<?php
-// $Id: spam.php,v 1.87 2007/01/02 05:57:51 henoheno Exp $
-// Copyright (C) 2006 PukiWiki Developers Team
+// $Id: spam.php,v 1.125 2007/03/10 01:30:14 henoheno Exp $
+// Copyright (C) 2006-2007 PukiWiki Developers Team
// License: GPL v2 or (at your option) any later version
-
+//
// Functions for Concept-work of spam-uri metrics
+//
// (PHP 4 >= 4.3.0): preg_match_all(PREG_OFFSET_CAPTURE): $method['uri_XXX'] related feature
-// (PHP 4 >= 4.2.0): var_export(): mail-reporting and dump related
if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
// ---------------------
+// Compat etc
+
+// (PHP 4 >= 4.2.0): var_export(): mail-reporting and dump related
+if (! function_exists('var_export')) {
+ function var_export() {
+ return 'var_export() is not found on this server' . "\n";
+ }
+}
+
+// (PHP 4 >= 4.2.0): preg_grep() enables invert option
+function preg_grep_invert($pattern = '//', $input = array())
+{
+ static $invert;
+ if (! isset($invert)) $invert = defined('PREG_GREP_INVERT');
+
+ if ($invert) {
+ return preg_grep($pattern, $input, PREG_GREP_INVERT);
+ } else {
+ $result = preg_grep($pattern, $input);
+ if ($result) {
+ return array_diff($input, preg_grep($pattern, $input));
+ } else {
+ return $input;
+ }
+ }
+}
+
+// ---------------------
// URI pickup
// Return an array of URIs in the $string
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
- '[^\s<>"\'\[\]:/\#?&\\\]+' . // FQDN: foo.example.org
+ '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' . // IPv4(dot-decimal): 001.22.3.44
+ '[a-z0-9.-]+' . // hostname(FQDN) : foo.example.org
')' .
'(?::([0-9]*))?' . // 4: Port
'((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
unset($array[$uri]);
continue;
}
- $_uri['host'] = strtolower($_uri['host']);
+ $_uri['host'] = host_normalize($_uri['host']);
$_uri['port'] = port_normalize($_uri['port'], $_uri['scheme'], FALSE);
$_uri['path'] = path_normalize($_uri['path']);
if ($preserve_rawuri) $_uri['rawuri'] = & $_uri[0];
// NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
// [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
// [OK] http://victim.example.org/http://nasty.example.org
+// TODO: link.toolbot.com, urlx.org
function spam_uri_pickup_preprocess($string = '')
{
if (! is_string($string)) return '';
// Domain exposure (See _preg_replace_callback_domain_exposure())
$string = preg_replace_callback(
array(
- // Something Google: http://www.google.com/supported_domains
- '#(http)://([a-z0-9.]+\.google\.[a-z]{2,3}(?:\.[a-z]{2})?)/' .
- '([a-z0-9?=&.%_+-]+)' . // ?query=foo+
- '\bsite:([a-z0-9.%_-]+)' . // site:nasty.example.com
+ '#(http)://' .
+ '(' .
+ // 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' .
+
+ ')' .
+ '/' .
+ '([a-z0-9?=&.%_/\'\\\+-]+)' . // path/?query=foo+bar+
+ '\bsite:([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' . // site:nasty.example.com
//'()' . // Preserve or remove?
'#i',
),
return $scheme;
}
+// Hostname normlization
+// www.foo => www.foo ('foo' seems TLD)
+// www.foo.bar => foo.bar
+// www.10.20 => www.10.20 (Invalid hostname)
+// NOTE:
+// 'www' is mostly used as traditional hostname of WWW server.
+// 'www.foo.bar' may be identical with 'foo.bar'.
+function host_normalize($host = '')
+{
+ $host = strtolower($host);
+
+ $matches = array();
+ if (preg_match('/^www\.(.+\.[a-z]+)$/', $host, $matches)) {
+ return $matches[1];
+ } else {
+ return $host;
+ }
+}
+
// Port normalization: Suppress the (redundant) default port
// HTTP://example.org:80/ => http://example.org/
// HTTP://example.org:8080/ => http://example.org:8080/
// ---------------------
// Part One : Checker
+// Rough implementation of globbing
+//
+// USAGE: $regex = '/^' . generate_glob_regex('*.txt', '/') . '$/i';
+//
function generate_glob_regex($string = '', $divider = '/')
{
static $from = array(
// 23 => ']',
);
- if (is_array($string)) {
- // Recurse
- return '(?:' .
- implode('|', // OR
- array_map('generate_glob_regex',
- $string,
- array_pad(array(), count($string), $divider)
- )
- ) .
- ')';
+ $string = str_replace($from, $mid, $string); // Hide
+ $string = preg_quote($string, $divider);
+ $string = str_replace($mid, $to, $string); // Unhide
+
+ return $string;
+}
+
+// Rough hostname checker
+// [OK] 192.168.
+// TODO: Strict digit, 0x, CIDR, IPv6
+function is_ip($string = '')
+{
+ if (preg_match('/^' .
+ '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' .
+ '(?:[0-9]{1,3}\.){1,3}' . '$/',
+ $string)) {
+ return 4; // Seems IPv4(dot-decimal)
} else {
- $string = str_replace($from, $mid, $string); // Hide
- $string = preg_quote($string, $divider);
- $string = str_replace($mid, $to, $string); // Unhide
- return $string;
+ return 0; // Seems not IP
}
}
-// TODO: Ignore list
-// TODO: preg_grep() ?
-// TODO: Multi list
-function is_badhost($hosts = '', $asap = TRUE)
+// Generate host (FQDN, IPv4, ...) regex
+// 'localhost' : Matches with 'localhost' only
+// 'example.org' : Matches with 'example.org', and 'www.example.org'
+// '.example.org' : Matches with ALL FQDN ended with '.example.org'
+// '*.example.org' : Almost the same of '.example.org' except 'www.example.org'
+// '10.20.30.40' : Matches with IPv4 address '10.20.30.40' only
+// '192168.' : Matches with all IPv4 hosts started with '192.'
+// TODO: IPv4, CIDR?, IPv6
+function generate_host_regex($string = '', $divider = '/')
{
- static $regex;
-
- if (! isset($regex)) {
- $regex = array();
- $regex['badhost'] = array();
-
- // Sample
- if (TRUE) {
- $blocklist['badhost'] = array(
- //'*', // Deny all uri
- //'10.20.*.*', // 10.20.example.com also matches
- //'*.blogspot.com', // Blog services subdomains
- //array('blogspot.com', '*.blogspot.com')
-
- // Viral/Buzz marketers' site, trying to make people
- // as commercial Wiki spammers
- // http://pukiwiki.sourceforge.jp/image/2006-12-16_wikiviral_pressblog.gif
- array('pressblog.jp', '*.pressblog.jp'),
- );
- foreach ($blocklist['badhost'] as $part) {
- $_part = is_array($part) ? implode('/', $part) : $part;
- $regex['badhost'][$_part] = '/^' . generate_glob_regex($part) . '$/i';
- }
+ if (mb_strpos($string, '.') === FALSE)
+ return generate_glob_regex($string, $divider);
+
+ $result = '';
+ if (is_ip($string)) {
+ // IPv4
+ return generate_glob_regex($string, $divider);
+ } else {
+ // FQDN or something
+ $part = explode('.', $string, 2);
+ if ($part[0] == '') {
+ $part[0] = '(?:.*\.)?'; // And all related FQDN
+ } else if ($part[0] == '*') {
+ $part[0] = '.*\.'; // All subdomains/hosts only
+ } else {
+ return generate_glob_regex($string, $divider);
}
+ $part[1] = generate_glob_regex($part[1], $divider);
+ return implode('', $part);
+ }
+}
- // Load
+function get_blocklist($list = '')
+{
+ static $regexs;
+
+ if (! isset($regexs)) {
+ $regexs = array();
if (file_exists(SPAM_INI_FILE)) {
$blocklist = array();
- require(SPAM_INI_FILE);
- foreach ($blocklist['badhost'] as $part) {
- $_part = is_array($part) ? implode('/', $part) : $part;
- $regex['badhost'][$_part] = '/^' . generate_glob_regex($part) . '$/i';
+ include(SPAM_INI_FILE);
+ // $blocklist['badhost'] = array(
+ // '*.blogspot.com', // Blog services's subdomains (only)
+ // 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
+ // );
+ foreach(array('goodhost', 'badhost') as $_list) {
+ if (! isset($blocklist[$list])) continue;
+ foreach ($blocklist[$_list] as $key => $value) {
+ if (is_array($value)) {
+ $regexs[$_list][$key] = array();
+ foreach($value as $_key => $_value) {
+ if (is_string($_key)) {
+ $regexs[$_list][$key][$_key] = $_value; // A regex
+ } else {
+ $regexs[$_list][$key][$_value] =
+ '/^' . generate_host_regex($_value, '/') . '$/i';
+ }
+ }
+ } else {
+ if (is_string($key)) {
+ $regexs[$_list][$key] = $value; // A regex
+ } else {
+ $regexs[$_list][$value] =
+ '/^' . generate_host_regex($value, '/') . '$/i';
+ }
+ }
+ }
}
}
}
+ if ($list == '') {
+ return $regexs; // ALL
+ } else if (isset($regexs[$list])) {
+ return $regexs[$list];
+ } else {
+ return array();
+ }
+}
+
+function is_badhost($hosts = array(), $asap = TRUE, & $remains)
+{
$result = array();
if (! is_array($hosts)) $hosts = array($hosts);
+ foreach(array_keys($hosts) as $key) {
+ if (! is_string($hosts[$key])) unset($hosts[$key]);
+ }
+ if (empty($hosts)) return $result;
- foreach($hosts as $host) {
- if (! is_string($host)) $host = '';
- foreach ($regex['badhost'] as $part => $_regex) {
- if (preg_match($_regex, $host)) {
- if (! isset($result[$part])) $result[$part] = array();
- $result[$part][] = $host;
- if ($asap) {
- return $result;
- } else {
- break;
+ foreach (get_blocklist('goodhost') as $regex) {
+ $hosts = preg_grep_invert($regex, $hosts);
+ }
+ if (empty($hosts)) return $result;
+
+ $tmp = array();
+ foreach (get_blocklist('badhost') as $label => $regex) {
+ if (is_array($regex)) {
+ $result[$label] = array();
+ foreach($regex as $_label => $_regex) {
+ $_group = preg_grep($_regex, $hosts);
+ if ($_group) {
+ $result[$label][$_label] = $_group;
+ $hosts = array_diff($hosts, $_group);
+ if ($asap) break;
}
}
+ if (empty($result[$label])) unset($result[$label]);
+ } else {
+ $_group = preg_grep($regex, $hosts);
+ if ($_group) {
+ $result[$label] = $_group;
+ $hosts = array_diff($hosts, $result[$label]);
+ if ($asap) break;
+ }
}
}
+ $remains = $hosts;
+
return $result;
}
$positive = array(
// Thresholds
'quantity' => 8 * $times, // Allow N URIs
- 'non_uniqhost' => 7 * $times, // Allow N duped (and normalized) Hosts
- 'non_uniquri' => 3 * $times, // Allow N duped (and normalized) URIs
+ 'non_uniqhost' => 3 * $times, // Allow N duped (and normalized) Hosts
+ //'non_uniquri'=> 3 * $times, // Allow N duped (and normalized) URIs
// Areas
'area_anchor' => $t_area, // Using <a href> HTML tag
),
'is_spam' => array(),
'method' => & $method,
+ 'remains' => array(),
+ 'error' => array(),
);
$sum = & $progress['sum'];
$is_spam = & $progress['is_spam'];
+ $remains = & $progress['remains'];
+ $error = & $progress['error'];
$asap = isset($method['asap']);
- // Return if ...
+ // Recurse
if (is_array($target)) {
foreach($target as $str) {
// Recurse
$_progress = check_uri_spam($str, $method);
$_sum = & $_progress['sum'];
$_is_spam = & $_progress['is_spam'];
+ $_remains = & $_progress['remains'];
+ $_error = & $_progress['error'];
foreach (array_keys($_sum) as $key) {
$sum[$key] += $_sum[$key];
}
- foreach(array_keys($_is_spam) as $key) {
+ foreach (array_keys($_is_spam) as $key) {
if (is_array($_is_spam[$key])) {
// Marge keys (badhost)
foreach(array_keys($_is_spam[$key]) as $_key) {
$is_spam[$key] = TRUE;
}
}
+ foreach ($_remains as $key=>$value) {
+ foreach ($value as $_key=>$_value) {
+ if (is_int($_key)) {
+ $remains[$key][] = $_value;
+ } else {
+ $remains[$key][$_key] = $_value;
+ }
+ }
+ }
+ if (! empty($_error)) $error += $_error;
if ($asap && $is_spam) break;
}
return $progress;
}
// Return if ...
- if ($asap && $is_spam) {
- return $progress;
- }
- // URI Init
+ if ($asap && $is_spam) return $progress;
+
+ // URI: Pickup
$pickups = spam_uri_pickup($target, $method);
- if (empty($pickups)) {
- return $progress;
- }
+ //$remains['uri_pickup'] = & $pickups;
+
+ // Return if ...
+ if (empty($pickups)) return $progress;
// URI: Check quantity
$sum['quantity'] += count($pickups);
}
// Return if ...
- if ($asap && $is_spam) {
- return $progress;
- }
+ if ($asap && $is_spam) return $progress;
// Host: Uniqueness (uniq / non-uniq)
$hosts = array();
foreach ($pickups as $pickup) $hosts[] = & $pickup['host'];
$hosts = array_unique($hosts);
+ //$remains['uniqhost'] = & $hosts;
$sum['uniqhost'] += count($hosts);
if ((! $asap || ! $is_spam) && isset($method['non_uniqhost'])) {
$sum['non_uniqhost'] = $sum['quantity'] - $sum['uniqhost'];
}
// Return if ...
- if ($asap && $is_spam) {
- return $progress;
- }
+ if ($asap && $is_spam) return $progress;
// URI: Bad host
if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
- $badhost = is_badhost($hosts, $asap);
+ $__remains = array();
+ $badhost = is_badhost($hosts, $asap, $__remains);
+ if (! $asap) {
+ if ($__remains) {
+ $remains['badhost'] = array();
+ foreach ($__remains as $value) {
+ $remains['badhost'][$value] = TRUE;
+ }
+ }
+ }
+ unset($__remains);
if (! empty($badhost)) {
$sum['badhost'] += array_count_leaves($badhost);
foreach(array_keys($badhost) as $keys) {
foreach($progress['is_spam']['badhost'] as $glob=>$number) {
$badhost[] = $glob . '(' . $number . ')';
}
- $summary['BADHOST'] = implode(', ', $badhost);
+ $summary['DETAIL_BADHOST'] = implode(', ', $badhost);
+ }
+ if (! $asap && $progress['remains']['badhost']) {
+ $count = count($progress['remains']['badhost']);
+ $summary['DETAIL_NEUTRAL_HOST'] = $count .
+ ' (' .
+ preg_replace(
+ '/[^, a-z0-9.-]/i', '',
+ implode(', ', array_keys($progress['remains']['badhost']))
+ ) .
+ ')';
}
$summary['COMMENT'] = $action;
$summary['PAGE'] = '[blocked] ' . (is_pagename($page) ? $page : '');
$summary['URI'] = get_script_uri() . '?' . rawurlencode($page);
$summary['USER_AGENT'] = TRUE;
$summary['REMOTE_ADDR'] = TRUE;
- pkwk_mail_notify($notify_subject, var_export($target, TRUE), $summary);
+ pkwk_mail_notify($notify_subject, var_export($target, TRUE), $summary, TRUE);
}
?>