2 // $Id: spam.php,v 1.198 2007/07/28 12:47:28 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, $input));
42 // ---------------------
45 // Very roughly, shrink the lines of var_export()
46 // NOTE: If the same data exists, it must be corrupted.
47 function var_export_shrink($expression, $return = FALSE, $ignore_numeric_keys = FALSE)
49 $result = var_export($expression, TRUE);
51 $result = preg_replace(
52 // Remove a newline and spaces
53 '# => \n *array \(#', ' => array (',
57 if ($ignore_numeric_keys) {
58 $result =preg_replace(
59 // Remove numeric keys
60 '#^( *)[0-9]+ => #m', '$1',
73 // Reverse $string with specified delimiter
74 function delimiter_reverse($string = 'foo.bar.example.com', $from_delim = '.', $to_delim = '.')
76 if (! is_string($string) || ! is_string($from_delim) || ! is_string($to_delim))
79 // com.example.bar.foo
80 return implode($to_delim, array_reverse(explode($from_delim, $string)));
84 function ksort_by_domain(& $array)
87 foreach(array_keys($array) as $key) {
88 $sort[delimiter_reverse($key)] = $key;
90 ksort($sort, SORT_STRING);
92 foreach($sort as $key) {
93 $result[$key] = & $array[$key];
98 // Roughly strings(1) using PCRE
99 // This function is useful to:
100 // * Reduce the size of data, from removing unprintable binary data
101 // * Detect _bare_strings_ from binary data
103 // http://www.freebsd.org/cgi/man.cgi?query=strings (Man-page of GNU strings)
104 // http://www.pcre.org/pcre.txt
105 // Note: mb_ereg_replace() is one of mbstring extension's functions
106 // and need to init its encoding.
107 function strings($binary = '', $min_len = 4, $ignore_space = FALSE, $multibyte = FALSE)
110 $binary = (is_array($binary) || $binary === TRUE) ? '' : strval($binary);
112 $regex = $ignore_space ?
113 '[^[:graph:] \t\n]+' : // Remove "\0" etc, and readable spaces
114 '[^[:graph:][:space:]]+'; // Preserve readable spaces if possible
116 $binary = $multibyte ?
117 mb_ereg_replace($regex, "\n", $binary) :
118 preg_replace('/' . $regex . '/s', "\n", $binary);
121 $binary = preg_replace(
136 // The last character seems "\n" or not
137 $br = (! empty($binary) && $binary[strlen($binary) - 1] == "\n") ? "\n" : '';
139 $min_len = min(1024, intval($min_len));
140 $regex = '/^.{' . $min_len . ',}/S';
141 $binary = implode("\n", preg_grep($regex, explode("\n", $binary))) . $br;
148 // ---------------------
151 // Count leaves (A leaf = value that is not an array, or an empty array)
152 function array_count_leaves($array = array(), $count_empty = FALSE)
154 if (! is_array($array) || (empty($array) && $count_empty)) return 1;
158 foreach ($array as $part) {
159 $count += array_count_leaves($part, $count_empty);
165 // Similar to PHP array_merge_leaves(), except strictly preserving keys as string
166 function array_merge_leaves($array1, $array2, $sort_keys = TRUE)
169 $is_array1 = is_array($array1);
170 $is_array2 = is_array($array2);
177 } else if ($is_array2) {
180 return $array2; // Not array ($array1 is overwritten)
183 $keys_all = array_merge(array_keys($array1), array_keys($array2));
184 if ($sort_keys) sort($keys_all, SORT_STRING);
187 foreach($keys_all as $key) {
188 $isset1 = isset($array1[$key]);
189 $isset2 = isset($array2[$key]);
190 if ($isset1 && $isset2) {
192 $result[$key] = array_merge_leaves($array1[$key], $array2[$key], $sort_keys);
193 } else if ($isset1) {
194 $result[$key] = & $array1[$key];
196 $result[$key] = & $array2[$key];
202 // An array-leaves to a flat array
203 function array_flat_leaves($array, $unique = TRUE)
205 if (! is_array($array)) return $array;
208 foreach(array_keys($array) as $key) {
209 if (is_array($array[$key])) {
211 foreach(array_flat_leaves($array[$key]) as $_value) {
215 $tmp[] = & $array[$key];
219 return $unique ? array_values(array_unique($tmp)) : $tmp;
222 // $array['something'] => $array['wanted']
223 function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '')
225 if (! is_array($array) || ! is_array($keys)) return FALSE;
227 // Nondestructive test
229 foreach(array_keys($keys) as $from)
230 if (! isset($array[$from]))
233 foreach($keys as $from => $to) {
234 if ($from === $to) continue;
235 if (! $force || isset($array[$from])) {
236 $array[$to] = & $array[$from];
237 unset($array[$from]);
239 $array[$to] = $default;
246 // Remove redundant values from array()
247 function array_unique_recursive($array = array())
249 if (! is_array($array)) return $array;
252 foreach($array as $key => $value){
253 if (is_array($value)) {
254 $array[$key] = array_unique_recursive($value);
256 if (isset($tmp[$value])) {
268 // ---------------------
269 // Part One : Checker
271 // Rough implementation of globbing
273 // USAGE: $regex = '/^' . generate_glob_regex('*.txt', '/') . '$/i';
275 function generate_glob_regex($string = '', $divider = '/')
277 static $from = array(
280 // 22 => '[', // Maybe cause regex compilation error (e.g. '[]')
296 if (! is_string($string)) return '';
298 $string = str_replace($from, $mid, $string); // Hide
299 $string = preg_quote($string, $divider);
300 $string = str_replace($mid, $to, $string); // Unhide
305 // Generate host (FQDN, IPv4, ...) regex
306 // 'localhost' : Matches with 'localhost' only
307 // 'example.org' : Matches with 'example.org' only (See host_normalize() about 'www')
308 // '.example.org' : Matches with ALL FQDN ended with '.example.org'
309 // '*.example.org' : Almost the same of '.example.org' except 'www.example.org'
310 // '10.20.30.40' : Matches with IPv4 address '10.20.30.40' only
311 // [TODO] '192.' : Matches with all IPv4 hosts started with '192.'
312 // TODO: IPv4, CIDR?, IPv6
313 function generate_host_regex($string = '', $divider = '/')
315 if (! is_string($string)) return '';
317 if (mb_strpos($string, '.') === FALSE)
318 return generate_glob_regex($string, $divider);
321 if (is_ip($string)) {
323 return generate_glob_regex($string, $divider);
326 $part = explode('.', $string, 2);
327 if ($part[0] == '') {
328 $part[0] = '(?:.*\.)?'; // And all related FQDN
329 } else if ($part[0] == '*') {
330 $part[0] = '.*\.'; // All subdomains/hosts only
332 return generate_glob_regex($string, $divider);
334 $part[1] = generate_glob_regex($part[1], $divider);
335 return implode('', $part);
339 // Rough hostname checker
341 // TODO: Strict digit, 0x, CIDR, IPv6
342 function is_ip($string = '')
344 if (preg_match('/^' .
345 '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' .
346 '(?:[0-9]{1,3}\.){1,3}' . '$/',
348 return 4; // Seems IPv4(dot-decimal)
350 return 0; // Seems not IP
354 function get_blocklist($list = '')
358 if ($list === NULL) {
359 $regexes = NULL; // Unset
363 if (! isset($regexes)) {
365 if (file_exists(SPAM_INI_FILE)) {
366 $blocklist = array();
367 include(SPAM_INI_FILE);
368 // $blocklist['badhost'] = array(
369 // '*.blogspot.com', // Blog services's subdomains (only)
370 // 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
372 if (isset($blocklist['list'])) {
373 $regexes['list'] = & $blocklist['list'];
376 $blocklist['list'] = array(
381 foreach(array_keys($blocklist['list']) as $_list) {
382 if (! isset($blocklist[$_list])) continue;
383 foreach ($blocklist[$_list] as $key => $value) {
384 if (is_array($value)) {
385 $regexes[$_list][$key] = array();
386 foreach($value as $_key => $_value) {
387 get_blocklist_add($regexes[$_list][$key], $_key, $_value);
390 get_blocklist_add($regexes[$_list], $key, $value);
393 unset($blocklist[$_list]);
399 return $regexes; // ALL
400 } else if (isset($regexes[$list])) {
401 return $regexes[$list];
407 // Subroutine of get_blocklist()
408 function get_blocklist_add(& $array, $key = 0, $value = '*.example.org')
410 if (is_string($key)) {
411 $array[$key] = & $value; // Treat $value as a regex
413 $array[$value] = '/^' . generate_host_regex($value, '/') . '$/i';
417 // Blocklist metrics: Separate $host, to $blocked and not blocked
418 function blocklist_distiller(& $hosts, $keys = array('goodhost', 'badhost'), $asap = FALSE)
420 if (! is_array($hosts)) $hosts = array($hosts);
421 if (! is_array($keys)) $keys = array($keys);
423 $list = get_blocklist('list');
426 foreach($keys as $key){
427 foreach (get_blocklist($key) as $label => $regex) {
428 if (is_array($regex)) {
429 foreach($regex as $_label => $_regex) {
430 $group = preg_grep($_regex, $hosts);
432 $hosts = array_diff($hosts, $group);
433 $blocked[$key][$label][$_label] = $group;
434 if ($asap && $list[$key]) break;
438 $group = preg_grep($regex, $hosts);
440 $hosts = array_diff($hosts, $group);
441 $blocked[$key][$label] = $group;
442 if ($asap && $list[$key]) break;
452 // ---------------------
455 // Default (enabled) methods and thresholds (for content insertion)
456 function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
458 $times = intval($times);
459 $t_area = intval($t_area);
463 'quantity' => 8 * $times, // Allow N URIs
464 'non_uniqhost' => 3 * $times, // Allow N duped (and normalized) Hosts
465 //'non_uniquri'=> 3 * $times, // Allow N duped (and normalized) URIs
468 'area_anchor' => $t_area, // Using <a href> HTML tag
469 'area_bbcode' => $t_area, // Using [url] or [link] BBCode
470 //'uri_anchor' => $t_area, // URI inside <a href> HTML tag
471 //'uri_bbcode' => $t_area, // URI inside [url] or [link] BBCode
476 //'asap' => TRUE, // Quit or return As Soon As Possible
477 'uniqhost' => TRUE, // Show uniq host (at block notification mail)
478 'badhost' => TRUE, // Check badhost
484 // Remove non-$positive values
485 foreach (array_keys($positive) as $key) {
486 if ($positive[$key] < 0) unset($positive[$key]);
489 return $positive + $bool;
492 // Simple/fast spam check
493 function check_uri_spam($target = '', $method = array())
498 // Theme to do => Dummy, optional value, or optional array()
500 //'uniqhost' => TRUE,
501 //'non_uniqhost'=> 3,
502 //'non_uniquri' => 3,
504 //'area_anchor' => 0,
505 //'area_bbcode' => 0,
510 // Theme => Volume found (int)
513 // Flag. If someting defined here,
514 // one or more spam will be included
519 //'category' => array(
529 $sum = & $progress['sum'];
530 $is_spam = & $progress['is_spam'];
531 $progress['method'] = & $method; // Argument
532 $blocked = & $progress['blocked'];
533 $hosts = & $progress['hosts'];
534 $asap = isset($method['asap']);
537 if (! is_array($method) || empty($method)) {
538 $method = check_uri_spam_method();
540 foreach(array_keys($method) as $key) {
541 if (! isset($sum[$key])) $sum[$key] = 0;
543 if (! isset($sum['quantity'])) $sum['quantity'] = 0;
545 if (is_array($target)) {
546 foreach($target as $str) {
547 if (! is_string($str)) continue;
549 $_progress = check_uri_spam($str, $method); // Recurse
552 $_sum = & $_progress['sum'];
553 foreach (array_keys($_sum) as $key) {
554 if (! isset($sum[$key])) {
555 $sum[$key] = & $_sum[$key];
557 $sum[$key] += $_sum[$key];
562 $_is_spam = & $_progress['is_spam'];
563 foreach (array_keys($_is_spam) as $key) {
564 $is_spam[$key] = TRUE;
567 if ($asap && $is_spam) break;
570 $blocked = array_merge_leaves($blocked, $_progress['blocked'], FALSE);
571 $hosts = array_merge_leaves($hosts, $_progress['hosts']), FALSE;
575 $blocked = array_unique_recursive($blocked);
576 $hosts = array_unique_recursive($hosts);
578 // Recount $sum['badhost']
579 $sum['badhost'] = array_count_leaves($blocked);
584 // Area: There's HTML anchor tag
585 if ((! $asap || ! $is_spam) && isset($method['area_anchor'])) {
586 $key = 'area_anchor';
587 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
588 $result = area_pickup($target, array($key => TRUE) + $_asap);
590 $sum[$key] = $result[$key];
591 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
592 $is_spam[$key] = TRUE;
597 // Area: There's 'BBCode' linking tag
598 if ((! $asap || ! $is_spam) && isset($method['area_bbcode'])) {
599 $key = 'area_bbcode';
600 $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
601 $result = area_pickup($target, array($key => TRUE) + $_asap);
603 $sum[$key] = $result[$key];
604 if (isset($method[$key]) && $sum[$key] > $method[$key]) {
605 $is_spam[$key] = TRUE;
611 if ($asap && $is_spam) return $progress;
614 $pickups = uri_pickup_normalize(spam_uri_pickup($target, $method));
617 if (empty($pickups)) return $progress;
619 // URI: Check quantity
620 $sum['quantity'] += count($pickups);
622 if ((! $asap || ! $is_spam) && isset($method['quantity']) &&
623 $sum['quantity'] > $method['quantity']) {
624 $is_spam['quantity'] = TRUE;
627 // URI: used inside HTML anchor tag pair
628 if ((! $asap || ! $is_spam) && isset($method['uri_anchor'])) {
630 foreach($pickups as $pickup) {
631 if (isset($pickup['area'][$key])) {
632 $sum[$key] += $pickup['area'][$key];
633 if(isset($method[$key]) &&
634 $sum[$key] > $method[$key]) {
635 $is_spam[$key] = TRUE;
636 if ($asap && $is_spam) break;
638 if ($asap && $is_spam) break;
643 // URI: used inside 'BBCode' pair
644 if ((! $asap || ! $is_spam) && isset($method['uri_bbcode'])) {
646 foreach($pickups as $pickup) {
647 if (isset($pickup['area'][$key])) {
648 $sum[$key] += $pickup['area'][$key];
649 if(isset($method[$key]) &&
650 $sum[$key] > $method[$key]) {
651 $is_spam[$key] = TRUE;
652 if ($asap && $is_spam) break;
654 if ($asap && $is_spam) break;
659 // URI: Uniqueness (and removing non-uniques)
660 if ((! $asap || ! $is_spam) && isset($method['non_uniquri'])) {
663 foreach (array_keys($pickups) as $key) {
664 $uris[$key] = uri_pickup_implode($pickups[$key]);
666 $count = count($uris);
667 $uris = array_unique($uris);
668 $sum['non_uniquri'] += $count - count($uris);
669 if ($sum['non_uniquri'] > $method['non_uniquri']) {
670 $is_spam['non_uniquri'] = TRUE;
672 if (! $asap || ! $is_spam) {
673 foreach (array_diff(array_keys($pickups),
674 array_keys($uris)) as $remove) {
675 unset($pickups[$remove]);
682 if ($asap && $is_spam) return $progress;
684 // Host: Uniqueness (uniq / non-uniq)
685 foreach ($pickups as $pickup) $hosts[] = & $pickup['host'];
686 $hosts = array_unique($hosts);
687 $sum['uniqhost'] += count($hosts);
688 if ((! $asap || ! $is_spam) && isset($method['non_uniqhost'])) {
689 $sum['non_uniqhost'] = $sum['quantity'] - $sum['uniqhost'];
690 if ($sum['non_uniqhost'] > $method['non_uniqhost']) {
691 $is_spam['non_uniqhost'] = TRUE;
696 if ($asap && $is_spam) return $progress;
698 // URI: Bad host (Separate good/bad hosts from $hosts)
699 if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
702 $list = get_blocklist('list');
703 $blocked = blocklist_distiller($hosts, array_keys($list), $asap);
704 foreach($list as $key=>$type){
705 if (! $type) unset($blocked[$key]); // Ignore goodhost etc
709 if (! empty($blocked)) $is_spam['badhost'] = TRUE;
715 // ---------------------
718 // Summarize $progress (blocked only)
719 function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
722 $tmp = array_keys($progress['is_spam']);
725 $method = & $progress['method'];
726 if (isset($progress['sum'])) {
727 foreach ($progress['sum'] as $key => $value) {
728 if (isset($method[$key]) && $value) {
729 $tmp[] = $key . '(' . $value . ')';
735 return implode(', ', $tmp);
738 function summarize_detail_badhost($progress = array())
740 if (! isset($progress['blocked']) || empty($progress['blocked'])) return '';
744 foreach($progress['blocked'] as $list => $lvalue) {
745 foreach($lvalue as $group => $gvalue) {
746 $flat = implode(', ', array_flat_leaves($gvalue));
747 if ($flat === $group) {
748 $blocked[$list][] = $flat;
750 $blocked[$list][$group] = $flat;
756 // From: 'A-1' => array('ie.to')
757 // To: 'A-1' => 'ie.to'
758 foreach($blocked as $list => $lvalue) {
759 if (is_array($lvalue) &&
760 count($lvalue) == 1 &&
761 is_numeric(key($lvalue))) {
762 $blocked[$list] = current($lvalue);
766 return var_export_shrink($blocked, TRUE, TRUE);
769 function summarize_detail_newtral($progress = array())
771 if (! isset($progress['hosts']) ||
772 ! is_array($progress['hosts']) ||
773 empty($progress['hosts'])) return '';
775 // Generate a responsible $trie
777 foreach($progress['hosts'] as $value) {
778 // 'A.foo.bar.example.com'
779 $resp = whois_responsibility($value); // 'example.com'
781 // One or more test, or do nothing here
782 $resp = strval($value);
785 $rest = rtrim(substr($value, 0, - strlen($resp)), '.'); // 'A.foo.bar'
787 $trie = array_merge_leaves($trie, array($resp => array($rest => NULL)), FALSE);
790 // Format: var_export_shrink() -like output
792 ksort_by_domain($trie);
793 foreach(array_keys($trie) as $key) {
794 ksort_by_domain($trie[$key]);
795 if (count($trie[$key]) == 1 && key($trie[$key]) == '') {
796 // Just one 'responsibility.example.com'
797 $result[] = ' \'' . $key . '\',';
799 // One subdomain-or-host, or several ones
801 foreach(array_keys($trie[$key]) as $sub) {
805 $subs[] = $sub . '.' . $key;
808 $result[] = ' \'' . $key . '\' => \'' . implode(', ', $subs) . '\',';
814 implode("\n", $result) . "\n" .
819 // Check responsibility-root of the FQDN
820 // 'foo.bar.example.com' => 'example.com' (.com has the last whois for it)
821 // 'foo.bar.example.au' => 'example.au' (.au has the last whois for it)
822 // 'foo.bar.example.edu.au' => 'example.edu.au' (.edu.au has the last whois for it)
823 // 'foo.bar.example.act.edu.au' => 'example.act.edu.au' (.act.edu.au has the last whois for it)
824 function whois_responsibility($fqdn = 'foo.bar.example.com', $parent = FALSE, $implicit = TRUE)
828 if ($fqdn === NULL) {
829 $domain = NULL; // Unset
832 if (! is_string($fqdn)) return '';
834 if (is_ip($fqdn)) return $fqdn;
836 if (! isset($domain)) {
838 if (file_exists(DOMAIN_INI_FILE)) {
839 include(DOMAIN_INI_FILE); // Set
844 $dcursor = & $domain;
845 $array = array_reverse(explode('.', $fqdn));
848 if (! isset($array[$i])) break;
849 $acursor = $array[$i];
850 if (is_array($dcursor) && isset($dcursor[$acursor])) {
851 $result[] = & $array[$i];
852 $dcursor = & $dcursor[$acursor];
854 if (! $parent && isset($acursor)) {
855 $result[] = & $array[$i]; // Whois servers must know this subdomain
862 // Implicit responsibility: Top-Level-Domains must not be yours
863 // 'bar.foo.something' => 'foo.something'
864 if ($implicit && count($result) == 1 && count($array) > 1) {
865 $result[] = & $array[1];
868 return $result ? implode('.', array_reverse($result)) : '';
872 // ---------------------
876 function spam_dispose()
879 whois_responsibility(NULL);
882 // Common bahavior for blocking
883 // NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
884 function spam_exit($mode = '', $data = array())
893 echo('<pre>' . "\n");
894 echo htmlspecialchars(var_export($data, TRUE));
895 echo('</pre>' . "\n");
899 if ($exit) exit; // Force exit
903 // ---------------------
907 // Simple/fast spam filter ($target: 'a string' or an array())
908 function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $exitmode = '')
910 $progress = check_uri_spam($target, $method);
912 if (empty($progress['is_spam'])) {
916 // TODO: detect encoding from $target for mbstring functions
918 // foreach(array_keys($target) as $key) {
919 // $tmp[strings($key, 0, FALSE, TRUE)] = strings($target[$key], 0, FALSE, TRUE); // Removing "\0" etc
923 pkwk_spamnotify($action, $page, $target, $progress, $method);
924 spam_exit($exitmode, $progress);
928 // ---------------------
931 // Mail to administrator(s)
932 function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array(), $method = array())
934 global $notify, $notify_subject;
936 if (! $notify) return;
938 $asap = isset($method['asap']);
940 $summary['ACTION'] = 'Blocked by: ' . summarize_spam_progress($progress, TRUE);
942 $summary['METRICS'] = summarize_spam_progress($progress);
945 $tmp = summarize_detail_badhost($progress);
946 if ($tmp != '') $summary['DETAIL_BADHOST'] = $tmp;
948 $tmp = summarize_detail_newtral($progress);
949 if (! $asap && $tmp != '') $summary['DETAIL_NEUTRAL_HOST'] = $tmp;
951 $summary['COMMENT'] = $action;
952 $summary['PAGE'] = '[blocked] ' . (is_pagename($page) ? $page : '');
953 $summary['URI'] = get_script_uri() . '?' . rawurlencode($page);
954 $summary['USER_AGENT'] = TRUE;
955 $summary['REMOTE_ADDR'] = TRUE;
956 pkwk_mail_notify($notify_subject, var_export($target, TRUE), $summary, TRUE);