<?php
-// $Id: spam.php,v 1.157 2007/05/05 13:58:39 henoheno Exp $
+// $Id: spam.php,v 1.195 2007/06/29 15:35:53 henoheno Exp $
// Copyright (C) 2006-2007 PukiWiki Developers Team
// License: GPL v2 or (at your option) any later version
//
//
// (PHP 4 >= 4.3.0): preg_match_all(PREG_OFFSET_CAPTURE): $method['uri_XXX'] related feature
-if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
+if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
+if (! defined('DOMAIN_INI_FILE')) define('DOMAIN_INI_FILE', 'domain.ini.php');
// ---------------------
// Compat etc
// NOTE: If the same data exists, it must be corrupted.
function var_export_shrink($expression, $return = FALSE, $ignore_numeric_keys = FALSE)
{
- $result =preg_replace(
+ $result = var_export($expression, TRUE);
+
+ $result = preg_replace(
// Remove a newline and spaces
'# => \n *array \(#', ' => array (',
- var_export($expression, TRUE)
+ $result
);
if ($ignore_numeric_keys) {
// References:
// http://www.freebsd.org/cgi/man.cgi?query=strings (Man-page of GNU strings)
// http://www.pcre.org/pcre.txt
-function strings($binary = '', $min_len = 4, $ignore_space = FALSE)
+// Note: mb_ereg_replace() is one of mbstring extension's functions
+// and need to init its encoding.
+function strings($binary = '', $min_len = 4, $ignore_space = FALSE, $multibyte = FALSE)
{
+ // String only
+ $binary = (is_array($binary) || $binary === TRUE) ? '' : strval($binary);
+
+ $regex = $ignore_space ?
+ '[^[:graph:] \t\n]+' : // Remove "\0" etc, and readable spaces
+ '[^[:graph:][:space:]]+'; // Preserve readable spaces if possible
+
+ $binary = $multibyte ?
+ mb_ereg_replace($regex, "\n", $binary) :
+ preg_replace('/' . $regex . '/s', "\n", $binary);
+
if ($ignore_space) {
$binary = preg_replace(
array(
- '/(?:[^[:graph:] \t\n]|[\r])+/s',
'/[ \t]{2,}/',
'/^[ \t]/m',
'/[ \t]$/m',
),
array(
- "\n",
' ',
'',
''
),
$binary);
- } else {
- $binary = preg_replace('/(?:[^[:graph:][:space:]]|[\r])+/s', "\n", $binary);
}
if ($min_len > 1) {
+ // The last character seems "\n" or not
+ $br = (! empty($binary) && $binary[strlen($binary) - 1] == "\n") ? "\n" : '';
+
$min_len = min(1024, intval($min_len));
- $binary =
- implode("\n",
- preg_grep('/^.{' . $min_len . ',}/S',
- explode("\n", $binary)
- )
- );
+ $regex = '/^.{' . $min_len . ',}/S';
+ $binary = implode("\n", preg_grep($regex, explode("\n", $binary))) . $br;
}
return $binary;
preg_match_all(
// scheme://userinfo@host:port/path/or/pathinfo/maybefile.and?query=string#fragment
// Refer RFC3986 (Regex below is not strict)
- '#(\b[a-z][a-z0-9.+-]{1,8}):/+' . // 1: Scheme
+ '#(\b[a-z][a-z0-9.+-]{1,8}):[/\\\]+' . // 1: Scheme
'(?:' .
'([^\s<>"\'\[\]/\#?@]*)' . // 2: Userinfo (Username)
'@)?' .
// 3: Host
'\[[0-9a-f:.]+\]' . '|' . // IPv6([colon-hex and dot]): RFC2732
'(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' . // IPv4(dot-decimal): 001.22.3.44
- '[a-z0-9.-]+' . // hostname(FQDN) : foo.example.org
+ '[a-z0-9_-][a-z0-9_.-]+[a-z0-9_-]' . // hostname(FQDN) : foo.example.org
')' .
'(?::([0-9]*))?' . // 4: Port
'((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
// ---------------------
// Spam-uri pickup
-// Domain exposure callback (See spam_uri_pickup_preprocess())
+// Preprocess: Removing uninterest part for URI detection
+function spam_uri_removing_hocus_pocus($binary = '', $method = array())
+{
+ $length = 4 ; // 'http'(1) and '://'(2) and 'fqdn'(1)
+ if (is_array($method)) {
+ // '<a'(2) or 'href='(5) or '>'(1) or '</a>'(4)
+ // '[uri'(4) or ']'(1) or '[/uri]'(6)
+ if (isset($method['area_anchor']) || isset($method['uri_anchor']) ||
+ isset($method['area_bbcode']) || isset($method['uri_bbcode']))
+ $length = 1; // Seems not effective
+ }
+
+ // Removing sequential spaces and too short lines
+ $binary = strings($binary, $length, TRUE, FALSE); // Multibyte NOT needed
+
+ // Remove words (has no '<>[]:') between spaces
+ $binary = preg_replace('/[ \t][\w.,()\ \t]+[ \t]/', ' ', $binary);
+
+ return $binary;
+}
+
+// Preprocess: Domain exposure callback (See spam_uri_pickup_preprocess())
// http://victim.example.org/?foo+site:nasty.example.com+bar
// => http://nasty.example.com/?refer=victim.example.org
// NOTE: 'refer=' is not so good for (at this time).
// [OK] http://victim.example.org/nasty.example.org
// [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
// [OK] http://victim.example.org/http://nasty.example.org
-function spam_uri_pickup_preprocess($string = '')
+function spam_uri_pickup_preprocess($string = '', $method = array())
{
if (! is_string($string)) return '';
- $string = rawurldecode($string);
+ $string = spam_uri_removing_hocus_pocus(rawurldecode($string), $method);
+ //var_dump(htmlspecialchars($string));
// Domain exposure (simple)
// http://victim.example.org/nasty.example.org/path#frag
// => http://nasty.example.org/?refer=victim.example.org and original
$string = preg_replace(
- '#http://' .
+ '#h?ttp://' .
'(' .
'ime\.nu' . '|' . // 2ch.net
'ime\.st' . '|' . // 2ch.net
$string
);
- // Domain exposure (See _preg_replace_callback_domain_exposure())
+ // Domain exposure (gate-big5)
+ // http://victim.example.org/gate/big5/nasty.example.org/path
+ // => http://nasty.example.org/?refer=victim.example.org and original
+ $string = preg_replace(
+ '#h?ttp://' .
+ '(' .
+ 'big5.51job.com' . '|' .
+ 'big5.china.com' . '|' .
+ 'big5.xinhuanet.com' . '|' .
+ ')' .
+ '/gate/big5' .
+ '/([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' .
+ '#i', // nasty.example.org
+ 'http://$2/?refer=$1 $0', // Preserve $0 or remove?
+ $string
+ );
+
+ // Domain exposure (site:) See _preg_replace_callback_domain_exposure()
$string = preg_replace_callback(
array(
- '#(http)://' .
+ '#(h?ttp)://' . // 1:Scheme
+ // 2:Host
'(' .
+ '(?:[a-z0-9_.-]+\.)?[a-z0-9_-]+\.[a-z0-9_-]+' .
// Something Google: http://www.google.com/supported_domains
- '(?:[a-z0-9.]+\.)?google\.[a-z]{2,3}(?:\.[a-z]{2})?' .
- '|' .
- // AltaVista
- '(?:[a-z0-9.]+\.)?altavista.com' .
-
+ // AltaVista: http://es.altavista.com/web/results?q=site%3Anasty.example.org+foobar
+ // Live Search: search.live.com
+ // MySpace: http://sads.myspace.com/Modules/Search/Pages/Search.aspx?_snip_&searchString=site:nasty.example.org
+ // (also searchresults.myspace.com)
+ // alltheweb.com
+ // search.bbc.co.uk
+ // search.orange.co.uk
+ // ...
')' .
'/' .
- '([a-z0-9?=&.%_/\'\\\+-]+)' . // path/?query=foo+bar+
- '\bsite:([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' . // site:nasty.example.com
- //'()' . // Preserve or remove?
+ '([a-z0-9?=&.%_/\'\\\+-]+)' . // 3:path/?query=foo+bar+
+ '\bsite:([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' . // 4:site:nasty.example.com
+ '()' . // 5:Preserve or remove?
'#i',
),
'_preg_replace_callback_domain_exposure',
$method = check_uri_spam_method();
}
- $string = spam_uri_pickup_preprocess($string);
+ $string = spam_uri_pickup_preprocess($string, $method);
$array = uri_pickup($string);
function get_blocklist($list = '')
{
- static $regexs;
+ static $regexes;
- if (! isset($regexs)) {
- $regexs = array();
+ if ($list === NULL) {
+ $regexes = NULL; // Unset
+ return array();
+ }
+
+ if (! isset($regexes)) {
+ $regexes = array();
if (file_exists(SPAM_INI_FILE)) {
$blocklist = array();
include(SPAM_INI_FILE);
// 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
// );
if (isset($blocklist['list'])) {
- $regexs['list'] = & $blocklist['list'];
+ $regexes['list'] = & $blocklist['list'];
} else {
// Default
$blocklist['list'] = array(
if (! isset($blocklist[$_list])) continue;
foreach ($blocklist[$_list] as $key => $value) {
if (is_array($value)) {
- $regexs[$_list][$key] = array();
+ $regexes[$_list][$key] = array();
foreach($value as $_key => $_value) {
- get_blocklist_add($regexs[$_list][$key], $_key, $_value);
+ get_blocklist_add($regexes[$_list][$key], $_key, $_value);
}
} else {
- get_blocklist_add($regexs[$_list], $key, $value);
+ get_blocklist_add($regexes[$_list], $key, $value);
}
}
unset($blocklist[$_list]);
}
}
- if ($list == '') {
- return $regexs; // ALL
- } else if (isset($regexs[$list])) {
- return $regexs[$list];
- } else {
+ if ($list === '') {
+ return $regexes; // ALL
+ } else if (isset($regexes[$list])) {
+ return $regexes[$list];
+ } else {
return array();
}
}
return $blocked;
}
-// Simple example for badhost (not used now)
-function is_badhost($hosts = array(), $asap = TRUE, $bool = TRUE)
-{
- $list = get_blocklist('list');
- $blocked = blocklist_distiller($hosts, array_keys($list), $asap);
- foreach($list as $key=>$type){
- if (! $type) unset($blocked[$key]); // Ignore goodhost etc
- }
-
- return $bool ? ! empty($blocked) : $blocked;
-}
-
// Default (enabled) methods and thresholds (for content insertion)
function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
{
foreach(array_keys($method) as $key) {
if (! isset($sum[$key])) $sum[$key] = 0;
}
+ if (! isset($sum['quantity'])) $sum['quantity'] = 0;
if (is_array($target)) {
foreach($target as $str) {
if ($asap && $is_spam) break;
// Merge only
- $blocked = array_merge_leaves($blocked, $_progress['blocked'], FALSE, FALSE);
- $hosts = array_merge_leaves($hosts, $_progress['hosts'], FALSE, FALSE);
+ $blocked = array_merge_recursive($blocked, $_progress['blocked']);
+ $hosts = array_merge_recursive($hosts, $_progress['hosts']);
}
// Unique values
$blocked = array_unique_recursive($blocked);
$hosts = array_unique_recursive($hosts);
- // Renumber numeric keys
- array_renumber_numeric_keys($blocked);
- array_renumber_numeric_keys($hosts);
-
// Recount $sum['badhost']
$sum['badhost'] = array_count_leaves($blocked);
return $count;
}
-// Merge two leaves' value
-function array_merge_leaves(& $array1, & $array2, $unique_values = TRUE, $renumber_numeric = TRUE)
-{
- $array = array_merge_recursive($array1, $array2);
-
- // Redundant values (and keys) are vanished
- if ($unique_values) $array = array_unique_recursive($array);
-
- // All NUMERIC keys are always renumbered from 0
- if ($renumber_numeric) array_renumber_numeric_keys($array);
-
- return $array;
-}
-
// An array-leaves to a flat array
function array_flat_leaves($array, $unique = TRUE)
{
return $unique ? array_values(array_unique($tmp)) : $tmp;
}
+// An array() to an array leaf
+function array_leaf($array = array('A', 'B', 'C.D'), $stem = FALSE, $edge = TRUE)
+{
+ if (! is_array($array)) return $array;
+
+ $leaf = array();
+ $tmp = & $leaf;
+ foreach($array as $arg) {
+ if (! is_string($arg) && ! is_int($arg)) continue;
+ $tmp[$arg] = array();
+ $parent = & $tmp;
+ $tmp = & $tmp[$arg];
+ }
+ if ($stem) {
+ $parent[key($parent)] = & $edge;
+ } else {
+ $parent = key($parent);
+ }
+
+ return $leaf; // array('A' => array('B' => 'C.D'))
+}
+
+
// ---------------------
// Reporting
-// TODO: Don't show unused $method!
// Summarize $progress (blocked only)
function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
{
foreach($progress['blocked'] as $list => $lvalue) {
foreach($lvalue as $group => $gvalue) {
$flat = implode(', ', array_flat_leaves($gvalue));
- if ($flat == $group) {
+ if ($flat === $group) {
$blocked[$list][] = $flat;
} else {
$blocked[$list][$group] = $flat;
! is_array($progress['hosts']) ||
empty($progress['hosts'])) return '';
- // Sort by domain
- $tmp = array();
+ // Generate a responsible $trie
+ $trie = array();
foreach($progress['hosts'] as $value) {
- $tmp[delimiter_reverse($value)] = $value;
+ // 'A.foo.bar.example.com'
+ $resp = whois_responsibility($value); // 'example.com'
+ if (empty($resp)) {
+ // One or more test, or do nothing here
+ $resp = strval($value);
+ $rest = '';
+ } else {
+ $rest = rtrim(substr($value, 0, - strlen($resp)), '.'); // 'A.foo.bar'
+ }
+ $trie = array_merge_recursive($trie, array($resp => array($rest => NULL)));
+ }
+
+ // Format: var_export_shrink() -like output
+ $result = array();
+ ksort_by_domain($trie);
+ foreach(array_keys($trie) as $key) {
+ ksort_by_domain($trie[$key]);
+ if (count($trie[$key]) == 1 && key($trie[$key]) == '') {
+ // Just one 'responsibility.example.com'
+ $result[] = ' \'' . $key . '\',';
+ } else {
+ // One subdomain-or-host, or several ones
+ $subs = array();
+ foreach(array_keys($trie[$key]) as $sub) {
+ if ($sub == '') {
+ $subs[] = $key;
+ } else {
+ $subs[] = $sub . '.' . $key;
+ }
+ }
+ $result[] = ' \'' . $key . '\' => \'' . implode(', ', $subs) . '\',';
+ }
+ unset($trie[$key]);
}
- ksort($tmp);
+ return
+ 'array (' . "\n" .
+ implode("\n", $result) . "\n" .
+ ')';
+}
- return count($tmp) . ' (' .implode(', ', $tmp) . ')';
+// ksort() by domain
+function ksort_by_domain(& $array)
+{
+ $sort = array();
+ foreach(array_keys($array) as $key) {
+ $sort[delimiter_reverse($key)] = $key;
+ }
+ ksort($sort, SORT_STRING);
+ $result = array();
+ foreach($sort as $key) {
+ $result[$key] = & $array[$key];
+ }
+ $array = $result;
+}
+
+// Check responsibility-root of the FQDN
+// 'foo.bar.example.com' => 'example.com' (.com has the last whois for it)
+// 'foo.bar.example.au' => 'example.au' (.au has the last whois for it)
+// 'foo.bar.example.edu.au' => 'example.edu.au' (.edu.au has the last whois for it)
+// 'foo.bar.example.act.edu.au' => 'example.act.edu.au' (.act.edu.au has the last whois for it)
+function whois_responsibility($fqdn = 'foo.bar.example.com', $parent = FALSE, $implicit = TRUE)
+{
+ static $domain;
+
+ if ($fqdn === NULL) {
+ $domain = NULL; // Unset
+ return '';
+ }
+ if (! is_string($fqdn)) return '';
+
+ if (is_ip($fqdn)) return $fqdn;
+
+ if (! isset($domain)) {
+ $domain = array();
+ if (file_exists(DOMAIN_INI_FILE)) {
+ include(DOMAIN_INI_FILE); // Set
+ }
+ }
+
+ $result = array();
+ $dcursor = & $domain;
+ $array = array_reverse(explode('.', $fqdn));
+ $i = 0;
+ while(TRUE) {
+ if (! isset($array[$i])) break;
+ $acursor = $array[$i];
+ if (is_array($dcursor) && isset($dcursor[$acursor])) {
+ $result[] = & $array[$i];
+ $dcursor = & $dcursor[$acursor];
+ } else {
+ if (! $parent && isset($acursor)) {
+ $result[] = & $array[$i]; // Whois servers must know this subdomain
+ }
+ break;
+ }
+ ++$i;
+ }
+
+ // Implicit responsibility: Top-Level-Domains must not be yours
+ // 'bar.foo.something' => 'foo.something'
+ if ($implicit && count($result) == 1 && count($array) > 1) {
+ $result[] = & $array[1];
+ }
+
+ return $result ? implode('.', array_reverse($result)) : '';
}
// ---------------------
// Exit
+// Freeing memories
+function spam_dispose()
+{
+ get_blocklist(NULL);
+ whois_responsibility(NULL);
+}
+
// Common bahavior for blocking
// NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
function spam_exit($mode = '', $data = array())
{
+ $exit = TRUE;
+
switch ($mode) {
- case '': echo("\n"); break;
+ case '':
+ echo("\n");
+ break;
case 'dump':
echo('<pre>' . "\n");
echo htmlspecialchars(var_export($data, TRUE));
break;
};
- // Force exit
- exit;
+ if ($exit) exit; // Force exit
}
{
$progress = check_uri_spam($target, $method);
- if (! empty($progress['is_spam'])) {
- // Mail to administrator(s)
- pkwk_spamnotify($action, $page, $target, $progress, $method);
+ if (empty($progress['is_spam'])) {
+ spam_dispose();
+ } else {
- // Exit
+// TODO: detect encoding from $target for mbstring functions
+// $tmp = array();
+// foreach(array_keys($target) as $key) {
+// $tmp[strings($key, 0, FALSE, TRUE)] = strings($target[$key], 0, FALSE, TRUE); // Removing "\0" etc
+// }
+// $target = & $tmp;
+
+ pkwk_spamnotify($action, $page, $target, $progress, $method);
spam_exit($exitmode, $progress);
}
}
OSDN Git Service
