2 // $Id: spam.php,v 1.108 2007/01/11 16:21:25 henoheno Exp $
3 // Copyright (C) 2006-2007 PukiWiki Developers Team
4 // License: GPL v2 or (at your option) any later version
5 // Functions for Concept-work of spam-uri metrics
6 // (PHP 4 >= 4.3.0): preg_match_all(PREG_OFFSET_CAPTURE): $method['uri_XXX'] related feature
8 if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
10 // ---------------------
13 // (PHP 4 >= 4.2.0): var_export(): mail-reporting and dump related
14 if (! function_exists('var_export')) {
15 function var_export() {
16 return 'var_export() is not found on this server' . "\n";
20 // (PHP 4 >= 4.2.0): preg_grep() enables invert option
21 function preg_grep_invert($pattern = '//', $input = array())
24 if (! isset($invert)) $invert = defined('PREG_GREP_INVERT');
27 return preg_grep($pattern, $input, PREG_GREP_INVERT);
29 $result = preg_grep($pattern, $input);
31 return array_diff($input, preg_grep($pattern, $input));
38 // ---------------------
41 // Return an array of URIs in the $string
42 // [OK] http://nasty.example.org#nasty_string
43 // [OK] http://nasty.example.org:80/foo/xxx#nasty_string/bar
44 // [OK] ftp://nasty.example.org:80/dfsdfs
45 // [OK] ftp://cnn.example.com&story=breaking_news@10.0.0.1/top_story.htm (from RFC3986)
46 function uri_pickup($string = '', $normalize = TRUE,
47 $preserve_rawuri = FALSE, $preserve_chunk = TRUE)
49 // Not available for: IDN(ignored)
52 // scheme://userinfo@host:port/path/or/pathinfo/maybefile.and?query=string#fragment
53 // Refer RFC3986 (Regex below is not strict)
54 '#(\b[a-z][a-z0-9.+-]{1,8})://' . // 1: Scheme
56 '([^\s<>"\'\[\]/\#?@]*)' . // 2: Userinfo (Username)
60 '\[[0-9a-f:.]+\]' . '|' . // IPv6([colon-hex and dot]): RFC2732
61 '(?:[0-9]{1-3}\.){3}[0-9]{1-3}' . '|' . // IPv4(dot-decimal): 001.22.3.44
62 '[a-z0-9.-]+' . // hostname(FQDN) : foo.example.org
64 '(?::([0-9]*))?' . // 4: Port
65 '((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
66 '([^\s<>"\'\[\]\#?]+)?' . // 6: File?
67 '(?:\?([^\s<>"\'\[\]\#]+))?' . // 7: Query string
68 '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 8: Fragment
70 $string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE
74 static $parts = array(
75 1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port',
76 5 => 'path', 6 => 'file', 7 => 'query', 8 => 'fragment'
79 foreach(array_keys($array) as $uri) {
80 $_uri = & $array[$uri];
81 array_rename_keys($_uri, $parts, TRUE, $default);
83 $offset = $_uri['scheme'][1]; // Scheme's offset
84 foreach(array_keys($_uri) as $part) {
85 // Remove offsets for each part
86 $_uri[$part] = & $_uri[$part][0];
90 $_uri['scheme'] = scheme_normalize($_uri['scheme']);
91 if ($_uri['scheme'] === '') {
95 $_uri['host'] = strtolower($_uri['host']);
96 $_uri['port'] = port_normalize($_uri['port'], $_uri['scheme'], FALSE);
97 $_uri['path'] = path_normalize($_uri['path']);
98 if ($preserve_rawuri) $_uri['rawuri'] = & $_uri[0];
101 //$_uri['uri'] = uri_array_implode($_uri);
103 $_uri['uri'] = & $_uri[0]; // Raw
105 unset($_uri[0]); // Matched string itself
106 if (! $preserve_chunk) {
119 // Area offset for area_measure()
120 $_uri['area']['offset'] = $offset;
126 // Destructive normalize of URI array
127 // NOTE: Give me the uri_pickup() result with chunks
128 function uri_array_normalize(& $pickups, $preserve = TRUE)
130 if (! is_array($pickups)) return $pickups;
132 foreach (array_keys($pickups) as $key) {
133 $_key = & $pickups[$key];
134 $_key['path'] = isset($_key['path']) ? strtolower($_key['path']) : '';
135 $_key['file'] = isset($_key['file']) ? file_normalize($_key['file']) : '';
136 $_key['query'] = isset($_key['query']) ? query_normalize(strtolower($_key['query']), TRUE) : '';
137 $_key['fragment'] = (isset($_key['fragment']) && $preserve) ?
138 strtolower($_key['fragment']) : ''; // Just ignore
144 // An URI array => An URI (See uri_pickup())
145 function uri_array_implode($uri = array())
147 if (empty($uri) || ! is_array($uri)) return NULL;
150 if (isset($uri['scheme']) && $uri['scheme'] !== '') {
151 $tmp[] = & $uri['scheme'];
154 if (isset($uri['userinfo']) && $uri['userinfo'] !== '') {
155 $tmp[] = & $uri['userinfo'];
158 if (isset($uri['host']) && $uri['host'] !== '') {
159 $tmp[] = & $uri['host'];
161 if (isset($uri['port']) && $uri['port'] !== '') {
163 $tmp[] = & $uri['port'];
165 if (isset($uri['path']) && $uri['path'] !== '') {
166 $tmp[] = & $uri['path'];
168 if (isset($uri['file']) && $uri['file'] !== '') {
169 $tmp[] = & $uri['file'];
171 if (isset($uri['query']) && $uri['query'] !== '') {
173 $tmp[] = & $uri['query'];
175 if (isset($uri['fragment']) && $uri['fragment'] !== '') {
177 $tmp[] = & $uri['fragment'];
180 return implode('', $tmp);
183 // $array['something'] => $array['wanted']
184 function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '')
186 if (! is_array($array) || ! is_array($keys)) return FALSE;
188 // Nondestructive test
190 foreach(array_keys($keys) as $from)
191 if (! isset($array[$from]))
194 foreach($keys as $from => $to) {
195 if ($from === $to) continue;
196 if (! $force || isset($array[$from])) {
197 $array[$to] = & $array[$from];
198 unset($array[$from]);
200 $array[$to] = $default;
207 // ---------------------
210 // Pickup all of markup areas
211 function area_pickup($string = '', $method = array())
214 if (empty($method)) return $area;
216 // Anchor tag pair by preg_match and preg_match_all()
218 // [OK] <a href= >Good site!</a>
219 // [OK] <a href= "#" >test</a>
220 // [OK] <a href="http://nasty.example.com">visit http://nasty.example.com/</a>
221 // [OK] <a href=\'http://nasty.example.com/\' >discount foobar</a>
222 // [NG] <a href="http://ng.example.com">visit http://ng.example.com _not_ended_
223 $regex = '#<a\b[^>]*\bhref\b[^>]*>.*?</a\b[^>]*(>)#i';
224 if (isset($method['area_anchor'])) {
226 $count = isset($method['asap']) ?
227 preg_match($regex, $string) :
228 preg_match_all($regex, $string, $areas);
229 if (! empty($count)) $area['area_anchor'] = $count;
231 if (isset($method['uri_anchor'])) {
233 preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
234 foreach(array_keys($areas) as $_area) {
235 $areas[$_area] = array(
236 $areas[$_area][0][1], // Area start (<a href>)
237 $areas[$_area][1][1], // Area end (</a>)
240 if (! empty($areas)) $area['uri_anchor'] = $areas;
243 // phpBB's "BBCode" pair by preg_match and preg_match_all()
245 // [OK] [url]http://nasty.example.com/[/url]
246 // [OK] [link]http://nasty.example.com/[/link]
247 // [OK] [url=http://nasty.example.com]visit http://nasty.example.com/[/url]
248 // [OK] [link http://nasty.example.com/]buy something[/link]
249 $regex = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#i';
250 if (isset($method['area_bbcode'])) {
252 $count = isset($method['asap']) ?
253 preg_match($regex, $string) :
254 preg_match_all($regex, $string, $areas, PREG_SET_ORDER);
255 if (! empty($count)) $area['area_bbcode'] = $count;
257 if (isset($method['uri_bbcode'])) {
259 preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
260 foreach(array_keys($areas) as $_area) {
261 $areas[$_area] = array(
262 $areas[$_area][0][1], // Area start ([url])
263 $areas[$_area][2][1], // Area end ([/url])
266 if (! empty($areas)) $area['uri_bbcode'] = $areas;
269 // Various Wiki syntax
270 // [text_or_uri>text_or_uri]
271 // [text_or_uri:text_or_uri]
272 // [text_or_uri|text_or_uri]
273 // [text_or_uri->text_or_uri]
274 // [text_or_uri text_or_uri] // MediaWiki
275 // MediaWiki: [http://nasty.example.com/ visit http://nasty.example.com/]
280 // If in doubt, it's a little doubtful
281 // if (Area => inside <= Area) $brief += -1
282 function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset')
284 if (! is_array($areas) || ! is_array($array)) return;
286 $areas_keys = array_keys($areas);
287 foreach(array_keys($array) as $u_index) {
288 $offset = isset($array[$u_index][$o_key]) ?
289 intval($array[$u_index][$o_key]) : 0;
290 foreach($areas_keys as $a_index) {
291 if (isset($array[$u_index][$a_key])) {
292 $offset_s = intval($areas[$a_index][0]);
293 $offset_e = intval($areas[$a_index][1]);
294 // [Area => inside <= Area]
295 if ($offset_s < $offset && $offset < $offset_e) {
296 $array[$u_index][$a_key] += $belief;
303 // ---------------------
306 // Domain exposure callback (See spam_uri_pickup_preprocess())
307 // http://victim.example.org/?foo+site:nasty.example.com+bar
308 // => http://nasty.example.com/?refer=victim.example.org
309 // NOTE: 'refer=' is not so good for (at this time).
310 // Consider about using IP address of the victim, try to avoid that.
311 function _preg_replace_callback_domain_exposure($matches = array())
315 // Preserve the victim URI as a complicity or ...
316 if (isset($matches[5])) {
318 $matches[1] . '://' . // scheme
319 $matches[2] . '/' . // victim.example.org
320 $matches[3]; // The rest of all (before victim)
324 if (isset($matches[4])) {
326 $matches[1] . '://' . // scheme
327 $matches[4] . // nasty.example.com
328 '/?refer=' . strtolower($matches[2]) . // victim.example.org
335 // Preprocess: rawurldecode() and adding space(s) and something
336 // to detect/count some URIs _if possible_
337 // NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
338 // [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
339 // [OK] http://victim.example.org/http://nasty.example.org
340 function spam_uri_pickup_preprocess($string = '')
342 if (! is_string($string)) return '';
344 $string = rawurldecode($string);
346 // Domain exposure (See _preg_replace_callback_domain_exposure())
347 $string = preg_replace_callback(
349 // Something Google: http://www.google.com/supported_domains
350 '#(http)://((?:[a-z0-9.]+\.)?google\.[a-z]{2,3}(?:\.[a-z]{2})?)/' .
351 '([a-z0-9?=&.%_+-]+)' . // ?query=foo+bar+
352 '\bsite:([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' . // site:nasty.example.com
353 //'()' . // Preserve or remove?
356 '_preg_replace_callback_domain_exposure',
360 // URI exposure (uriuri => uri uri)
361 $string = preg_replace(
363 '#(?<! )(?:https?|ftp):/#i',
364 // '#[a-z][a-z0-9.+-]{1,8}://#i',
365 // '#[a-z][a-z0-9.+-]{1,8}://#i'
374 // Main function of spam-uri pickup
375 function spam_uri_pickup($string = '', $method = array())
377 if (! is_array($method) || empty($method)) {
378 $method = check_uri_spam_method();
381 $string = spam_uri_pickup_preprocess($string);
383 $array = uri_pickup($string);
385 // Area elevation of URIs, for '(especially external)link' intension
386 if (! empty($array)) {
388 if (isset($method['uri_anchor'])) $_method['uri_anchor'] = & $method['uri_anchor'];
389 if (isset($method['uri_bbcode'])) $_method['uri_bbcode'] = & $method['uri_bbcode'];
390 $areas = area_pickup($string, $_method, TRUE);
391 if (! empty($areas)) {
392 $area_shadow = array();
393 foreach (array_keys($array) as $key) {
394 $area_shadow[$key] = & $array[$key]['area'];
395 foreach (array_keys($_method) as $_key) {
396 $area_shadow[$key][$_key] = 0;
399 foreach (array_keys($_method) as $_key) {
400 if (isset($areas[$_key])) {
401 area_measure($areas[$_key], $area_shadow, 1, $_key);
407 // Remove 'offset's for area_measure()
408 foreach(array_keys($array) as $key)
409 unset($array[$key]['area']['offset']);
415 // ---------------------
418 // Scheme normalization: Renaming the schemes
419 // snntp://example.org => nntps://example.org
420 // NOTE: Keep the static lists simple. See also port_normalize().
421 function scheme_normalize($scheme = '', $considerd_harmfull = TRUE)
423 // Abbreviations considerable they don't have link intension
424 static $abbrevs = array(
429 // Alias => normalized
430 static $aliases = array(
440 $scheme = strtolower(trim($scheme));
441 if (isset($abbrevs[$scheme])) {
442 if ($considerd_harmfull) {
443 $scheme = $abbrevs[$scheme];
448 if (isset($aliases[$scheme])) $scheme = $aliases[$scheme];
453 // Port normalization: Suppress the (redundant) default port
454 // HTTP://example.org:80/ => http://example.org/
455 // HTTP://example.org:8080/ => http://example.org:8080/
456 // HTTPS://example.org:443/ => https://example.org/
457 function port_normalize($port, $scheme, $scheme_normalize = TRUE)
459 // Schemes that users _maybe_ want to add protocol-handlers
460 // to their web browsers. (and attackers _maybe_ want to use ...)
461 // Reference: http://www.iana.org/assignments/port-numbers
462 static $array = array(
463 // scheme => default port
490 if ($port === '') return $port;
492 if ($scheme_normalize) $scheme = scheme_normalize($scheme);
493 if (isset($array[$scheme]) && $port == $array[$scheme])
494 $port = ''; // Ignore the defaults
499 // Path normalization
500 // http://example.org => http://example.org/
501 // http://example.org#hoge => http://example.org/#hoge
502 // http://example.org/path/a/b/./c////./d => http://example.org/path/a/b/c/d
503 // http://example.org/path/../../a/../back => http://example.org/back
504 function path_normalize($path = '', $divider = '/', $addroot = TRUE)
506 if (! is_string($path) || $path == '')
507 return $addroot ? $divider : '';
510 $last = ($path[strlen($path) - 1] == $divider) ? $divider : '';
511 $array = explode($divider, $path);
514 foreach(array_keys($array) as $key) {
515 if ($array[$key] == '' || $array[$key] == '.')
520 foreach($array as $value) {
521 if ($value == '..') {
524 array_push($tmp, $value);
529 $path = $addroot ? $divider : '';
530 if (! empty($array)) $path .= implode($divider, $array) . $last;
535 // DirectoryIndex normalize (Destructive and rough)
536 function file_normalize($string = 'index.html.en')
538 static $array = array(
539 'index' => TRUE, // Some system can omit the suffix
541 'index.html' => TRUE,
542 'index.shtml' => TRUE,
545 'index.php3' => TRUE,
546 'index.php4' => TRUE,
547 //'index.pl' => TRUE,
548 //'index.py' => TRUE,
549 //'index.rb' => TRUE,
551 'default.htm' => TRUE,
552 'default.html' => TRUE,
553 'default.asp' => TRUE,
554 'default.aspx' => TRUE,
557 // Content-negothiation filter:
558 // Roughly removing ISO 639 -like
559 // 2-letter suffixes (See RFC3066)
561 if (preg_match('/(.*)\.[a-z][a-z](?:-[a-z][a-z])?$/i', $string, $matches)) {
562 $_string = $matches[1];
564 $_string = & $string;
567 if (isset($array[strtolower($_string)])) {
574 // Sort query-strings if possible (Destructive and rough)
575 // [OK] &&&&f=d&b&d&c&a=0dd => a=0dd&b&c&d&f=d
576 // [OK] nothing==&eg=dummy&eg=padding&eg=foobar => eg=foobar
577 function query_normalize($string = '', $equal = FALSE, $equal_cutempty = TRUE)
579 $array = explode('&', $string);
581 // Remove '&' paddings
582 foreach(array_keys($array) as $key) {
583 if ($array[$key] == '') {
588 // Consider '='-sepalated input and paddings
590 $equals = $not_equals = array();
591 foreach ($array as $part) {
592 if (strpos($part, '=') === FALSE) {
593 $not_equals[] = $part;
595 list($key, $value) = explode('=', $part, 2);
596 $value = ltrim($value, '=');
597 if (! $equal_cutempty || $value != '') {
598 $equals[$key] = $value;
603 $array = & $not_equals;
604 foreach ($equals as $key => $value) {
605 $array[] = $key . '=' . $value;
611 return implode('&', $array);
614 // ---------------------
615 // Part One : Checker
617 function generate_glob_regex($string = '', $divider = '/')
619 static $from = array(
622 // 22 => '[', // Maybe cause regex compilation error (e.g. '[]')
638 if (is_array($string)) {
642 array_map('generate_glob_regex',
644 array_pad(array(), count($string), $divider)
649 $string = str_replace($from, $mid, $string); // Hide
650 $string = preg_quote($string, $divider);
651 $string = str_replace($mid, $to, $string); // Unhide
656 function get_blocklist($list = '')
660 if (! isset($regex)) {
665 $blocklist['badhost'] = array(
666 //'*', // Deny all uri
667 //'10.20.*.*', // 10.20.example.com also matches
668 //'*.blogspot.com', // Blog services subdomains
669 //array('blogspot.com', '*.blogspot.com')
671 foreach ($blocklist['badhost'] as $part) {
672 $_part = is_array($part) ? implode('/', $part) : $part;
673 $regex['badhost'][$_part] = '/^' . generate_glob_regex($part) . '$/i';
678 if (file_exists(SPAM_INI_FILE)) {
679 $blocklist = array();
680 require(SPAM_INI_FILE);
681 foreach(array('goodhost', 'badhost') as $key) {
682 if (! isset($blocklist[$key])) continue;
683 foreach ($blocklist[$key] as $part) {
684 $_part = is_array($part) ? implode('/', $part) : $part;
685 $regex[$key][$_part] = '/^' . generate_glob_regex($part) . '$/i';
693 } else if (isset($regex[$list])) {
694 return $regex[$list];
700 function is_badhost($hosts = array(), $asap = TRUE, & $remains)
703 if (! is_array($hosts)) $hosts = array($hosts);
704 foreach(array_keys($hosts) as $key) {
705 if (! is_string($hosts[$key])) unset($hosts[$key]);
707 if (empty($hosts)) return $result;
709 foreach (get_blocklist('goodhost') as $_regex) {
710 $hosts = preg_grep_invert($_regex, $hosts);
712 if (empty($hosts)) return $result;
715 foreach (get_blocklist('badhost') as $part => $_regex) {
716 $result[$part] = preg_grep($_regex, $hosts);
717 if (empty($result[$part])) {
718 unset($result[$part]);
720 $hosts = array_diff($hosts, $result[$part]);
730 // Default (enabled) methods and thresholds (for content insertion)
731 function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
733 $times = intval($times);
734 $t_area = intval($t_area);
738 'quantity' => 8 * $times, // Allow N URIs
739 'non_uniqhost' => 3 * $times, // Allow N duped (and normalized) Hosts
740 //'non_uniquri'=> 3 * $times, // Allow N duped (and normalized) URIs
743 'area_anchor' => $t_area, // Using <a href> HTML tag
744 'area_bbcode' => $t_area, // Using [url] or [link] BBCode
745 //'uri_anchor' => $t_area, // URI inside <a href> HTML tag
746 //'uri_bbcode' => $t_area, // URI inside [url] or [link] BBCode
751 //'asap' => TRUE, // Quit or return As Soon As Possible
752 'uniqhost' => TRUE, // Show uniq host (at block notification mail)
753 'badhost' => TRUE, // Check badhost
759 // Remove non-$positive values
760 foreach (array_keys($positive) as $key) {
761 if ($positive[$key] < 0) unset($positive[$key]);
764 return $positive + $bool;
767 // Simple/fast spam check
768 function check_uri_spam($target = '', $method = array())
770 if (! is_array($method) || empty($method)) {
771 $method = check_uri_spam_method();
785 'is_spam' => array(),
786 'method' => & $method,
787 'remains' => array(),
789 $sum = & $progress['sum'];
790 $is_spam = & $progress['is_spam'];
791 $remains = & $progress['remains'];
792 $asap = isset($method['asap']);
795 if (is_array($target)) {
796 foreach($target as $str) {
798 $_progress = check_uri_spam($str, $method);
799 $_sum = & $_progress['sum'];
800 $_is_spam = & $_progress['is_spam'];
801 $_remains = & $_progress['remains'];
802 foreach (array_keys($_sum) as $key) {
803 $sum[$key] += $_sum[$key];
805 foreach (array_keys($_is_spam) as $key) {
806 if (is_array($_is_spam[$key])) {
807 // Marge keys (badhost)
808 foreach(array_keys($_is_spam[$key]) as $_key) {
809 if (! isset($is_spam[$key][$_key])) {
810 $is_spam[$key][$_key] = $_is_spam[$key][$_key];
812 $is_spam[$key][$_key] += $_is_spam[$key][$_key];
816 $is_spam[$key] = TRUE;
819 foreach ($_remains as $key=>$value) {
820 foreach ($value as $_key=>$_value) {
821 $remains[$key][$_key] = $_value;
824 if ($asap && $is_spam) break;
829 // Area: There's HTML anchor tag
830 if ((! $asap || ! $is_spam) && isset($method['area_anchor'])) {
831 $key = 'area_anchor';
832 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
833 $result = area_pickup($target, array($key => TRUE) + $_asap);
835 $sum[$key] = $result[$key];
836 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
837 $is_spam[$key] = TRUE;
842 // Area: There's 'BBCode' linking tag
843 if ((! $asap || ! $is_spam) && isset($method['area_bbcode'])) {
844 $key = 'area_bbcode';
845 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
846 $result = area_pickup($target, array($key => TRUE) + $_asap);
848 $sum[$key] = $result[$key];
849 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
850 $is_spam[$key] = TRUE;
856 if ($asap && $is_spam) {
860 $pickups = spam_uri_pickup($target, $method);
861 if (empty($pickups)) {
865 // URI: Check quantity
866 $sum['quantity'] += count($pickups);
868 if ((! $asap || ! $is_spam) && isset($method['quantity']) &&
869 $sum['quantity'] > $method['quantity']) {
870 $is_spam['quantity'] = TRUE;
873 // URI: used inside HTML anchor tag pair
874 if ((! $asap || ! $is_spam) && isset($method['uri_anchor'])) {
876 foreach($pickups as $pickup) {
877 if (isset($pickup['area'][$key])) {
878 $sum[$key] += $pickup['area'][$key];
879 if(isset($method[$key]) &&
880 $sum[$key] > $method[$key]) {
881 $is_spam[$key] = TRUE;
882 if ($asap && $is_spam) break;
884 if ($asap && $is_spam) break;
889 // URI: used inside 'BBCode' pair
890 if ((! $asap || ! $is_spam) && isset($method['uri_bbcode'])) {
892 foreach($pickups as $pickup) {
893 if (isset($pickup['area'][$key])) {
894 $sum[$key] += $pickup['area'][$key];
895 if(isset($method[$key]) &&
896 $sum[$key] > $method[$key]) {
897 $is_spam[$key] = TRUE;
898 if ($asap && $is_spam) break;
900 if ($asap && $is_spam) break;
905 // URI: Uniqueness (and removing non-uniques)
906 if ((! $asap || ! $is_spam) && isset($method['non_uniquri'])) {
908 // Destructive normalize of URIs
909 uri_array_normalize($pickups);
912 foreach (array_keys($pickups) as $key) {
913 $uris[$key] = uri_array_implode($pickups[$key]);
915 $count = count($uris);
916 $uris = array_unique($uris);
917 $sum['non_uniquri'] += $count - count($uris);
918 if ($sum['non_uniquri'] > $method['non_uniquri']) {
919 $is_spam['non_uniquri'] = TRUE;
921 if (! $asap || ! $is_spam) {
922 foreach (array_diff(array_keys($pickups),
923 array_keys($uris)) as $remove) {
924 unset($pickups[$remove]);
931 if ($asap && $is_spam) {
935 // Host: Uniqueness (uniq / non-uniq)
937 foreach ($pickups as $pickup) $hosts[] = & $pickup['host'];
938 $hosts = array_unique($hosts);
939 $sum['uniqhost'] += count($hosts);
940 if ((! $asap || ! $is_spam) && isset($method['non_uniqhost'])) {
941 $sum['non_uniqhost'] = $sum['quantity'] - $sum['uniqhost'];
942 if ($sum['non_uniqhost'] > $method['non_uniqhost']) {
943 $is_spam['non_uniqhost'] = TRUE;
948 if ($asap && $is_spam) {
953 if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
954 $__remains = array();
956 $badhost = is_badhost($hosts, $asap, $__remains);
958 $badhost = is_badhost($hosts, $asap, $__remains);
960 $progress['remains']['badhost'] = array();
961 foreach ($__remains as $value) {
962 $progress['remains']['badhost'][$value] = TRUE;
967 if (! empty($badhost)) {
968 $sum['badhost'] += array_count_leaves($badhost);
969 foreach(array_keys($badhost) as $keys) {
970 $is_spam['badhost'][$keys] =
971 array_count_leaves($badhost[$keys]);
981 function array_count_leaves($array = array(), $count_empty_array = FALSE)
983 if (! is_array($array) || (empty($array) && $count_empty_array))
988 foreach ($array as $part) {
989 $result += array_count_leaves($part, $count_empty_array);
994 // ---------------------
997 // TODO: Don't show unused $method!
998 // Summarize $progress (blocked only)
999 function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
1002 $tmp = array_keys($progress['is_spam']);
1005 $method = & $progress['method'];
1006 if (isset($progress['sum'])) {
1007 foreach ($progress['sum'] as $key => $value) {
1008 if (isset($method[$key])) {
1009 $tmp[] = $key . '(' . $value . ')';
1015 return implode(', ', $tmp);
1018 // ---------------------
1021 // Common bahavior for blocking
1022 // NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
1023 function spam_exit($mode = '', $data = array())
1026 case '': echo("\n"); break;
1028 echo('<pre>' . "\n");
1029 echo htmlspecialchars(var_export($data, TRUE));
1030 echo('</pre>' . "\n");
1039 // ---------------------
1042 // TODO: Record them
1043 // Simple/fast spam filter ($target: 'a string' or an array())
1044 function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $exitmode = '')
1046 $progress = check_uri_spam($target, $method);
1048 if (! empty($progress['is_spam'])) {
1049 // Mail to administrator(s)
1050 pkwk_spamnotify($action, $page, $target, $progress, $method);
1053 spam_exit($exitmode, $progress);
1057 // ---------------------
1058 // PukiWiki original
1060 // Mail to administrator(s)
1061 function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array(), $method = array())
1063 global $notify, $notify_subject;
1065 if (! $notify) return;
1067 $asap = isset($method['asap']);
1069 $summary['ACTION'] = 'Blocked by: ' . summarize_spam_progress($progress, TRUE);
1071 $summary['METRICS'] = summarize_spam_progress($progress);
1073 if (isset($progress['is_spam']['badhost'])) {
1075 foreach($progress['is_spam']['badhost'] as $glob=>$number) {
1076 $badhost[] = $glob . '(' . $number . ')';
1078 $summary['DETAIL_BADHOST'] = implode(', ', $badhost);
1080 if (! $asap && $progress['remains']['badhost']) {
1081 $count = count($progress['remains']['badhost']);
1082 $summary['DETAIL_NEUTRAL_HOST'] = $count .
1085 '/[^, a-z0-9.-]/i', '',
1086 implode(', ', array_keys($progress['remains']['badhost']))
1090 $summary['COMMENT'] = $action;
1091 $summary['PAGE'] = '[blocked] ' . (is_pagename($page) ? $page : '');
1092 $summary['URI'] = get_script_uri() . '?' . rawurlencode($page);
1093 $summary['USER_AGENT'] = TRUE;
1094 $summary['REMOTE_ADDR'] = TRUE;
1095 pkwk_mail_notify($notify_subject, var_export($target, TRUE), $summary);