2 // $Id: spam.php,v 1.192 2007/06/25 15:12:22 henoheno Exp $
3 // Copyright (C) 2006-2007 PukiWiki Developers Team
4 // License: GPL v2 or (at your option) any later version
6 // Functions for Concept-work of spam-uri metrics
8 // (PHP 4 >= 4.3.0): preg_match_all(PREG_OFFSET_CAPTURE): $method['uri_XXX'] related feature
10 if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
11 if (! defined('DOMAIN_INI_FILE')) define('DOMAIN_INI_FILE', 'domain.ini.php');
13 // ---------------------
16 // (PHP 4 >= 4.2.0): var_export(): mail-reporting and dump related
17 if (! function_exists('var_export')) {
18 function var_export() {
19 return 'var_export() is not found on this server' . "\n";
23 // (PHP 4 >= 4.2.0): preg_grep() enables invert option
24 function preg_grep_invert($pattern = '//', $input = array())
27 if (! isset($invert)) $invert = defined('PREG_GREP_INVERT');
30 return preg_grep($pattern, $input, PREG_GREP_INVERT);
32 $result = preg_grep($pattern, $input);
34 return array_diff($input, preg_grep($pattern, $input));
43 // Very roughly, shrink the lines of var_export()
44 // NOTE: If the same data exists, it must be corrupted.
45 function var_export_shrink($expression, $return = FALSE, $ignore_numeric_keys = FALSE)
47 $result = var_export($expression, TRUE);
49 $result = preg_replace(
50 // Remove a newline and spaces
51 '# => \n *array \(#', ' => array (',
55 if ($ignore_numeric_keys) {
56 $result =preg_replace(
57 // Remove numeric keys
58 '#^( *)[0-9]+ => #m', '$1',
71 // Remove redundant values from array()
72 function array_unique_recursive($array = array())
74 if (! is_array($array)) return $array;
77 foreach($array as $key => $value){
78 if (is_array($value)) {
79 $array[$key] = array_unique_recursive($value);
81 if (isset($tmp[$value])) {
92 // Renumber all numeric keys from 0
93 function array_renumber_numeric_keys(& $array)
95 if (! is_array($array)) return $array;
99 foreach($array as $key => $value){
100 if (is_array($value)) array_renumber_numeric_keys($array[$key]); // Recurse
101 if (is_numeric($key)) $tmp[$key] = ++$count;
103 array_rename_keys($array, $tmp);
108 // Roughly strings(1) using PCRE
109 // This function is useful to:
110 // * Reduce the size of data, from removing unprintable binary data
111 // * Detect _bare_strings_ from binary data
113 // http://www.freebsd.org/cgi/man.cgi?query=strings (Man-page of GNU strings)
114 // http://www.pcre.org/pcre.txt
115 // Note: mb_ereg_replace() is one of mbstring extension's functions
116 // and need to init its encoding.
117 function strings($binary = '', $min_len = 4, $ignore_space = FALSE, $multibyte = FALSE)
120 $binary = (is_array($binary) || $binary === TRUE) ? '' : strval($binary);
122 $regex = $ignore_space ?
123 '[^[:graph:] \t\n]+' : // Remove "\0" etc, and readable spaces
124 '[^[:graph:][:space:]]+'; // Preserve readable spaces if possible
126 $binary = $multibyte ?
127 mb_ereg_replace($regex, "\n", $binary) :
128 preg_replace('/' . $regex . '/s', "\n", $binary);
131 $binary = preg_replace(
146 // The last character seems "\n" or not
147 $br = (! empty($binary) && $binary[strlen($binary) - 1] == "\n") ? "\n" : '';
149 $min_len = min(1024, intval($min_len));
150 $regex = '/^.{' . $min_len . ',}/S';
151 $binary = implode("\n", preg_grep($regex, explode("\n", $binary))) . $br;
157 // Reverse $string with specified delimiter
158 function delimiter_reverse($string = 'foo.bar.example.com', $from_delim = '.', $to_delim = '.')
160 if (! is_string($string) || ! is_string($from_delim) || ! is_string($to_delim))
163 // com.example.bar.foo
164 return implode($to_delim, array_reverse(explode($from_delim, $string)));
168 // ---------------------
171 // Return an array of URIs in the $string
172 // [OK] http://nasty.example.org#nasty_string
173 // [OK] http://nasty.example.org:80/foo/xxx#nasty_string/bar
174 // [OK] ftp://nasty.example.org:80/dfsdfs
175 // [OK] ftp://cnn.example.com&story=breaking_news@10.0.0.1/top_story.htm (from RFC3986)
176 function uri_pickup($string = '')
178 if (! is_string($string)) return array();
180 // Not available for: IDN(ignored)
183 // scheme://userinfo@host:port/path/or/pathinfo/maybefile.and?query=string#fragment
184 // Refer RFC3986 (Regex below is not strict)
185 '#(\b[a-z][a-z0-9.+-]{1,8}):[/\\\]+' . // 1: Scheme
187 '([^\s<>"\'\[\]/\#?@]*)' . // 2: Userinfo (Username)
191 '\[[0-9a-f:.]+\]' . '|' . // IPv6([colon-hex and dot]): RFC2732
192 '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' . // IPv4(dot-decimal): 001.22.3.44
193 '[a-z0-9_-][a-z0-9_.-]+[a-z0-9_-]' . // hostname(FQDN) : foo.example.org
195 '(?::([0-9]*))?' . // 4: Port
196 '((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
197 '([^\s<>"\'\[\]\#?]+)?' . // 6: File?
198 '(?:\?([^\s<>"\'\[\]\#]+))?' . // 7: Query string
199 '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 8: Fragment
201 $string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE
205 static $parts = array(
206 1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port',
207 5 => 'path', 6 => 'file', 7 => 'query', 8 => 'fragment'
209 $default = array('');
210 foreach(array_keys($array) as $uri) {
211 $_uri = & $array[$uri];
212 array_rename_keys($_uri, $parts, TRUE, $default);
213 $offset = $_uri['scheme'][1]; // Scheme's offset = URI's offset
214 foreach(array_keys($_uri) as $part) {
215 $_uri[$part] = & $_uri[$part][0]; // Remove offsets
219 foreach(array_keys($array) as $uri) {
220 $_uri = & $array[$uri];
221 if ($_uri['scheme'] === '') {
222 unset($array[$uri]); // Considererd harmless
225 unset($_uri[0]); // Matched string itself
226 $_uri['area']['offset'] = $offset; // Area offset for area_measure()
232 // Normalize an array of URI arrays
233 // NOTE: Give me the uri_pickup() results
234 function uri_pickup_normalize(& $pickups, $destructive = TRUE)
236 if (! is_array($pickups)) return $pickups;
239 foreach (array_keys($pickups) as $key) {
240 $_key = & $pickups[$key];
241 $_key['scheme'] = isset($_key['scheme']) ? scheme_normalize($_key['scheme']) : '';
242 $_key['host'] = isset($_key['host']) ? host_normalize($_key['host']) : '';
243 $_key['port'] = isset($_key['port']) ? port_normalize($_key['port'], $_key['scheme'], FALSE) : '';
244 $_key['path'] = isset($_key['path']) ? strtolower(path_normalize($_key['path'])) : '';
245 $_key['file'] = isset($_key['file']) ? file_normalize($_key['file']) : '';
246 $_key['query'] = isset($_key['query']) ? query_normalize($_key['query']) : '';
247 $_key['fragment'] = isset($_key['fragment']) ? strtolower($_key['fragment']) : '';
250 foreach (array_keys($pickups) as $key) {
251 $_key = & $pickups[$key];
252 $_key['scheme'] = isset($_key['scheme']) ? scheme_normalize($_key['scheme']) : '';
253 $_key['host'] = isset($_key['host']) ? strtolower($_key['host']) : '';
254 $_key['port'] = isset($_key['port']) ? port_normalize($_key['port'], $_key['scheme'], FALSE) : '';
255 $_key['path'] = isset($_key['path']) ? path_normalize($_key['path']) : '';
262 // An URI array => An URI (See uri_pickup())
264 // $pickups = uri_pickup('a string include some URIs');
266 // foreach (array_keys($pickups) as $key) {
267 // $uris[$key] = uri_pickup_implode($pickups[$key]);
269 function uri_pickup_implode($uri = array())
271 if (empty($uri) || ! is_array($uri)) return NULL;
274 if (isset($uri['scheme']) && $uri['scheme'] !== '') {
275 $tmp[] = & $uri['scheme'];
278 if (isset($uri['userinfo']) && $uri['userinfo'] !== '') {
279 $tmp[] = & $uri['userinfo'];
282 if (isset($uri['host']) && $uri['host'] !== '') {
283 $tmp[] = & $uri['host'];
285 if (isset($uri['port']) && $uri['port'] !== '') {
287 $tmp[] = & $uri['port'];
289 if (isset($uri['path']) && $uri['path'] !== '') {
290 $tmp[] = & $uri['path'];
292 if (isset($uri['file']) && $uri['file'] !== '') {
293 $tmp[] = & $uri['file'];
295 if (isset($uri['query']) && $uri['query'] !== '') {
297 $tmp[] = & $uri['query'];
299 if (isset($uri['fragment']) && $uri['fragment'] !== '') {
301 $tmp[] = & $uri['fragment'];
304 return implode('', $tmp);
307 // $array['something'] => $array['wanted']
308 function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '')
310 if (! is_array($array) || ! is_array($keys)) return FALSE;
312 // Nondestructive test
314 foreach(array_keys($keys) as $from)
315 if (! isset($array[$from]))
318 foreach($keys as $from => $to) {
319 if ($from === $to) continue;
320 if (! $force || isset($array[$from])) {
321 $array[$to] = & $array[$from];
322 unset($array[$from]);
324 $array[$to] = $default;
331 // ---------------------
334 // Pickup all of markup areas
335 function area_pickup($string = '', $method = array())
338 if (empty($method)) return $area;
340 // Anchor tag pair by preg_match and preg_match_all()
342 // [OK] <a href= >Good site!</a>
343 // [OK] <a href= "#" >test</a>
344 // [OK] <a href="http://nasty.example.com">visit http://nasty.example.com/</a>
345 // [OK] <a href=\'http://nasty.example.com/\' >discount foobar</a>
346 // [NG] <a href="http://ng.example.com">visit http://ng.example.com _not_ended_
347 $regex = '#<a\b[^>]*\bhref\b[^>]*>.*?</a\b[^>]*(>)#is';
348 if (isset($method['area_anchor'])) {
350 $count = isset($method['asap']) ?
351 preg_match($regex, $string) :
352 preg_match_all($regex, $string, $areas);
353 if (! empty($count)) $area['area_anchor'] = $count;
355 if (isset($method['uri_anchor'])) {
357 preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
358 foreach(array_keys($areas) as $_area) {
359 $areas[$_area] = array(
360 $areas[$_area][0][1], // Area start (<a href>)
361 $areas[$_area][1][1], // Area end (</a>)
364 if (! empty($areas)) $area['uri_anchor'] = $areas;
367 // phpBB's "BBCode" pair by preg_match and preg_match_all()
369 // [OK] [url]http://nasty.example.com/[/url]
370 // [OK] [link]http://nasty.example.com/[/link]
371 // [OK] [url=http://nasty.example.com]visit http://nasty.example.com/[/url]
372 // [OK] [link http://nasty.example.com/]buy something[/link]
373 $regex = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#is';
374 if (isset($method['area_bbcode'])) {
376 $count = isset($method['asap']) ?
377 preg_match($regex, $string) :
378 preg_match_all($regex, $string, $areas, PREG_SET_ORDER);
379 if (! empty($count)) $area['area_bbcode'] = $count;
381 if (isset($method['uri_bbcode'])) {
383 preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
384 foreach(array_keys($areas) as $_area) {
385 $areas[$_area] = array(
386 $areas[$_area][0][1], // Area start ([url])
387 $areas[$_area][2][1], // Area end ([/url])
390 if (! empty($areas)) $area['uri_bbcode'] = $areas;
393 // Various Wiki syntax
394 // [text_or_uri>text_or_uri]
395 // [text_or_uri:text_or_uri]
396 // [text_or_uri|text_or_uri]
397 // [text_or_uri->text_or_uri]
398 // [text_or_uri text_or_uri] // MediaWiki
399 // MediaWiki: [http://nasty.example.com/ visit http://nasty.example.com/]
404 // If in doubt, it's a little doubtful
405 // if (Area => inside <= Area) $brief += -1
406 function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset')
408 if (! is_array($areas) || ! is_array($array)) return;
410 $areas_keys = array_keys($areas);
411 foreach(array_keys($array) as $u_index) {
412 $offset = isset($array[$u_index][$o_key]) ?
413 intval($array[$u_index][$o_key]) : 0;
414 foreach($areas_keys as $a_index) {
415 if (isset($array[$u_index][$a_key])) {
416 $offset_s = intval($areas[$a_index][0]);
417 $offset_e = intval($areas[$a_index][1]);
418 // [Area => inside <= Area]
419 if ($offset_s < $offset && $offset < $offset_e) {
420 $array[$u_index][$a_key] += $belief;
427 // ---------------------
430 // Domain exposure callback (See spam_uri_pickup_preprocess())
431 // http://victim.example.org/?foo+site:nasty.example.com+bar
432 // => http://nasty.example.com/?refer=victim.example.org
433 // NOTE: 'refer=' is not so good for (at this time).
434 // Consider about using IP address of the victim, try to avoid that.
435 function _preg_replace_callback_domain_exposure($matches = array())
439 // Preserve the victim URI as a complicity or ...
440 if (isset($matches[5])) {
442 $matches[1] . '://' . // scheme
443 $matches[2] . '/' . // victim.example.org
444 $matches[3]; // The rest of all (before victim)
448 if (isset($matches[4])) {
450 $matches[1] . '://' . // scheme
451 $matches[4] . // nasty.example.com
452 '/?refer=' . strtolower($matches[2]) . // victim.example.org
459 // Preprocess: Removing uninterest part for URI detection
460 function spam_uri_removing_hocus_pocus($binary = '', $method = array())
462 $length = 4 ; // 'http'(1) and '://'(2) and 'fqdn'(1)
463 if (is_array($method)) {
464 // '<a'(2) or 'href='(5) or '>'(1) or '</a>'(4)
465 // '[uri'(4) or ']'(1) or '[/uri]'(6)
466 if (isset($method['area_anchor']) || isset($method['uri_anchor']) ||
467 isset($method['area_bbcode']) || isset($method['uri_bbcode']))
468 $length = 1; // Seems not effective
471 // Removing sequential spaces and too short lines
472 $binary = strings($binary, $length, TRUE, FALSE); // Multibyte NOT needed
474 // Remove words (has no '<>[]:') between spaces
475 $binary = preg_replace('/[ \t][\w.,()\ \t]+[ \t]/', ' ', $binary);
480 // Preprocess: rawurldecode() and adding space(s) and something
481 // to detect/count some URIs _if possible_
482 // NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
483 // [OK] http://victim.example.org/?site:nasty.example.org
484 // [OK] http://victim.example.org/nasty.example.org
485 // [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
486 // [OK] http://victim.example.org/http://nasty.example.org
487 function spam_uri_pickup_preprocess($string = '', $method = array())
489 if (! is_string($string)) return '';
491 $string = spam_uri_removing_hocus_pocus(rawurldecode($string), $method);
492 //var_dump(htmlspecialchars($string));
494 // Domain exposure (simple)
495 // http://victim.example.org/nasty.example.org/path#frag
496 // => http://nasty.example.org/?refer=victim.example.org and original
497 $string = preg_replace(
500 'ime\.nu' . '|' . // 2ch.net
501 'ime\.st' . '|' . // 2ch.net
502 'link\.toolbot\.com' . '|' .
505 '/([a-z0-9.%_-]+\.[a-z0-9.%_-]+)#i', // nasty.example.org
506 'http://$2/?refer=$1 $0', // Preserve $0 or remove?
510 // Domain exposure (gate-big5)
511 // http://victim.example.org/gate/big5/nasty.example.org/path
512 // => http://nasty.example.org/?refer=victim.example.org and original
513 $string = preg_replace(
516 'big5.51job.com' . '|' .
517 'big5.china.com' . '|' .
518 'big5.xinhuanet.com' . '|' .
521 '/([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' .
522 '#i', // nasty.example.org
523 'http://$2/?refer=$1 $0', // Preserve $0 or remove?
527 // Domain exposure (See _preg_replace_callback_domain_exposure())
528 $string = preg_replace_callback(
532 // Something Google: http://www.google.com/supported_domains
533 '(?:[a-z0-9.]+\.)?google\.[a-z]{2,3}(?:\.[a-z]{2})?' .
536 '(?:[a-z0-9.]+\.)?altavista.com' .
540 '([a-z0-9?=&.%_/\'\\\+-]+)' . // path/?query=foo+bar+
541 '\bsite:([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' . // site:nasty.example.com
542 //'()' . // Preserve or remove?
545 '_preg_replace_callback_domain_exposure',
549 // URI exposure (uriuri => uri uri)
550 $string = preg_replace(
552 '#(?<! )(?:https?|ftp):/#i',
553 // '#[a-z][a-z0-9.+-]{1,8}://#i',
554 // '#[a-z][a-z0-9.+-]{1,8}://#i'
563 // Main function of spam-uri pickup,
564 // A wrapper function of uri_pickup()
565 function spam_uri_pickup($string = '', $method = array())
567 if (! is_array($method) || empty($method)) {
568 $method = check_uri_spam_method();
571 $string = spam_uri_pickup_preprocess($string, $method);
573 $array = uri_pickup($string);
575 // Area elevation of URIs, for '(especially external)link' intension
576 if (! empty($array)) {
578 if (isset($method['uri_anchor'])) $_method['uri_anchor'] = & $method['uri_anchor'];
579 if (isset($method['uri_bbcode'])) $_method['uri_bbcode'] = & $method['uri_bbcode'];
580 $areas = area_pickup($string, $_method, TRUE);
581 if (! empty($areas)) {
582 $area_shadow = array();
583 foreach (array_keys($array) as $key) {
584 $area_shadow[$key] = & $array[$key]['area'];
585 foreach (array_keys($_method) as $_key) {
586 $area_shadow[$key][$_key] = 0;
589 foreach (array_keys($_method) as $_key) {
590 if (isset($areas[$_key])) {
591 area_measure($areas[$_key], $area_shadow, 1, $_key);
597 // Remove 'offset's for area_measure()
598 foreach(array_keys($array) as $key)
599 unset($array[$key]['area']['offset']);
605 // ---------------------
608 // Scheme normalization: Renaming the schemes
609 // snntp://example.org => nntps://example.org
610 // NOTE: Keep the static lists simple. See also port_normalize().
611 function scheme_normalize($scheme = '', $abbrevs_harmfull = TRUE)
613 // Abbreviations they have no intention of link
614 static $abbrevs = array(
619 // Aliases => normalized ones
620 static $aliases = array(
630 if (! is_string($scheme)) return '';
632 $scheme = strtolower($scheme);
633 if (isset($abbrevs[$scheme])) {
634 $scheme = $abbrevs_harmfull ? $abbrevs[$scheme] : '';
636 if (isset($aliases[$scheme])) {
637 $scheme = $aliases[$scheme];
643 // Hostname normlization (Destructive)
644 // www.foo => www.foo ('foo' seems TLD)
645 // www.foo.bar => foo.bar
646 // www.10.20 => www.10.20 (Invalid hostname)
648 // 'www' is mostly used as traditional hostname of WWW server.
649 // 'www.foo.bar' may be identical with 'foo.bar'.
650 function host_normalize($host = '')
652 if (! is_string($host)) return '';
654 $host = strtolower($host);
656 if (preg_match('/^www\.(.+\.[a-z]+)$/', $host, $matches)) {
663 // Port normalization: Suppress the (redundant) default port
664 // HTTP://example.org:80/ => http://example.org/
665 // HTTP://example.org:8080/ => http://example.org:8080/
666 // HTTPS://example.org:443/ => https://example.org/
667 function port_normalize($port, $scheme, $scheme_normalize = FALSE)
669 // Schemes that users _maybe_ want to add protocol-handlers
670 // to their web browsers. (and attackers _maybe_ want to use ...)
671 // Reference: http://www.iana.org/assignments/port-numbers
672 static $array = array(
673 // scheme => default port
699 // intval() converts '0-1' to '0', so preg_match() rejects these invalid ones
700 if (! is_numeric($port) || $port < 0 || preg_match('/[^0-9]/i', $port))
703 $port = intval($port);
704 if ($scheme_normalize) $scheme = scheme_normalize($scheme);
705 if (isset($array[$scheme]) && $port == $array[$scheme])
706 $port = ''; // Ignore the defaults
711 // Path normalization
712 // http://example.org => http://example.org/
713 // http://example.org#hoge => http://example.org/#hoge
714 // http://example.org/path/a/b/./c////./d => http://example.org/path/a/b/c/d
715 // http://example.org/path/../../a/../back => http://example.org/back
716 function path_normalize($path = '', $divider = '/', $add_root = TRUE)
718 if (! is_string($divider)) return is_string($path) ? $path : '';
721 $first_div = & $divider;
725 if (! is_string($path) || $path == '') return $first_div;
727 if (strpos($path, $divider, strlen($path) - strlen($divider)) === FALSE) {
730 $last_div = & $divider;
733 $array = explode($divider, $path);
735 // Remove paddings ('//' and '/./')
736 foreach(array_keys($array) as $key) {
737 if ($array[$key] == '' || $array[$key] == '.') {
742 // Remove back-tracks ('/../')
744 foreach($array as $value) {
745 if ($value == '..') {
748 array_push($tmp, $value);
756 return $first_div . implode($divider, $array) . $last_div;
760 // DirectoryIndex normalize (Destructive and rough)
761 // TODO: sample.en.ja.html.gz => sample.html
762 function file_normalize($file = 'index.html.en')
764 static $simple_defaults = array(
765 'default.htm' => TRUE,
766 'default.html' => TRUE,
767 'default.asp' => TRUE,
768 'default.aspx' => TRUE,
769 'index' => TRUE, // Some system can omit the suffix
772 static $content_suffix = array(
773 // index.xxx, sample.xxx
788 static $language_suffix = array(
789 // Reference: Apache 2.0.59 'AddLanguage' default
818 // Reference: Apache 2.0.59 default 'index.html' variants
824 static $charset_suffix = array(
825 // Reference: Apache 2.0.59 'AddCharset' default
826 'iso8859-1' => TRUE, // ISO-8859-1
827 'latin1' => TRUE, // ISO-8859-1
828 'iso8859-2' => TRUE, // ISO-8859-2
829 'latin2' => TRUE, // ISO-8859-2
830 'cen' => TRUE, // ISO-8859-2
831 'iso8859-3' => TRUE, // ISO-8859-3
832 'latin3' => TRUE, // ISO-8859-3
833 'iso8859-4' => TRUE, // ISO-8859-4
834 'latin4' => TRUE, // ISO-8859-4
835 'iso8859-5' => TRUE, // ISO-8859-5
836 'latin5' => TRUE, // ISO-8859-5
837 'cyr' => TRUE, // ISO-8859-5
838 'iso-ru' => TRUE, // ISO-8859-5
839 'iso8859-6' => TRUE, // ISO-8859-6
840 'latin6' => TRUE, // ISO-8859-6
841 'arb' => TRUE, // ISO-8859-6
842 'iso8859-7' => TRUE, // ISO-8859-7
843 'latin7' => TRUE, // ISO-8859-7
844 'grk' => TRUE, // ISO-8859-7
845 'iso8859-8' => TRUE, // ISO-8859-8
846 'latin8' => TRUE, // ISO-8859-8
847 'heb' => TRUE, // ISO-8859-8
848 'iso8859-9' => TRUE, // ISO-8859-9
849 'latin9' => TRUE, // ISO-8859-9
850 'trk' => TRUE, // ISO-8859-9
851 'iso2022-jp'=> TRUE, // ISO-2022-JP
852 'jis' => TRUE, // ISO-2022-JP
853 'iso2022-kr'=> TRUE, // ISO-2022-KR
854 'kis' => TRUE, // ISO-2022-KR
855 'iso2022-cn'=> TRUE, // ISO-2022-CN
856 'cis' => TRUE, // ISO-2022-CN
858 'cp-1251' => TRUE, // ru, WINDOWS-1251
859 'win-1251' => TRUE, // ru, WINDOWS-1251
860 'cp866' => TRUE, // ru
861 'koi8-r' => TRUE, // ru, KOI8-r
862 'koi8-ru' => TRUE, // ru, KOI8-r
863 'koi8-uk' => TRUE, // ru, KOI8-ru
864 'ua' => TRUE, // ru, KOI8-ru
865 'ucs2' => TRUE, // ru, ISO-10646-UCS-2
866 'ucs4' => TRUE, // ru, ISO-10646-UCS-4
869 // Reference: Apache 2.0.59 default 'index.html' variants
874 // May uncompress by web browsers on the fly
875 // Must be at the last of the filename
876 // Reference: Apache 2.0.59 'AddEncoding'
877 static $encoding_suffix = array(
882 if (! is_string($file)) return '';
883 $_file = strtolower($file);
884 if (isset($simple_defaults[$_file])) return '';
887 // Roughly removing language/character-set/encoding suffixes
889 // * Apache 2 document about 'Content-negotiaton', 'mod_mime' and 'mod_negotiation'
890 // http://httpd.apache.org/docs/2.0/content-negotiation.html
891 // http://httpd.apache.org/docs/2.0/mod/mod_mime.html
892 // http://httpd.apache.org/docs/2.0/mod/mod_negotiation.html
893 // * http://www.iana.org/assignments/character-sets
894 // * RFC3066: Tags for the Identification of Languages
895 // http://www.ietf.org/rfc/rfc3066.txt
896 // * ISO 639: codes of 'language names'
897 $suffixes = explode('.', $_file);
898 $body = array_shift($suffixes);
900 // Remove the last .gz/.z
901 $last_key = end(array_keys($suffixes));
902 if (isset($encoding_suffix[$suffixes[$last_key]])) {
903 unset($suffixes[$last_key]);
906 // Cut language and charset suffixes
907 foreach($suffixes as $key => $value){
908 if (isset($language_suffix[$value]) || isset($charset_suffix[$value])) {
909 unset($suffixes[$key]);
912 if (empty($suffixes)) return $body;
915 $count = count($suffixes);
917 $current = current($suffixes);
918 if ($body == 'index' && $count == 1 && isset($content_suffix[$current])) return '';
923 // Sort query-strings if possible (Destructive and rough)
924 // [OK] &&&&f=d&b&d&c&a=0dd => a=0dd&b&c&d&f=d
925 // [OK] nothing==&eg=dummy&eg=padding&eg=foobar => eg=foobar
926 function query_normalize($string = '', $equal = TRUE, $equal_cutempty = TRUE, $stortolower = TRUE)
928 if (! is_string($string)) return '';
929 if ($stortolower) $string = strtolower($string);
931 $array = explode('&', $string);
933 // Remove '&' paddings
934 foreach(array_keys($array) as $key) {
935 if ($array[$key] == '') {
940 // Consider '='-sepalated input and paddings
942 $equals = $not_equals = array();
943 foreach ($array as $part) {
944 if (strpos($part, '=') === FALSE) {
945 $not_equals[] = $part;
947 list($key, $value) = explode('=', $part, 2);
948 $value = ltrim($value, '=');
949 if (! $equal_cutempty || $value != '') {
950 $equals[$key] = $value;
955 $array = & $not_equals;
956 foreach ($equals as $key => $value) {
957 $array[] = $key . '=' . $value;
963 return implode('&', $array);
966 // ---------------------
967 // Part One : Checker
969 // Rough implementation of globbing
971 // USAGE: $regex = '/^' . generate_glob_regex('*.txt', '/') . '$/i';
973 function generate_glob_regex($string = '', $divider = '/')
975 static $from = array(
978 // 22 => '[', // Maybe cause regex compilation error (e.g. '[]')
994 if (! is_string($string)) return '';
996 $string = str_replace($from, $mid, $string); // Hide
997 $string = preg_quote($string, $divider);
998 $string = str_replace($mid, $to, $string); // Unhide
1003 // Rough hostname checker
1005 // TODO: Strict digit, 0x, CIDR, IPv6
1006 function is_ip($string = '')
1008 if (preg_match('/^' .
1009 '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' .
1010 '(?:[0-9]{1,3}\.){1,3}' . '$/',
1012 return 4; // Seems IPv4(dot-decimal)
1014 return 0; // Seems not IP
1018 // Generate host (FQDN, IPv4, ...) regex
1019 // 'localhost' : Matches with 'localhost' only
1020 // 'example.org' : Matches with 'example.org' only (See host_normalize() about 'www')
1021 // '.example.org' : Matches with ALL FQDN ended with '.example.org'
1022 // '*.example.org' : Almost the same of '.example.org' except 'www.example.org'
1023 // '10.20.30.40' : Matches with IPv4 address '10.20.30.40' only
1024 // [TODO] '192.' : Matches with all IPv4 hosts started with '192.'
1025 // TODO: IPv4, CIDR?, IPv6
1026 function generate_host_regex($string = '', $divider = '/')
1028 if (! is_string($string)) return '';
1030 if (mb_strpos($string, '.') === FALSE)
1031 return generate_glob_regex($string, $divider);
1034 if (is_ip($string)) {
1036 return generate_glob_regex($string, $divider);
1038 // FQDN or something
1039 $part = explode('.', $string, 2);
1040 if ($part[0] == '') {
1041 $part[0] = '(?:.*\.)?'; // And all related FQDN
1042 } else if ($part[0] == '*') {
1043 $part[0] = '.*\.'; // All subdomains/hosts only
1045 return generate_glob_regex($string, $divider);
1047 $part[1] = generate_glob_regex($part[1], $divider);
1048 return implode('', $part);
1052 function get_blocklist($list = '')
1056 if ($list === NULL) {
1057 $regexes = NULL; // Unset
1061 if (! isset($regexes)) {
1063 if (file_exists(SPAM_INI_FILE)) {
1064 $blocklist = array();
1065 include(SPAM_INI_FILE);
1066 // $blocklist['badhost'] = array(
1067 // '*.blogspot.com', // Blog services's subdomains (only)
1068 // 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
1070 if (isset($blocklist['list'])) {
1071 $regexes['list'] = & $blocklist['list'];
1074 $blocklist['list'] = array(
1075 'goodhost' => FALSE,
1079 foreach(array_keys($blocklist['list']) as $_list) {
1080 if (! isset($blocklist[$_list])) continue;
1081 foreach ($blocklist[$_list] as $key => $value) {
1082 if (is_array($value)) {
1083 $regexes[$_list][$key] = array();
1084 foreach($value as $_key => $_value) {
1085 get_blocklist_add($regexes[$_list][$key], $_key, $_value);
1088 get_blocklist_add($regexes[$_list], $key, $value);
1091 unset($blocklist[$_list]);
1097 return $regexes; // ALL
1098 } else if (isset($regexes[$list])) {
1099 return $regexes[$list];
1105 // Subroutine of get_blocklist()
1106 function get_blocklist_add(& $array, $key = 0, $value = '*.example.org')
1108 if (is_string($key)) {
1109 $array[$key] = & $value; // Treat $value as a regex
1111 $array[$value] = '/^' . generate_host_regex($value, '/') . '$/i';
1115 // Blocklist metrics: Separate $host, to $blocked and not blocked
1116 function blocklist_distiller(& $hosts, $keys = array('goodhost', 'badhost'), $asap = FALSE)
1118 if (! is_array($hosts)) $hosts = array($hosts);
1119 if (! is_array($keys)) $keys = array($keys);
1121 $list = get_blocklist('list');
1124 foreach($keys as $key){
1125 foreach (get_blocklist($key) as $label => $regex) {
1126 if (is_array($regex)) {
1127 foreach($regex as $_label => $_regex) {
1128 $group = preg_grep($_regex, $hosts);
1130 $hosts = array_diff($hosts, $group);
1131 $blocked[$key][$label][$_label] = $group;
1132 if ($asap && $list[$key]) break;
1136 $group = preg_grep($regex, $hosts);
1138 $hosts = array_diff($hosts, $group);
1139 $blocked[$key][$label] = $group;
1140 if ($asap && $list[$key]) break;
1149 // Default (enabled) methods and thresholds (for content insertion)
1150 function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
1152 $times = intval($times);
1153 $t_area = intval($t_area);
1157 'quantity' => 8 * $times, // Allow N URIs
1158 'non_uniqhost' => 3 * $times, // Allow N duped (and normalized) Hosts
1159 //'non_uniquri'=> 3 * $times, // Allow N duped (and normalized) URIs
1162 'area_anchor' => $t_area, // Using <a href> HTML tag
1163 'area_bbcode' => $t_area, // Using [url] or [link] BBCode
1164 //'uri_anchor' => $t_area, // URI inside <a href> HTML tag
1165 //'uri_bbcode' => $t_area, // URI inside [url] or [link] BBCode
1170 //'asap' => TRUE, // Quit or return As Soon As Possible
1171 'uniqhost' => TRUE, // Show uniq host (at block notification mail)
1172 'badhost' => TRUE, // Check badhost
1178 // Remove non-$positive values
1179 foreach (array_keys($positive) as $key) {
1180 if ($positive[$key] < 0) unset($positive[$key]);
1183 return $positive + $bool;
1186 // Simple/fast spam check
1187 function check_uri_spam($target = '', $method = array())
1192 // Theme to do => Dummy, optional value, or optional array()
1194 //'uniqhost' => TRUE,
1195 //'non_uniqhost'=> 3,
1196 //'non_uniquri' => 3,
1197 //'badhost' => TRUE,
1198 //'area_anchor' => 0,
1199 //'area_bbcode' => 0,
1200 //'uri_anchor' => 0,
1201 //'uri_bbcode' => 0,
1204 // Theme => Volume found (int)
1207 // Flag. If someting defined here,
1208 // one or more spam will be included
1213 //'category' => array(
1218 // Hosts not blocked
1223 $sum = & $progress['sum'];
1224 $is_spam = & $progress['is_spam'];
1225 $progress['method'] = & $method; // Argument
1226 $blocked = & $progress['blocked'];
1227 $hosts = & $progress['hosts'];
1228 $asap = isset($method['asap']);
1231 if (! is_array($method) || empty($method)) {
1232 $method = check_uri_spam_method();
1234 foreach(array_keys($method) as $key) {
1235 if (! isset($sum[$key])) $sum[$key] = 0;
1237 if (! isset($sum['quantity'])) $sum['quantity'] = 0;
1239 if (is_array($target)) {
1240 foreach($target as $str) {
1241 if (! is_string($str)) continue;
1243 $_progress = check_uri_spam($str, $method); // Recurse
1246 $_sum = & $_progress['sum'];
1247 foreach (array_keys($_sum) as $key) {
1248 if (! isset($sum[$key])) {
1249 $sum[$key] = & $_sum[$key];
1251 $sum[$key] += $_sum[$key];
1256 $_is_spam = & $_progress['is_spam'];
1257 foreach (array_keys($_is_spam) as $key) {
1258 $is_spam[$key] = TRUE;
1261 if ($asap && $is_spam) break;
1264 $blocked = array_merge_recursive($blocked, $_progress['blocked']);
1265 $hosts = array_merge_recursive($hosts, $_progress['hosts']);
1269 $blocked = array_unique_recursive($blocked);
1270 $hosts = array_unique_recursive($hosts);
1272 // Recount $sum['badhost']
1273 $sum['badhost'] = array_count_leaves($blocked);
1278 // Area: There's HTML anchor tag
1279 if ((! $asap || ! $is_spam) && isset($method['area_anchor'])) {
1280 $key = 'area_anchor';
1281 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
1282 $result = area_pickup($target, array($key => TRUE) + $_asap);
1284 $sum[$key] = $result[$key];
1285 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
1286 $is_spam[$key] = TRUE;
1291 // Area: There's 'BBCode' linking tag
1292 if ((! $asap || ! $is_spam) && isset($method['area_bbcode'])) {
1293 $key = 'area_bbcode';
1294 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
1295 $result = area_pickup($target, array($key => TRUE) + $_asap);
1297 $sum[$key] = $result[$key];
1298 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
1299 $is_spam[$key] = TRUE;
1305 if ($asap && $is_spam) return $progress;
1308 $pickups = uri_pickup_normalize(spam_uri_pickup($target, $method));
1311 if (empty($pickups)) return $progress;
1313 // URI: Check quantity
1314 $sum['quantity'] += count($pickups);
1316 if ((! $asap || ! $is_spam) && isset($method['quantity']) &&
1317 $sum['quantity'] > $method['quantity']) {
1318 $is_spam['quantity'] = TRUE;
1321 // URI: used inside HTML anchor tag pair
1322 if ((! $asap || ! $is_spam) && isset($method['uri_anchor'])) {
1323 $key = 'uri_anchor';
1324 foreach($pickups as $pickup) {
1325 if (isset($pickup['area'][$key])) {
1326 $sum[$key] += $pickup['area'][$key];
1327 if(isset($method[$key]) &&
1328 $sum[$key] > $method[$key]) {
1329 $is_spam[$key] = TRUE;
1330 if ($asap && $is_spam) break;
1332 if ($asap && $is_spam) break;
1337 // URI: used inside 'BBCode' pair
1338 if ((! $asap || ! $is_spam) && isset($method['uri_bbcode'])) {
1339 $key = 'uri_bbcode';
1340 foreach($pickups as $pickup) {
1341 if (isset($pickup['area'][$key])) {
1342 $sum[$key] += $pickup['area'][$key];
1343 if(isset($method[$key]) &&
1344 $sum[$key] > $method[$key]) {
1345 $is_spam[$key] = TRUE;
1346 if ($asap && $is_spam) break;
1348 if ($asap && $is_spam) break;
1353 // URI: Uniqueness (and removing non-uniques)
1354 if ((! $asap || ! $is_spam) && isset($method['non_uniquri'])) {
1357 foreach (array_keys($pickups) as $key) {
1358 $uris[$key] = uri_pickup_implode($pickups[$key]);
1360 $count = count($uris);
1361 $uris = array_unique($uris);
1362 $sum['non_uniquri'] += $count - count($uris);
1363 if ($sum['non_uniquri'] > $method['non_uniquri']) {
1364 $is_spam['non_uniquri'] = TRUE;
1366 if (! $asap || ! $is_spam) {
1367 foreach (array_diff(array_keys($pickups),
1368 array_keys($uris)) as $remove) {
1369 unset($pickups[$remove]);
1376 if ($asap && $is_spam) return $progress;
1378 // Host: Uniqueness (uniq / non-uniq)
1379 foreach ($pickups as $pickup) $hosts[] = & $pickup['host'];
1380 $hosts = array_unique($hosts);
1381 $sum['uniqhost'] += count($hosts);
1382 if ((! $asap || ! $is_spam) && isset($method['non_uniqhost'])) {
1383 $sum['non_uniqhost'] = $sum['quantity'] - $sum['uniqhost'];
1384 if ($sum['non_uniqhost'] > $method['non_uniqhost']) {
1385 $is_spam['non_uniqhost'] = TRUE;
1390 if ($asap && $is_spam) return $progress;
1392 // URI: Bad host (Separate good/bad hosts from $hosts)
1393 if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
1396 $list = get_blocklist('list');
1397 $blocked = blocklist_distiller($hosts, array_keys($list), $asap);
1398 foreach($list as $key=>$type){
1399 if (! $type) unset($blocked[$key]); // Ignore goodhost etc
1403 if (! empty($blocked)) $is_spam['badhost'] = TRUE;
1409 // Count leaves (A leaf = value that is not an array, or an empty array)
1410 function array_count_leaves($array = array(), $count_empty = FALSE)
1412 if (! is_array($array) || (empty($array) && $count_empty)) return 1;
1416 foreach ($array as $part) {
1417 $count += array_count_leaves($part, $count_empty);
1422 // An array-leaves to a flat array
1423 function array_flat_leaves($array, $unique = TRUE)
1425 if (! is_array($array)) return $array;
1428 foreach(array_keys($array) as $key) {
1429 if (is_array($array[$key])) {
1431 foreach(array_flat_leaves($array[$key]) as $_value) {
1435 $tmp[] = & $array[$key];
1439 return $unique ? array_values(array_unique($tmp)) : $tmp;
1442 // An array() to an array leaf
1443 function array_leaf($array = array('A', 'B', 'C.D'), $stem = FALSE, $edge = TRUE)
1445 if (! is_array($array)) return $array;
1449 foreach($array as $arg) {
1450 if (! is_string($arg) && ! is_int($arg)) continue;
1451 $tmp[$arg] = array();
1453 $tmp = & $tmp[$arg];
1456 $parent[key($parent)] = & $edge;
1458 $parent = key($parent);
1461 return $leaf; // array('A' => array('B' => 'C.D'))
1465 // ---------------------
1468 // Summarize $progress (blocked only)
1469 function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
1472 $tmp = array_keys($progress['is_spam']);
1475 $method = & $progress['method'];
1476 if (isset($progress['sum'])) {
1477 foreach ($progress['sum'] as $key => $value) {
1478 if (isset($method[$key]) && $value) {
1479 $tmp[] = $key . '(' . $value . ')';
1485 return implode(', ', $tmp);
1488 function summarize_detail_badhost($progress = array())
1490 if (! isset($progress['blocked']) || empty($progress['blocked'])) return '';
1494 foreach($progress['blocked'] as $list => $lvalue) {
1495 foreach($lvalue as $group => $gvalue) {
1496 $flat = implode(', ', array_flat_leaves($gvalue));
1497 if ($flat === $group) {
1498 $blocked[$list][] = $flat;
1500 $blocked[$list][$group] = $flat;
1506 // From: 'A-1' => array('ie.to')
1507 // To: 'A-1' => 'ie.to'
1508 foreach($blocked as $list => $lvalue) {
1509 if (is_array($lvalue) &&
1510 count($lvalue) == 1 &&
1511 is_numeric(key($lvalue))) {
1512 $blocked[$list] = current($lvalue);
1516 return var_export_shrink($blocked, TRUE, TRUE);
1519 function summarize_detail_newtral($progress = array())
1521 if (! isset($progress['hosts']) ||
1522 ! is_array($progress['hosts']) ||
1523 empty($progress['hosts'])) return '';
1525 // Generate a responsible $trie
1527 foreach($progress['hosts'] as $value) {
1528 // 'A.foo.bar.example.com'
1529 $resp = whois_responsibility($value); // 'example.com'
1531 // One or more test, or do nothing here
1532 $resp = strval($value);
1535 $rest = rtrim(substr($value, 0, - strlen($resp)), '.'); // 'A.foo.bar'
1537 $trie = array_merge_recursive($trie, array($resp => array($rest => NULL)));
1540 // Format: var_export_shrink() -like output
1542 ksort_by_domain($trie);
1543 foreach(array_keys($trie) as $key) {
1544 ksort_by_domain($trie[$key]);
1545 if (count($trie[$key]) == 1 && key($trie[$key]) == '') {
1546 // Just one 'responsibility.example.com'
1547 $result[] = ' \'' . $key . '\',';
1549 // One subdomain-or-host, or several ones
1551 foreach(array_keys($trie[$key]) as $sub) {
1555 $subs[] = $sub . '.' . $key;
1558 $result[] = ' \'' . $key . '\' => \'' . implode(', ', $subs) . '\',';
1564 implode("\n", $result) . "\n" .
1568 // ksort() by domain
1569 function ksort_by_domain(& $array)
1572 foreach(array_keys($array) as $key) {
1573 $sort[delimiter_reverse($key)] = $key;
1575 ksort($sort, SORT_STRING);
1577 foreach($sort as $key) {
1578 $result[$key] = & $array[$key];
1583 // Check responsibility-root of the FQDN
1584 // 'foo.bar.example.com' => 'example.com' (.com has the last whois for it)
1585 // 'foo.bar.example.au' => 'example.au' (.au has the last whois for it)
1586 // 'foo.bar.example.edu.au' => 'example.edu.au' (.edu.au has the last whois for it)
1587 // 'foo.bar.example.act.edu.au' => 'example.act.edu.au' (.act.edu.au has the last whois for it)
1588 function whois_responsibility($fqdn = 'foo.bar.example.com', $parent = FALSE, $implicit = TRUE)
1592 if ($fqdn === NULL) {
1593 $domain = NULL; // Unset
1596 if (! is_string($fqdn)) return '';
1598 if (is_ip($fqdn)) return $fqdn;
1600 if (! isset($domain)) {
1602 if (file_exists(DOMAIN_INI_FILE)) {
1603 include(DOMAIN_INI_FILE); // Set
1608 $dcursor = & $domain;
1609 $array = array_reverse(explode('.', $fqdn));
1612 if (! isset($array[$i])) break;
1613 $acursor = $array[$i];
1614 if (is_array($dcursor) && isset($dcursor[$acursor])) {
1615 $result[] = & $array[$i];
1616 $dcursor = & $dcursor[$acursor];
1618 if (! $parent && isset($acursor)) {
1619 $result[] = & $array[$i]; // Whois servers must know this subdomain
1626 // Implicit responsibility: Top-Level-Domains must not be yours
1627 // 'bar.foo.something' => 'foo.something'
1628 if ($implicit && count($result) == 1 && count($array) > 1) {
1629 $result[] = & $array[1];
1632 return $result ? implode('.', array_reverse($result)) : '';
1636 // ---------------------
1640 function spam_dispose()
1642 get_blocklist(NULL);
1643 whois_responsibility(NULL);
1646 // Common bahavior for blocking
1647 // NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
1648 function spam_exit($mode = '', $data = array())
1657 echo('<pre>' . "\n");
1658 echo htmlspecialchars(var_export($data, TRUE));
1659 echo('</pre>' . "\n");
1663 if ($exit) exit; // Force exit
1667 // ---------------------
1670 // TODO: Record them
1671 // Simple/fast spam filter ($target: 'a string' or an array())
1672 function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $exitmode = '')
1674 $progress = check_uri_spam($target, $method);
1676 if (empty($progress['is_spam'])) {
1680 // TODO: detect encoding from $target for mbstring functions
1682 // foreach(array_keys($target) as $key) {
1683 // $tmp[strings($key, 0, FALSE, TRUE)] = strings($target[$key], 0, FALSE, TRUE); // Removing "\0" etc
1685 // $target = & $tmp;
1687 pkwk_spamnotify($action, $page, $target, $progress, $method);
1688 spam_exit($exitmode, $progress);
1692 // ---------------------
1693 // PukiWiki original
1695 // Mail to administrator(s)
1696 function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array(), $method = array())
1698 global $notify, $notify_subject;
1700 if (! $notify) return;
1702 $asap = isset($method['asap']);
1704 $summary['ACTION'] = 'Blocked by: ' . summarize_spam_progress($progress, TRUE);
1706 $summary['METRICS'] = summarize_spam_progress($progress);
1709 $tmp = summarize_detail_badhost($progress);
1710 if ($tmp != '') $summary['DETAIL_BADHOST'] = $tmp;
1712 $tmp = summarize_detail_newtral($progress);
1713 if (! $asap && $tmp != '') $summary['DETAIL_NEUTRAL_HOST'] = $tmp;
1715 $summary['COMMENT'] = $action;
1716 $summary['PAGE'] = '[blocked] ' . (is_pagename($page) ? $page : '');
1717 $summary['URI'] = get_script_uri() . '?' . rawurlencode($page);
1718 $summary['USER_AGENT'] = TRUE;
1719 $summary['REMOTE_ADDR'] = TRUE;
1720 pkwk_mail_notify($notify_subject, var_export($target, TRUE), $summary, TRUE);