2 // $Id: spam.php,v 1.98 2007/01/03 10:54:55 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
7 // (PHP 4 >= 4.3.0): preg_match_all(PREG_OFFSET_CAPTURE): $method['uri_XXX'] related feature
9 if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
11 // ---------------------
14 // (PHP 4 >= 4.2.0): var_export(): mail-reporting and dump related
15 if (! function_exists('var_export')) {
16 function var_export() {
17 return 'var_export() is not found' . "\n";
21 // (PHP 4 >= 4.2.0): preg_grep() enables invert option
22 function preg_grep_invert($pattern = '//', $input = array())
25 if (! isset($invert)) $invert = defined('PREG_GREP_INVERT');
28 return preg_grep($pattern, $input, PREG_GREP_INVERT);
30 $result = preg_grep($pattern, $input);
32 return array_diff($input, preg_grep($pattern, $input));
39 // ---------------------
42 // Return an array of URIs in the $string
43 // [OK] http://nasty.example.org#nasty_string
44 // [OK] http://nasty.example.org:80/foo/xxx#nasty_string/bar
45 // [OK] ftp://nasty.example.org:80/dfsdfs
46 // [OK] ftp://cnn.example.com&story=breaking_news@10.0.0.1/top_story.htm (from RFC3986)
47 function uri_pickup($string = '', $normalize = TRUE,
48 $preserve_rawuri = FALSE, $preserve_chunk = TRUE)
50 // Not available for: IDN(ignored)
53 // scheme://userinfo@host:port/path/or/pathinfo/maybefile.and?query=string#fragment
54 // Refer RFC3986 (Regex below is not strict)
55 '#(\b[a-z][a-z0-9.+-]{1,8})://' . // 1: Scheme
57 '([^\s<>"\'\[\]/\#?@]*)' . // 2: Userinfo (Username)
61 '\[[0-9a-f:.]+\]' . '|' . // IPv6([colon-hex and dot]): RFC2732
62 '(?:[0-9]{1-3}\.){3}[0-9]{1-3}' . '|' . // IPv4(dot-decimal): 001.22.3.44
63 '[a-z0-9.-]+' . // hostname(FQDN) : foo.example.org
65 '(?::([0-9]*))?' . // 4: Port
66 '((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
67 '([^\s<>"\'\[\]\#?]+)?' . // 6: File?
68 '(?:\?([^\s<>"\'\[\]\#]+))?' . // 7: Query string
69 '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 8: Fragment
71 $string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE
75 static $parts = array(
76 1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port',
77 5 => 'path', 6 => 'file', 7 => 'query', 8 => 'fragment'
80 foreach(array_keys($array) as $uri) {
81 $_uri = & $array[$uri];
82 array_rename_keys($_uri, $parts, TRUE, $default);
84 $offset = $_uri['scheme'][1]; // Scheme's offset
85 foreach(array_keys($_uri) as $part) {
86 // Remove offsets for each part
87 $_uri[$part] = & $_uri[$part][0];
91 $_uri['scheme'] = scheme_normalize($_uri['scheme']);
92 if ($_uri['scheme'] === '') {
96 $_uri['host'] = strtolower($_uri['host']);
97 $_uri['port'] = port_normalize($_uri['port'], $_uri['scheme'], FALSE);
98 $_uri['path'] = path_normalize($_uri['path']);
99 if ($preserve_rawuri) $_uri['rawuri'] = & $_uri[0];
102 //$_uri['uri'] = uri_array_implode($_uri);
104 $_uri['uri'] = & $_uri[0]; // Raw
106 unset($_uri[0]); // Matched string itself
107 if (! $preserve_chunk) {
120 // Area offset for area_measure()
121 $_uri['area']['offset'] = $offset;
127 // Destructive normalize of URI array
128 // NOTE: Give me the uri_pickup() result with chunks
129 function uri_array_normalize(& $pickups, $preserve = TRUE)
131 if (! is_array($pickups)) return $pickups;
133 foreach (array_keys($pickups) as $key) {
134 $_key = & $pickups[$key];
135 $_key['path'] = isset($_key['path']) ? strtolower($_key['path']) : '';
136 $_key['file'] = isset($_key['file']) ? file_normalize($_key['file']) : '';
137 $_key['query'] = isset($_key['query']) ? query_normalize(strtolower($_key['query']), TRUE) : '';
138 $_key['fragment'] = (isset($_key['fragment']) && $preserve) ?
139 strtolower($_key['fragment']) : ''; // Just ignore
145 // An URI array => An URI (See uri_pickup())
146 function uri_array_implode($uri = array())
148 if (empty($uri) || ! is_array($uri)) return NULL;
151 if (isset($uri['scheme']) && $uri['scheme'] !== '') {
152 $tmp[] = & $uri['scheme'];
155 if (isset($uri['userinfo']) && $uri['userinfo'] !== '') {
156 $tmp[] = & $uri['userinfo'];
159 if (isset($uri['host']) && $uri['host'] !== '') {
160 $tmp[] = & $uri['host'];
162 if (isset($uri['port']) && $uri['port'] !== '') {
164 $tmp[] = & $uri['port'];
166 if (isset($uri['path']) && $uri['path'] !== '') {
167 $tmp[] = & $uri['path'];
169 if (isset($uri['file']) && $uri['file'] !== '') {
170 $tmp[] = & $uri['file'];
172 if (isset($uri['query']) && $uri['query'] !== '') {
174 $tmp[] = & $uri['query'];
176 if (isset($uri['fragment']) && $uri['fragment'] !== '') {
178 $tmp[] = & $uri['fragment'];
181 return implode('', $tmp);
184 // $array['something'] => $array['wanted']
185 function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '')
187 if (! is_array($array) || ! is_array($keys)) return FALSE;
189 // Nondestructive test
191 foreach(array_keys($keys) as $from)
192 if (! isset($array[$from]))
195 foreach($keys as $from => $to) {
196 if ($from === $to) continue;
197 if (! $force || isset($array[$from])) {
198 $array[$to] = & $array[$from];
199 unset($array[$from]);
201 $array[$to] = $default;
208 // ---------------------
211 // Pickup all of markup areas
212 function area_pickup($string = '', $method = array())
215 if (empty($method)) return $area;
217 // Anchor tag pair by preg_match and preg_match_all()
219 // [OK] <a href= >Good site!</a>
220 // [OK] <a href= "#" >test</a>
221 // [OK] <a href="http://nasty.example.com">visit http://nasty.example.com/</a>
222 // [OK] <a href=\'http://nasty.example.com/\' >discount foobar</a>
223 // [NG] <a href="http://ng.example.com">visit http://ng.example.com _not_ended_
224 $regex = '#<a\b[^>]*\bhref\b[^>]*>.*?</a\b[^>]*(>)#i';
225 if (isset($method['area_anchor'])) {
227 $count = isset($method['asap']) ?
228 preg_match($regex, $string) :
229 preg_match_all($regex, $string, $areas);
230 if (! empty($count)) $area['area_anchor'] = $count;
232 if (isset($method['uri_anchor'])) {
234 preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
235 foreach(array_keys($areas) as $_area) {
236 $areas[$_area] = array(
237 $areas[$_area][0][1], // Area start (<a href>)
238 $areas[$_area][1][1], // Area end (</a>)
241 if (! empty($areas)) $area['uri_anchor'] = $areas;
244 // phpBB's "BBCode" pair by preg_match and preg_match_all()
246 // [OK] [url]http://nasty.example.com/[/url]
247 // [OK] [link]http://nasty.example.com/[/link]
248 // [OK] [url=http://nasty.example.com]visit http://nasty.example.com/[/url]
249 // [OK] [link http://nasty.example.com/]buy something[/link]
250 $regex = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#i';
251 if (isset($method['area_bbcode'])) {
253 $count = isset($method['asap']) ?
254 preg_match($regex, $string) :
255 preg_match_all($regex, $string, $areas, PREG_SET_ORDER);
256 if (! empty($count)) $area['area_bbcode'] = $count;
258 if (isset($method['uri_bbcode'])) {
260 preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
261 foreach(array_keys($areas) as $_area) {
262 $areas[$_area] = array(
263 $areas[$_area][0][1], // Area start ([url])
264 $areas[$_area][2][1], // Area end ([/url])
267 if (! empty($areas)) $area['uri_bbcode'] = $areas;
270 // Various Wiki syntax
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]
275 // [text_or_uri text_or_uri] // MediaWiki
276 // MediaWiki: [http://nasty.example.com/ visit http://nasty.example.com/]
281 // If in doubt, it's a little doubtful
282 // if (Area => inside <= Area) $brief += -1
283 function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset')
285 if (! is_array($areas) || ! is_array($array)) return;
287 $areas_keys = array_keys($areas);
288 foreach(array_keys($array) as $u_index) {
289 $offset = isset($array[$u_index][$o_key]) ?
290 intval($array[$u_index][$o_key]) : 0;
291 foreach($areas_keys as $a_index) {
292 if (isset($array[$u_index][$a_key])) {
293 $offset_s = intval($areas[$a_index][0]);
294 $offset_e = intval($areas[$a_index][1]);
295 // [Area => inside <= Area]
296 if ($offset_s < $offset && $offset < $offset_e) {
297 $array[$u_index][$a_key] += $belief;
304 // ---------------------
307 // Domain exposure callback (See spam_uri_pickup_preprocess())
308 // http://victim.example.org/?foo+site:nasty.example.com+bar
309 // => http://nasty.example.com/?refer=victim.example.org
310 // NOTE: 'refer=' is not so good for (at this time).
311 // Consider about using IP address of the victim, try to avoid that.
312 function _preg_replace_callback_domain_exposure($matches = array())
316 // Preserve the victim URI as a complicity or ...
317 if (isset($matches[5])) {
319 $matches[1] . '://' . // scheme
320 $matches[2] . '/' . // victim.example.org
321 $matches[3]; // The rest of all (before victim)
325 if (isset($matches[4])) {
327 $matches[1] . '://' . // scheme
328 $matches[4] . // nasty.example.com
329 '/?refer=' . strtolower($matches[2]) . // victim.example.org
336 // Preprocess: rawurldecode() and adding space(s) and something
337 // to detect/count some URIs _if possible_
338 // NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
339 // [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
340 // [OK] http://victim.example.org/http://nasty.example.org
341 function spam_uri_pickup_preprocess($string = '')
343 if (! is_string($string)) return '';
345 $string = rawurldecode($string);
347 // Domain exposure (See _preg_replace_callback_domain_exposure())
348 $string = preg_replace_callback(
350 // Something Google: http://www.google.com/supported_domains
351 '#(http)://([a-z0-9.]+\.google\.[a-z]{2,3}(?:\.[a-z]{2})?)/' .
352 '([a-z0-9?=&.%_+-]+)' . // ?query=foo+
353 '\bsite:([a-z0-9.%_-]+)' . // site:nasty.example.com
354 //'()' . // Preserve or remove?
357 '_preg_replace_callback_domain_exposure',
361 // URI exposure (uriuri => uri uri)
362 $string = preg_replace(
364 '#(?<! )(?:https?|ftp):/#i',
365 // '#[a-z][a-z0-9.+-]{1,8}://#i',
366 // '#[a-z][a-z0-9.+-]{1,8}://#i'
375 // Main function of spam-uri pickup
376 function spam_uri_pickup($string = '', $method = array())
378 if (! is_array($method) || empty($method)) {
379 $method = check_uri_spam_method();
382 $string = spam_uri_pickup_preprocess($string);
384 $array = uri_pickup($string);
386 // Area elevation of URIs, for '(especially external)link' intension
387 if (! empty($array)) {
389 if (isset($method['uri_anchor'])) $_method['uri_anchor'] = & $method['uri_anchor'];
390 if (isset($method['uri_bbcode'])) $_method['uri_bbcode'] = & $method['uri_bbcode'];
391 $areas = area_pickup($string, $_method, TRUE);
392 if (! empty($areas)) {
393 $area_shadow = array();
394 foreach (array_keys($array) as $key) {
395 $area_shadow[$key] = & $array[$key]['area'];
396 foreach (array_keys($_method) as $_key) {
397 $area_shadow[$key][$_key] = 0;
400 foreach (array_keys($_method) as $_key) {
401 if (isset($areas[$_key])) {
402 area_measure($areas[$_key], $area_shadow, 1, $_key);
408 // Remove 'offset's for area_measure()
409 foreach(array_keys($array) as $key)
410 unset($array[$key]['area']['offset']);
416 // ---------------------
419 // Scheme normalization: Renaming the schemes
420 // snntp://example.org => nntps://example.org
421 // NOTE: Keep the static lists simple. See also port_normalize().
422 function scheme_normalize($scheme = '', $considerd_harmfull = TRUE)
424 // Abbreviations considerable they don't have link intension
425 static $abbrevs = array(
430 // Alias => normalized
431 static $aliases = array(
441 $scheme = strtolower(trim($scheme));
442 if (isset($abbrevs[$scheme])) {
443 if ($considerd_harmfull) {
444 $scheme = $abbrevs[$scheme];
449 if (isset($aliases[$scheme])) $scheme = $aliases[$scheme];
454 // Port normalization: Suppress the (redundant) default port
455 // HTTP://example.org:80/ => http://example.org/
456 // HTTP://example.org:8080/ => http://example.org:8080/
457 // HTTPS://example.org:443/ => https://example.org/
458 function port_normalize($port, $scheme, $scheme_normalize = TRUE)
460 // Schemes that users _maybe_ want to add protocol-handlers
461 // to their web browsers. (and attackers _maybe_ want to use ...)
462 // Reference: http://www.iana.org/assignments/port-numbers
463 static $array = array(
464 // scheme => default port
491 if ($port === '') return $port;
493 if ($scheme_normalize) $scheme = scheme_normalize($scheme);
494 if (isset($array[$scheme]) && $port == $array[$scheme])
495 $port = ''; // Ignore the defaults
500 // Path normalization
501 // http://example.org => http://example.org/
502 // http://example.org#hoge => http://example.org/#hoge
503 // http://example.org/path/a/b/./c////./d => http://example.org/path/a/b/c/d
504 // http://example.org/path/../../a/../back => http://example.org/back
505 function path_normalize($path = '', $divider = '/', $addroot = TRUE)
507 if (! is_string($path) || $path == '')
508 return $addroot ? $divider : '';
511 $last = ($path[strlen($path) - 1] == $divider) ? $divider : '';
512 $array = explode($divider, $path);
515 foreach(array_keys($array) as $key) {
516 if ($array[$key] == '' || $array[$key] == '.')
521 foreach($array as $value) {
522 if ($value == '..') {
525 array_push($tmp, $value);
530 $path = $addroot ? $divider : '';
531 if (! empty($array)) $path .= implode($divider, $array) . $last;
536 // DirectoryIndex normalize (Destructive and rough)
537 function file_normalize($string = 'index.html.en')
539 static $array = array(
540 'index' => TRUE, // Some system can omit the suffix
542 'index.html' => TRUE,
543 'index.shtml' => TRUE,
546 'index.php3' => TRUE,
547 'index.php4' => TRUE,
548 //'index.pl' => TRUE,
549 //'index.py' => TRUE,
550 //'index.rb' => TRUE,
552 'default.htm' => TRUE,
553 'default.html' => TRUE,
554 'default.asp' => TRUE,
555 'default.aspx' => TRUE,
558 // Content-negothiation filter:
559 // Roughly removing ISO 639 -like
560 // 2-letter suffixes (See RFC3066)
562 if (preg_match('/(.*)\.[a-z][a-z](?:-[a-z][a-z])?$/i', $string, $matches)) {
563 $_string = $matches[1];
565 $_string = & $string;
568 if (isset($array[strtolower($_string)])) {
575 // Sort query-strings if possible (Destructive and rough)
576 // [OK] &&&&f=d&b&d&c&a=0dd => a=0dd&b&c&d&f=d
577 // [OK] nothing==&eg=dummy&eg=padding&eg=foobar => eg=foobar
578 function query_normalize($string = '', $equal = FALSE, $equal_cutempty = TRUE)
580 $array = explode('&', $string);
582 // Remove '&' paddings
583 foreach(array_keys($array) as $key) {
584 if ($array[$key] == '') {
589 // Consider '='-sepalated input and paddings
591 $equals = $not_equals = array();
592 foreach ($array as $part) {
593 if (strpos($part, '=') === FALSE) {
594 $not_equals[] = $part;
596 list($key, $value) = explode('=', $part, 2);
597 $value = ltrim($value, '=');
598 if (! $equal_cutempty || $value != '') {
599 $equals[$key] = $value;
604 $array = & $not_equals;
605 foreach ($equals as $key => $value) {
606 $array[] = $key . '=' . $value;
612 return implode('&', $array);
615 // ---------------------
616 // Part One : Checker
618 function generate_glob_regex($string = '', $divider = '/')
620 static $from = array(
623 // 22 => '[', // Maybe cause regex compilation error (e.g. '[]')
639 if (is_array($string)) {
643 array_map('generate_glob_regex',
645 array_pad(array(), count($string), $divider)
650 $string = str_replace($from, $mid, $string); // Hide
651 $string = preg_quote($string, $divider);
652 $string = str_replace($mid, $to, $string); // Unhide
657 function get_blocklist($list = '')
661 if (! isset($regex)) {
666 $blocklist['badhost'] = array(
667 //'*', // Deny all uri
668 //'10.20.*.*', // 10.20.example.com also matches
669 //'*.blogspot.com', // Blog services subdomains
670 //array('blogspot.com', '*.blogspot.com')
672 foreach ($blocklist['badhost'] as $part) {
673 $_part = is_array($part) ? implode('/', $part) : $part;
674 $regex['badhost'][$_part] = '/^' . generate_glob_regex($part) . '$/i';
679 if (file_exists(SPAM_INI_FILE)) {
680 $blocklist = array();
681 require(SPAM_INI_FILE);
682 foreach(array('goodhost', 'badhost') as $key) {
683 if (! isset($blocklist[$key])) continue;
684 foreach ($blocklist[$key] as $part) {
685 $_part = is_array($part) ? implode('/', $part) : $part;
686 $regex[$key][$_part] = '/^' . generate_glob_regex($part) . '$/i';
694 } else if (isset($regex[$list])) {
695 return $regex[$list];
701 function is_badhost($hosts = array(), $asap = TRUE, & $remains)
704 if (! is_array($hosts)) $hosts = array($hosts);
705 foreach(array_keys($hosts) as $key) {
706 if (! is_string($hosts[$key])) unset($hosts[$key]);
708 if (empty($hosts)) return $result;
710 foreach (get_blocklist('goodhost') as $_regex) {
711 $hosts = preg_grep_invert($_regex, $hosts);
713 if (empty($hosts)) return $result;
716 foreach (get_blocklist('badhost') as $part => $_regex) {
717 $result[$part] = preg_grep($_regex, $hosts);
718 if (empty($result[$part])) {
719 unset($result[$part]);
721 $hosts = array_diff($hosts, $result[$part]);
731 // Default (enabled) methods and thresholds (for content insertion)
732 function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
734 $times = intval($times);
735 $t_area = intval($t_area);
739 'quantity' => 8 * $times, // Allow N URIs
740 'non_uniqhost' => 3 * $times, // Allow N duped (and normalized) Hosts
741 'non_uniquri' => 0 * $times, // Allow N duped (and normalized) URIs
744 'area_anchor' => $t_area, // Using <a href> HTML tag
745 'area_bbcode' => $t_area, // Using [url] or [link] BBCode
746 //'uri_anchor' => $t_area, // URI inside <a href> HTML tag
747 //'uri_bbcode' => $t_area, // URI inside [url] or [link] BBCode
752 //'asap' => TRUE, // Quit or return As Soon As Possible
753 'uniqhost' => TRUE, // Show uniq host (at block notification mail)
754 'badhost' => TRUE, // Check badhost
760 // Remove non-$positive values
761 foreach (array_keys($positive) as $key) {
762 if ($positive[$key] < 0) unset($positive[$key]);
765 return $positive + $bool;
768 // Simple/fast spam check
769 function check_uri_spam($target = '', $method = array())
771 if (! is_array($method) || empty($method)) {
772 $method = check_uri_spam_method();
786 'is_spam' => array(),
787 'method' => & $method,
788 'remains' => array(),
790 $sum = & $progress['sum'];
791 $is_spam = & $progress['is_spam'];
792 $remains = & $progress['remains'];
793 $asap = isset($method['asap']);
796 if (is_array($target)) {
797 foreach($target as $str) {
799 $_progress = check_uri_spam($str, $method);
800 $_sum = & $_progress['sum'];
801 $_is_spam = & $_progress['is_spam'];
802 $_remains = & $_progress['remains'];
803 foreach (array_keys($_sum) as $key) {
804 $sum[$key] += $_sum[$key];
806 foreach (array_keys($_is_spam) as $key) {
807 if (is_array($_is_spam[$key])) {
808 // Marge keys (badhost)
809 foreach(array_keys($_is_spam[$key]) as $_key) {
810 if (! isset($is_spam[$key][$_key])) {
811 $is_spam[$key][$_key] = $_is_spam[$key][$_key];
813 $is_spam[$key][$_key] += $_is_spam[$key][$_key];
817 $is_spam[$key] = TRUE;
820 foreach ($_remains as $key=>$value) {
821 foreach ($value as $_key=>$_value) {
822 $remains[$key][$_key] = $_value;
825 if ($asap && $is_spam) break;
830 // Area: There's HTML anchor tag
831 if ((! $asap || ! $is_spam) && isset($method['area_anchor'])) {
832 $key = 'area_anchor';
833 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
834 $result = area_pickup($target, array($key => TRUE) + $_asap);
836 $sum[$key] = $result[$key];
837 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
838 $is_spam[$key] = TRUE;
843 // Area: There's 'BBCode' linking tag
844 if ((! $asap || ! $is_spam) && isset($method['area_bbcode'])) {
845 $key = 'area_bbcode';
846 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
847 $result = area_pickup($target, array($key => TRUE) + $_asap);
849 $sum[$key] = $result[$key];
850 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
851 $is_spam[$key] = TRUE;
857 if ($asap && $is_spam) {
861 $pickups = spam_uri_pickup($target, $method);
862 if (empty($pickups)) {
866 // URI: Check quantity
867 $sum['quantity'] += count($pickups);
869 if ((! $asap || ! $is_spam) && isset($method['quantity']) &&
870 $sum['quantity'] > $method['quantity']) {
871 $is_spam['quantity'] = TRUE;
874 // URI: used inside HTML anchor tag pair
875 if ((! $asap || ! $is_spam) && isset($method['uri_anchor'])) {
877 foreach($pickups as $pickup) {
878 if (isset($pickup['area'][$key])) {
879 $sum[$key] += $pickup['area'][$key];
880 if(isset($method[$key]) &&
881 $sum[$key] > $method[$key]) {
882 $is_spam[$key] = TRUE;
883 if ($asap && $is_spam) break;
885 if ($asap && $is_spam) break;
890 // URI: used inside 'BBCode' pair
891 if ((! $asap || ! $is_spam) && isset($method['uri_bbcode'])) {
893 foreach($pickups as $pickup) {
894 if (isset($pickup['area'][$key])) {
895 $sum[$key] += $pickup['area'][$key];
896 if(isset($method[$key]) &&
897 $sum[$key] > $method[$key]) {
898 $is_spam[$key] = TRUE;
899 if ($asap && $is_spam) break;
901 if ($asap && $is_spam) break;
906 // URI: Uniqueness (and removing non-uniques)
907 if ((! $asap || ! $is_spam) && isset($method['non_uniquri'])) {
909 // Destructive normalize of URIs
910 uri_array_normalize($pickups);
913 foreach (array_keys($pickups) as $key) {
914 $uris[$key] = uri_array_implode($pickups[$key]);
916 $count = count($uris);
917 $uris = array_unique($uris);
918 $sum['non_uniquri'] += $count - count($uris);
919 if ($sum['non_uniquri'] > $method['non_uniquri']) {
920 $is_spam['non_uniquri'] = TRUE;
922 if (! $asap || ! $is_spam) {
923 foreach (array_diff(array_keys($pickups),
924 array_keys($uris)) as $remove) {
925 unset($pickups[$remove]);
932 if ($asap && $is_spam) {
936 // Host: Uniqueness (uniq / non-uniq)
938 foreach ($pickups as $pickup) $hosts[] = & $pickup['host'];
939 $hosts = array_unique($hosts);
940 $sum['uniqhost'] += count($hosts);
941 if ((! $asap || ! $is_spam) && isset($method['non_uniqhost'])) {
942 $sum['non_uniqhost'] = $sum['quantity'] - $sum['uniqhost'];
943 if ($sum['non_uniqhost'] > $method['non_uniqhost']) {
944 $is_spam['non_uniqhost'] = TRUE;
949 if ($asap && $is_spam) {
954 if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
956 $badhost = is_badhost($hosts, $asap);
958 $__remains = array();
959 $badhost = is_badhost($hosts, $asap, $__remains);
961 $progress['remains']['badhost'] = array();
962 foreach ($__remains as $value) {
963 $progress['remains']['badhost'][$value] = TRUE;
968 if (! empty($badhost)) {
969 $sum['badhost'] += array_count_leaves($badhost);
970 foreach(array_keys($badhost) as $keys) {
971 $is_spam['badhost'][$keys] =
972 array_count_leaves($badhost[$keys]);
982 function array_count_leaves($array = array(), $count_empty_array = FALSE)
984 if (! is_array($array) || (empty($array) && $count_empty_array))
989 foreach ($array as $part) {
990 $result += array_count_leaves($part, $count_empty_array);
995 // ---------------------
998 // TODO: Don't show unused $method!
999 // Summarize $progress (blocked only)
1000 function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
1003 $tmp = array_keys($progress['is_spam']);
1006 $method = & $progress['method'];
1007 if (isset($progress['sum'])) {
1008 foreach ($progress['sum'] as $key => $value) {
1009 if (isset($method[$key])) {
1010 $tmp[] = $key . '(' . $value . ')';
1016 return implode(', ', $tmp);
1019 // ---------------------
1022 // Common bahavior for blocking
1023 // NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
1024 function spam_exit($mode = '', $data = array())
1027 case '': echo("\n"); break;
1029 echo('<pre>' . "\n");
1030 echo htmlspecialchars(var_export($data, TRUE));
1031 echo('</pre>' . "\n");
1040 // ---------------------
1043 // TODO: Record them
1044 // Simple/fast spam filter ($target: 'a string' or an array())
1045 function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $exitmode = '')
1047 $progress = check_uri_spam($target, $method);
1049 if (! empty($progress['is_spam'])) {
1050 // Mail to administrator(s)
1051 pkwk_spamnotify($action, $page, $target, $progress, $method);
1054 spam_exit($exitmode, $progress);
1058 // ---------------------
1059 // PukiWiki original
1061 // Mail to administrator(s)
1062 function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array(), $method = array())
1064 global $notify, $notify_subject;
1066 if (! $notify) return;
1068 $asap = isset($method['asap']);
1070 $summary['ACTION'] = 'Blocked by: ' . summarize_spam_progress($progress, TRUE);
1072 $summary['METRICS'] = summarize_spam_progress($progress);
1074 if (isset($progress['is_spam']['badhost'])) {
1076 foreach($progress['is_spam']['badhost'] as $glob=>$number) {
1077 $badhost[] = $glob . '(' . $number . ')';
1079 $summary['DETAIL_BADHOST'] = implode(', ', $badhost);
1081 if (! $asap && $progress['remains']['badhost']) {
1082 $summary['DETAIL_NEUTRAL_HOST'] =
1084 '/[^, a-z0-9.-]/i', '',
1085 implode(', ', array_keys($progress['remains']['badhost']))
1088 $summary['COMMENT'] = $action;
1089 $summary['PAGE'] = '[blocked] ' . (is_pagename($page) ? $page : '');
1090 $summary['URI'] = get_script_uri() . '?' . rawurlencode($page);
1091 $summary['USER_AGENT'] = TRUE;
1092 $summary['REMOTE_ADDR'] = TRUE;
1093 pkwk_mail_notify($notify_subject, var_export($target, TRUE), $summary);