4 htmLawed 1.1.8.1, 16 July 2009
5 Copyright Santosh Patnaik
7 A PHP Labware internal utility; www.bioinformatics.org/phplabware/internal_utilities/htmLawed
9 See htmLawed_README.txt/htm
12 function htmLawed($t, $C=1, $S=array()){
13 $C = is_array($C) ? $C : array();
14 if(!empty($C['valid_xhtml'])){
15 $C['elements'] = empty($C['elements']) ? '*-center-dir-font-isindex-menu-s-strike-u' : $C['elements'];
16 $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 2;
17 $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 2;
20 $e = array('a'=>1, 'abbr'=>1, 'acronym'=>1, 'address'=>1, 'applet'=>1, 'area'=>1, 'b'=>1, 'bdo'=>1, 'big'=>1, 'blockquote'=>1, 'br'=>1, 'button'=>1, 'caption'=>1, 'center'=>1, 'cite'=>1, 'code'=>1, 'col'=>1, 'colgroup'=>1, 'dd'=>1, 'del'=>1, 'dfn'=>1, 'dir'=>1, 'div'=>1, 'dl'=>1, 'dt'=>1, 'em'=>1, 'embed'=>1, 'fieldset'=>1, 'font'=>1, 'form'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'hr'=>1, 'i'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'ins'=>1, 'isindex'=>1, 'kbd'=>1, 'label'=>1, 'legend'=>1, 'li'=>1, 'map'=>1, 'menu'=>1, 'noscript'=>1, 'object'=>1, 'ol'=>1, 'optgroup'=>1, 'option'=>1, 'p'=>1, 'param'=>1, 'pre'=>1, 'q'=>1, 'rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1, 'ruby'=>1, 's'=>1, 'samp'=>1, 'script'=>1, 'select'=>1, 'small'=>1, 'span'=>1, 'strike'=>1, 'strong'=>1, 'sub'=>1, 'sup'=>1, 'table'=>1, 'tbody'=>1, 'td'=>1, 'textarea'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1, 'tt'=>1, 'u'=>1, 'ul'=>1, 'var'=>1); // 86/deprecated+embed+ruby
21 if(!empty($C['safe'])){
22 unset($e['applet'], $e['embed'], $e['iframe'], $e['object'], $e['script']);
24 $x = !empty($C['elements']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['elements']) : '*';
25 if($x == '-*'){$e = array();}
26 elseif(strpos($x, '*') === false){$e = array_flip(explode(',', $x));}
29 preg_match_all('`(?:^|-|\+)[^\-+]+?(?=-|\+|$)`', $x, $m, PREG_SET_ORDER);
30 for($i=count($m); --$i>=0;){$m[$i] = $m[$i][0];}
32 if($v[0] == '+'){$e[substr($v, 1)] = 1;}
33 if($v[0] == '-' && isset($e[($v = substr($v, 1))]) && !in_array('+'. $v, $m)){unset($e[$v]);}
39 $x = !empty($C['deny_attribute']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['deny_attribute']) : '';
40 $x = array_flip((isset($x[0]) && $x[0] == '*') ? explode('-', $x) : explode(',', $x. (!empty($C['safe']) ? ',on*' : '')));
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);
45 $C['deny_attribute'] = $x;
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 $C['schemes'] = array();
49 foreach(explode(';', str_replace(array(' ', "\t", "\r", "\n"), '', $x)) as $v){
50 $x = $x2 = null; list($x, $x2) = explode(':', $v, 2);
51 if($x2){$C['schemes'][$x] = array_flip(explode(',', $x2));}
53 if(!isset($C['schemes']['*'])){$C['schemes']['*'] = array('file'=>1, 'http'=>1, 'https'=>1,);}
54 if(!empty($C['safe']) && empty($C['schemes']['style'])){$C['schemes']['style'] = array('nil'=>1);}
55 $C['abs_url'] = isset($C['abs_url']) ? $C['abs_url'] : 0;
56 if(!isset($C['base_url']) or !preg_match('`^[a-zA-Z\d.+\-]+://[^/]+/(.+?/)?$`', $C['base_url'])){
57 $C['base_url'] = $C['abs_url'] = 0;
60 $C['and_mark'] = empty($C['and_mark']) ? 0 : 1;
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 $C['anti_mail_spam'] = isset($C['anti_mail_spam']) ? $C['anti_mail_spam'] : 0;
63 $C['balance'] = isset($C['balance']) ? (bool)$C['balance'] : 1;
64 $C['cdata'] = isset($C['cdata']) ? $C['cdata'] : (empty($C['safe']) ? 3 : 0);
65 $C['clean_ms_char'] = empty($C['clean_ms_char']) ? 0 : $C['clean_ms_char'];
66 $C['comment'] = isset($C['comment']) ? $C['comment'] : (empty($C['safe']) ? 3 : 0);
67 $C['css_expression'] = empty($C['css_expression']) ? 0 : 1;
68 $C['hexdec_entity'] = isset($C['hexdec_entity']) ? $C['hexdec_entity'] : 1;
69 $C['hook'] = (!empty($C['hook']) && function_exists($C['hook'])) ? $C['hook'] : 0;
70 $C['hook_tag'] = (!empty($C['hook_tag']) && function_exists($C['hook_tag'])) ? $C['hook_tag'] : 0;
71 $C['keep_bad'] = isset($C['keep_bad']) ? $C['keep_bad'] : 6;
72 $C['lc_std_val'] = isset($C['lc_std_val']) ? (bool)$C['lc_std_val'] : 1;
73 $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 1;
74 $C['named_entity'] = isset($C['named_entity']) ? (bool)$C['named_entity'] : 1;
75 $C['no_deprecated_attr'] = isset($C['no_deprecated_attr']) ? $C['no_deprecated_attr'] : 1;
76 $C['parent'] = isset($C['parent'][0]) ? strtolower($C['parent']) : 'body';
77 $C['show_setting'] = !empty($C['show_setting']) ? $C['show_setting'] : 0;
78 $C['style_pass'] = empty($C['style_pass']) ? 0 : 1;
79 $C['tidy'] = empty($C['tidy']) ? 0 : $C['tidy'];
80 $C['unique_ids'] = isset($C['unique_ids']) ? $C['unique_ids'] : 1;
81 $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 0;
83 if(isset($GLOBALS['C'])){$reC = $GLOBALS['C'];}
85 $S = is_array($S) ? $S : hl_spec($S);
86 if(isset($GLOBALS['S'])){$reS = $GLOBALS['S'];}
89 $t = preg_replace('`[\x00-\x08\x0b-\x0c\x0e-\x1f]`', '', $t);
90 if($C['clean_ms_char']){
91 $x = array("\x7f"=>'', "\x80"=>'€', "\x81"=>'', "\x83"=>'ƒ', "\x85"=>'…', "\x86"=>'†', "\x87"=>'‡', "\x88"=>'ˆ', "\x89"=>'‰', "\x8a"=>'Š', "\x8b"=>'‹', "\x8c"=>'Œ', "\x8d"=>'', "\x8e"=>'Ž', "\x8f"=>'', "\x90"=>'', "\x95"=>'•', "\x96"=>'–', "\x97"=>'—', "\x98"=>'˜', "\x99"=>'™', "\x9a"=>'š', "\x9b"=>'›', "\x9c"=>'œ', "\x9d"=>'', "\x9e"=>'ž', "\x9f"=>'Ÿ');
92 $x = $x + ($C['clean_ms_char'] == 1 ? array("\x82"=>'‚', "\x84"=>'„', "\x91"=>'‘', "\x92"=>'’', "\x93"=>'“', "\x94"=>'”') : array("\x82"=>'\'', "\x84"=>'"', "\x91"=>'\'', "\x92"=>'\'', "\x93"=>'"', "\x94"=>'"'));
95 if($C['cdata'] or $C['comment']){$t = preg_replace_callback('`<!(?:(?:--.*?--)|(?:\[CDATA\[.*?\]\]))>`sm', 'hl_cmtcd', $t);}
96 $t = preg_replace_callback('`&([A-Za-z][A-Za-z0-9]{1,30}|#(?:[0-9]{1,8}|[Xx][0-9A-Fa-f]{1,7}));`', 'hl_ent', str_replace('&', '&', $t));
97 if($C['unique_ids'] && !isset($GLOBALS['hl_Ids'])){$GLOBALS['hl_Ids'] = array();}
98 if($C['hook']){$t = $C['hook']($t, $C, $S);}
99 if($C['show_setting'] && preg_match('`^[a-z][a-z0-9_]*$`i', $C['show_setting'])){
100 $GLOBALS[$C['show_setting']] = array('config'=>$C, 'spec'=>$S, 'time'=>microtime());
103 $t = preg_replace_callback('`<(?:(?:\s|$)|(?:[^>]*(?:>|$)))|>`m', 'hl_tag', $t);
104 $t = $C['balance'] ? hl_bal($t, $C['keep_bad'], $C['parent']) : $t;
105 $t = (($C['cdata'] or $C['comment']) && strpos($t, "\x01") !== false) ? str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05"), array('', '', '&', '<', '>'), $t) : $t;
106 $t = $C['tidy'] ? hl_tidy($t, $C['tidy'], $C['parent']) : $t;
108 if(isset($reC)){$GLOBALS['C'] = $reC;}
109 if(isset($reS)){$GLOBALS['S'] = $reS;}
114 function hl_attrval($t, $p){
115 // check attr val against $S
116 $o = 1; $l = strlen($t);
117 foreach($p as $k=>$v){
119 case 'maxlen':if($l > $v){$o = 0;}
120 break; case 'minlen': if($l < $v){$o = 0;}
121 break; case 'maxval': if((float)($t) > $v){$o = 0;}
122 break; case 'minval': if((float)($t) < $v){$o = 0;}
123 break; case 'match': if(!preg_match($v, $t)){$o = 0;}
124 break; case 'nomatch': if(preg_match($v, $t)){$o = 0;}
127 foreach(explode('|', $v) as $n){if($t == $n){$m = 1; break;}}
129 break; case 'noneof':
131 foreach(explode('|', $v) as $n){if($t == $n){$m = 0; break;}}
138 return ($o ? $t : (isset($p['default']) ? $p['default'] : 0));
142 function hl_bal($t, $do=1, $in='div'){
145 $cB = array('blockquote'=>1, 'form'=>1, 'map'=>1, 'noscript'=>1); // Block
146 $cE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty
147 $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
148 $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
149 $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
150 $cN2 = array_keys($cN);
151 $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);
152 $cS = array('colgroup'=>array('col'=>1), 'dir'=>array('li'), '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
153 $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
154 $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
155 // block/inline type; ins & del both type; #pcdata: text
156 $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);
157 $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, 'param'=>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);
158 $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
159 $eO = array('area'=>1, 'caption'=>1, 'col'=>1, 'colgroup'=>1, 'dd'=>1, 'dt'=>1, 'legend'=>1, 'li'=>1, 'optgroup'=>1, 'option'=>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
162 // $in sets allowed child
163 $in = ((isset($eF[$in]) && $in != '#pcdata') or isset($eO[$in])) ? $in : 'div';
165 return (!$do ? '' : str_replace(array('<', '>'), array('<', '>'), $t));
167 if(isset($cS[$in])){$inOk = $cS[$in];}
168 elseif(isset($cI[$in])){$inOk = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
169 elseif(isset($cF[$in])){$inOk = $eF; unset($cI['del'], $cI['ins']);}
170 elseif(isset($cB[$in])){$inOk = $eB; unset($cI['del'], $cI['ins']);}
171 if(isset($cO[$in])){$inOk = $inOk + $cO[$in];}
172 if(isset($cN[$in])){$inOk = array_diff_assoc($inOk, $cN[$in]);}
174 $t = explode('<', $t);
175 $ok = $q = array(); // $q seq list of open non-empty ele
178 for($i=-1, $ci=count($t); ++$i<$ci;){
179 // allowed $ok in parent $p
183 if(isset($cS[$p])){$ok = $cS[$p];}
184 elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
185 elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);}
186 elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);}
187 if(isset($cO[$p])){$ok = $ok + $cO[$p];}
188 if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);}
189 }else{$ok = $inOk; unset($cI['del'], $cI['ins']);}
190 // bad tags, & ele content
191 if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){
192 echo '<', $s, $e, $a, '>';
195 if($do < 3 or isset($ok['#pcdata'])){echo $x;}
196 elseif(strpos($x, "\x02\x04")){
197 foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){
198 echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : ''));
200 }elseif($do > 4){echo preg_replace('`\S`', '', $x);}
203 if(!preg_match('`^(/?)([a-zA-Z1-6]+)([^>]*)>(.*)`sm', $t[$i], $r)){$x = $t[$i]; continue;}
204 $s = null; $e = null; $a = null; $x = null; list($all, $s, $e, $a, $x) = $r;
207 if(isset($cE[$e]) or !in_array($e, $q)){continue;} // Empty/unopen
208 if($p == $e){array_pop($q); echo '</', $e, '>'; unset($e); continue;} // Last open
209 $add = ''; // Nesting - close open tags that need to be
210 for($j=-1, $cj=count($q); ++$j<$cj;){
211 if(($d = array_pop($q)) == $e){break;}
212 else{$add .= "</{$d}>";}
214 echo $add, '</', $e, '>'; unset($e); continue;
217 // $cB ele needs $eB ele as child
218 if(isset($cB[$e]) && strlen(trim($x))){
219 $t[$i] = "{$e}{$a}>";
220 array_splice($t, $i+1, 0, 'div>'. $x); unset($e, $x); ++$ci; --$i; continue;
222 if((($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql)) && !isset($eB[$e]) && !isset($ok[$e])){
223 array_splice($t, $i, 0, 'div>'); unset($e, $x); ++$ci; --$i; continue;
225 // if no open ele, $in = parent; mostly immediate parent-child relation should hold
226 if(!$ql or !isset($eN[$e]) or !array_intersect($q, $cN2)){
228 if($ql && isset($cT[$p])){echo '</', array_pop($q), '>'; unset($e, $x); --$i;}
231 if(!isset($cE[$e])){$q[] = $e;}
232 echo '<', $e, $a, '>'; unset($e); continue;
234 // specific parent-child
235 if(isset($cS[$p][$e])){
236 if(!isset($cE[$e])){$q[] = $e;}
237 echo '<', $e, $a, '>'; unset($e); continue;
242 for($k=-1, $kc=count($q); ++$k<$kc;){
245 if(isset($cS[$d])){$q2[] = $d; continue;}
246 $ok2 = isset($cI[$d]) ? $eI : $eF;
247 if(isset($cO[$d])){$ok2 = $ok2 + $cO[$d];}
248 if(isset($cN[$d])){$ok2 = array_diff_assoc($ok2, $cN[$d]);}
249 if(!isset($ok2[$e])){
250 if(!$k && !isset($inOk[$e])){continue 2;}
252 for(;++$k<$kc;){$add = "</{$q[$k]}>{$add}";}
258 if(!isset($cE[$e])){$q[] = $e;}
259 echo $add, '<', $e, $a, '>'; unset($e); continue;
266 if(isset($cS[$p])){$ok = $cS[$p];}
267 elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;}
268 elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);}
269 elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);}
270 if(isset($cO[$p])){$ok = $ok + $cO[$p];}
271 if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);}
272 }else{$ok = $inOk; unset($cI['del'], $cI['ins']);}
273 if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){
274 echo '<', $s, $e, $a, '>';
277 if(strlen(trim($x)) && (($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql))){
278 echo '<div>', $x, '</div>';
280 elseif($do < 3 or isset($ok['#pcdata'])){echo $x;}
281 elseif(strpos($x, "\x02\x04")){
282 foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){
283 echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : ''));
285 }elseif($do > 4){echo preg_replace('`\S`', '', $x);}
287 while(!empty($q) && ($e = array_pop($q))){echo '</', $e, '>';}
288 $o = ob_get_contents();
294 function hl_cmtcd($t){
295 // comment/CDATA sec handler
299 if(!$C['comment']){return $t;}
300 if($C['comment'] == 1){return '';}
301 if(substr(($t = preg_replace('`--+`', '-', substr($t, 4, -3))), -1) != ' '){$t .= ' ';}
302 $t = $C['comment'] == 2 ? str_replace(array('&', '<', '>'), array('&', '<', '>'), $t) : $t;
303 $t = "\x01\x02\x04!--$t--\x05\x02\x01";
305 if(!$C['cdata']){return $t;}
306 if($C['cdata'] == 1){return '';}
307 $t = substr($t, 1, -1);
308 $t = $C['cdata'] == 2 ? str_replace(array('&', '<', '>'), array('&', '<', '>'), $t) : $t;
309 $t = "\x01\x01\x04$t\x05\x01\x01";
311 return str_replace(array('&', '<', '>'), array("\x03", "\x04", "\x05"), $t);
319 static $U = array('quot'=>1,'amp'=>1,'lt'=>1,'gt'=>1);
320 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');
322 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;'. $t)). ';';
324 if(($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 $n > 1114111))){
325 return ($C['and_mark'] ? "\x06" : '&'). "amp;#{$t};";
327 return ($C['and_mark'] ? "\x06" : '&'). '#'. (((ctype_digit($t) && $C['hexdec_entity'] < 2) or !$C['hexdec_entity']) ? $n : 'x'. dechex($n)). ';';
331 function hl_prot($p, $c=null){
335 if($c == null){$c = 'style'; $b = $p[1]; $a = $p[3]; $p = trim($p[2]);}
336 $c = isset($C['schemes'][$c]) ? $C['schemes'][$c] : $C['schemes']['*'];
337 if(isset($c['*']) or !strcspn($p, '#?;')){return "{$b}{$p}{$a}";} // All ok, frag, query, param
338 if(preg_match('`^([a-z\d\-+.&#; ]+?)(:|&#(58|x3a);|%3a|\\\\0{0,4}3a).`i', $p, $m) && !isset($c[strtolower($m[1])])){ // Denied prot
339 return "{$b}denied:{$p}{$a}";
342 if($C['abs_url'] == -1 && strpos($p, $C['base_url']) === 0){ // Make url rel
343 $p = substr($p, strlen($C['base_url']));
344 }elseif(empty($m[1])){ // Make URL abs
345 if(substr($p, 0, 2) == '//'){$p = substr($C['base_url'], 0, strpos($C['base_url'], ':')+1). $p;}
346 elseif($p[0] == '/'){$p = preg_replace('`(^.+?://[^/]+)(.*)`', '$1', $C['base_url']). $p;}
347 elseif(strcspn($p, './')){$p = $C['base_url']. $p;}
349 preg_match('`^([a-zA-Z\d\-+.]+://[^/]+)(.*)`', $C['base_url'], $m);
350 $p = preg_replace('`(?<=/)\./`', '', $m[2]. $p);
351 while(preg_match('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', $p)){
352 $p = preg_replace('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', '', $p);
358 return "{$b}{$p}{$a}";
362 function hl_regex($p){
364 if(empty($p)){return 0;}
365 if($t = ini_get('track_errors')){$o = isset($php_errormsg) ? $php_errormsg : null;}
366 else{ini_set('track_errors', 1);}
367 unset($php_errormsg);
368 if(($d = ini_get('display_errors'))){ini_set('display_errors', 0);}
370 if($d){ini_set('display_errors', 1);}
371 $r = isset($php_errormsg) ? 0 : 1;
372 if($t){$php_errormsg = isset($o) ? $o : null;}
373 else{ini_set('track_errors', 0);}
378 function hl_spec($t){
381 $t = str_replace(array("\t", "\r", "\n", ' '), '', preg_replace('/"(?>(`.|[^"])*)"/sme', 'substr(str_replace(array(";", "|", "~", " ", ",", "/", "(", ")", \'`"\'), array("\x01", "\x02", "\x03", "\x04", "\x05", "\x06", "\x07", "\x08", "\""), "$0"), 1, -1)', trim($t)));
382 for($i = count(($t = explode(';', $t))); --$i>=0;){
384 if(empty($w) or ($e = strpos($w, '=')) === false or !strlen(($a = substr($w, $e+1)))){continue;}
386 foreach(explode(',', $a) as $v){
387 if(!preg_match('`^([a-z:\-\*]+)(?:\((.*?)\))?`i', $v, $m)){continue;}
388 if(($x = strtolower($m[1])) == '-*'){$n['*'] = 1; continue;}
389 if($x[0] == '-'){$n[substr($x, 1)] = 1; continue;}
390 if(!isset($m[2])){$y[$x] = 1; continue;}
391 foreach(explode('/', $m[2]) as $m){
392 if(empty($m) or ($p = strpos($m, '=')) == 0 or $p < 5){$y[$x] = 1; continue;}
393 $y[$x][strtolower(substr($m, 0, $p))] = str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x06", "\x07", "\x08"), array(";", "|", "~", " ", ",", "/", "(", ")"), substr($m, $p+1));
395 if(isset($y[$x]['match']) && !hl_regex($y[$x]['match'])){unset($y[$x]['match']);}
396 if(isset($y[$x]['nomatch']) && !hl_regex($y[$x]['nomatch'])){unset($y[$x]['nomatch']);}
398 if(!count($y) && !count($n)){continue;}
399 foreach(explode(',', substr($w, 0, $e)) as $v){
400 if(!strlen(($v = strtolower($v)))){continue;}
401 if(count($y)){$s[$v] = $y;}
402 if(count($n)){$s[$v]['n'] = $n;}
410 // tag/attribute handler
414 if($t == '< '){return '< ';}
415 if($t == '>'){return '>';}
416 if(!preg_match('`^<(/?)([a-zA-Z][a-zA-Z1-6]*)([^>]*?)\s?>$`m', $t, $m)){
417 return str_replace(array('<', '>'), array('<', '>'), $t);
418 }elseif(!isset($C['elements'][($e = strtolower($m[2]))])){
419 return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('<', '>'), $t) : '');
422 $a = str_replace(array("\xad", "\n", "\r", "\t"), ' ', trim($m[3]));
423 if(strpos($a, '&') !== false){
424 str_replace(array('­', '­', '­'), ' ', $a);
427 static $eD = array('applet'=>1, 'center'=>1, 'dir'=>1, 'embed'=>1, 'font'=>1, 'isindex'=>1, 'menu'=>1, 's'=>1, 'strike'=>1, 'u'=>1); // Deprecated
428 if($C['make_tag_strict'] && isset($eD[$e])){
429 $trt = hl_tag2($e, $a, $C['make_tag_strict']);
430 if(!$e){return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('<', '>'), $t) : '');}
433 static $eE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty ele
435 return (!isset($eE[$e]) ? "</$e>" : (($C['keep_bad'])%2 ? str_replace(array('<', '>'), array('<', '>'), $t) : ''));
439 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), '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
440 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
441 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
442 static $aNU = array('class'=>array('param'=>1, 'script'=>1), 'dir'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'id'=>array('script'=>1), 'lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'xml:lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'onclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'ondblclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeydown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeypress'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeyup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousedown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousemove'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseout'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseover'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'style'=>array('param'=>1, 'script'=>1), 'title'=>array('param'=>1, 'script'=>1)); // Univ & exceptions
444 if($C['lc_std_val']){
445 // predef attr vals for $eAL & $aNE ele
446 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);
447 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);
448 $lcase = isset($eAL[$e]) ? 1 : 0;
452 if($C['no_deprecated_attr']){
453 // dep attr:applicable ele
454 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));
455 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);
456 $depTr = isset($eAD[$e]) ? 1 : 0;
460 if(strpos($a, "\x01") !== false){$a = preg_replace('`\x01[^\x01]*\x01`', '', $a);} // No comment/CDATA sec
461 $mode = 0; $a = trim($a, ' /'); $aA = array();
466 if(preg_match('`^[a-zA-Z][\-a-zA-Z:]+`', $a, $m)){
467 $nm = strtolower($m[0]);
468 $w = $mode = 1; $a = ltrim(substr_replace($a, '', 0, strlen($m[0])));
471 if($a[0] == '='){ // =
472 $w = 1; $mode = 2; $a = ltrim($a, '= ');
474 $w = 1; $mode = 0; $a = ltrim($a);
477 break; case 2: // Val
478 if(preg_match('`^"[^"]*"`', $a, $m) or preg_match("`^'[^']*'`", $a, $m) or preg_match("`^\s*[^\s\"']+`", $a, $m)){
479 $m = $m[0]; $w = 1; $mode = 0; $a = ltrim(substr_replace($a, '', 0, strlen($m)));
480 $aA[$nm] = trim(($m[0] == '"' or $m[0] == '\'') ? substr($m, 1, -1) : $m);
484 if($w == 0){ // Parse errs, deal with space, " & '
485 $a = preg_replace('`^(?:"[^"]*("|$)|\'[^\']*(\'|$)|\S)*\s*`', '', $a);
489 if($mode == 1){$aA[$nm] = '';}
493 $rl = isset($S[$e]) ? $S[$e] : array();
494 $a = array(); $nfr = 0;
495 foreach($aA as $k=>$v){
496 if(((isset($C['deny_attribute']['*']) ? isset($C['deny_attribute'][$k]) : !isset($C['deny_attribute'][$k])) or isset($rl[$k])) && ((!isset($rl['n'][$k]) && !isset($rl['n']['*'])) or isset($rl[$k])) && (isset($aN[$k][$e]) or (isset($aNU[$k]) && !isset($aNU[$k][$e])))){
497 if(isset($aNE[$k])){$v = $k;}
498 elseif(!empty($lcase) && (($e != 'button' or $e != 'input') or $k == 'type')){ // Rather loose but ?not cause issues
499 $v = (isset($aNL[($v2 = strtolower($v))])) ? $v2 : $v;
501 if($k == 'style' && !$C['style_pass']){
502 if(false !== strpos($v, '&#')){
503 static $sC = array(' '=>' ', ' '=>' ', 'E'=>'e', 'E'=>'e', 'e'=>'e', 'e'=>'e', 'X'=>'x', 'X'=>'x', 'x'=>'x', 'x'=>'x', 'P'=>'p', 'P'=>'p', 'p'=>'p', 'p'=>'p', 'S'=>'s', 'S'=>'s', 's'=>'s', 's'=>'s', 'I'=>'i', 'I'=>'i', 'i'=>'i', 'i'=>'i', 'O'=>'o', 'O'=>'o', 'o'=>'o', 'o'=>'o', 'N'=>'n', 'N'=>'n', 'n'=>'n', 'n'=>'n', 'U'=>'u', 'U'=>'u', 'u'=>'u', 'u'=>'u', 'R'=>'r', 'R'=>'r', 'r'=>'r', 'r'=>'r', 'L'=>'l', 'L'=>'l', 'l'=>'l', 'l'=>'l', '('=>'(', '('=>'(', ')'=>')', ')'=>')', ' '=>':', ' '=>':', '"'=>'"', '"'=>'"', '''=>"'", '''=>"'", '/'=>'/', '/'=>'/', '*'=>'*', '*'=>'*', '\'=>'\\', '\'=>'\\');
506 $v = preg_replace_callback('`(url(?:\()(?: )*(?:\'|"|&(?:quot|apos);)?)(.+)((?:\'|"|&(?:quot|apos);)?(?: )*(?:\)))`iS', 'hl_prot', $v);
507 $v = !$C['css_expression'] ? preg_replace('`expression`i', ' ', preg_replace('`\\\\\S|(/|(%2f))(\*|(%2a))`i', ' ', $v)) : $v;
508 }elseif(isset($aNP[$k]) or strpos($k, 'src') !== false or $k[0] == 'o'){
509 $v = hl_prot($v, $k);
510 if($k == 'href'){ // X-spam
511 if($C['anti_mail_spam'] && strpos($v, 'mailto:') === 0){
512 $v = str_replace('@', htmlspecialchars($C['anti_mail_spam']), $v);
513 }elseif($C['anti_link_spam']){
514 $r1 = $C['anti_link_spam'][1];
515 if(!empty($r1) && preg_match($r1, $v)){continue;}
516 $r0 = $C['anti_link_spam'][0];
517 if(!empty($r0) && preg_match($r0, $v)){
518 if(isset($a['rel'])){
519 if(!preg_match('`\bnofollow\b`i', $a['rel'])){$a['rel'] .= ' nofollow';}
520 }elseif(isset($aA['rel'])){
521 if(!preg_match('`\bnofollow\b`i', $aA['rel'])){$nfr = 1;}
522 }else{$a['rel'] = 'nofollow';}
527 if(isset($rl[$k]) && is_array($rl[$k]) && ($v = hl_attrval($v, $rl[$k])) === 0){continue;}
528 $a[$k] = str_replace('"', '"', $v);
531 if($nfr){$a['rel'] = isset($a['rel']) ? $a['rel']. ' nofollow' : 'nofollow';}
534 static $eAR = array('area'=>array('alt'=>'area'), 'bdo'=>array('dir'=>'ltr'), 'form'=>array('action'=>''), 'img'=>array('src'=>'', 'alt'=>'image'), 'map'=>array('name'=>''), 'optgroup'=>array('label'=>''), 'param'=>array('name'=>''), 'script'=>array('type'=>'text/javascript'), 'textarea'=>array('rows'=>'10', 'cols'=>'50'));
536 foreach($eAR[$e] as $k=>$v){
537 if(!isset($a[$k])){$a[$k] = isset($v[0]) ? $v : $k;}
544 foreach($a as $k=>$v){
545 if($k == 'style' or !isset($aND[$k][$e])){continue;}
548 if($e == 'img' && ($v == 'left' or $v == 'right')){$c[] = 'float: '. $v;}
549 elseif(($e == 'div' or $e == 'table') && $v == 'center'){$c[] = 'margin: auto';}
550 else{$c[] = 'text-align: '. $v;}
551 }elseif($k == 'bgcolor'){
552 unset($a['bgcolor']);
553 $c[] = 'background-color: '. $v;
554 }elseif($k == 'border'){
555 unset($a['border']); $c[] = "border: {$v}px";
556 }elseif($k == 'bordercolor'){
557 unset($a['bordercolor']); $c[] = 'border-color: '. $v;
558 }elseif($k == 'clear'){
559 unset($a['clear']); $c[] = 'clear: '. ($v != 'all' ? $v : 'both');
560 }elseif($k == 'compact'){
561 unset($a['compact']); $c[] = 'font-size: 85%';
562 }elseif($k == 'height' or $k == 'width'){
563 unset($a[$k]); $c[] = $k. ': '. ($v[0] != '*' ? $v. (ctype_digit($v) ? 'px' : '') : 'auto');
564 }elseif($k == 'hspace'){
565 unset($a['hspace']); $c[] = "margin-left: {$v}px; margin-right: {$v}px";
566 }elseif($k == 'language' && !isset($a['type'])){
567 unset($a['language']);
568 $a['type'] = 'text/'. strtolower($v);
569 }elseif($k == 'name'){
570 if($C['no_deprecated_attr'] == 2 or ($e != 'a' && $e != 'map')){unset($a['name']);}
571 if(!isset($a['id']) && preg_match('`[a-zA-Z][a-zA-Z\d.:_\-]*`', $v)){$a['id'] = $v;}
572 }elseif($k == 'noshade'){
573 unset($a['noshade']); $c[] = 'border-style: none; border: 0; background-color: gray; color: gray';
574 }elseif($k == 'nowrap'){
575 unset($a['nowrap']); $c[] = 'white-space: nowrap';
576 }elseif($k == 'size'){
577 unset($a['size']); $c[] = 'size: '. $v. 'px';
578 }elseif($k == 'start' or $k == 'value'){
580 }elseif($k == 'type'){
582 static $ol_type = array('i'=>'lower-roman', 'I'=>'upper-roman', 'a'=>'lower-latin', 'A'=>'upper-latin', '1'=>'decimal');
583 $c[] = 'list-style-type: '. (isset($ol_type[$v]) ? $ol_type[$v] : 'decimal');
584 }elseif($k == 'vspace'){
585 unset($a['vspace']); $c[] = "margin-top: {$v}px; margin-bottom: {$v}px";
589 $c = implode('; ', $c);
590 $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $c. ';': $c. ';';
594 if($C['unique_ids'] && isset($a['id'])){
595 if(!preg_match('`^[A-Za-z][A-Za-z0-9_\-.:]*$`', ($id = $a['id'])) or (isset($GLOBALS['hl_Ids'][$id]) && $C['unique_ids'] == 1)){unset($a['id']);
597 while(isset($GLOBALS['hl_Ids'][$id])){$id = $C['unique_ids']. $id;}
598 $GLOBALS['hl_Ids'][($a['id'] = $id)] = 1;
602 if($C['xml:lang'] && isset($a['lang'])){
603 $a['xml:lang'] = isset($a['xml:lang']) ? $a['xml:lang'] : $a['lang'];
604 if($C['xml:lang'] == 2){unset($a['lang']);}
606 // for transformed tag
608 $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $trt : $trt;
610 // return with empty ele /
611 if(empty($C['hook_tag'])){
613 foreach($a as $k=>$v){$aA .= " {$k}=\"{$v}\"";}
614 return "<{$e}{$aA}". (isset($eE[$e]) ? ' /' : ''). '>';
616 else{return $C['hook_tag']($e, $a);}
620 function hl_tag2(&$e, &$a, $t=1){
622 if($e == 'center'){$e = 'div'; return 'text-align: center;';}
623 if($e == 'dir' or $e == 'menu'){$e = 'ul'; return '';}
624 if($e == 's' or $e == 'strike'){$e = 'span'; return 'text-decoration: line-through;';}
625 if($e == 'u'){$e = 'span'; return 'text-decoration: underline;';}
626 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%');
629 if(preg_match('`face\s*=\s*(\'|")([^=]+?)\\1`i', $a, $m) or preg_match('`face\s*=\s*([^"])(\S+)`i', $a, $m)){
630 $a2 .= ' font-family: '. str_replace('"', '\'', trim($m[2])). ';';
632 if(preg_match('`color\s*=\s*(\'|")?(.+?)(\\1|\s|$)`i', $a, $m)){
633 $a2 .= ' color: '. trim($m[2]). ';';
635 if(preg_match('`size\s*=\s*(\'|")?(.+?)(\\1|\s|$)`i', $a, $m) && isset($fs[($m = trim($m[2]))])){
636 $a2 .= ' font-size: '. $fs[$m]. ';';
638 $e = 'span'; return ltrim($a2);
640 if($t == 2){$e = 0; return 0;}
645 function hl_tidy($t, $w, $p){
647 if(strpos(' pre,script,textarea', "$p,")){return $t;}
648 $t = str_replace(' </', '</', preg_replace(array('`(<\w[^>]*(?<!/)>)\s+`', '`\s+`', '`(<\w[^>]*(?<!/)>) `'), array(' $1', ' ', '$1'), 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];'), $t)));
649 if(($w = strtolower($w)) == -1){
650 return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t);
652 $s = strpos(" $w", 't') ? "\t" : ' ';
653 $s = preg_match('`\d`', $w, $m) ? str_repeat($s, $m[0]) : str_repeat($s, ($s == "\t" ? 1 : 2));
654 $n = preg_match('`[ts]([1-9])`', $w, $m) ? $m[1] : 0;
656 $b = array('button'=>1, 'input'=>1, 'option'=>1);
657 $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);
658 $d = array('address'=>1, 'blockquote'=>1, 'center'=>1, 'colgroup'=>1, 'dir'=>1, 'div'=>1, 'dl'=>1, 'fieldset'=>1, 'form'=>1, 'hr'=>1, 'iframe'=>1, 'map'=>1, 'menu'=>1, 'noscript'=>1, 'ol'=>1, 'optgroup'=>1, 'rbc'=>1, 'rtc'=>1, 'ruby'=>1, 'script'=>1, 'select'=>1, 'table'=>1, 'tfoot'=>1, 'thead'=>1, 'tr'=>1, 'ul'=>1);
660 if(isset($d[$p])){echo str_repeat($s, ++$n);}
661 $t = explode('<', $t);
662 echo ltrim(array_shift($t));
663 for($i=-1, $j=count($t); ++$i<$j;){
664 $r = ''; list($e, $r) = explode('>', $t[$i]);
665 $x = $e[0] == '/' ? 0 : (substr($e, -1) == '/' ? 1 : ($e[0] != '!' ? 2 : -1));
666 $y = !$x ? ltrim($e, '/') : ($x > 0 ? substr($e, 0, strcspn($e, ' ')) : 0);
669 if(!$x){echo "\n", str_repeat($s, --$n), "$e\n", str_repeat($s, $n);}
670 else{echo "\n", str_repeat($s, $n), "$e\n", str_repeat($s, ($x != 1 ? ++$n : $n));}
671 echo ltrim($r); continue;
673 $f = "\n". str_repeat($s, $n);
675 if(!$x){echo $e, $f, ltrim($r);}
676 else{echo $f, $e, $r;}
677 }elseif(isset($b[$y])){echo $f, $e, $r;
678 }elseif(isset($a[$y])){echo $e, $f, ltrim($r);
679 }elseif(!$y){echo $f, $e, $f, ltrim($r);
682 $t = preg_replace('`[\n]\s*?[\n]+`', "\n", ob_get_contents());
684 if(($l = strpos(" $w", 'r') ? (strpos(" $w", 'n') ? "\r\n" : "\r") : 0)){
685 $t = str_replace("\n", $l, $t);
687 return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t);
691 function hl_version(){
697 function kses($t, $h, $p=array('http', 'https', 'ftp', 'news', 'nntp', 'telnet', 'gopher', 'mailto')){
699 foreach($h as $k=>$v){
700 $h[$k]['n']['*'] = 1;
702 $C['cdata'] = $C['comment'] = $C['make_tag_strict'] = $C['no_deprecated_attr'] = $C['unique_ids'] = 0;
704 $C['elements'] = count($h) ? strtolower(implode(',', array_keys($h))) : '-*';
705 $C['hook'] = 'kses_hook';
706 $C['schemes'] = '*:'. implode(',', $p);
707 return htmLawed($t, $C, $h);
711 function kses_hook($t, &$C, &$S){