htmLawed.php 47 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723
  1. <?php
  2. /*
  3. htmLawed 1.1.19, 19 January 2015
  4. Copyright Santosh Patnaik
  5. Dual licensed with LGPL 3 and GPL 2+
  6. A PHP Labware internal utility; www.bioinformatics.org/phplabware/internal_utilities/htmLawed
  7. See htmLawed_README.txt/htm
  8. */
  9. function htmLawed($t, $C=1, $S=array()){
  10. $C = is_array($C) ? $C : array();
  11. if(!empty($C['valid_xhtml'])){
  12. $C['elements'] = empty($C['elements']) ? '*-center-dir-font-isindex-menu-s-strike-u' : $C['elements'];
  13. $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 2;
  14. $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 2;
  15. }
  16. // config eles
  17. $e = array('a'=>1, 'abbr'=>1, 'acronym'=>1, 'address'=>1, 'applet'=>1, 'area'=>1, 'b'=>1, 'bdo'=>1, 'big'=>1, 'blockquote'=>1, 'br'=>1, 'button'=>1, 'caption'=>1, 'center'=>1, 'cite'=>1, 'code'=>1, 'col'=>1, 'colgroup'=>1, 'dd'=>1, 'del'=>1, 'dfn'=>1, 'dir'=>1, 'div'=>1, 'dl'=>1, 'dt'=>1, 'em'=>1, 'embed'=>1, 'fieldset'=>1, 'font'=>1, 'form'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'hr'=>1, 'i'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'ins'=>1, 'isindex'=>1, 'kbd'=>1, 'label'=>1, 'legend'=>1, 'li'=>1, 'map'=>1, 'menu'=>1, 'noscript'=>1, 'object'=>1, 'ol'=>1, 'optgroup'=>1, 'option'=>1, 'p'=>1, 'param'=>1, 'pre'=>1, 'q'=>1, 'rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1, 'ruby'=>1, 's'=>1, 'samp'=>1, 'script'=>1, 'select'=>1, 'small'=>1, 'span'=>1, 'strike'=>1, 'strong'=>1, 'sub'=>1, 'sup'=>1, 'table'=>1, 'tbody'=>1, 'td'=>1, 'textarea'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1, 'tt'=>1, 'u'=>1, 'ul'=>1, 'var'=>1); // 86/deprecated+embed+ruby
  18. if(!empty($C['safe'])){
  19. unset($e['applet'], $e['embed'], $e['iframe'], $e['object'], $e['script']);
  20. }
  21. $x = !empty($C['elements']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['elements']) : '*';
  22. if($x == '-*'){$e = array();}
  23. elseif(strpos($x, '*') === false){$e = array_flip(explode(',', $x));}
  24. else{
  25. if(isset($x[1])){
  26. preg_match_all('`(?:^|-|\+)[^\-+]+?(?=-|\+|$)`', $x, $m, PREG_SET_ORDER);
  27. for($i=count($m); --$i>=0;){$m[$i] = $m[$i][0];}
  28. foreach($m as $v){
  29. if($v[0] == '+'){$e[substr($v, 1)] = 1;}
  30. if($v[0] == '-' && isset($e[($v = substr($v, 1))]) && !in_array('+'. $v, $m)){unset($e[$v]);}
  31. }
  32. }
  33. }
  34. $C['elements'] =& $e;
  35. // config attrs
  36. $x = !empty($C['deny_attribute']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['deny_attribute']) : '';
  37. $x = array_flip((isset($x[0]) && $x[0] == '*') ? explode('-', $x) : explode(',', $x. (!empty($C['safe']) ? ',on*' : '')));
  38. if(isset($x['on*'])){
  39. unset($x['on*']);
  40. $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);
  41. }
  42. $C['deny_attribute'] = $x;
  43. // config URL
  44. $x = (isset($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';
  45. $C['schemes'] = array();
  46. foreach(explode(';', str_replace(array(' ', "\t", "\r", "\n"), '', $x)) as $v){
  47. $x = $x2 = null; list($x, $x2) = explode(':', $v, 2);
  48. if($x2){$C['schemes'][$x] = array_flip(explode(',', $x2));}
  49. }
  50. if(!isset($C['schemes']['*'])){$C['schemes']['*'] = array('file'=>1, 'http'=>1, 'https'=>1,);}
  51. if(!empty($C['safe']) && empty($C['schemes']['style'])){$C['schemes']['style'] = array('!'=>1);}
  52. $C['abs_url'] = isset($C['abs_url']) ? $C['abs_url'] : 0;
  53. if(!isset($C['base_url']) or !preg_match('`^[a-zA-Z\d.+\-]+://[^/]+/(.+?/)?$`', $C['base_url'])){
  54. $C['base_url'] = $C['abs_url'] = 0;
  55. }
  56. // config rest
  57. $C['and_mark'] = empty($C['and_mark']) ? 0 : 1;
  58. $C['anti_link_spam'] = (isset($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;
  59. $C['anti_mail_spam'] = isset($C['anti_mail_spam']) ? $C['anti_mail_spam'] : 0;
  60. $C['balance'] = isset($C['balance']) ? (bool)$C['balance'] : 1;
  61. $C['cdata'] = isset($C['cdata']) ? $C['cdata'] : (empty($C['safe']) ? 3 : 0);
  62. $C['clean_ms_char'] = empty($C['clean_ms_char']) ? 0 : $C['clean_ms_char'];
  63. $C['comment'] = isset($C['comment']) ? $C['comment'] : (empty($C['safe']) ? 3 : 0);
  64. $C['css_expression'] = empty($C['css_expression']) ? 0 : 1;
  65. $C['direct_list_nest'] = empty($C['direct_list_nest']) ? 0 : 1;
  66. $C['hexdec_entity'] = isset($C['hexdec_entity']) ? $C['hexdec_entity'] : 1;
  67. $C['hook'] = (!empty($C['hook']) && function_exists($C['hook'])) ? $C['hook'] : 0;
  68. $C['hook_tag'] = (!empty($C['hook_tag']) && function_exists($C['hook_tag'])) ? $C['hook_tag'] : 0;
  69. $C['keep_bad'] = isset($C['keep_bad']) ? $C['keep_bad'] : 6;
  70. $C['lc_std_val'] = isset($C['lc_std_val']) ? (bool)$C['lc_std_val'] : 1;
  71. $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 1;
  72. $C['named_entity'] = isset($C['named_entity']) ? (bool)$C['named_entity'] : 1;
  73. $C['no_deprecated_attr'] = isset($C['no_deprecated_attr']) ? $C['no_deprecated_attr'] : 1;
  74. $C['parent'] = isset($C['parent'][0]) ? strtolower($C['parent']) : 'body';
  75. $C['show_setting'] = !empty($C['show_setting']) ? $C['show_setting'] : 0;
  76. $C['style_pass'] = empty($C['style_pass']) ? 0 : 1;
  77. $C['tidy'] = empty($C['tidy']) ? 0 : $C['tidy'];
  78. $C['unique_ids'] = isset($C['unique_ids']) ? $C['unique_ids'] : 1;
  79. $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 0;
  80. if(isset($GLOBALS['C'])){$reC = $GLOBALS['C'];}
  81. $GLOBALS['C'] = $C;
  82. $S = is_array($S) ? $S : hl_spec($S);
  83. if(isset($GLOBALS['S'])){$reS = $GLOBALS['S'];}
  84. $GLOBALS['S'] = $S;
  85. $t = preg_replace('`[\x00-\x08\x0b-\x0c\x0e-\x1f]`', '', $t);
  86. if($C['clean_ms_char']){
  87. $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;');
  88. $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"=>'"'));
  89. $t = strtr($t, $x);
  90. }
  91. if($C['cdata'] or $C['comment']){$t = preg_replace_callback('`<!(?:(?:--.*?--)|(?:\[CDATA\[.*?\]\]))>`sm', 'hl_cmtcd', $t);}
  92. $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;', $t));
  93. if($C['unique_ids'] && !isset($GLOBALS['hl_Ids'])){$GLOBALS['hl_Ids'] = array();}
  94. if($C['hook']){$t = $C['hook']($t, $C, $S);}
  95. if($C['show_setting'] && preg_match('`^[a-z][a-z0-9_]*$`i', $C['show_setting'])){
  96. $GLOBALS[$C['show_setting']] = array('config'=>$C, 'spec'=>$S, 'time'=>microtime());
  97. }
  98. // main
  99. $t = preg_replace_callback('`<(?:(?:\s|$)|(?:[^>]*(?:>|$)))|>`m', 'hl_tag', $t);
  100. $t = $C['balance'] ? hl_bal($t, $C['keep_bad'], $C['parent']) : $t;
  101. $t = (($C['cdata'] or $C['comment']) && strpos($t, "\x01") !== false) ? str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05"), array('', '', '&', '<', '>'), $t) : $t;
  102. $t = $C['tidy'] ? hl_tidy($t, $C['tidy'], $C['parent']) : $t;
  103. unset($C, $e);
  104. if(isset($reC)){$GLOBALS['C'] = $reC;}
  105. if(isset($reS)){$GLOBALS['S'] = $reS;}
  106. return $t;
  107. // eof
  108. }
  109. function hl_attrval($t, $p){
  110. // check attr val against $S
  111. $o = 1; $l = strlen($t);
  112. foreach($p as $k=>$v){
  113. switch($k){
  114. case 'maxlen':if($l > $v){$o = 0;}
  115. break; case 'minlen': if($l < $v){$o = 0;}
  116. break; case 'maxval': if((float)($t) > $v){$o = 0;}
  117. break; case 'minval': if((float)($t) < $v){$o = 0;}
  118. break; case 'match': if(!preg_match($v, $t)){$o = 0;}
  119. break; case 'nomatch': if(preg_match($v, $t)){$o = 0;}
  120. break; case 'oneof':
  121. $m = 0;
  122. foreach(explode('|', $v) as $n){if($t == $n){$m = 1; break;}}
  123. $o = $m;
  124. break; case 'noneof':
  125. $m = 1;
  126. foreach(explode('|', $v) as $n){if($t == $n){$m = 0; break;}}
  127. $o = $m;
  128. break; default:
  129. break;
  130. }
  131. if(!$o){break;}
  132. }
  133. return ($o ? $t : (isset($p['default']) ? $p['default'] : 0));
  134. // eof
  135. }
  136. function hl_bal($t, $do=1, $in='div'){
  137. // balance tags
  138. // by content
  139. $cB = array('blockquote'=>1, 'form'=>1, 'map'=>1, 'noscript'=>1); // Block
  140. $cE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty
  141. $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
  142. $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
  143. $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
  144. $cN2 = array_keys($cN);
  145. $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);
  146. $cS = array('colgroup'=>array('col'=>1), 'dir'=>array('li'=>1), 'dl'=>array('dd'=>1, 'dt'=>1), 'menu'=>array('li'=>1), 'ol'=>array('li'=>1), 'optgroup'=>array('option'=>1), 'option'=>array('#pcdata'=>1), 'rbc'=>array('rb'=>1), 'rp'=>array('#pcdata'=>1), 'rtc'=>array('rt'=>1), 'ruby'=>array('rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1), 'select'=>array('optgroup'=>1, 'option'=>1), 'script'=>array('#pcdata'=>1), 'table'=>array('caption'=>1, 'col'=>1, 'colgroup'=>1, 'tfoot'=>1, 'tbody'=>1, 'tr'=>1, 'thead'=>1), 'tbody'=>array('tr'=>1), 'tfoot'=>array('tr'=>1), 'textarea'=>array('#pcdata'=>1), 'thead'=>array('tr'=>1), 'tr'=>array('td'=>1, 'th'=>1), 'ul'=>array('li'=>1)); // Specific - immediate parent-child
  147. if($GLOBALS['C']['direct_list_nest']){$cS['ol'] = $cS['ul'] += array('ol'=>1, 'ul'=>1);}
  148. $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
  149. $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
  150. // block/inline type; ins & del both type; #pcdata: text
  151. $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);
  152. $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);
  153. $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
  154. $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
  155. $eF = $eB + $eI;
  156. // $in sets allowed child
  157. $in = ((isset($eF[$in]) && $in != '#pcdata') or isset($eO[$in])) ? $in : 'div';
  158. if(isset($cE[$in])){
  159. return (!$do ? '' : str_replace(array('<', '>'), array('&lt;', '&gt;'), $t));
  160. }
  161. if(isset($cS[$in])){$inOk = $cS[$in];}
  162. elseif(isset($cI[$in])){$inOk = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
  163. elseif(isset($cF[$in])){$inOk = $eF; unset($cI['del'], $cI['ins']);}
  164. elseif(isset($cB[$in])){$inOk = $eB; unset($cI['del'], $cI['ins']);}
  165. if(isset($cO[$in])){$inOk = $inOk + $cO[$in];}
  166. if(isset($cN[$in])){$inOk = array_diff_assoc($inOk, $cN[$in]);}
  167. $t = explode('<', $t);
  168. $ok = $q = array(); // $q seq list of open non-empty ele
  169. ob_start();
  170. for($i=-1, $ci=count($t); ++$i<$ci;){
  171. // allowed $ok in parent $p
  172. if($ql = count($q)){
  173. $p = array_pop($q);
  174. $q[] = $p;
  175. if(isset($cS[$p])){$ok = $cS[$p];}
  176. elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
  177. elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);}
  178. elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);}
  179. if(isset($cO[$p])){$ok = $ok + $cO[$p];}
  180. if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);}
  181. }else{$ok = $inOk; unset($cI['del'], $cI['ins']);}
  182. // bad tags, & ele content
  183. if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){
  184. echo '&lt;', $s, $e, $a, '&gt;';
  185. }
  186. if(isset($x[0])){
  187. if(strlen(trim($x)) && (($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql))){
  188. echo '<div>', $x, '</div>';
  189. }
  190. elseif($do < 3 or isset($ok['#pcdata'])){echo $x;}
  191. elseif(strpos($x, "\x02\x04")){
  192. foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){
  193. echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : ''));
  194. }
  195. }elseif($do > 4){echo preg_replace('`\S`', '', $x);}
  196. }
  197. // get markup
  198. if(!preg_match('`^(/?)([a-z1-6]+)([^>]*)>(.*)`sm', $t[$i], $r)){$x = $t[$i]; continue;}
  199. $s = null; $e = null; $a = null; $x = null; list($all, $s, $e, $a, $x) = $r;
  200. // close tag
  201. if($s){
  202. if(isset($cE[$e]) or !in_array($e, $q)){continue;} // Empty/unopen
  203. if($p == $e){array_pop($q); echo '</', $e, '>'; unset($e); continue;} // Last open
  204. $add = ''; // Nesting - close open tags that need to be
  205. for($j=-1, $cj=count($q); ++$j<$cj;){
  206. if(($d = array_pop($q)) == $e){break;}
  207. else{$add .= "</{$d}>";}
  208. }
  209. echo $add, '</', $e, '>'; unset($e); continue;
  210. }
  211. // open tag
  212. // $cB ele needs $eB ele as child
  213. if(isset($cB[$e]) && strlen(trim($x))){
  214. $t[$i] = "{$e}{$a}>";
  215. array_splice($t, $i+1, 0, 'div>'. $x); unset($e, $x); ++$ci; --$i; continue;
  216. }
  217. if((($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql)) && !isset($eB[$e]) && !isset($ok[$e])){
  218. array_splice($t, $i, 0, 'div>'); unset($e, $x); ++$ci; --$i; continue;
  219. }
  220. // if no open ele, $in = parent; mostly immediate parent-child relation should hold
  221. if(!$ql or !isset($eN[$e]) or !array_intersect($q, $cN2)){
  222. if(!isset($ok[$e])){
  223. if($ql && isset($cT[$p])){echo '</', array_pop($q), '>'; unset($e, $x); --$i;}
  224. continue;
  225. }
  226. if(!isset($cE[$e])){$q[] = $e;}
  227. echo '<', $e, $a, '>'; unset($e); continue;
  228. }
  229. // specific parent-child
  230. if(isset($cS[$p][$e])){
  231. if(!isset($cE[$e])){$q[] = $e;}
  232. echo '<', $e, $a, '>'; unset($e); continue;
  233. }
  234. // nesting
  235. $add = '';
  236. $q2 = array();
  237. for($k=-1, $kc=count($q); ++$k<$kc;){
  238. $d = $q[$k];
  239. $ok2 = array();
  240. if(isset($cS[$d])){$q2[] = $d; continue;}
  241. $ok2 = isset($cI[$d]) ? $eI : $eF;
  242. if(isset($cO[$d])){$ok2 = $ok2 + $cO[$d];}
  243. if(isset($cN[$d])){$ok2 = array_diff_assoc($ok2, $cN[$d]);}
  244. if(!isset($ok2[$e])){
  245. if(!$k && !isset($inOk[$e])){continue 2;}
  246. $add = "</{$d}>";
  247. for(;++$k<$kc;){$add = "</{$q[$k]}>{$add}";}
  248. break;
  249. }
  250. else{$q2[] = $d;}
  251. }
  252. $q = $q2;
  253. if(!isset($cE[$e])){$q[] = $e;}
  254. echo $add, '<', $e, $a, '>'; unset($e); continue;
  255. }
  256. // end
  257. if($ql = count($q)){
  258. $p = array_pop($q);
  259. $q[] = $p;
  260. if(isset($cS[$p])){$ok = $cS[$p];}
  261. elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
  262. elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);}
  263. elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);}
  264. if(isset($cO[$p])){$ok = $ok + $cO[$p];}
  265. if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);}
  266. }else{$ok = $inOk; unset($cI['del'], $cI['ins']);}
  267. if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){
  268. echo '&lt;', $s, $e, $a, '&gt;';
  269. }
  270. if(isset($x[0])){
  271. if(strlen(trim($x)) && (($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql))){
  272. echo '<div>', $x, '</div>';
  273. }
  274. elseif($do < 3 or isset($ok['#pcdata'])){echo $x;}
  275. elseif(strpos($x, "\x02\x04")){
  276. foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){
  277. echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : ''));
  278. }
  279. }elseif($do > 4){echo preg_replace('`\S`', '', $x);}
  280. }
  281. while(!empty($q) && ($e = array_pop($q))){echo '</', $e, '>';}
  282. $o = ob_get_contents();
  283. ob_end_clean();
  284. return $o;
  285. // eof
  286. }
  287. function hl_cmtcd($t){
  288. // comment/CDATA sec handler
  289. $t = $t[0];
  290. global $C;
  291. if(!($v = $C[$n = $t[3] == '-' ? 'comment' : 'cdata'])){return $t;}
  292. if($v == 1){return '';}
  293. if($n == 'comment'){
  294. if(substr(($t = preg_replace('`--+`', '-', substr($t, 4, -3))), -1) != ' '){$t .= ' ';}
  295. }
  296. else{$t = substr($t, 1, -1);}
  297. $t = $v == 2 ? str_replace(array('&', '<', '>'), array('&amp;', '&lt;', '&gt;'), $t) : $t;
  298. return str_replace(array('&', '<', '>'), array("\x03", "\x04", "\x05"), ($n == 'comment' ? "\x01\x02\x04!--$t--\x05\x02\x01" : "\x01\x01\x04$t\x05\x01\x01"));
  299. // eof
  300. }
  301. function hl_ent($t){
  302. // entitity handler
  303. global $C;
  304. $t = $t[1];
  305. static $U = array('quot'=>1,'amp'=>1,'lt'=>1,'gt'=>1);
  306. 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');
  307. if($t[0] != '#'){
  308. 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;'. $t)). ';';
  309. }
  310. if(($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 $n > 1114111))){
  311. return ($C['and_mark'] ? "\x06" : '&'). "amp;#{$t};";
  312. }
  313. return ($C['and_mark'] ? "\x06" : '&'). '#'. (((ctype_digit($t) && $C['hexdec_entity'] < 2) or !$C['hexdec_entity']) ? $n : 'x'. dechex($n)). ';';
  314. // eof
  315. }
  316. function hl_prot($p, $c=null){
  317. // check URL scheme
  318. global $C;
  319. $b = $a = '';
  320. if($c == null){$c = 'style'; $b = $p[1]; $a = $p[3]; $p = trim($p[2]);}
  321. $c = isset($C['schemes'][$c]) ? $C['schemes'][$c] : $C['schemes']['*'];
  322. static $d = 'denied:';
  323. if(isset($c['!']) && substr($p, 0, 7) != $d){$p = "$d$p";}
  324. if(isset($c['*']) or !strcspn($p, '#?;') or (substr($p, 0, 7) == $d)){return "{$b}{$p}{$a}";} // All ok, frag, query, param
  325. if(preg_match('`^([^:?[@!$()*,=/\'\]]+?)(:|&#(58|x3a);|%3a|\\\\0{0,4}3a).`i', $p, $m) && !isset($c[strtolower($m[1])])){ // Denied prot
  326. return "{$b}{$d}{$p}{$a}";
  327. }
  328. if($C['abs_url']){
  329. if($C['abs_url'] == -1 && strpos($p, $C['base_url']) === 0){ // Make url rel
  330. $p = substr($p, strlen($C['base_url']));
  331. }elseif(empty($m[1])){ // Make URL abs
  332. if(substr($p, 0, 2) == '//'){$p = substr($C['base_url'], 0, strpos($C['base_url'], ':')+1). $p;}
  333. elseif($p[0] == '/'){$p = preg_replace('`(^.+?://[^/]+)(.*)`', '$1', $C['base_url']). $p;}
  334. elseif(strcspn($p, './')){$p = $C['base_url']. $p;}
  335. else{
  336. preg_match('`^([a-zA-Z\d\-+.]+://[^/]+)(.*)`', $C['base_url'], $m);
  337. $p = preg_replace('`(?<=/)\./`', '', $m[2]. $p);
  338. while(preg_match('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', $p)){
  339. $p = preg_replace('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', '', $p);
  340. }
  341. $p = $m[1]. $p;
  342. }
  343. }
  344. }
  345. return "{$b}{$p}{$a}";
  346. // eof
  347. }
  348. function hl_regex($p){
  349. // ?regex
  350. if(empty($p)){return 0;}
  351. if($t = ini_get('track_errors')){$o = isset($php_errormsg) ? $php_errormsg : null;}
  352. else{ini_set('track_errors', 1);}
  353. unset($php_errormsg);
  354. if(($d = ini_get('display_errors'))){ini_set('display_errors', 0);}
  355. preg_match($p, '');
  356. if($d){ini_set('display_errors', 1);}
  357. $r = isset($php_errormsg) ? 0 : 1;
  358. if($t){$php_errormsg = isset($o) ? $o : null;}
  359. else{ini_set('track_errors', 0);}
  360. return $r;
  361. // eof
  362. }
  363. function hl_spec($t){
  364. // final $spec
  365. $s = array();
  366. $t = str_replace(array("\t", "\r", "\n", ' '), '', preg_replace_callback('/"(?>(`.|[^"])*)"/sm', create_function('$m', 'return substr(str_replace(array(";", "|", "~", " ", ",", "/", "(", ")", \'`"\'), array("\x01", "\x02", "\x03", "\x04", "\x05", "\x06", "\x07", "\x08", "\""), $m[0]), 1, -1);'), trim($t)));
  367. for($i = count(($t = explode(';', $t))); --$i>=0;){
  368. $w = $t[$i];
  369. if(empty($w) or ($e = strpos($w, '=')) === false or !strlen(($a = substr($w, $e+1)))){continue;}
  370. $y = $n = array();
  371. foreach(explode(',', $a) as $v){
  372. if(!preg_match('`^([a-z:\-\*]+)(?:\((.*?)\))?`i', $v, $m)){continue;}
  373. if(($x = strtolower($m[1])) == '-*'){$n['*'] = 1; continue;}
  374. if($x[0] == '-'){$n[substr($x, 1)] = 1; continue;}
  375. if(!isset($m[2])){$y[$x] = 1; continue;}
  376. foreach(explode('/', $m[2]) as $m){
  377. if(empty($m) or ($p = strpos($m, '=')) == 0 or $p < 5){$y[$x] = 1; continue;}
  378. $y[$x][strtolower(substr($m, 0, $p))] = str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x06", "\x07", "\x08"), array(";", "|", "~", " ", ",", "/", "(", ")"), substr($m, $p+1));
  379. }
  380. if(isset($y[$x]['match']) && !hl_regex($y[$x]['match'])){unset($y[$x]['match']);}
  381. if(isset($y[$x]['nomatch']) && !hl_regex($y[$x]['nomatch'])){unset($y[$x]['nomatch']);}
  382. }
  383. if(!count($y) && !count($n)){continue;}
  384. foreach(explode(',', substr($w, 0, $e)) as $v){
  385. if(!strlen(($v = strtolower($v)))){continue;}
  386. if(count($y)){$s[$v] = $y;}
  387. if(count($n)){$s[$v]['n'] = $n;}
  388. }
  389. }
  390. return $s;
  391. // eof
  392. }
  393. function hl_tag($t){
  394. // tag/attribute handler
  395. global $C;
  396. $t = $t[0];
  397. // invalid < >
  398. if($t == '< '){return '&lt; ';}
  399. if($t == '>'){return '&gt;';}
  400. if(!preg_match('`^<(/?)([a-zA-Z][a-zA-Z1-6]*)([^>]*?)\s?>$`m', $t, $m)){
  401. return str_replace(array('<', '>'), array('&lt;', '&gt;'), $t);
  402. }elseif(!isset($C['elements'][($e = strtolower($m[2]))])){
  403. return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('&lt;', '&gt;'), $t) : '');
  404. }
  405. // attr string
  406. $a = str_replace(array("\n", "\r", "\t"), ' ', trim($m[3]));
  407. // tag transform
  408. static $eD = array('applet'=>1, 'center'=>1, 'dir'=>1, 'embed'=>1, 'font'=>1, 'isindex'=>1, 'menu'=>1, 's'=>1, 'strike'=>1, 'u'=>1); // Deprecated
  409. if($C['make_tag_strict'] && isset($eD[$e])){
  410. $trt = hl_tag2($e, $a, $C['make_tag_strict']);
  411. if(!$e){return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('&lt;', '&gt;'), $t) : '');}
  412. }
  413. // close tag
  414. static $eE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty ele
  415. if(!empty($m[1])){
  416. return (!isset($eE[$e]) ? (empty($C['hook_tag']) ? "</$e>" : $C['hook_tag']($e)) : (($C['keep_bad'])%2 ? str_replace(array('<', '>'), array('&lt;', '&gt;'), $t) : ''));
  417. }
  418. // open tag & attr
  419. 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), '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
  420. static $aNE = array('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
  421. 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
  422. static $aNU = array('class'=>array('param'=>1, 'script'=>1), 'dir'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'id'=>array('script'=>1), 'lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'xml:lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'onclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'ondblclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeydown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeypress'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeyup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousedown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousemove'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseout'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseover'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'style'=>array('param'=>1, 'script'=>1), 'title'=>array('param'=>1, 'script'=>1)); // Univ & exceptions
  423. if($C['lc_std_val']){
  424. // predef attr vals for $eAL & $aNE ele
  425. 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);
  426. 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);
  427. $lcase = isset($eAL[$e]) ? 1 : 0;
  428. }
  429. $depTr = 0;
  430. if($C['no_deprecated_attr']){
  431. // dep attr:applicable ele
  432. 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));
  433. 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);
  434. $depTr = isset($eAD[$e]) ? 1 : 0;
  435. }
  436. // attr name-vals
  437. if(strpos($a, "\x01") !== false){$a = preg_replace('`\x01[^\x01]*\x01`', '', $a);} // No comment/CDATA sec
  438. $mode = 0; $a = trim($a, ' /'); $aA = array();
  439. while(strlen($a)){
  440. $w = 0;
  441. switch($mode){
  442. case 0: // Name
  443. if(preg_match('`^[a-zA-Z][\-a-zA-Z:]+`', $a, $m)){
  444. $nm = strtolower($m[0]);
  445. $w = $mode = 1; $a = ltrim(substr_replace($a, '', 0, strlen($m[0])));
  446. }
  447. break; case 1:
  448. if($a[0] == '='){ // =
  449. $w = 1; $mode = 2; $a = ltrim($a, '= ');
  450. }else{ // No val
  451. $w = 1; $mode = 0; $a = ltrim($a);
  452. $aA[$nm] = '';
  453. }
  454. break; case 2: // Val
  455. if(preg_match('`^((?:"[^"]*")|(?:\'[^\']*\')|(?:\s*[^\s"\']+))(.*)`', $a, $m)){
  456. $a = ltrim($m[2]); $m = $m[1]; $w = 1; $mode = 0;
  457. $aA[$nm] = trim(str_replace('<', '&lt;', ($m[0] == '"' or $m[0] == '\'') ? substr($m, 1, -1) : $m));
  458. }
  459. break;
  460. }
  461. if($w == 0){ // Parse errs, deal with space, " & '
  462. $a = preg_replace('`^(?:"[^"]*("|$)|\'[^\']*(\'|$)|\S)*\s*`', '', $a);
  463. $mode = 0;
  464. }
  465. }
  466. if($mode == 1){$aA[$nm] = '';}
  467. // clean attrs
  468. global $S;
  469. $rl = isset($S[$e]) ? $S[$e] : array();
  470. $a = array(); $nfr = 0;
  471. foreach($aA as $k=>$v){
  472. 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])){
  473. if(isset($aNE[$k])){$v = $k;}
  474. elseif(!empty($lcase) && (($e != 'button' or $e != 'input') or $k == 'type')){ // Rather loose but ?not cause issues
  475. $v = (isset($aNL[($v2 = strtolower($v))])) ? $v2 : $v;
  476. }
  477. if($k == 'style' && !$C['style_pass']){
  478. if(false !== strpos($v, '&#')){
  479. 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;'=>'\\');
  480. $v = strtr($v, $sC);
  481. }
  482. $v = preg_replace_callback('`(url(?:\()(?: )*(?:\'|"|&(?:quot|apos);)?)(.+?)((?:\'|"|&(?:quot|apos);)?(?: )*(?:\)))`iS', 'hl_prot', $v);
  483. $v = !$C['css_expression'] ? preg_replace('`expression`i', ' ', preg_replace('`\\\\\S|(/|(%2f))(\*|(%2a))`i', ' ', $v)) : $v;
  484. }elseif(isset($aNP[$k]) or strpos($k, 'src') !== false or $k[0] == 'o'){
  485. $v = str_replace("­", ' ', (strpos($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
  486. $v = hl_prot($v, $k);
  487. if($k == 'href'){ // X-spam
  488. if($C['anti_mail_spam'] && strpos($v, 'mailto:') === 0){
  489. $v = str_replace('@', htmlspecialchars($C['anti_mail_spam']), $v);
  490. }elseif($C['anti_link_spam']){
  491. $r1 = $C['anti_link_spam'][1];
  492. if(!empty($r1) && preg_match($r1, $v)){continue;}
  493. $r0 = $C['anti_link_spam'][0];
  494. if(!empty($r0) && preg_match($r0, $v)){
  495. if(isset($a['rel'])){
  496. if(!preg_match('`\bnofollow\b`i', $a['rel'])){$a['rel'] .= ' nofollow';}
  497. }elseif(isset($aA['rel'])){
  498. if(!preg_match('`\bnofollow\b`i', $aA['rel'])){$nfr = 1;}
  499. }else{$a['rel'] = 'nofollow';}
  500. }
  501. }
  502. }
  503. }
  504. if(isset($rl[$k]) && is_array($rl[$k]) && ($v = hl_attrval($v, $rl[$k])) === 0){continue;}
  505. $a[$k] = str_replace('"', '&quot;', $v);
  506. }
  507. }
  508. if($nfr){$a['rel'] = isset($a['rel']) ? $a['rel']. ' nofollow' : 'nofollow';}
  509. // rqd attr
  510. static $eAR = array('area'=>array('alt'=>'area'), 'bdo'=>array('dir'=>'ltr'), 'form'=>array('action'=>''), 'img'=>array('src'=>'', 'alt'=>'image'), 'map'=>array('name'=>''), 'optgroup'=>array('label'=>''), 'param'=>array('name'=>''), 'script'=>array('type'=>'text/javascript'), 'textarea'=>array('rows'=>'10', 'cols'=>'50'));
  511. if(isset($eAR[$e])){
  512. foreach($eAR[$e] as $k=>$v){
  513. if(!isset($a[$k])){$a[$k] = isset($v[0]) ? $v : $k;}
  514. }
  515. }
  516. // depr attrs
  517. if($depTr){
  518. $c = array();
  519. foreach($a as $k=>$v){
  520. if($k == 'style' or !isset($aND[$k][$e])){continue;}
  521. if($k == 'align'){
  522. unset($a['align']);
  523. if($e == 'img' && ($v == 'left' or $v == 'right')){$c[] = 'float: '. $v;}
  524. elseif(($e == 'div' or $e == 'table') && $v == 'center'){$c[] = 'margin: auto';}
  525. else{$c[] = 'text-align: '. $v;}
  526. }elseif($k == 'bgcolor'){
  527. unset($a['bgcolor']);
  528. $c[] = 'background-color: '. $v;
  529. }elseif($k == 'border'){
  530. unset($a['border']); $c[] = "border: {$v}px";
  531. }elseif($k == 'bordercolor'){
  532. unset($a['bordercolor']); $c[] = 'border-color: '. $v;
  533. }elseif($k == 'clear'){
  534. unset($a['clear']); $c[] = 'clear: '. ($v != 'all' ? $v : 'both');
  535. }elseif($k == 'compact'){
  536. unset($a['compact']); $c[] = 'font-size: 85%';
  537. }elseif($k == 'height' or $k == 'width'){
  538. unset($a[$k]); $c[] = $k. ': '. ($v[0] != '*' ? $v. (ctype_digit($v) ? 'px' : '') : 'auto');
  539. }elseif($k == 'hspace'){
  540. unset($a['hspace']); $c[] = "margin-left: {$v}px; margin-right: {$v}px";
  541. }elseif($k == 'language' && !isset($a['type'])){
  542. unset($a['language']);
  543. $a['type'] = 'text/'. strtolower($v);
  544. }elseif($k == 'name'){
  545. if($C['no_deprecated_attr'] == 2 or ($e != 'a' && $e != 'map')){unset($a['name']);}
  546. if(!isset($a['id']) && preg_match('`[a-zA-Z][a-zA-Z\d.:_\-]*`', $v)){$a['id'] = $v;}
  547. }elseif($k == 'noshade'){
  548. unset($a['noshade']); $c[] = 'border-style: none; border: 0; background-color: gray; color: gray';
  549. }elseif($k == 'nowrap'){
  550. unset($a['nowrap']); $c[] = 'white-space: nowrap';
  551. }elseif($k == 'size'){
  552. unset($a['size']); $c[] = 'size: '. $v. 'px';
  553. }elseif($k == 'start' or $k == 'value'){
  554. unset($a[$k]);
  555. }elseif($k == 'type'){
  556. unset($a['type']);
  557. static $ol_type = array('i'=>'lower-roman', 'I'=>'upper-roman', 'a'=>'lower-latin', 'A'=>'upper-latin', '1'=>'decimal');
  558. $c[] = 'list-style-type: '. (isset($ol_type[$v]) ? $ol_type[$v] : 'decimal');
  559. }elseif($k == 'vspace'){
  560. unset($a['vspace']); $c[] = "margin-top: {$v}px; margin-bottom: {$v}px";
  561. }
  562. }
  563. if(count($c)){
  564. $c = implode('; ', $c);
  565. $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $c. ';': $c. ';';
  566. }
  567. }
  568. // unique ID
  569. if($C['unique_ids'] && isset($a['id'])){
  570. if(!preg_match('`^[A-Za-z][A-Za-z0-9_\-.:]*$`', ($id = $a['id'])) or (isset($GLOBALS['hl_Ids'][$id]) && $C['unique_ids'] == 1)){unset($a['id']);
  571. }else{
  572. while(isset($GLOBALS['hl_Ids'][$id])){$id = $C['unique_ids']. $id;}
  573. $GLOBALS['hl_Ids'][($a['id'] = $id)] = 1;
  574. }
  575. }
  576. // xml:lang
  577. if($C['xml:lang'] && isset($a['lang'])){
  578. $a['xml:lang'] = isset($a['xml:lang']) ? $a['xml:lang'] : $a['lang'];
  579. if($C['xml:lang'] == 2){unset($a['lang']);}
  580. }
  581. // for transformed tag
  582. if(!empty($trt)){
  583. $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $trt : $trt;
  584. }
  585. // return with empty ele /
  586. if(empty($C['hook_tag'])){
  587. $aA = '';
  588. foreach($a as $k=>$v){$aA .= " {$k}=\"{$v}\"";}
  589. return "<{$e}{$aA}". (isset($eE[$e]) ? ' /' : ''). '>';
  590. }
  591. else{return $C['hook_tag']($e, $a);}
  592. // eof
  593. }
  594. function hl_tag2(&$e, &$a, $t=1){
  595. // transform tag
  596. if($e == 'center'){$e = 'div'; return 'text-align: center;';}
  597. if($e == 'dir' or $e == 'menu'){$e = 'ul'; return '';}
  598. if($e == 's' or $e == 'strike'){$e = 'span'; return 'text-decoration: line-through;';}
  599. if($e == 'u'){$e = 'span'; return 'text-decoration: underline;';}
  600. static $fs = array('0'=>'xx-small', '1'=>'xx-small', '2'=>'small', '3'=>'medium', '4'=>'large', '5'=>'x-large', '6'=>'xx-large', '7'=>'300%', '-1'=>'smaller', '-2'=>'60%', '+1'=>'larger', '+2'=>'150%', '+3'=>'200%', '+4'=>'300%');
  601. if($e == 'font'){
  602. $a2 = '';
  603. if(preg_match('`face\s*=\s*(\'|")([^=]+?)\\1`i', $a, $m) or preg_match('`face\s*=(\s*)(\S+)`i', $a, $m)){
  604. $a2 .= ' font-family: '. str_replace('"', '\'', trim($m[2])). ';';
  605. }
  606. if(preg_match('`color\s*=\s*(\'|")?(.+?)(\\1|\s|$)`i', $a, $m)){
  607. $a2 .= ' color: '. trim($m[2]). ';';
  608. }
  609. if(preg_match('`size\s*=\s*(\'|")?(.+?)(\\1|\s|$)`i', $a, $m) && isset($fs[($m = trim($m[2]))])){
  610. $a2 .= ' font-size: '. $fs[$m]. ';';
  611. }
  612. $e = 'span'; return ltrim($a2);
  613. }
  614. if($t == 2){$e = 0; return 0;}
  615. return '';
  616. // eof
  617. }
  618. function hl_tidy($t, $w, $p){
  619. // Tidy/compact HTM
  620. if(strpos(' pre,script,textarea', "$p,")){return $t;}
  621. $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];'), $t));
  622. if(($w = strtolower($w)) == -1){
  623. return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t);
  624. }
  625. $s = strpos(" $w", 't') ? "\t" : ' ';
  626. $s = preg_match('`\d`', $w, $m) ? str_repeat($s, $m[0]) : str_repeat($s, ($s == "\t" ? 1 : 2));
  627. $N = preg_match('`[ts]([1-9])`', $w, $m) ? $m[1] : 0;
  628. $a = array('br'=>1);
  629. $b = array('button'=>1, 'input'=>1, 'option'=>1, 'param'=>1);
  630. $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);
  631. $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);
  632. $T = explode('<', $t);
  633. $X = 1;
  634. while($X){
  635. $n = $N;
  636. $t = $T;
  637. ob_start();
  638. if(isset($d[$p])){echo str_repeat($s, ++$n);}
  639. echo ltrim(array_shift($t));
  640. for($i=-1, $j=count($t); ++$i<$j;){
  641. $r = ''; list($e, $r) = explode('>', $t[$i]);
  642. $x = $e[0] == '/' ? 0 : (substr($e, -1) == '/' ? 1 : ($e[0] != '!' ? 2 : -1));
  643. $y = !$x ? ltrim($e, '/') : ($x > 0 ? substr($e, 0, strcspn($e, ' ')) : 0);
  644. $e = "<$e>";
  645. if(isset($d[$y])){
  646. if(!$x){
  647. if($n){echo "\n", str_repeat($s, --$n), "$e\n", str_repeat($s, $n);}
  648. else{++$N; ob_end_clean(); continue 2;}
  649. }
  650. else{echo "\n", str_repeat($s, $n), "$e\n", str_repeat($s, ($x != 1 ? ++$n : $n));}
  651. echo $r; continue;
  652. }
  653. $f = "\n". str_repeat($s, $n);
  654. if(isset($c[$y])){
  655. if(!$x){echo $e, $f, $r;}
  656. else{echo $f, $e, $r;}
  657. }elseif(isset($b[$y])){echo $f, $e, $r;
  658. }elseif(isset($a[$y])){echo $e, $f, $r;
  659. }elseif(!$y){echo $f, $e, $f, $r;
  660. }else{echo $e, $r;}
  661. }
  662. $X = 0;
  663. }
  664. $t = str_replace(array("\n ", " \n"), "\n", preg_replace('`[\n]\s*?[\n]+`', "\n", ob_get_contents()));
  665. ob_end_clean();
  666. if(($l = strpos(" $w", 'r') ? (strpos(" $w", 'n') ? "\r\n" : "\r") : 0)){
  667. $t = str_replace("\n", $l, $t);
  668. }
  669. return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t);
  670. // eof
  671. }
  672. function hl_version(){
  673. // rel
  674. return '1.1.19';
  675. // eof
  676. }
  677. function kses($t, $h, $p=array('http', 'https', 'ftp', 'news', 'nntp', 'telnet', 'gopher', 'mailto')){
  678. // kses compat
  679. foreach($h as $k=>$v){
  680. $h[$k]['n']['*'] = 1;
  681. }
  682. $C['cdata'] = $C['comment'] = $C['make_tag_strict'] = $C['no_deprecated_attr'] = $C['unique_ids'] = 0;
  683. $C['keep_bad'] = 1;
  684. $C['elements'] = count($h) ? strtolower(implode(',', array_keys($h))) : '-*';
  685. $C['hook'] = 'kses_hook';
  686. $C['schemes'] = '*:'. implode(',', $p);
  687. return htmLawed($t, $C, $h);
  688. // eof
  689. }
  690. function kses_hook($t, &$C, &$S){
  691. // kses compat
  692. return $t;
  693. // eof
  694. }