<?php
-// $Id: spam.php,v 1.23 2006/11/21 13:42:40 henoheno Exp $
+// $Id: spam.php,v 1.32 2006/11/25 02:37:21 henoheno Exp $
// Copyright (C) 2006 PukiWiki Developers Team
// License: GPL v2 or (at your option) any later version
'([^\s<>"\'\[\]\#]+)?' . // 6: File and query string
'(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 7: Fragment
'#i',
- $string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
+ $string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE
+ );
//var_dump(recursive_map('htmlspecialchars', $array));
// Shrink $array
static $parts = array(
- 1 => 'scheme', 2 => 'userinfo',
- 3 => 'host', 4 => 'port', 5 => 'path', 6 => 'file', 7 => 'fragment');
+ 1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port',
+ 5 => 'path', 6 => 'file', 7 => 'fragment'
+ );
$default = array('');
foreach(array_keys($array) as $uri) {
array_rename_keys($array[$uri], $parts, TRUE, $default);
if ($normalize) {
$array[$uri]['scheme'] = scheme_normalize($array[$uri]['scheme']);
+ //if ($array[$uri]['scheme'] === '') {
+ // // Ignore
+ // unset ($array[$uri]);
+ // continue;
+ //}
+
$array[$uri]['host'] = strtolower($array[$uri]['host']);
$array[$uri]['port'] = port_normalize($array[$uri]['port'], $array[$uri]['scheme'], FALSE);
$array[$uri]['path'] = path_normalize($array[$uri]['path']);
$array[$uri]['path'],
$array[$uri]['file'],
$array[$uri]['fragment']
- );
+ );
}
$array[$uri]['offset'] = $offset;
return $array;
}
+// Domain exposure callback (See spam_uri_pickup_preprocess())
+// http://victim.example.org/?foo+site:nasty.example.com+bar
+// => http://nasty.example.com/?refer=victim.example.org
+// NOTE: 'refer=' is not so good for (at this time).
+// Consider about using IP address of the victim, try to avoid that.
+function _preg_replace_callback_domain_exposure($matches = array())
+{
+ $result = '';
+
+ // Preserve the victim URI as a complicity or ...
+ if (isset($matches[5])) {
+ $result =
+ $matches[1] . '://' . // scheme
+ $matches[2] . '/' . // victim.example.org
+ $matches[3]; // The rest of all (before victim)
+ }
+
+ // Flipped URI
+ $result =
+ $matches[1] . '://' . // scheme
+ $matches[4] . // nasty.example.com
+ '/?refer=' . strtolower($matches[2]) . // victim.example.org
+ ' ' . $result;
+
+ return $result;
+}
+
// Preprocess: rawurldecode() and adding space(s) to detect/count some URIs _if possible_
// NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
// [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
// [OK] http://victim.example.org/http://nasty.example.org
function spam_uri_pickup_preprocess($string = '')
{
- if (is_string($string)) {
- return preg_replace(
- array(
- '#(?:https?|ftp):/#',
- '#\b[a-z][a-z0-9.+-]{1,8}://#i',
- '#[a-z][a-z0-9.+-]{1,8}://#i'
- ),
- ' $0',
- rawurldecode($string)
- );
- } else {
- return '';
- }
+ if (! is_string($string)) return '';
+
+ $string = rawurldecode($string);
+
+ // 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
+ '()' . // Preserve?
+ '#i',
+ ),
+ '_preg_replace_callback_domain_exposure',
+ $string
+ );
+
+ // URI exposure (uriuri => uri uri)
+ $string = preg_replace(
+ array(
+ '#(?<! )(?:https?|ftp):/#',
+ // '#[a-z][a-z0-9.+-]{1,8}://#i',
+ // '#[a-z][a-z0-9.+-]{1,8}://#i'
+ ),
+ ' $0',
+ $string
+ );
+
+ return $string;
}
// TODO: Area selection (Check BBCode only, check anchor only, check ...)
{
$string = spam_uri_pickup_preprocess($string);
- $array = uri_pickup($string, FALSE, TRUE, FALSE);
+ $array = uri_pickup($string);
// Area elevation for '(especially external)link' intension
if (! empty($array)) {
// ---------------------
// Part Two
-// Scheme normalization: Rename the schemes
+// Scheme normalization: Renaming the schemes
// snntp://example.org => nntps://example.org
-// NOTE: Keep the static list simple. See also port_normalize().
-function scheme_normalize($scheme = '')
+// NOTE: Keep the static lists simple. See also port_normalize().
+function scheme_normalize($scheme = '', $considerd_harmfull = TRUE)
{
+ // Abbreviations considerable they don't have link intension
+ static $abbrevs = array(
+ 'ttp' => 'http',
+ 'ttps' => 'https',
+ );
+
+ // Alias => normalized
static $aliases = array(
- // alias => normalized
'pop' => 'pop3',
'news' => 'nntp',
'imap4' => 'imap',
);
$scheme = strtolower(trim($scheme));
+ if (isset($abbrevs[$scheme])) {
+ if ($considerd_harmfull) {
+ $scheme = $abbrevs[$scheme];
+ } else {
+ $scheme = '';
+ }
+ }
if (isset($aliases[$scheme])) $scheme = $aliases[$scheme];
return $scheme;
// ---------------------
// Part One : Checker
+function generate_glob_regex($string = '', $divider = '/')
+{
+ static $from = array(
+ 0 => '*',
+ 1 => '?',
+ 2 => '\[',
+ 3 => '\]',
+ 4 => '[',
+ 5 => ']',
+ );
+ static $mid = array(
+ 0 => '_AST_',
+ 1 => '_QUE_',
+ 2 => '_eRBR_',
+ 3 => '_eLBR_',
+ 4 => '_RBR_',
+ 5 => '_LBR_',
+ );
+ static $to = array(
+ 0 => '.*',
+ 1 => '.',
+ 2 => '\[',
+ 3 => '\]',
+ 4 => '[',
+ 5 => ']',
+ );
+
+ $string = str_replace($from, $mid, $string); // Hide
+ $string = preg_quote($string, $divider);
+ $string = str_replace($mid, $to, $string); // Unhide
+
+ return $string;
+}
+
+// TODO: Ignore list
+// TODO: require_or_include_once(another file)
+function is_badhost($host = '')
+{
+ static $blocklist_regex;
+
+ if (! isset($blocklist_regex)) {
+ $blocklist_regex = array();
+ $blocklist = array(
+ // Deny all uri
+ //'*',
+
+ // IP address or ...
+ //'10.20.*.*', // 10.20.example.com also matches
+ //'\[1\]',
+
+ // Too much malicious sub-domains
+ '*.blogspot.com',
+
+ // 2006-11 dev
+ 'wwwtahoo.com',
+
+ // 2006-11 dev
+ '*.infogami.com',
+
+ // 2006/11/19 17:50 dev
+ '*.google0site.org',
+ '*.bigpricesearch.org',
+ '*.osfind.org',
+ '*.bablomira.biz',
+ );
+ foreach ($blocklist as $part) {
+ $blocklist_regex[] = '#^' . generate_glob_regex($part, '#') . '$#';
+ }
+ }
+
+ $host = strtolower($host);
+ $result = FALSE;
+ foreach ($blocklist_regex as $regex) {
+ if (preg_match($regex, $host)) {
+ $result = TRUE;
+ break;
+ }
+ }
+
+ return $result;
+}
+
+// TODO return TRUE or FALSE!
// Simple/fast spam check
function is_uri_spam($target = '')
{
}
}
}
+
+ foreach ($pickups as $pickup) {
+ if (is_badhost($pickup['host'])) {
+ $is_spam = TRUE;
+ break;
+ }
+ }
}
}
// ---------------------
+// TODO: Separate check-part(s) and mail part
// TODO: Multi-metrics (uri, host, user-agent, ...)
// TODO: Mail to administrator with more measurement data?
// Simple/fast spam filter ($target: 'a string' or an array())