diff options
author | Nicolas LÅ“uillet <nicolas@loeuillet.org> | 2015-01-19 16:53:59 +0100 |
---|---|---|
committer | Nicolas LÅ“uillet <nicolas@loeuillet.org> | 2015-01-19 16:53:59 +0100 |
commit | 9e7f6caf03b90076ba7b448aa7f11d40f584045f (patch) | |
tree | 074658bb13da08de1fd747f01999c5baf0b61080 | |
parent | 820d81aa614f0a6c7e52b049c39ef167c9debd51 (diff) | |
download | wallabag-9e7f6caf03b90076ba7b448aa7f11d40f584045f.tar.gz wallabag-9e7f6caf03b90076ba7b448aa7f11d40f584045f.tar.zst wallabag-9e7f6caf03b90076ba7b448aa7f11d40f584045f.zip |
htmlawed via composer
l--------- | bin/twig-gettext-extractor | 1 | ||||
-rw-r--r-- | composer.json | 1 | ||||
-rw-r--r-- | composer.lock | 42 | ||||
-rw-r--r-- | inc/3rdparty/libraries/htmLawed/htmLawed.php | 728 |
4 files changed, 43 insertions, 729 deletions
diff --git a/bin/twig-gettext-extractor b/bin/twig-gettext-extractor new file mode 120000 index 00000000..0a8959b0 --- /dev/null +++ b/bin/twig-gettext-extractor | |||
@@ -0,0 +1 @@ | |||
../vendor/umpirsky/twig-gettext-extractor/twig-gettext-extractor \ No newline at end of file | |||
diff --git a/composer.json b/composer.json index b29082df..47458df4 100644 --- a/composer.json +++ b/composer.json | |||
@@ -58,6 +58,7 @@ | |||
58 | "robmorgan/phinx": "*", | 58 | "robmorgan/phinx": "*", |
59 | "tecnick.com/tcpdf": "dev-master", | 59 | "tecnick.com/tcpdf": "dev-master", |
60 | "simplepie/simplepie": "dev-master", | 60 | "simplepie/simplepie": "dev-master", |
61 | "htmlawed/htmlawed": "dev-master", | ||
61 | "wallabag/PHP-Flash-Messages": "dev-master", | 62 | "wallabag/PHP-Flash-Messages": "dev-master", |
62 | "wallabag/kriss_php5": "dev-master", | 63 | "wallabag/kriss_php5": "dev-master", |
63 | "wallabag/pagination": "dev-master", | 64 | "wallabag/pagination": "dev-master", |
diff --git a/composer.lock b/composer.lock index d5dce6ab..8d538cd2 100644 --- a/composer.lock +++ b/composer.lock | |||
@@ -4,7 +4,7 @@ | |||
4 | "Read more about it at http://getcomposer.org/doc/01-basic-usage.md#composer-lock-the-lock-file", | 4 | "Read more about it at http://getcomposer.org/doc/01-basic-usage.md#composer-lock-the-lock-file", |
5 | "This file is @generated automatically" | 5 | "This file is @generated automatically" |
6 | ], | 6 | ], |
7 | "hash": "bc1406a8a4ef54f0927a48c0b7f5eeb1", | 7 | "hash": "4caad70325b5cf76eb8d4011bae06cbb", |
8 | "packages": [ | 8 | "packages": [ |
9 | { | 9 | { |
10 | "name": "ezyang/htmlpurifier", | 10 | "name": "ezyang/htmlpurifier", |
@@ -258,6 +258,45 @@ | |||
258 | "time": "2014-10-12 19:18:40" | 258 | "time": "2014-10-12 19:18:40" |
259 | }, | 259 | }, |
260 | { | 260 | { |
261 | "name": "htmlawed/htmlawed", | ||
262 | "version": "dev-master", | ||
263 | "source": { | ||
264 | "type": "git", | ||
265 | "url": "https://github.com/kesar/HTMLawed.git", | ||
266 | "reference": "9d292af5f4c288aa68f38b87f5d88c8214f5f233" | ||
267 | }, | ||
268 | "dist": { | ||
269 | "type": "zip", | ||
270 | "url": "https://api.github.com/repos/kesar/HTMLawed/zipball/9d292af5f4c288aa68f38b87f5d88c8214f5f233", | ||
271 | "reference": "9d292af5f4c288aa68f38b87f5d88c8214f5f233", | ||
272 | "shasum": "" | ||
273 | }, | ||
274 | "require": { | ||
275 | "php": ">4.4.0" | ||
276 | }, | ||
277 | "type": "library", | ||
278 | "autoload": { | ||
279 | "files": [ | ||
280 | "htmLawed.php" | ||
281 | ] | ||
282 | }, | ||
283 | "notification-url": "https://packagist.org/downloads/", | ||
284 | "license": [ | ||
285 | "GPL-2.0+", | ||
286 | "LGPL-3.0" | ||
287 | ], | ||
288 | "description": "htmLawed - Process text with HTML markup to make it more compliant with HTML standards and administrative policies", | ||
289 | "homepage": "http://www.bioinformatics.org/phplabware/internal_utilities/htmLawed/", | ||
290 | "keywords": [ | ||
291 | "HTMLtidy", | ||
292 | "html", | ||
293 | "sanitize", | ||
294 | "strip", | ||
295 | "tags" | ||
296 | ], | ||
297 | "time": "2014-01-05 13:06:52" | ||
298 | }, | ||
299 | { | ||
261 | "name": "mgargano/simplehtmldom", | 300 | "name": "mgargano/simplehtmldom", |
262 | "version": "dev-master", | 301 | "version": "dev-master", |
263 | "source": { | 302 | "source": { |
@@ -2345,6 +2384,7 @@ | |||
2345 | "mgargano/simplehtmldom": 20, | 2384 | "mgargano/simplehtmldom": 20, |
2346 | "tecnick.com/tcpdf": 20, | 2385 | "tecnick.com/tcpdf": 20, |
2347 | "simplepie/simplepie": 20, | 2386 | "simplepie/simplepie": 20, |
2387 | "htmlawed/htmlawed": 20, | ||
2348 | "wallabag/php-flash-messages": 20, | 2388 | "wallabag/php-flash-messages": 20, |
2349 | "wallabag/kriss_php5": 20, | 2389 | "wallabag/kriss_php5": 20, |
2350 | "wallabag/pagination": 20, | 2390 | "wallabag/pagination": 20, |
diff --git a/inc/3rdparty/libraries/htmLawed/htmLawed.php b/inc/3rdparty/libraries/htmLawed/htmLawed.php deleted file mode 100644 index 9a62aca5..00000000 --- a/inc/3rdparty/libraries/htmLawed/htmLawed.php +++ /dev/null | |||
@@ -1,728 +0,0 @@ | |||
1 | <?php | ||
2 | |||
3 | /* | ||
4 | htmLawed 1.1.14, 8 August 2012 | ||
5 | OOP code, 8 August 2012 | ||
6 | Copyright Santosh Patnaik | ||
7 | Dual LGPL v3 and GPL v2+ license | ||
8 | A PHP Labware internal utility; www.bioinformatics.org/phplabware/internal_utilities/htmLawed | ||
9 | |||
10 | See htmLawed_README.txt/htm | ||
11 | */ | ||
12 | |||
13 | class htmLawed{ | ||
14 | // begin class | ||
15 | public static function hl($t, $C=1, $S=array()){ | ||
16 | $C = is_array($C) ? $C : array(); | ||
17 | if(!empty($C['valid_xhtml'])){ | ||
18 | $C['elements'] = empty($C['elements']) ? '*-center-dir-font-isindex-menu-s-strike-u' : $C['elements']; | ||
19 | $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 2; | ||
20 | $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 2; | ||
21 | } | ||
22 | // config eles | ||
23 | $e = array('a'=>1, 'abbr'=>1, 'acronym'=>1, 'address'=>1, 'applet'=>1, 'area'=>1, 'b'=>1, 'bdo'=>1, 'big'=>1, 'blockquote'=>1, 'br'=>1, 'button'=>1, 'caption'=>1, 'center'=>1, 'cite'=>1, 'code'=>1, 'col'=>1, 'colgroup'=>1, 'dd'=>1, 'del'=>1, 'dfn'=>1, 'dir'=>1, 'div'=>1, 'dl'=>1, 'dt'=>1, 'em'=>1, 'embed'=>1, 'fieldset'=>1, 'font'=>1, 'form'=>1, 'h1'=>1, 'h2'=>1, 'h3'=>1, 'h4'=>1, 'h5'=>1, 'h6'=>1, 'hr'=>1, 'i'=>1, 'iframe'=>1, 'img'=>1, 'input'=>1, 'ins'=>1, 'isindex'=>1, 'kbd'=>1, 'label'=>1, 'legend'=>1, 'li'=>1, 'map'=>1, 'menu'=>1, 'noscript'=>1, 'object'=>1, 'ol'=>1, 'optgroup'=>1, 'option'=>1, 'p'=>1, 'param'=>1, 'pre'=>1, 'q'=>1, 'rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1, 'ruby'=>1, 's'=>1, 'samp'=>1, 'script'=>1, 'select'=>1, 'small'=>1, 'span'=>1, 'strike'=>1, 'strong'=>1, 'sub'=>1, 'sup'=>1, 'table'=>1, 'tbody'=>1, 'td'=>1, 'textarea'=>1, 'tfoot'=>1, 'th'=>1, 'thead'=>1, 'tr'=>1, 'tt'=>1, 'u'=>1, 'ul'=>1, 'var'=>1); // 86/deprecated+embed+ruby | ||
24 | if(!empty($C['safe'])){ | ||
25 | unset($e['applet'], $e['embed'], $e['iframe'], $e['object'], $e['script']); | ||
26 | } | ||
27 | $x = !empty($C['elements']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['elements']) : '*'; | ||
28 | if($x == '-*'){$e = array();} | ||
29 | elseif(strpos($x, '*') === false){$e = array_flip(explode(',', $x));} | ||
30 | else{ | ||
31 | if(isset($x[1])){ | ||
32 | preg_match_all('`(?:^|-|\+)[^\-+]+?(?=-|\+|$)`', $x, $m, PREG_SET_ORDER); | ||
33 | for($i=count($m); --$i>=0;){$m[$i] = $m[$i][0];} | ||
34 | foreach($m as $v){ | ||
35 | if($v[0] == '+'){$e[substr($v, 1)] = 1;} | ||
36 | if($v[0] == '-' && isset($e[($v = substr($v, 1))]) && !in_array('+'. $v, $m)){unset($e[$v]);} | ||
37 | } | ||
38 | } | ||
39 | } | ||
40 | $C['elements'] =& $e; | ||
41 | // config attrs | ||
42 | $x = !empty($C['deny_attribute']) ? str_replace(array("\n", "\r", "\t", ' '), '', $C['deny_attribute']) : ''; | ||
43 | $x = array_flip((isset($x[0]) && $x[0] == '*') ? explode('-', $x) : explode(',', $x. (!empty($C['safe']) ? ',on*' : ''))); | ||
44 | if(isset($x['on*'])){ | ||
45 | unset($x['on*']); | ||
46 | $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); | ||
47 | } | ||
48 | $C['deny_attribute'] = $x; | ||
49 | // config URL | ||
50 | $x = (isset($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'; | ||
51 | $C['schemes'] = array(); | ||
52 | foreach(explode(';', str_replace(array(' ', "\t", "\r", "\n"), '', $x)) as $v){ | ||
53 | $x = $x2 = null; list($x, $x2) = explode(':', $v, 2); | ||
54 | if($x2){$C['schemes'][$x] = array_flip(explode(',', $x2));} | ||
55 | } | ||
56 | if(!isset($C['schemes']['*'])){$C['schemes']['*'] = array('file'=>1, 'http'=>1, 'https'=>1,);} | ||
57 | if(!empty($C['safe']) && empty($C['schemes']['style'])){$C['schemes']['style'] = array('!'=>1);} | ||
58 | $C['abs_url'] = isset($C['abs_url']) ? $C['abs_url'] : 0; | ||
59 | if(!isset($C['base_url']) or !preg_match('`^[a-zA-Z\d.+\-]+://[^/]+/(.+?/)?$`', $C['base_url'])){ | ||
60 | $C['base_url'] = $C['abs_url'] = 0; | ||
61 | } | ||
62 | // config rest | ||
63 | $C['and_mark'] = empty($C['and_mark']) ? 0 : 1; | ||
64 | $C['anti_link_spam'] = (isset($C['anti_link_spam']) && is_array($C['anti_link_spam']) && count($C['anti_link_spam']) == 2 && (empty($C['anti_link_spam'][0]) or htmLawed::hl_regex($C['anti_link_spam'][0])) && (empty($C['anti_link_spam'][1]) or htmLawed::hl_regex($C['anti_link_spam'][1]))) ? $C['anti_link_spam'] : 0; | ||
65 | $C['anti_mail_spam'] = isset($C['anti_mail_spam']) ? $C['anti_mail_spam'] : 0; | ||
66 | $C['balance'] = isset($C['balance']) ? (bool)$C['balance'] : 1; | ||
67 | $C['cdata'] = isset($C['cdata']) ? $C['cdata'] : (empty($C['safe']) ? 3 : 0); | ||
68 | $C['clean_ms_char'] = empty($C['clean_ms_char']) ? 0 : $C['clean_ms_char']; | ||
69 | $C['comment'] = isset($C['comment']) ? $C['comment'] : (empty($C['safe']) ? 3 : 0); | ||
70 | $C['css_expression'] = empty($C['css_expression']) ? 0 : 1; | ||
71 | $C['direct_list_nest'] = empty($C['direct_list_nest']) ? 0 : 1; | ||
72 | $C['hexdec_entity'] = isset($C['hexdec_entity']) ? $C['hexdec_entity'] : 1; | ||
73 | $C['hook'] = (!empty($C['hook']) && function_exists($C['hook'])) ? $C['hook'] : 0; | ||
74 | $C['hook_tag'] = (!empty($C['hook_tag']) && function_exists($C['hook_tag'])) ? $C['hook_tag'] : 0; | ||
75 | $C['keep_bad'] = isset($C['keep_bad']) ? $C['keep_bad'] : 6; | ||
76 | $C['lc_std_val'] = isset($C['lc_std_val']) ? (bool)$C['lc_std_val'] : 1; | ||
77 | $C['make_tag_strict'] = isset($C['make_tag_strict']) ? $C['make_tag_strict'] : 1; | ||
78 | $C['named_entity'] = isset($C['named_entity']) ? (bool)$C['named_entity'] : 1; | ||
79 | $C['no_deprecated_attr'] = isset($C['no_deprecated_attr']) ? $C['no_deprecated_attr'] : 1; | ||
80 | $C['parent'] = isset($C['parent'][0]) ? strtolower($C['parent']) : 'body'; | ||
81 | $C['show_setting'] = !empty($C['show_setting']) ? $C['show_setting'] : 0; | ||
82 | $C['style_pass'] = empty($C['style_pass']) ? 0 : 1; | ||
83 | $C['tidy'] = empty($C['tidy']) ? 0 : $C['tidy']; | ||
84 | $C['unique_ids'] = isset($C['unique_ids']) ? $C['unique_ids'] : 1; | ||
85 | $C['xml:lang'] = isset($C['xml:lang']) ? $C['xml:lang'] : 0; | ||
86 | |||
87 | if(isset($GLOBALS['C'])){$reC = $GLOBALS['C'];} | ||
88 | $GLOBALS['C'] = $C; | ||
89 | $S = is_array($S) ? $S : htmLawed::hl_spec($S); | ||
90 | if(isset($GLOBALS['S'])){$reS = $GLOBALS['S'];} | ||
91 | $GLOBALS['S'] = $S; | ||
92 | |||
93 | $t = preg_replace('`[\x00-\x08\x0b-\x0c\x0e-\x1f]`', '', $t); | ||
94 | if($C['clean_ms_char']){ | ||
95 | $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'); | ||
96 | $x = $x + ($C['clean_ms_char'] == 1 ? array("\x82"=>'‚', "\x84"=>'„', "\x91"=>'‘', "\x92"=>'’', "\x93"=>'“', "\x94"=>'”') : array("\x82"=>'\'', "\x84"=>'"', "\x91"=>'\'', "\x92"=>'\'', "\x93"=>'"', "\x94"=>'"')); | ||
97 | $t = strtr($t, $x); | ||
98 | } | ||
99 | if($C['cdata'] or $C['comment']){$t = preg_replace_callback('`<!(?:(?:--.*?--)|(?:\[CDATA\[.*?\]\]))>`sm', 'htmLawed::hl_cmtcd', $t);} | ||
100 | $t = preg_replace_callback('`&([A-Za-z][A-Za-z0-9]{1,30}|#(?:[0-9]{1,8}|[Xx][0-9A-Fa-f]{1,7}));`', 'htmLawed::hl_ent', str_replace('&', '&', $t)); | ||
101 | if($C['unique_ids'] && !isset($GLOBALS['hl_Ids'])){$GLOBALS['hl_Ids'] = array();} | ||
102 | if($C['hook']){$t = $C['hook']($t, $C, $S);} | ||
103 | if($C['show_setting'] && preg_match('`^[a-z][a-z0-9_]*$`i', $C['show_setting'])){ | ||
104 | $GLOBALS[$C['show_setting']] = array('config'=>$C, 'spec'=>$S, 'time'=>microtime()); | ||
105 | } | ||
106 | // main | ||
107 | $t = preg_replace_callback('`<(?:(?:\s|$)|(?:[^>]*(?:>|$)))|>`m', 'htmLawed::hl_tag', $t); | ||
108 | $t = $C['balance'] ? htmLawed::hl_bal($t, $C['keep_bad'], $C['parent']) : $t; | ||
109 | $t = (($C['cdata'] or $C['comment']) && strpos($t, "\x01") !== false) ? str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05"), array('', '', '&', '<', '>'), $t) : $t; | ||
110 | $t = $C['tidy'] ? htmLawed::hl_tidy($t, $C['tidy'], $C['parent']) : $t; | ||
111 | unset($C, $e); | ||
112 | if(isset($reC)){$GLOBALS['C'] = $reC;} | ||
113 | if(isset($reS)){$GLOBALS['S'] = $reS;} | ||
114 | return $t; | ||
115 | // eof | ||
116 | } | ||
117 | |||
118 | public static function hl_attrval($t, $p){ | ||
119 | // check attr val against $S | ||
120 | $o = 1; $l = strlen($t); | ||
121 | foreach($p as $k=>$v){ | ||
122 | switch($k){ | ||
123 | case 'maxlen':if($l > $v){$o = 0;} | ||
124 | break; case 'minlen': if($l < $v){$o = 0;} | ||
125 | break; case 'maxval': if((float)($t) > $v){$o = 0;} | ||
126 | break; case 'minval': if((float)($t) < $v){$o = 0;} | ||
127 | break; case 'match': if(!preg_match($v, $t)){$o = 0;} | ||
128 | break; case 'nomatch': if(preg_match($v, $t)){$o = 0;} | ||
129 | break; case 'oneof': | ||
130 | $m = 0; | ||
131 | foreach(explode('|', $v) as $n){if($t == $n){$m = 1; break;}} | ||
132 | $o = $m; | ||
133 | break; case 'noneof': | ||
134 | $m = 1; | ||
135 | foreach(explode('|', $v) as $n){if($t == $n){$m = 0; break;}} | ||
136 | $o = $m; | ||
137 | break; default: | ||
138 | break; | ||
139 | } | ||
140 | if(!$o){break;} | ||
141 | } | ||
142 | return ($o ? $t : (isset($p['default']) ? $p['default'] : 0)); | ||
143 | // eof | ||
144 | } | ||
145 | |||
146 | public static function hl_bal($t, $do=1, $in='div'){ | ||
147 | // balance tags | ||
148 | // by content | ||
149 | $cB = array('blockquote'=>1, 'form'=>1, 'map'=>1, 'noscript'=>1); // Block | ||
150 | $cE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty | ||
151 | $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 | ||
152 | $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 | ||
153 | $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 | ||
154 | $cN2 = array_keys($cN); | ||
155 | $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); | ||
156 | $cS = array('colgroup'=>array('col'=>1), 'dir'=>array('li'=>1), 'dl'=>array('dd'=>1, 'dt'=>1), 'menu'=>array('li'=>1), 'ol'=>array('li'=>1), 'optgroup'=>array('option'=>1), 'option'=>array('#pcdata'=>1), 'rbc'=>array('rb'=>1), 'rp'=>array('#pcdata'=>1), 'rtc'=>array('rt'=>1), 'ruby'=>array('rb'=>1, 'rbc'=>1, 'rp'=>1, 'rt'=>1, 'rtc'=>1), 'select'=>array('optgroup'=>1, 'option'=>1), 'script'=>array('#pcdata'=>1), 'table'=>array('caption'=>1, 'col'=>1, 'colgroup'=>1, 'tfoot'=>1, 'tbody'=>1, 'tr'=>1, 'thead'=>1), 'tbody'=>array('tr'=>1), 'tfoot'=>array('tr'=>1), 'textarea'=>array('#pcdata'=>1), 'thead'=>array('tr'=>1), 'tr'=>array('td'=>1, 'th'=>1), 'ul'=>array('li'=>1)); // Specific - immediate parent-child | ||
157 | if($GLOBALS['C']['direct_list_nest']){$cS['ol'] = $cS['ul'] += array('ol'=>1, 'ul'=>1);} | ||
158 | $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 | ||
159 | $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 | ||
160 | // block/inline type; ins & del both type; #pcdata: text | ||
161 | $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); | ||
162 | $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); | ||
163 | $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 | ||
164 | $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 | ||
165 | $eF = $eB + $eI; | ||
166 | |||
167 | // $in sets allowed child | ||
168 | $in = ((isset($eF[$in]) && $in != '#pcdata') or isset($eO[$in])) ? $in : 'div'; | ||
169 | if(isset($cE[$in])){ | ||
170 | return (!$do ? '' : str_replace(array('<', '>'), array('<', '>'), $t)); | ||
171 | } | ||
172 | if(isset($cS[$in])){$inOk = $cS[$in];} | ||
173 | elseif(isset($cI[$in])){$inOk = $eI; $cI['del'] = 1; $cI['ins'] = 1;} | ||
174 | elseif(isset($cF[$in])){$inOk = $eF; unset($cI['del'], $cI['ins']);} | ||
175 | elseif(isset($cB[$in])){$inOk = $eB; unset($cI['del'], $cI['ins']);} | ||
176 | if(isset($cO[$in])){$inOk = $inOk + $cO[$in];} | ||
177 | if(isset($cN[$in])){$inOk = array_diff_assoc($inOk, $cN[$in]);} | ||
178 | |||
179 | $t = explode('<', $t); | ||
180 | $ok = $q = array(); // $q seq list of open non-empty ele | ||
181 | ob_start(); | ||
182 | |||
183 | for($i=-1, $ci=count($t); ++$i<$ci;){ | ||
184 | // allowed $ok in parent $p | ||
185 | if($ql = count($q)){ | ||
186 | $p = array_pop($q); | ||
187 | $q[] = $p; | ||
188 | if(isset($cS[$p])){$ok = $cS[$p];} | ||
189 | elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;} | ||
190 | elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);} | ||
191 | elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);} | ||
192 | if(isset($cO[$p])){$ok = $ok + $cO[$p];} | ||
193 | if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);} | ||
194 | }else{$ok = $inOk; unset($cI['del'], $cI['ins']);} | ||
195 | // bad tags, & ele content | ||
196 | if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){ | ||
197 | echo '<', $s, $e, $a, '>'; | ||
198 | } | ||
199 | if(isset($x[0])){ | ||
200 | if(strlen(trim($x)) && (($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql))){ | ||
201 | echo '<div>', $x, '</div>'; | ||
202 | } | ||
203 | elseif($do < 3 or isset($ok['#pcdata'])){echo $x;} | ||
204 | elseif(strpos($x, "\x02\x04")){ | ||
205 | foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){ | ||
206 | echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : '')); | ||
207 | } | ||
208 | }elseif($do > 4){echo preg_replace('`\S`', '', $x);} | ||
209 | } | ||
210 | // get markup | ||
211 | if(!preg_match('`^(/?)([a-z1-6]+)([^>]*)>(.*)`sm', $t[$i], $r)){$x = $t[$i]; continue;} | ||
212 | $s = null; $e = null; $a = null; $x = null; list($all, $s, $e, $a, $x) = $r; | ||
213 | // close tag | ||
214 | if($s){ | ||
215 | if(isset($cE[$e]) or !in_array($e, $q)){continue;} // Empty/unopen | ||
216 | if($p == $e){array_pop($q); echo '</', $e, '>'; unset($e); continue;} // Last open | ||
217 | $add = ''; // Nesting - close open tags that need to be | ||
218 | for($j=-1, $cj=count($q); ++$j<$cj;){ | ||
219 | if(($d = array_pop($q)) == $e){break;} | ||
220 | else{$add .= "</{$d}>";} | ||
221 | } | ||
222 | echo $add, '</', $e, '>'; unset($e); continue; | ||
223 | } | ||
224 | // open tag | ||
225 | // $cB ele needs $eB ele as child | ||
226 | if(isset($cB[$e]) && strlen(trim($x))){ | ||
227 | $t[$i] = "{$e}{$a}>"; | ||
228 | array_splice($t, $i+1, 0, 'div>'. $x); unset($e, $x); ++$ci; --$i; continue; | ||
229 | } | ||
230 | if((($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql)) && !isset($eB[$e]) && !isset($ok[$e])){ | ||
231 | array_splice($t, $i, 0, 'div>'); unset($e, $x); ++$ci; --$i; continue; | ||
232 | } | ||
233 | // if no open ele, $in = parent; mostly immediate parent-child relation should hold | ||
234 | if(!$ql or !isset($eN[$e]) or !array_intersect($q, $cN2)){ | ||
235 | if(!isset($ok[$e])){ | ||
236 | if($ql && isset($cT[$p])){echo '</', array_pop($q), '>'; unset($e, $x); --$i;} | ||
237 | continue; | ||
238 | } | ||
239 | if(!isset($cE[$e])){$q[] = $e;} | ||
240 | echo '<', $e, $a, '>'; unset($e); continue; | ||
241 | } | ||
242 | // specific parent-child | ||
243 | if(isset($cS[$p][$e])){ | ||
244 | if(!isset($cE[$e])){$q[] = $e;} | ||
245 | echo '<', $e, $a, '>'; unset($e); continue; | ||
246 | } | ||
247 | // nesting | ||
248 | $add = ''; | ||
249 | $q2 = array(); | ||
250 | for($k=-1, $kc=count($q); ++$k<$kc;){ | ||
251 | $d = $q[$k]; | ||
252 | $ok2 = array(); | ||
253 | if(isset($cS[$d])){$q2[] = $d; continue;} | ||
254 | $ok2 = isset($cI[$d]) ? $eI : $eF; | ||
255 | if(isset($cO[$d])){$ok2 = $ok2 + $cO[$d];} | ||
256 | if(isset($cN[$d])){$ok2 = array_diff_assoc($ok2, $cN[$d]);} | ||
257 | if(!isset($ok2[$e])){ | ||
258 | if(!$k && !isset($inOk[$e])){continue 2;} | ||
259 | $add = "</{$d}>"; | ||
260 | for(;++$k<$kc;){$add = "</{$q[$k]}>{$add}";} | ||
261 | break; | ||
262 | } | ||
263 | else{$q2[] = $d;} | ||
264 | } | ||
265 | $q = $q2; | ||
266 | if(!isset($cE[$e])){$q[] = $e;} | ||
267 | echo $add, '<', $e, $a, '>'; unset($e); continue; | ||
268 | } | ||
269 | |||
270 | // end | ||
271 | if($ql = count($q)){ | ||
272 | $p = array_pop($q); | ||
273 | $q[] = $p; | ||
274 | if(isset($cS[$p])){$ok = $cS[$p];} | ||
275 | elseif(isset($cI[$p])){$ok = $eI; $cI['del'] = 1; $cI['ins'] = 1;} | ||
276 | elseif(isset($cF[$p])){$ok = $eF; unset($cI['del'], $cI['ins']);} | ||
277 | elseif(isset($cB[$p])){$ok = $eB; unset($cI['del'], $cI['ins']);} | ||
278 | if(isset($cO[$p])){$ok = $ok + $cO[$p];} | ||
279 | if(isset($cN[$p])){$ok = array_diff_assoc($ok, $cN[$p]);} | ||
280 | }else{$ok = $inOk; unset($cI['del'], $cI['ins']);} | ||
281 | if(isset($e) && ($do == 1 or (isset($ok['#pcdata']) && ($do == 3 or $do == 5)))){ | ||
282 | echo '<', $s, $e, $a, '>'; | ||
283 | } | ||
284 | if(isset($x[0])){ | ||
285 | if(strlen(trim($x)) && (($ql && isset($cB[$p])) or (isset($cB[$in]) && !$ql))){ | ||
286 | echo '<div>', $x, '</div>'; | ||
287 | } | ||
288 | elseif($do < 3 or isset($ok['#pcdata'])){echo $x;} | ||
289 | elseif(strpos($x, "\x02\x04")){ | ||
290 | foreach(preg_split('`(\x01\x02[^\x01\x02]+\x02\x01)`', $x, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY) as $v){ | ||
291 | echo (substr($v, 0, 2) == "\x01\x02" ? $v : ($do > 4 ? preg_replace('`\S`', '', $v) : '')); | ||
292 | } | ||
293 | }elseif($do > 4){echo preg_replace('`\S`', '', $x);} | ||
294 | } | ||
295 | while(!empty($q) && ($e = array_pop($q))){echo '</', $e, '>';} | ||
296 | $o = ob_get_contents(); | ||
297 | ob_end_clean(); | ||
298 | return $o; | ||
299 | // eof | ||
300 | } | ||
301 | |||
302 | public static function hl_cmtcd($t){ | ||
303 | // comment/CDATA sec handler | ||
304 | $t = $t[0]; | ||
305 | global $C; | ||
306 | if(!($v = $C[$n = $t[3] == '-' ? 'comment' : 'cdata'])){return $t;} | ||
307 | if($v == 1){return '';} | ||
308 | if($n == 'comment'){ | ||
309 | if(substr(($t = preg_replace('`--+`', '-', substr($t, 4, -3))), -1) != ' '){$t .= ' ';} | ||
310 | } | ||
311 | else{$t = substr($t, 1, -1);} | ||
312 | $t = $v == 2 ? str_replace(array('&', '<', '>'), array('&', '<', '>'), $t) : $t; | ||
313 | return str_replace(array('&', '<', '>'), array("\x03", "\x04", "\x05"), ($n == 'comment' ? "\x01\x02\x04!--$t--\x05\x02\x01" : "\x01\x01\x04$t\x05\x01\x01")); | ||
314 | // eof | ||
315 | } | ||
316 | |||
317 | public static function hl_ent($t){ | ||
318 | // entitity handler | ||
319 | global $C; | ||
320 | $t = $t[1]; | ||
321 | static $U = array('quot'=>1,'amp'=>1,'lt'=>1,'gt'=>1); | ||
322 | 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'); | ||
323 | if($t[0] != '#'){ | ||
324 | 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;'. $t)). ';'; | ||
325 | } | ||
326 | if(($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 $n > 1114111))){ | ||
327 | return ($C['and_mark'] ? "\x06" : '&'). "amp;#{$t};"; | ||
328 | } | ||
329 | return ($C['and_mark'] ? "\x06" : '&'). '#'. (((ctype_digit($t) && $C['hexdec_entity'] < 2) or !$C['hexdec_entity']) ? $n : 'x'. dechex($n)). ';'; | ||
330 | // eof | ||
331 | } | ||
332 | |||
333 | public static function hl_prot($p, $c=null){ | ||
334 | // check URL scheme | ||
335 | global $C; | ||
336 | $b = $a = ''; | ||
337 | if($c == null){$c = 'style'; $b = $p[1]; $a = $p[3]; $p = trim($p[2]);} | ||
338 | $c = isset($C['schemes'][$c]) ? $C['schemes'][$c] : $C['schemes']['*']; | ||
339 | static $d = 'denied:'; | ||
340 | if(isset($c['!']) && substr($p, 0, 7) != $d){$p = "$d$p";} | ||
341 | if(isset($c['*']) or !strcspn($p, '#?;') or (substr($p, 0, 7) == $d)){return "{$b}{$p}{$a}";} // All ok, frag, query, param | ||
342 | if(preg_match('`^([a-z\d\-+.&#; ]+?)(:|&#(58|x3a);|%3a|\\\\0{0,4}3a).`i', $p, $m) && !isset($c[strtolower($m[1])])){ // Denied prot | ||
343 | return "{$b}{$d}{$p}{$a}"; | ||
344 | } | ||
345 | if($C['abs_url']){ | ||
346 | if($C['abs_url'] == -1 && strpos($p, $C['base_url']) === 0){ // Make url rel | ||
347 | $p = substr($p, strlen($C['base_url'])); | ||
348 | }elseif(empty($m[1])){ // Make URL abs | ||
349 | if(substr($p, 0, 2) == '//'){$p = substr($C['base_url'], 0, strpos($C['base_url'], ':')+1). $p;} | ||
350 | elseif($p[0] == '/'){$p = preg_replace('`(^.+?://[^/]+)(.*)`', '$1', $C['base_url']). $p;} | ||
351 | elseif(strcspn($p, './')){$p = $C['base_url']. $p;} | ||
352 | else{ | ||
353 | preg_match('`^([a-zA-Z\d\-+.]+://[^/]+)(.*)`', $C['base_url'], $m); | ||
354 | $p = preg_replace('`(?<=/)\./`', '', $m[2]. $p); | ||
355 | while(preg_match('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', $p)){ | ||
356 | $p = preg_replace('`(?<=/)([^/]{3,}|[^/.]+?|\.[^/.]|[^/.]\.)/\.\./`', '', $p); | ||
357 | } | ||
358 | $p = $m[1]. $p; | ||
359 | } | ||
360 | } | ||
361 | } | ||
362 | return "{$b}{$p}{$a}"; | ||
363 | // eof | ||
364 | } | ||
365 | |||
366 | public static function hl_regex($p){ | ||
367 | // ?regex | ||
368 | if(empty($p)){return 0;} | ||
369 | if($t = ini_get('track_errors')){$o = isset($php_errormsg) ? $php_errormsg : null;} | ||
370 | else{ini_set('track_errors', 1);} | ||
371 | unset($php_errormsg); | ||
372 | if(($d = ini_get('display_errors'))){ini_set('display_errors', 0);} | ||
373 | preg_match($p, ''); | ||
374 | if($d){ini_set('display_errors', 1);} | ||
375 | $r = isset($php_errormsg) ? 0 : 1; | ||
376 | if($t){$php_errormsg = isset($o) ? $o : null;} | ||
377 | else{ini_set('track_errors', 0);} | ||
378 | return $r; | ||
379 | // eof | ||
380 | } | ||
381 | |||
382 | public static function hl_spec($t){ | ||
383 | // final $spec | ||
384 | $s = array(); | ||
385 | $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))); | ||
386 | for($i = count(($t = explode(';', $t))); --$i>=0;){ | ||
387 | $w = $t[$i]; | ||
388 | if(empty($w) or ($e = strpos($w, '=')) === false or !strlen(($a = substr($w, $e+1)))){continue;} | ||
389 | $y = $n = array(); | ||
390 | foreach(explode(',', $a) as $v){ | ||
391 | if(!preg_match('`^([a-z:\-\*]+)(?:\((.*?)\))?`i', $v, $m)){continue;} | ||
392 | if(($x = strtolower($m[1])) == '-*'){$n['*'] = 1; continue;} | ||
393 | if($x[0] == '-'){$n[substr($x, 1)] = 1; continue;} | ||
394 | if(!isset($m[2])){$y[$x] = 1; continue;} | ||
395 | foreach(explode('/', $m[2]) as $m){ | ||
396 | if(empty($m) or ($p = strpos($m, '=')) == 0 or $p < 5){$y[$x] = 1; continue;} | ||
397 | $y[$x][strtolower(substr($m, 0, $p))] = str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x06", "\x07", "\x08"), array(";", "|", "~", " ", ",", "/", "(", ")"), substr($m, $p+1)); | ||
398 | } | ||
399 | if(isset($y[$x]['match']) && !htmLawed::hl_regex($y[$x]['match'])){unset($y[$x]['match']);} | ||
400 | if(isset($y[$x]['nomatch']) && !htmLawed::hl_regex($y[$x]['nomatch'])){unset($y[$x]['nomatch']);} | ||
401 | } | ||
402 | if(!count($y) && !count($n)){continue;} | ||
403 | foreach(explode(',', substr($w, 0, $e)) as $v){ | ||
404 | if(!strlen(($v = strtolower($v)))){continue;} | ||
405 | if(count($y)){$s[$v] = $y;} | ||
406 | if(count($n)){$s[$v]['n'] = $n;} | ||
407 | } | ||
408 | } | ||
409 | return $s; | ||
410 | // eof | ||
411 | } | ||
412 | |||
413 | public static function hl_tag($t){ | ||
414 | // tag/attribute handler | ||
415 | global $C; | ||
416 | $t = $t[0]; | ||
417 | // invalid < > | ||
418 | if($t == '< '){return '< ';} | ||
419 | if($t == '>'){return '>';} | ||
420 | if(!preg_match('`^<(/?)([a-zA-Z][a-zA-Z1-6]*)([^>]*?)\s?>$`m', $t, $m)){ | ||
421 | return str_replace(array('<', '>'), array('<', '>'), $t); | ||
422 | }elseif(!isset($C['elements'][($e = strtolower($m[2]))])){ | ||
423 | return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('<', '>'), $t) : ''); | ||
424 | } | ||
425 | // attr string | ||
426 | $a = str_replace(array("\n", "\r", "\t"), ' ', trim($m[3])); | ||
427 | // tag transform | ||
428 | static $eD = array('applet'=>1, 'center'=>1, 'dir'=>1, 'embed'=>1, 'font'=>1, 'isindex'=>1, 'menu'=>1, 's'=>1, 'strike'=>1, 'u'=>1); // Deprecated | ||
429 | if($C['make_tag_strict'] && isset($eD[$e])){ | ||
430 | $trt = htmLawed::hl_tag2($e, $a, $C['make_tag_strict']); | ||
431 | if(!$e){return (($C['keep_bad']%2) ? str_replace(array('<', '>'), array('<', '>'), $t) : '');} | ||
432 | } | ||
433 | // close tag | ||
434 | static $eE = array('area'=>1, 'br'=>1, 'col'=>1, 'embed'=>1, 'hr'=>1, 'img'=>1, 'input'=>1, 'isindex'=>1, 'param'=>1); // Empty ele | ||
435 | if(!empty($m[1])){ | ||
436 | return (!isset($eE[$e]) ? (empty($C['hook_tag']) ? "</$e>" : $C['hook_tag']($e)) : (($C['keep_bad'])%2 ? str_replace(array('<', '>'), array('<', '>'), $t) : '')); | ||
437 | } | ||
438 | |||
439 | // open tag & attr | ||
440 | 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 | ||
441 | 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 | ||
442 | 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 | ||
443 | static $aNU = array('class'=>array('param'=>1, 'script'=>1), 'dir'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'id'=>array('script'=>1), 'lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'xml:lang'=>array('applet'=>1, 'br'=>1, 'iframe'=>1, 'param'=>1, 'script'=>1), 'onclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'ondblclick'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeydown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeypress'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onkeyup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousedown'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmousemove'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseout'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseover'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'onmouseup'=>array('applet'=>1, 'bdo'=>1, 'br'=>1, 'font'=>1, 'iframe'=>1, 'isindex'=>1, 'param'=>1, 'script'=>1), 'style'=>array('param'=>1, 'script'=>1), 'title'=>array('param'=>1, 'script'=>1)); // Univ & exceptions | ||
444 | |||
445 | if($C['lc_std_val']){ | ||
446 | // predef attr vals for $eAL & $aNE ele | ||
447 | 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); | ||
448 | 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); | ||
449 | $lcase = isset($eAL[$e]) ? 1 : 0; | ||
450 | } | ||
451 | |||
452 | $depTr = 0; | ||
453 | if($C['no_deprecated_attr']){ | ||
454 | // dep attr:applicable ele | ||
455 | 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)); | ||
456 | 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); | ||
457 | $depTr = isset($eAD[$e]) ? 1 : 0; | ||
458 | } | ||
459 | |||
460 | // attr name-vals | ||
461 | if(strpos($a, "\x01") !== false){$a = preg_replace('`\x01[^\x01]*\x01`', '', $a);} // No comment/CDATA sec | ||
462 | $mode = 0; $a = trim($a, ' /'); $aA = array(); | ||
463 | while(strlen($a)){ | ||
464 | $w = 0; | ||
465 | switch($mode){ | ||
466 | case 0: // Name | ||
467 | if(preg_match('`^[a-zA-Z][\-a-zA-Z:]+`', $a, $m)){ | ||
468 | $nm = strtolower($m[0]); | ||
469 | $w = $mode = 1; $a = ltrim(substr_replace($a, '', 0, strlen($m[0]))); | ||
470 | } | ||
471 | break; case 1: | ||
472 | if($a[0] == '='){ // = | ||
473 | $w = 1; $mode = 2; $a = ltrim($a, '= '); | ||
474 | }else{ // No val | ||
475 | $w = 1; $mode = 0; $a = ltrim($a); | ||
476 | $aA[$nm] = ''; | ||
477 | } | ||
478 | break; case 2: // Val | ||
479 | if(preg_match('`^((?:"[^"]*")|(?:\'[^\']*\')|(?:\s*[^\s"\']+))(.*)`', $a, $m)){ | ||
480 | $a = ltrim($m[2]); $m = $m[1]; $w = 1; $mode = 0; | ||
481 | $aA[$nm] = trim(($m[0] == '"' or $m[0] == '\'') ? substr($m, 1, -1) : $m); | ||
482 | } | ||
483 | break; | ||
484 | } | ||
485 | if($w == 0){ // Parse errs, deal with space, " & ' | ||
486 | $a = preg_replace('`^(?:"[^"]*("|$)|\'[^\']*(\'|$)|\S)*\s*`', '', $a); | ||
487 | $mode = 0; | ||
488 | } | ||
489 | } | ||
490 | if($mode == 1){$aA[$nm] = '';} | ||
491 | |||
492 | // clean attrs | ||
493 | global $S; | ||
494 | $rl = isset($S[$e]) ? $S[$e] : array(); | ||
495 | $a = array(); $nfr = 0; | ||
496 | foreach($aA as $k=>$v){ | ||
497 | 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])){ | ||
498 | if(isset($aNE[$k])){$v = $k;} | ||
499 | elseif(!empty($lcase) && (($e != 'button' or $e != 'input') or $k == 'type')){ // Rather loose but ?not cause issues | ||
500 | $v = (isset($aNL[($v2 = strtolower($v))])) ? $v2 : $v; | ||
501 | } | ||
502 | if($k == 'style' && !$C['style_pass']){ | ||
503 | if(false !== strpos($v, '&#')){ | ||
504 | 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', '('=>'(', '('=>'(', ')'=>')', ')'=>')', ' '=>':', ' '=>':', '"'=>'"', '"'=>'"', '''=>"'", '''=>"'", '/'=>'/', '/'=>'/', '*'=>'*', '*'=>'*', '\'=>'\\', '\'=>'\\'); | ||
505 | $v = strtr($v, $sC); | ||
506 | } | ||
507 | $v = preg_replace_callback('`(url(?:\()(?: )*(?:\'|"|&(?:quot|apos);)?)(.+?)((?:\'|"|&(?:quot|apos);)?(?: )*(?:\)))`iS', 'htmLawed::hl_prot', $v); | ||
508 | $v = !$C['css_expression'] ? preg_replace('`expression`i', ' ', preg_replace('`\\\\\S|(/|(%2f))(\*|(%2a))`i', ' ', $v)) : $v; | ||
509 | }elseif(isset($aNP[$k]) or strpos($k, 'src') !== false or $k[0] == 'o'){ | ||
510 | $v = str_replace("\xad", ' ', (strpos($v, '&') !== false ? str_replace(array('­', '­', '­'), ' ', $v) : $v)); | ||
511 | $v = htmLawed::hl_prot($v, $k); | ||
512 | if($k == 'href'){ // X-spam | ||
513 | if($C['anti_mail_spam'] && strpos($v, 'mailto:') === 0){ | ||
514 | $v = str_replace('@', htmlspecialchars($C['anti_mail_spam']), $v); | ||
515 | }elseif($C['anti_link_spam']){ | ||
516 | $r1 = $C['anti_link_spam'][1]; | ||
517 | if(!empty($r1) && preg_match($r1, $v)){continue;} | ||
518 | $r0 = $C['anti_link_spam'][0]; | ||
519 | if(!empty($r0) && preg_match($r0, $v)){ | ||
520 | if(isset($a['rel'])){ | ||
521 | if(!preg_match('`\bnofollow\b`i', $a['rel'])){$a['rel'] .= ' nofollow';} | ||
522 | }elseif(isset($aA['rel'])){ | ||
523 | if(!preg_match('`\bnofollow\b`i', $aA['rel'])){$nfr = 1;} | ||
524 | }else{$a['rel'] = 'nofollow';} | ||
525 | } | ||
526 | } | ||
527 | } | ||
528 | } | ||
529 | if(isset($rl[$k]) && is_array($rl[$k]) && ($v = htmLawed::hl_attrval($v, $rl[$k])) === 0){continue;} | ||
530 | $a[$k] = str_replace('"', '"', $v); | ||
531 | } | ||
532 | } | ||
533 | if($nfr){$a['rel'] = isset($a['rel']) ? $a['rel']. ' nofollow' : 'nofollow';} | ||
534 | |||
535 | // rqd attr | ||
536 | static $eAR = array('area'=>array('alt'=>'area'), 'bdo'=>array('dir'=>'ltr'), 'form'=>array('action'=>''), 'img'=>array('src'=>'', 'alt'=>'image'), 'map'=>array('name'=>''), 'optgroup'=>array('label'=>''), 'param'=>array('name'=>''), 'script'=>array('type'=>'text/javascript'), 'textarea'=>array('rows'=>'10', 'cols'=>'50')); | ||
537 | if(isset($eAR[$e])){ | ||
538 | foreach($eAR[$e] as $k=>$v){ | ||
539 | if(!isset($a[$k])){$a[$k] = isset($v[0]) ? $v : $k;} | ||
540 | } | ||
541 | } | ||
542 | |||
543 | // depr attrs | ||
544 | if($depTr){ | ||
545 | $c = array(); | ||
546 | foreach($a as $k=>$v){ | ||
547 | if($k == 'style' or !isset($aND[$k][$e])){continue;} | ||
548 | if($k == 'align'){ | ||
549 | unset($a['align']); | ||
550 | if($e == 'img' && ($v == 'left' or $v == 'right')){$c[] = 'float: '. $v;} | ||
551 | elseif(($e == 'div' or $e == 'table') && $v == 'center'){$c[] = 'margin: auto';} | ||
552 | else{$c[] = 'text-align: '. $v;} | ||
553 | }elseif($k == 'bgcolor'){ | ||
554 | unset($a['bgcolor']); | ||
555 | $c[] = 'background-color: '. $v; | ||
556 | }elseif($k == 'border'){ | ||
557 | unset($a['border']); $c[] = "border: {$v}px"; | ||
558 | }elseif($k == 'bordercolor'){ | ||
559 | unset($a['bordercolor']); $c[] = 'border-color: '. $v; | ||
560 | }elseif($k == 'clear'){ | ||
561 | unset($a['clear']); $c[] = 'clear: '. ($v != 'all' ? $v : 'both'); | ||
562 | }elseif($k == 'compact'){ | ||
563 | unset($a['compact']); $c[] = 'font-size: 85%'; | ||
564 | }elseif($k == 'height' or $k == 'width'){ | ||
565 | unset($a[$k]); $c[] = $k. ': '. ($v[0] != '*' ? $v. (ctype_digit($v) ? 'px' : '') : 'auto'); | ||
566 | }elseif($k == 'hspace'){ | ||
567 | unset($a['hspace']); $c[] = "margin-left: {$v}px; margin-right: {$v}px"; | ||
568 | }elseif($k == 'language' && !isset($a['type'])){ | ||
569 | unset($a['language']); | ||
570 | $a['type'] = 'text/'. strtolower($v); | ||
571 | }elseif($k == 'name'){ | ||
572 | if($C['no_deprecated_attr'] == 2 or ($e != 'a' && $e != 'map')){unset($a['name']);} | ||
573 | if(!isset($a['id']) && preg_match('`[a-zA-Z][a-zA-Z\d.:_\-]*`', $v)){$a['id'] = $v;} | ||
574 | }elseif($k == 'noshade'){ | ||
575 | unset($a['noshade']); $c[] = 'border-style: none; border: 0; background-color: gray; color: gray'; | ||
576 | }elseif($k == 'nowrap'){ | ||
577 | unset($a['nowrap']); $c[] = 'white-space: nowrap'; | ||
578 | }elseif($k == 'size'){ | ||
579 | unset($a['size']); $c[] = 'size: '. $v. 'px'; | ||
580 | }elseif($k == 'start' or $k == 'value'){ | ||
581 | unset($a[$k]); | ||
582 | }elseif($k == 'type'){ | ||
583 | unset($a['type']); | ||
584 | static $ol_type = array('i'=>'lower-roman', 'I'=>'upper-roman', 'a'=>'lower-latin', 'A'=>'upper-latin', '1'=>'decimal'); | ||
585 | $c[] = 'list-style-type: '. (isset($ol_type[$v]) ? $ol_type[$v] : 'decimal'); | ||
586 | }elseif($k == 'vspace'){ | ||
587 | unset($a['vspace']); $c[] = "margin-top: {$v}px; margin-bottom: {$v}px"; | ||
588 | } | ||
589 | } | ||
590 | if(count($c)){ | ||
591 | $c = implode('; ', $c); | ||
592 | $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $c. ';': $c. ';'; | ||
593 | } | ||
594 | } | ||
595 | // unique ID | ||
596 | if($C['unique_ids'] && isset($a['id'])){ | ||
597 | if(!preg_match('`^[A-Za-z][A-Za-z0-9_\-.:]*$`', ($id = $a['id'])) or (isset($GLOBALS['hl_Ids'][$id]) && $C['unique_ids'] == 1)){unset($a['id']); | ||
598 | }else{ | ||
599 | while(isset($GLOBALS['hl_Ids'][$id])){$id = $C['unique_ids']. $id;} | ||
600 | $GLOBALS['hl_Ids'][($a['id'] = $id)] = 1; | ||
601 | } | ||
602 | } | ||
603 | // xml:lang | ||
604 | if($C['xml:lang'] && isset($a['lang'])){ | ||
605 | $a['xml:lang'] = isset($a['xml:lang']) ? $a['xml:lang'] : $a['lang']; | ||
606 | if($C['xml:lang'] == 2){unset($a['lang']);} | ||
607 | } | ||
608 | // for transformed tag | ||
609 | if(!empty($trt)){ | ||
610 | $a['style'] = isset($a['style']) ? rtrim($a['style'], ' ;'). '; '. $trt : $trt; | ||
611 | } | ||
612 | // return with empty ele / | ||
613 | if(empty($C['hook_tag'])){ | ||
614 | $aA = ''; | ||
615 | foreach($a as $k=>$v){$aA .= " {$k}=\"{$v}\"";} | ||
616 | return "<{$e}{$aA}". (isset($eE[$e]) ? ' /' : ''). '>'; | ||
617 | } | ||
618 | else{return $C['hook_tag']($e, $a);} | ||
619 | // eof | ||
620 | } | ||
621 | |||
622 | public static function hl_tag2(&$e, &$a, $t=1){ | ||
623 | // transform tag | ||
624 | if($e == 'center'){$e = 'div'; return 'text-align: center;';} | ||
625 | if($e == 'dir' or $e == 'menu'){$e = 'ul'; return '';} | ||
626 | if($e == 's' or $e == 'strike'){$e = 'span'; return 'text-decoration: line-through;';} | ||
627 | if($e == 'u'){$e = 'span'; return 'text-decoration: underline;';} | ||
628 | static $fs = array('0'=>'xx-small', '1'=>'xx-small', '2'=>'small', '3'=>'medium', '4'=>'large', '5'=>'x-large', '6'=>'xx-large', '7'=>'300%', '-1'=>'smaller', '-2'=>'60%', '+1'=>'larger', '+2'=>'150%', '+3'=>'200%', '+4'=>'300%'); | ||
629 | if($e == 'font'){ | ||
630 | $a2 = ''; | ||
631 | if(preg_match('`face\s*=\s*(\'|")([^=]+?)\\1`i', $a, $m) or preg_match('`face\s*=(\s*)(\S+)`i', $a, $m)){ | ||
632 | $a2 .= ' font-family: '. str_replace('"', '\'', trim($m[2])). ';'; | ||
633 | } | ||
634 | if(preg_match('`color\s*=\s*(\'|")?(.+?)(\\1|\s|$)`i', $a, $m)){ | ||
635 | $a2 .= ' color: '. trim($m[2]). ';'; | ||
636 | } | ||
637 | if(preg_match('`size\s*=\s*(\'|")?(.+?)(\\1|\s|$)`i', $a, $m) && isset($fs[($m = trim($m[2]))])){ | ||
638 | $a2 .= ' font-size: '. $fs[$m]. ';'; | ||
639 | } | ||
640 | $e = 'span'; return ltrim($a2); | ||
641 | } | ||
642 | if($t == 2){$e = 0; return 0;} | ||
643 | return ''; | ||
644 | // eof | ||
645 | } | ||
646 | |||
647 | public static function hl_tidy($t, $w, $p){ | ||
648 | // Tidy/compact HTM | ||
649 | if(strpos(' pre,script,textarea', "$p,")){return $t;} | ||
650 | $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];'), $t))); | ||
651 | if(($w = strtolower($w)) == -1){ | ||
652 | return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t); | ||
653 | } | ||
654 | $s = strpos(" $w", 't') ? "\t" : ' '; | ||
655 | $s = preg_match('`\d`', $w, $m) ? str_repeat($s, $m[0]) : str_repeat($s, ($s == "\t" ? 1 : 2)); | ||
656 | $N = preg_match('`[ts]([1-9])`', $w, $m) ? $m[1] : 0; | ||
657 | $a = array('br'=>1); | ||
658 | $b = array('button'=>1, 'input'=>1, 'option'=>1); | ||
659 | $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); | ||
660 | $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); | ||
661 | $T = explode('<', $t); | ||
662 | $X = 1; | ||
663 | while($X){ | ||
664 | $n = $N; | ||
665 | $t = $T; | ||
666 | ob_start(); | ||
667 | if(isset($d[$p])){echo str_repeat($s, ++$n);} | ||
668 | echo ltrim(array_shift($t)); | ||
669 | for($i=-1, $j=count($t); ++$i<$j;){ | ||
670 | $r = ''; list($e, $r) = explode('>', $t[$i]); | ||
671 | $x = $e[0] == '/' ? 0 : (substr($e, -1) == '/' ? 1 : ($e[0] != '!' ? 2 : -1)); | ||
672 | $y = !$x ? ltrim($e, '/') : ($x > 0 ? substr($e, 0, strcspn($e, ' ')) : 0); | ||
673 | $e = "<$e>"; | ||
674 | if(isset($d[$y])){ | ||
675 | if(!$x){ | ||
676 | if($n){echo "\n", str_repeat($s, --$n), "$e\n", str_repeat($s, $n);} | ||
677 | else{++$N; ob_end_clean(); continue 2;} | ||
678 | } | ||
679 | else{echo "\n", str_repeat($s, $n), "$e\n", str_repeat($s, ($x != 1 ? ++$n : $n));} | ||
680 | echo ltrim($r); continue; | ||
681 | } | ||
682 | $f = "\n". str_repeat($s, $n); | ||
683 | if(isset($c[$y])){ | ||
684 | if(!$x){echo $e, $f, ltrim($r);} | ||
685 | else{echo $f, $e, $r;} | ||
686 | }elseif(isset($b[$y])){echo $f, $e, $r; | ||
687 | }elseif(isset($a[$y])){echo $e, $f, ltrim($r); | ||
688 | }elseif(!$y){echo $f, $e, $f, ltrim($r); | ||
689 | }else{echo $e, $r;} | ||
690 | } | ||
691 | $X = 0; | ||
692 | } | ||
693 | $t = preg_replace('`[\n]\s*?[\n]+`', "\n", ob_get_contents()); | ||
694 | ob_end_clean(); | ||
695 | if(($l = strpos(" $w", 'r') ? (strpos(" $w", 'n') ? "\r\n" : "\r") : 0)){ | ||
696 | $t = str_replace("\n", $l, $t); | ||
697 | } | ||
698 | return str_replace(array("\x01", "\x02", "\x03", "\x04", "\x05", "\x07"), array('<', '>', "\n", "\r", "\t", ' '), $t); | ||
699 | // eof | ||
700 | } | ||
701 | |||
702 | public static function hl_version(){ | ||
703 | // rel | ||
704 | return '1.1.14'; | ||
705 | // eof | ||
706 | } | ||
707 | |||
708 | public static function kses($t, $h, $p=array('http', 'https', 'ftp', 'news', 'nntp', 'telnet', 'gopher', 'mailto')){ | ||
709 | // kses compat | ||
710 | foreach($h as $k=>$v){ | ||
711 | $h[$k]['n']['*'] = 1; | ||
712 | } | ||
713 | $C['cdata'] = $C['comment'] = $C['make_tag_strict'] = $C['no_deprecated_attr'] = $C['unique_ids'] = 0; | ||
714 | $C['keep_bad'] = 1; | ||
715 | $C['elements'] = count($h) ? strtolower(implode(',', array_keys($h))) : '-*'; | ||
716 | $C['hook'] = 'htmLawed::kses_hook'; | ||
717 | $C['schemes'] = '*:'. implode(',', $p); | ||
718 | return htmLawed::hl($t, $C, $h); | ||
719 | // eof | ||
720 | } | ||
721 | |||
722 | public static function kses_hook($t, &$C, &$S){ | ||
723 | // kses compat | ||
724 | return $t; | ||
725 | // eof | ||
726 | } | ||
727 | // end class | ||
728 | } \ No newline at end of file | ||