrave'=>'217', 'Uacute'=>'218', 'Ucirc'=>'219', 'Uuml'=>'220', 'Yacute'=>'221', 'THORN'=>'222', 'szlig'=>'223', 'agrave'=>'224', 'aacute'=>'225', 'acirc'=>'226', 'atilde'=>'227', 'auml'=>'228', 'aring'=>'229', 'aelig'=>'230', 'ccedil'=>'231', 'egrave'=>'232', 'eacute'=>'233', 'ecirc'=>'234', 'euml'=>'235', 'igrave'=>'236', 'iacute'=>'237', 'icirc'=>'238', 'iuml'=>'239', 'eth'=>'240', 'ntilde'=>'241', 'ograve'=>'242', 'oacute'=>'243', 'ocirc'=>'244', 'otilde'=>'245', 'ouml'=>'246', 'divide'=>'247', 'oslash'=>'248', 'ugrave'=>'249', 'uacute'=>'250', 'ucirc'=>'251', 'uuml'=>'252', 'yacute'=>'253', 'thorn'=>'254', 'yuml'=>'255');
    333. 

  333. if($t[0] != '#'){
    334. 

  334.  return ($C['and_mark'] ? "\x06" : '&'). (isset($U[$t]) ? $t : (isset($N[$t]) ? (!$C['named_entity'] ? '#'. ($C['hexdec_entity'] > 1 ? 'x'. dechex($N[$t]) : $N[$t]) : $t) : 'amp;'. $t)). ';';
    335. 

  335. }
    336. 

  336. if(($n = ctype_digit($t = substr($t, 1)) ? intval($t) : hexdec(substr($t, 1))) < 9 or ($n > 13 && $n < 32) or $n == 11 or $n == 12 or ($n > 126 && $n < 160 && $n != 133) or ($n > 55295 && ($n < 57344 or ($n > 64975 && $n < 64992) or $n == 65534 or $n == 65535 or $n > 1114111))){
    337. 

  337.  return ($C['and_mark'] ? "\x06" : '&'). "amp;#{$t};";
    338. 

  338. }
    339. 

  339. return ($C['and_mark'] ? "\x06" : '&'). '#'. (((ctype_digit($t) && $C['hexdec_entity'] < 2) or !$C['hexdec_entity']) ? $n : 'x'. dechex($n)). ';';
    340. 

  340. // eof
    341. 

  341. }
    342. 

  342. 

  343. function hl_prot($p, $c=null){
    344. 

  344. // check URL scheme
    345. 

  345. global $C;
    346. 

  346. $b = $a = '';
    347. 

  347. if($c == null){$c = 'style'; $b = $p[1]; $a = $p[3]; $p = trim($p[2]);}
    348. 

  348. $c = isset($C['schemes'][$c]) ? $C['schemes'][$c] : $C['schemes']['*'];
    349. 

  349. static $d = 'denied:';
    350. 

  350. if(isset($c['!']) && substr($p, 0, 7) != $d){$p = "$d$p";}
    351. 

  351. if(isset($c['*']) or !strcspn($p, '#?;') or (substr($p, 0, 7) == $d)){return "{$b}{$p}{$a}";} // All ok, frag, query, param
    352. 

  352. if(preg_match('`^([^:?[@!$()*,=/\'\]]+?)(:|&#(58|x3a);|%3a|\\\\0{0,4}3a).`i', $p, $m) && !isset($c[strtolower($m[1])])){ // Denied prot
    353. 

  353.  return "{$b}{$d}{$p}{$a}";
    354. 

  354. }
    355. 

  355. if($C['abs_url']){
    356. 

  356.  if($C['abs_url'] == -1 && strpos($p, $C['base_url']) === 0){ // Make url rel
    357. 

  357.   $p = substr($p, strlen($C['base_url']));
    358. 

  358.  }elseif(empty($m[1])){ // Make URL abs
    359. 

  359.   if(substr($p, 0, 2) == '//'){$p = substr($C['base_url'], 0, strpos($C['base_url'], ':')+1). $p;}
    360. 

  360.   elseif($p[0] == '/'){$p = preg_replace('`(^.+?://[^/]+)(.*)`', '$1', $C['base_url']). $p;}
    361. 

  361.   elseif(strcspn($p, './')){$p = $C['base_url']. $p;}
    362. 

  362.   else{
    363. 

  363.    preg_match('`^([a-zA-Z\d\-+.]+://[^/]+)(.*)`', $C['base_url'], $m);
    364. 

  364.    $p = preg_replace('`(?<=/)\./`', '', $m[2]. $p);
    365. 

  365.    while(preg_match('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', $p)){
    366. 

  366.     $p = preg_replace('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', '', $p);
    367. 

  367.    }
    368. 

  368.    $p = $m[1]. $p;
    369. 

  369.   }
    370. 

  370.  }
    371. 

  371. }
    372. 

  372. return "{$b}{$p}{$a}";
    373. 

  373. // eof
    374. 

  374. }
    375. 

  375. 

  376. function hl_regex($p){
    377. 

  377. // ?regex
    378. 

  378. if(empty($p)){return 0;}
    379. 

  379. if($t = ini_get('track_errors')){$o = isset($php_errormsg) ? $php_errormsg : null;}
    380. 

  380. else{ini_set('track_errors', 1);}
    381. 

  381. unset($php_errormsg);
    382. 

  382. if(($d = ini_get('display_errors'))){ini_set('display_errors', 0);}
    383. 

  383. preg_match($p, '');
    384. 

  384. if($d){ini_set('display_errors', 1);}
    385. 

  385. $r = isset($php_errormsg) ? 0 : 1;
    386. 

  386. if($t){$php_errormsg = isset($o) ? $o : null;}
    387. 

  387. else{ini_set('track_errors', 0);}
    388. 

  388. return $r;
    389. 

  389. // eof
    390. 

  390. }
    391. 

  391. 

  392. function hl_spec($t){
    393. 

  393. // final $spec
    394. 

  394. $s = array();
    395. 

  395. $t = str_replace(array("\t", "\r", "\n", ' '), '', preg_replace_callback('/"(?>(`.|[^"])*)"/sm', create_function('$m', 'return substr(str_replace(array(";", "|", "~", " ", ",", "/", "(", ")", \'`"\'), array("\x01", "\x02", "\x03", "\x04", "\x05", "\x06", "\x07", "\x08", "\""), $m[0]), 1, -1);'), trim($t))); 
    396. 

  396. for($i = count(($t = explode(';', $t))); --$i>=0;){
    397. 

  397.  $w = $t[$i];
    398. 

  398.  if(empty($w) or ($e = strpos($w, '=')) === false or !strlen(($a =  substr($w, $e+1)))){continue;}
    399. 

  399.  $y = $n = array();
    400. 

  400.  foreach(explode(',', $a) as $v){
    401. 

  401.   if(!preg_match('`^([a-z:\-\*]+)(?:\((.*?)\))?`i', $v, $m)){continue;}
    402. 

  402.   if(($x = strtolower($m[1])) == '-*'){$n['*'] = 1; continue;}
    403. 

  403.   if($x[0] == '-'){$n[substr($x, 1)] = 1; continue;}
    404. 

  404.   if(!isset($m[2])){$y[$x] = 1; continue;}
    405. 

  405.   foreach(explode('/', $m[2]) as $m){
    406. 

  406.    if(empty($m) or ($p = strpos($m, '=')) == 0 or $p < 5){$y[$x] = 1; continue;}
    407. 

  407.    $y[$x][strtolower(substr($m, 0, $p))] = str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x06", "\x07", "\x08"), array(";", "|", "~", " ", ",", "/", "(", ")"), substr($m, $p+1));
    408. 

  408.   }
    409. 

  409.   if(isset($y[$x]['match']) && !hl_regex($y[$x]['match'])){unset($y[$x]['match']);}
    410. 

  410.   if(isset($y[$x]['nomatch']) && !hl_regex($y[$x]['nomatch'])){unset($y[$x]['nomatch']);}
    411. 

  411.  }
    412. 

  412.  if(!count($y) && !count($n)){continue;}
    413. 

  413.  foreach(explode(',', substr($w, 0, $e)) as $v){
    414. 

  414.   if(!strlen(($v = strtolower($v)))){continue;}
    415. 

  415.   if(count($y)){$s[$v] = $y;}
    416. 

  416.   if(count($n)){$s[$v]['n'] = $n;}
    417. 

  417.  }
    418. 

  418. }
    419. 

  419. return $s;
    420. 

  420. // eof
    421. 

  421. }
    422. 

  422. 

  423. function hl_tag($t){
    424. 

  424. // tag/attribute handler
    425. 

  425. global $C;
    426. 

  426. $t = $t[0];
    427. 

  427. // invalid < >
    428. 

  428. if($t == '< '){return '&lt; ';}
    429. 

  429. if($t == '>'){return '&gt;';}
    430. 

  430. if(!preg_match('`^<(/?)([a-zA-Z][a-zA-Z1-6]*)([^>]*?)\s?>$`m', $t, $m)){
    431. 

  431.  return str_replace(array('<', '>'), array('&lt;', '&gt;'), $t);
    432. 

  432. }elseif(!isset($C['elements'][($e = strtolower($m[2]))])){
    433. 

  433.  return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('&lt;', '&gt;'), $t) : '');
    434. 

  434. }
    435. 

  435. // attr string
    436. 

  436. $a = str_replace(array("\n", "\r", "\t"), ' ', trim($m[3]));
    437. 

  437. // tag transform
    438. 

  438. static $eD = array('applet'=>1, 'center'=>1, 'dir'=>1, 'embed'=>1, 'font'=>1, 'isindex'=>1, 'menu'=>1, 's'=>1, 'strike'=>1, 'u'=>1); // Deprecated
    439. 

  439. if($C['make_tag_strict'] && isset($eD[$e])){
    440. 

  440.  $trt = hl_tag2($e, $a, $C['make_tag_strict']);
    441. 

  441.  if(!$e){return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('&lt;', '&gt;'), $t) : '');}
    442. 

  442. }
    443. 

  443. // close tag
    444. 

  444. static $eE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty ele
    445. 

  445. if(!empty($m[1])){
    446. 

  446.  return (!isset($eE[$e]) ? (empty($C['hook_tag']) ? "</$e>" : $C['hook_tag']($e)) : (($C['keep_bad'])%2 ? str_replace(array('<', '>'), array('&lt;', '&gt;'), $t) : ''));
    447. 

  447. }
    448. 

  448. // open tag & attr
    449. 

  449. static $aN = array('abbr'=>array('td'=>1, 'th'=>1), 'accept-charset'=>array('form'=>1), 'accept'=>array('form'=>1, 'input'=>1), 'accesskey'=>array('a'=>1, 'area'=>1, 'button'=>1, 'input'=>1, 'label'=>1, 'legend'=>1, 'textarea'=>1), 'action'=>array('form'=>1), 'align'=>array('caption'=>1, 'embed'=>1, 'applet'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'object'=>1, 'legend'=>1, 'table'=>1, 'hr'=>1, 'div'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'p'=>1, 'col'=>1, 'colgroup'=>1, 'tbody'=>1, 'td'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1), 'allowfullscreen'=>array('iframe'=>1), 'alt'=>array('applet'=>1, 'area'=>1, 'img'=>1, 'input'=>1), 'archive'=>array('applet'=>1, 'object'=>1), 'axis'=>array('td'=>1, 'th'=>1), 'bgcolor'=>array('embed'=>1, 'table'=>1, 'tr'=>1, 'td'=>1, 'th'=>1), 'border'=>array('table'=>1, 'img'=>1, 'object'=>1), 'bordercolor'=>array('table'=>1, 'td'=>1, 'tr'=>1), 'cellpadding'=>array('table'=>1), 'cellspacing'=>array('table'=>1), 'char'=>array('col'=>1, 'colgroup'=>1, 'tbody'=>1, 'td'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1), 'charoff'=>array('col'=>1, 'colgroup'=>1, 'tbody'=>1, 'td'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1), 'charset'=>array('a'=>1, 'script'=>1), 'checked'=>array('input'=>1), 'cite'=>array('blockquote'=>1, 'q'=>1, 'del'=>1, 'ins'=>1), 'classid'=>array('object'=>1), 'clear'=>array('br'=>1), 'code'=>array('applet'=>1), 'codebase'=>array('object'=>1, 'applet'=>1), 'codetype'=>array('object'=>1), 'color'=>array('font'=>1), 'cols'=>array('textarea'=>1), 'colspan'=>array('td'=>1, 'th'=>1), 'compact'=>array('dir'=>1, 'dl'=>1, 'menu'=>1, 'ol'=>1, 'ul'=>1), 'coords'=>array('area'=>1, 'a'=>1), 'data'=>array('object'=>1), 'datetime'=>array('del'=>1, 'ins'=>1), 'declare'=>array('object'=>1), 'defer'=>array('script'=>1), 'dir'=>array('bdo'=>1), 'disabled'=>array('button'=>1, 'input'=>1, 'optgroup'=>1, 'option'=>1, 'select'=>1, 'textarea'=>1), 'enctype'=>array('form'=>1), 'face'=>array('font'=>1), 'flashvars'=>array('embed'=>1), 'for'=>array('label'=>1), 'frame'=>array('table'=>1), 'frameborder'=>array('iframe'=>1), 'headers'=>array('td'=>1, 'th'=>1), 'height'=>array('embed'=>1, 'iframe'=>1, 'td'=>1, 'th'=>1, 'img'=>1, 'object'=>1, 'applet'=>1), 'href'=>array('a'=>1, 'area'=>1), 'hreflang'=>array('a'=>1), 'hspace'=>array('applet'=>1, 'img'=>1, 'object'=>1), 'ismap'=>array('img'=>1, 'input'=>1), 'label'=>array('option'=>1, 'optgroup'=>1), 'language'=>array('script'=>1), 'longdesc'=>array('img'=>1, 'iframe'=>1), 'marginheight'=>array('iframe'=>1), 'marginwidth'=>array('iframe'=>1), 'maxlength'=>array('input'=>1), 'method'=>array('form'=>1), 'model'=>array('embed'=>1), 'multiple'=>array('select'=>1), 'name'=>array('button'=>1, 'embed'=>1, 'textarea'=>1, 'applet'=>1, 'select'=>1, 'form'=>1, 'iframe'=>1, 'img'=>1, 'a'=>1, 'input'=>1, 'object'=>1, 'map'=>1, 'param'=>1), 'nohref'=>array('area'=>1), 'noshade'=>array('hr'=>1), 'nowrap'=>array('td'=>1, 'th'=>1), 'object'=>array('applet'=>1), 'onblur'=>array('a'=>1, 'area'=>1, 'button'=>1, 'input'=>1, 'label'=>1, 'select'=>1, 'textarea'=>1), 'onchange'=>array('input'=>1, 'select'=>1, 'textarea'=>1), 'onfocus'=>array('a'=>1, 'area'=>1, 'button'=>1, 'input'=>1, 'label'=>1, 'select'=>1, 'textarea'=>1), 'onreset'=>array('form'=>1), 'onselect'=>array('input'=>1, 'textarea'=>1), 'onsubmit'=>array('form'=>1), 'pluginspage'=>array('embed'=>1), 'pluginurl'=>array('embed'=>1), 'prompt'=>array('isindex'=>1), 'readonly'=>array('textarea'=>1, 'input'=>1), 'rel'=>array('a'=>1), 'rev'=>array('a'=>1), 'rows'=>array('textarea'=>1), 'rowspan'=>array('td'=>1, 'th'=>1), 'rules'=>array('table'=>1), 'scope'=>array('td'=>1, 'th'=>1), 'scrolling'=>array('iframe'=>1), 'selected'=>array('option'=>1), 'shape'=>array('area'=>1, 'a'=>1), 'size'=>array('hr'=>1, 'font'=>1, 'input'=>1, 'select'=>1), 'span'=>array('col'=>1, 'colgroup'=>1), 'src'=>array('embed'=>1, 'script'=>1, 'input'=>1, 'iframe'=>1, 'img'=>1), 'standby'=>array('object'=>1), 'start'=>array('ol'=>1), 'summary'=>array('table'=>1), 'tabindex'=>array('a'=>1, 'area'=>1, 'button'=>1, 'input'=>1, 'object'=>1, 'select'=>1, 'textarea'=>1), 'target'=>array('a'=>1, 'area'=>1, 'form'=>1), 'type'=>array('a'=>1, 'embed'=>1, 'object'=>1, 'param'=>1, 'script'=>1, 'input'=>1, 'li'=>1, 'ol'=>1, 'ul'=>1, 'button'=>1), 'usemap'=>array('img'=>1, 'input'=>1, 'object'=>1), 'valign'=>array('col'=>1, 'colgroup'=>1, 'tbody'=>1, 'td'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1), 'value'=>array('input'=>1, 'option'=>1, 'param'=>1, 'button'=>1, 'li'=>1), 'valuetype'=>array('param'=>1), 'vspace'=>array('applet'=>1, 'img'=>1, 'object'=>1), 'width'=>array('embed'=>1, 'hr'=>1, 'iframe'=>1, 'img'=>1, 'object'=>1, 'table'=>1, 'td'=>1, 'th'=>1, 'applet'=>1, 'col'=>1, 'colgroup'=>1, 'pre'=>1), 'wmode'=>array('embed'=>1), 'xml:space'=>array('pre'=>1, 'script'=>1, 'style'=>1)); // Ele-specific
    450. 

  450. static $aNE = array('allowfullscreen'=>1, 'checked'=>1, 'compact'=>1, 'declare'=>1, 'defer'=>1, 'disabled'=>1, 'ismap'=>1, 'multiple'=>1, 'nohref'=>1, 'noresize'=>1, 'noshade'=>1, 'nowrap'=>1, 'readonly'=>1, 'selected'=>1); // Empty
    451. 

  451. static $aNP = array('action'=>1, 'cite'=>1, 'classid'=>1, 'codebase'=>1, 'data'=>1, 'href'=>1, 'longdesc'=>1, 'model'=>1, 'pluginspage'=>1, 'pluginurl'=>1, 'usemap'=>1); // Need scheme check; excludes style, on* & src
    452. 

  452. static $aNU = array('class'=>array('param'=>1, 'script'=>1), 'dir'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'id'=>array('script'=>1), 'lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'xml:lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'onclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'ondblclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeydown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeypress'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeyup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousedown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousemove'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseout'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseover'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'style'=>array('param'=>1, 'script'=>1), 'title'=>array('param'=>1, 'script'=>1)); // Univ & exceptions
    453. 

  453. 

  454. if($C['lc_std_val']){
    455. 

  455.  // predef attr vals for $eAL & $aNE ele
    456. 

  456.  static $aNL = array('all'=>1, 'baseline'=>1, 'bottom'=>1, 'button'=>1, 'center'=>1, 'char'=>1, 'checkbox'=>1, 'circle'=>1, 'col'=>1, 'colgroup'=>1, 'cols'=>1, 'data'=>1, 'default'=>1, 'file'=>1, 'get'=>1, 'groups'=>1, 'hidden'=>1, 'image'=>1, 'justify'=>1, 'left'=>1, 'ltr'=>1, 'middle'=>1, 'none'=>1, 'object'=>1, 'password'=>1, 'poly'=>1, 'post'=>1, 'preserve'=>1, 'radio'=>1, 'rect'=>1, 'ref'=>1, 'reset'=>1, 'right'=>1, 'row'=>1, 'rowgroup'=>1, 'rows'=>1, 'rtl'=>1, 'submit'=>1, 'text'=>1, 'top'=>1);
    457. 

  457.  static $eAL = array('a'=>1, 'area'=>1, 'bdo'=>1, 'button'=>1, 'col'=>1, 'form'=>1, 'img'=>1, 'input'=>1, 'object'=>1, 'optgroup'=>1, 'option'=>1, 'param'=>1, 'script'=>1, 'select'=>1, 'table'=>1, 'td'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1, 'xml:space'=>1);
    458. 

  458.  $lcase = isset($eAL[$e]) ? 1 : 0;
    459. 

  459. }
    460. 

  460. 

  461. $depTr = 0;
    462. 

  462. if($C['no_deprecated_attr']){
    463. 

  463.  // dep attr:applicable ele
    464. 

  464.  static $aND = array('align'=>array('caption'=>1, 'div'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'legend'=>1, 'object'=>1, 'p'=>1, 'table'=>1), 'bgcolor'=>array('table'=>1, 'td'=>1, 'th'=>1, 'tr'=>1), 'border'=>array('img'=>1, 'object'=>1), 'bordercolor'=>array('table'=>1, 'td'=>1, 'tr'=>1), 'clear'=>array('br'=>1), 'compact'=>array('dl'=>1, 'ol'=>1, 'ul'=>1), 'height'=>array('td'=>1, 'th'=>1), 'hspace'=>array('img'=>1, 'object'=>1), 'language'=>array('script'=>1), 'name'=>array('a'=>1, 'form'=>1, 'iframe'=>1, 'img'=>1, 'map'=>1), 'noshade'=>array('hr'=>1), 'nowrap'=>array('td'=>1, 'th'=>1), 'size'=>array('hr'=>1), 'start'=>array('ol'=>1), 'type'=>array('li'=>1, 'ol'=>1, 'ul'=>1), 'value'=>array('li'=>1), 'vspace'=>array('img'=>1, 'object'=>1), 'width'=>array('hr'=>1, 'pre'=>1, 'td'=>1, 'th'=>1));
    465. 

  465.  static $eAD = array('a'=>1, 'br'=>1, 'caption'=>1, 'div'=>1, 'dl'=>1, 'form'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'hr'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'legend'=>1, 'li'=>1, 'map'=>1, 'object'=>1, 'ol'=>1, 'p'=>1, 'pre'=>1, 'script'=>1, 'table'=>1, 'td'=>1, 'th'=>1, 'tr'=>1, 'ul'=>1);
    466. 

  466.  $depTr = isset($eAD[$e]) ? 1 : 0;
    467. 

  467. }
    468. 

  468. 

  469. // attr name-vals
    470. 

  470. if(strpos($a, "\x01") !== false){$a = preg_replace('`\x01[^\x01]*\x01`', '', $a);} // No comment/CDATA sec
    471. 

  471. $mode = 0; $a = trim($a, ' /'); $aA = array();
    472. 

  472. while(strlen($a)){
    473. 

  473.  $w = 0;
    474. 

  474.  switch($mode){
    475. 

  475.   case 0: // Name
    476. 

  476.    if(preg_match('`^[a-zA-Z][\-a-zA-Z:]+`', $a, $m)){
    477. 

  477.     $nm = strtolower($m[0]);
    478. 

  478.     $w = $mode = 1; $a = ltrim(substr_replace($a, '', 0, strlen($m[0])));
    479. 

  479.    }
    480. 

  480.   break; case 1:
    481. 

  481.    if($a[0] == '='){ // =
    482. 

  482.     $w = 1; $mode = 2; $a = ltrim($a, '= ');
    483. 

  483.    }else{ // No val
    484. 

  484.     $w = 1; $mode = 0; $a = ltrim($a);
    485. 

  485.     $aA[$nm] = '';
    486. 

  486.    }
    487. 

  487.   break; case 2: // Val
    488. 

  488.    if(preg_match('`^((?:"[^"]*")|(?:\'[^\']*\')|(?:\s*[^\s"\']+))(.*)`', $a, $m)){
    489. 

  489.     $a = ltrim($m[2]); $m = $m[1]; $w = 1; $mode = 0;
    490. 

  490.     $aA[$nm] = trim(str_replace('<', '&lt;', ($m[0] == '"' or $m[0] == '\'') ? substr($m, 1, -1) : $m));
    491. 

  491.    }
    492. 

  492.   break;
    493. 

  493.  }
    494. 

  494.  if($w == 0){ // Parse errs, deal with space, " & '
    495. 

  495.   $a = preg_replace('`^(?:"[^"]*("|$)|\'[^\']*(\'|$)|\S)*\s*`', '', $a);
    496. 

  496.   $mode = 0;
    497. 

  497.  }
    498. 

  498. }
    499. 

  499. if($mode == 1){$aA[$nm] = '';}
    500. 

  500. 

  501. // clean attrs
    502. 

  502. global $S;
    503. 

  503. $rl = isset($S[$e]) ? $S[$e] : array();
    504. 

  504. $a = array(); $nfr = 0;
    505. 

  505. foreach($aA as $k=>$v){
    506. 

  506.   if(((isset($C['deny_attribute']['*']) ? isset($C['deny_attribute'][$k]) : !isset($C['deny_attribute'][$k])) && (isset($aN[$k][$e]) or (isset($aNU[$k]) && !isset($aNU[$k][$e]))) && !isset($rl['n'][$k]) && !isset($rl['n']['*'])) or isset($rl[$k])){
    507. 

  507.   if(isset($aNE[$k])){$v = $k;}
    508. 

  508.   elseif(!empty($lcase) && (($e != 'button' or $e != 'input') or $k == 'type')){ // Rather loose but ?not cause issues
    509. 

  509.    $v = (isset($aNL[($v2 = strtolower($v))])) ? $v2 : $v;
    510. 

  510.   }
    511. 

  511.   if($k == 'style' && !$C['style_pass']){
    512. 

  512.    if(false !== strpos($v, '&#')){
    513. 

  513.     static $sC = array('&#x20;'=>' ', '&#32;'=>' ', '&#x45;'=>'e', '&#69;'=>'e', '&#x65;'=>'e', '&#101;'=>'e', '&#x58;'=>'x', '&#88;'=>'x', '&#x78;'=>'x', '&#120;'=>'x', '&#x50;'=>'p', '&#80;'=>'p', '&#x70;'=>'p', '&#112;'=>'p', '&#x53;'=>'s', '&#83;'=>'s', '&#x73;'=>'s', '&#115;'=>'s', '&#x49;'=>'i', '&#73;'=>'i', '&#x69;'=>'i', '&#105;'=>'i', '&#x4f;'=>'o', '&#79;'=>'o', '&#x6f;'=>'o', '&#111;'=>'o', '&#x4e;'=>'n', '&#78;'=>'n', '&#x6e;'=>'n', '&#110;'=>'n', '&#x55;'=>'u', '&#85;'=>'u', '&#x75;'=>'u', '&#117;'=>'u', '&#x52;'=>'r', '&#82;'=>'r', '&#x72;'=>'r', '&#114;'=>'r', '&#x4c;'=>'l', '&#76;'=>'l', '&#x6c;'=>'l', '&#108;'=>'l', '&#x28;'=>'(', '&#40;'=>'(', '&#x29;'=>')', '&#41;'=>')', '&#x20;'=>':', '&#32;'=>':', '&#x22;'=>'"', '&#34;'=>'"', '&#x27;'=>"'", '&#39;'=>"'", '&#x2f;'=>'/', '&#47;'=>'/', '&#x2a;'=>'*', '&#42;'=>'*', '&#x5c;'=>'\\', '&#92;'=>'\\');
    514. 

  514.     $v = strtr($v, $sC);
    515. 

  515.    }
    516. 

  516.    $v = preg_replace_callback('`(url(?:\()(?: )*(?:\'|"|&(?:quot|apos);)?)(.+?)((?:\'|"|&(?:quot|apos);)?(?: )*(?:\)))`iS', 'hl_prot', $v);
    517. 

  517.    $v = !$C['css_expression'] ? preg_replace('`expression`i', ' ', preg_replace('`\\\\\S|(/|(%2f))(\*|(%2a))`i', ' ', $v)) : $v;
    518. 

  518.   }elseif(isset($aNP[$k]) or strpos($k, 'src') !== false or $k[0] == 'o'){
    519. 

  519.    $v = str_replace("­", ' ', (strpos($v, '&') !== false ? str_replace(array('&#xad;', '&#173;', '&shy;'), ' ', $v) : $v)); # double-quoted char is soft-hyphen; appears here as "­" or hyphen or something else depending on viewing software
    520. 

  520.    $v = hl_prot($v, $k);
    521. 

  521.    if($k == 'href'){ // X-spam
    522. 

  522.     if($C['anti_mail_spam'] && strpos($v, 'mailto:') === 0){
    523. 

  523.      $v = str_replace('@', htmlspecialchars($C['anti_mail_spam']), $v);
    524. 

  524.     }elseif($C['anti_link_spam']){
    525. 

  525.      $r1 = $C['anti_link_spam'][1];
    526. 

  526.      if(!empty($r1) && preg_match($r1, $v)){continue;}
    527. 

  527.      $r0 = $C['anti_link_spam'][0];
    528. 

  528.      if(!empty($r0) && preg_match($r0, $v)){
    529. 

  529.       if(isset($a['rel'])){
    530. 

  530.        if(!preg_match('`\bnofollow\b`i', $a['rel'])){$a['rel'] .= ' nofollow';}
    531. 

  531.       }elseif(isset($aA['rel'])){
    532. 

  532.        if(!preg_match('`\bnofollow\b`i', $aA['rel'])){$nfr = 1;}
    533. 

  533.       }else{$a['rel'] = 'nofollow';}
    534. 

  534.      }
    535. 

  535.     }
    536. 

  536.    }
    537. 

  537.   }
    538. 

  538.   if(isset($rl[$k]) && is_array($rl[$k]) && ($v = hl_attrval($v, $rl[$k])) === 0){continue;}
    539. 

  539.   $a[$k] = str_replace('"', '&quot;', $v);
    540. 

  540.  }
    541. 

  541. }
    542. 

  542. if($nfr){$a['rel'] = isset($a['rel']) ? $a['rel']. ' nofollow' : 'nofollow';}
    543. 

  543. 

  544. // rqd attr
    545. 

  545. static $eAR = array('area'=>array('alt'=>'area'), 'bdo'=>array('dir'=>'ltr'), 'form'=>array('action'=>''), 'img'=>array('src'=>'', 'alt'=>'image'), 'map'=>array('name'=>''), 'optgroup'=>array('label'=>''), 'param'=>array('name'=>''), 'script'=>array('type'=>'text/javascript'), 'textarea'=>array('rows'=>'10', 'cols'=>'50'));
    546. 

  546. if(isset($eAR[$e])){
    547. 

  547.  foreach($eAR[$e] as $k=>$v){
    548. 

  548.   if(!isset($a[$k])){$a[$k] = isset($v[0]) ? $v : $k;}
    549. 

  549.  }
    550. 

  550. }
    551. 

  551. 

  552. // depr attrs
    553. 

  553. if($depTr){
    554. 

  554.  $c = array();
    555. 

  555.  foreach($a as $k=>$v){
    556. 

  556.   if($k == 'style' or !isset($aND[$k][$e])){continue;}
    557. 

  557.   if($k == 'align'){
    558. 

  558.    unset($a['align']);
    559. 

  559.    if($e == 'img' && ($v == 'left' or $v == 'right')){$c[] = 'float: '. $v;}
    560. 

  560.    elseif(($e == 'div' or $e == 'table') && $v == 'center'){$c[] = 'margin: auto';}
    561. 

  561.    else{$c[] = 'text-align: '. $v;}
    562. 

  562.   }elseif($k == 'bgcolor'){
    563. 

  563.    unset($a['bgcolor']);
    564. 

  564.    $c[] = 'background-color: '. $v;
    565. 

  565.   }elseif($k == 'border'){
    566. 

  566.    unset($a['border']); $c[] = "border: {$v}px";
    567. 

  567.   }elseif($k == 'bordercolor'){
    568. 

  568.    unset($a['bordercolor']); $c[] = 'border-color: '. $v;
    569. 

  569.   }elseif($k == 'clear'){
    570. 

  570.    unset($a['clear']); $c[] = 'clear: '. ($v != 'all' ? $v : 'both');
    571. 

  571.   }elseif($k == 'compact'){
    572. 

  572.    unset($a['compact']); $c[] = 'font-size: 85%';
    573. 

  573.   }elseif($k == 'height' or $k == 'width'){
    574. 

  574.    unset($a[$k]); $c[] = $k. ': '. ($v[0] != '*' ? $v. (ctype_digit($v) ? 'px' : '') : 'auto');
    575. 

  575.   }elseif($k == 'hspace'){
    576. 

  576.    unset($a['hspace']); $c[] = "margin-left: {$v}px; margin-right: {$v}px";
    577. 

  577.   }elseif($k == 'language' && !isset($a['type'])){
    578. 

  578.    unset($a['language']);
    579. 

  579.    $a['type'] = 'text/'. strtolower($v);
    580. 

  580.   }elseif($k == 'name'){
    581. 

  581.    if($C['no_deprecated_attr'] == 2 or ($e != 'a' && $e != 'map')){unset($a['name']);}
    582. 

  582.    if(!isset($a['id']) && preg_match('`[a-zA-Z][a-zA-Z\d.:_\-]*`', $v)){$a['id'] = $v;}
    583. 

  583.   }elseif($k == 'noshade'){
    584. 

  584.    unset($a['noshade']); $c[] = 'border-style: none; border: 0; background-color: gray; color: gray';
    585. 

  585.   }elseif($k == 'nowrap'){
    586. 

  586.    unset($a['nowrap']); $c[] = 'white-space: nowrap';
    587. 

  587.   }elseif($k == 'size'){
    588. 

  588.    unset($a['size']); $c[] = 'size: '. $v. 'px';
    589. 

  589.   }elseif($k == 'start' or $k == 'value'){
    590. 

  590.    unset($a[$k]);
    591. 

  591.   }elseif($k == 'type'){
    592. 

  592.    unset($a['type']);
    593. 

  593.    static $ol_type = array('i'=>'lower-roman', 'I'=>'upper-roman', 'a'=>'lower-latin', 'A'=>'upper-latin', '1'=>'decimal');
    594. 

  594.    $c[] = 'list-style-type: '. (isset($ol_type[$v]) ? $ol_type[$v] : 'decimal');
    595. 

  595.   }elseif($k == 'vspace'){
    596. 

  596.    unset($a['vspace']); $c[] = "margin-top: {$v}px; margin-bottom: {$v}px";
    597. 

  597.   }
    598. 

  598.  }
    599. 

  599.  if(count($c)){
    600. 

  600.   $c = implode('; ', $c);
    601. 

  601.   $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $c. ';': $c. ';';
    602. 

  602.  }
    603. 

  603. }
    604. 

  604. // unique ID
    605. 

  605. if($C['unique_ids'] && isset($a['id'])){
    606. 

  606.  if(!preg_match('`^[A-Za-z][A-Za-z0-9_\-.:]*$`', ($id = $a['id'])) or (isset($GLOBALS['hl_Ids'][$id]) && $C['unique_ids'] == 1)){unset($a['id']);
    607. 

  607.  }else{
    608. 

  608.   while(isset($GLOBALS['hl_Ids'][$id])){$id = $C['unique_ids']. $id;}
    609. 

  609.   $GLOBALS['hl_Ids'][($a['id'] = $id)] = 1;
    610. 

  610.  }
    611. 

  611. }
    612. 

  612. // xml:lang
    613. 

  613. if($C['xml:lang'] && isset($a['lang'])){
    614. 

  614.  $a['xml:lang'] = isset($a['xml:lang']) ? $a['xml:lang'] : $a['lang'];
    615. 

  615.  if($C['xml:lang'] == 2){unset($a['lang']);}
    616. 

  616. }
    617. 

  617. // for transformed tag
    618. 

  618. if(!empty($trt)){
    619. 

  619.  $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $trt : $trt;
    620. 

  620. }
    621. 

  621. // return with empty ele /
    622. 

  622. if(empty($C['hook_tag'])){
    623. 

  623.  $aA = '';
    624. 

  624.  foreach($a as $k=>$v){$aA .= " {$k}=\"{$v}\"";}
    625. 

  625.  return "<{$e}{$aA}". (isset($eE[$e]) ? ' /' : ''). '>';
    626. 

  626. }
    627. 

  627. else{return $C['hook_tag']($e, $a);}
    628. 

  628. // eof
    629. 

  629. }
    630. 

  630. 

  631. function hl_tag2(&$e, &$a, $t=1){
    632. 

  632. // transform tag
    633. 

  633. if($e == 'center'){$e = 'div'; return 'text-align: center;';}
    634. 

  634. if($e == 'dir' or $e == 'menu'){$e = 'ul'; return '';}
    635. 

  635. if($e == 's' or $e == 'strike'){$e = 'span'; return 'text-decoration: line-through;';}
    636. 

  636. if($e == 'u'){$e = 'span'; return 'text-decoration: underline;';}
    637. 

  637. static $fs = array('0'=>'xx-small', '1'=>'xx-small', '2'=>'small', '3'=>'medium', '4'=>'large', '5'=>'x-large', '6'=>'xx-large', '7'=>'300%', '-1'=>'smaller', '-2'=>'60%', '+1'=>'larger', '+2'=>'150%', '+3'=>'200%', '+4'=>'300%');
    638. 

  638. if($e == 'font'){
    639. 

  639.  $a2 = '';
    640. 

  640.  while(preg_match('`(^|\s)(color|size)\s*=\s*(\'|")?(.+?)(\\3|\s|$)`i', $a, $m)){
    641. 

  641.   $a = str_replace($m[0], ' ', $a);
    642. 

  642.   $a2 .= strtolower($m[2]) == 'color' ? (' color: '. str_replace('"', '\'', trim($m[4])). ';') : (isset($fs[($m = trim($m[4]))]) ? ($a2 .= ' font-size: '. str_replace('"', '\'', $fs[$m]). ';') : '');
    643. 

  643.  }
    644. 

  644.  while(preg_match('`(^|\s)face\s*=\s*(\'|")?([^=]+?)\\2`i', $a, $m) or preg_match('`(^|\s)face\s*=(\s*)(\S+)`i', $a, $m)){
    645. 

  645.   $a = str_replace($m[0], ' ', $a);
    646. 

  646.   $a2 .= ' font-family: '. str_replace('"', '\'', trim($m[3])). ';';
    647. 

  647.  }
    648. 

  648.  $e = 'span'; return ltrim(str_replace('<', '', $a2));
    649. 

  649. }
    650. 

  650. if($t == 2){$e = 0; return 0;}
    651. 

  651. return '';
    652. 

  652. // eof
    653. 

  653. }
    654. 

  654. 

  655. function hl_tidy($t, $w, $p){
    656. 

  656. // Tidy/compact HTM
    657. 

  657. if(strpos(' pre,script,textarea', "$p,")){return $t;}
    658. 

  658. $t = preg_replace('`\s+`', ' ', preg_replace_callback(array('`(<(!\[CDATA\[))(.+?)(\]\]>)`sm', '`(<(!--))(.+?)(-->)`sm', '`(<(pre|script|textarea)[^>]*?>)(.+?)(</\2>)`sm'), create_function('$m', 'return $m[1]. str_replace(array("<", ">", "\n", "\r", "\t", " "), array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), $m[3]). $m[4];'), $t));
    659. 

  659. if(($w = strtolower($w)) == -1){
    660. 

  660.  return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t);
    661. 

  661. }
    662. 

  662. $s = strpos(" $w", 't') ? "\t" : ' ';
    663. 

  663. $s = preg_match('`\d`', $w, $m) ? str_repeat($s, $m[0]) : str_repeat($s, ($s == "\t" ? 1 : 2));
    664. 

  664. $N = preg_match('`[ts]([1-9])`', $w, $m) ? $m[1] : 0;
    665. 

  665. $a = array('br'=>1);
    666. 

  666. $b = array('button'=>1, 'input'=>1, 'option'=>1, 'param'=>1);
    667. 

  667. $c = array('caption'=>1, 'dd'=>1, 'dt'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'isindex'=>1, 'label'=>1, 'legend'=>1, 'li'=>1, 'object'=>1, 'p'=>1, 'pre'=>1, 'td'=>1, 'textarea'=>1, 'th'=>1);
    668. 

  668. $d = array('address'=>1, 'blockquote'=>1, 'center'=>1, 'colgroup'=>1, 'dir'=>1, 'div'=>1, 'dl'=>1, 'fieldset'=>1, 'form'=>1, 'hr'=>1, 'iframe'=>1, 'map'=>1, 'menu'=>1, 'noscript'=>1, 'ol'=>1, 'optgroup'=>1, 'rbc'=>1, 'rtc'=>1, 'ruby'=>1, 'script'=>1, 'select'=>1, 'table'=>1, 'tbody'=>1, 'tfoot'=>1, 'thead'=>1, 'tr'=>1, 'ul'=>1);
    669. 

  669. $T = explode('<', $t);
    670. 

  670. $X = 1;
    671. 

  671. while($X){
    672. 

  672.  $n = $N;
    673. 

  673.  $t = $T;
    674. 

  674.  ob_start();
    675. 

  675.  if(isset($d[$p])){echo str_repeat($s, ++$n);}
    676. 

  676.  echo ltrim(array_shift($t));
    677. 

  677.  for($i=-1, $j=count($t); ++$i<$j;){
    678. 

  678.   $r = ''; list($e, $r) = explode('>', $t[$i]);
    679. 

  679.   $x = $e[0] == '/' ? 0 : (substr($e, -1) == '/' ? 1 : ($e[0] != '!' ? 2 : -1));
    680. 

  680.   $y = !$x ? ltrim($e, '/') : ($x > 0 ? substr($e, 0, strcspn($e, ' ')) : 0);
    681. 

  681.   $e = "<$e>"; 
    682. 

  682.   if(isset($d[$y])){
    683. 

  683.    if(!$x){
    684. 

  684.     if($n){echo "\n", str_repeat($s, --$n), "$e\n", str_repeat($s, $n);}
    685. 

  685.     else{++$N; ob_end_clean(); continue 2;}
    686. 

  686.    }
    687. 

  687.    else{echo "\n", str_repeat($s, $n), "$e\n", str_repeat($s, ($x != 1 ? ++$n : $n));}
    688. 

  688.    echo $r; continue;
    689. 

  689.   }
    690. 

  690.   $f = "\n". str_repeat($s, $n);
    691. 

  691.   if(isset($c[$y])){
    692. 

  692.    if(!$x){echo $e, $f, $r;}
    693. 

  693.    else{echo $f, $e, $r;}
    694. 

  694.   }elseif(isset($b[$y])){echo $f, $e, $r;
    695. 

  695.   }elseif(isset($a[$y])){echo $e, $f, $r;
    696. 

  696.   }elseif(!$y){echo $f, $e, $f, $r;
    697. 

  697.   }else{echo $e, $r;}
    698. 

  698.  }
    699. 

  699.  $X = 0;
    700. 

  700. }
    701. 

  701. $t = str_replace(array("\n ", " \n"), "\n", preg_replace('`[\n]\s*?[\n]+`', "\n", ob_get_contents()));
    702. 

  702. ob_end_clean();
    703. 

  703. if(($l = strpos(" $w", 'r') ? (strpos(" $w", 'n') ? "\r\n" : "\r") : 0)){
    704. 

  704.  $t = str_replace("\n", $l, $t);
    705. 

  705. }
    706. 

  706. return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t);
    707. 

  707. // eof
    708. 

  708. }
    709. 

  709. 

  710. function hl_version(){
    711. 

  711. // rel
    712. 

  712. return '1.1.21';
    713. 

  713. // eof
    714. 

  714. }
    715. 

  715. 

  716. function kses($t, $h, $p=array('http', 'https', 'ftp', 'news', 'nntp', 'telnet', 'gopher', 'mailto')){
    717. 

  717. // kses compat
    718. 

  718. foreach($h as $k=>$v){
    719. 

  719.  $h[$k]['n']['*'] = 1;
    720. 

  720. }
    721. 

  721. $C['cdata'] = $C['comment'] = $C['make_tag_strict'] = $C['no_deprecated_attr'] = $C['unique_ids'] = 0;
    722. 

  722. $C['keep_bad'] = 1;
    723. 

  723. $C['elements'] = count($h) ? strtolower(implode(',', array_keys($h))) : '-*';
    724. 

  724. $C['hook'] = 'kses_hook';
    725. 

  725. $C['schemes'] = '*:'. implode(',', $p);
    726. 

  726. return htmLawed($t, $C, $h);
    727. 

  727. // eof
    728. 

  728. }
    729. 

  729. 

  730. function kses_hook($t, &$C, &$S){
    731. 

  731. // kses compat
    732. 

  732. return $t;
    733. 

  733. // eof
    734. 

  734. }
    735. 

  735.