<?php
-// $Id: spam.php,v 1.24 2006/11/22 13:19:21 henoheno Exp $
-// Copyright (C) 2006 PukiWiki Developers Team
+// $Id: spam.php,v 1.220 2011/01/24 14:46:45 henoheno Exp $
+// Copyright (C) 2006-2009 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
-// Return an array of URIs in the $string
-// [OK] http://nasty.example.org#nasty_string
-// [OK] http://nasty.example.org:80/foo/xxx#nasty_string/bar
-// [OK] ftp://nasty.example.org:80/dfsdfs
-// [OK] ftp://cnn.example.com&story=breaking_news@10.0.0.1/top_story.htm (from RFC3986)
-function uri_pickup($string = '', $normalize = TRUE,
- $preserve_rawuri = FALSE, $preserve_chunk = TRUE)
-{
- // Not available for: IDN(ignored)
- $array = array();
- 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
- '(?:' .
- '([^\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]*))?' . // 4: Port
- '((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
- '([^\s<>"\'\[\]\#]+)?' . // 6: File and query string
- '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 7: Fragment
- '#i',
- $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');
- $default = array('');
- foreach(array_keys($array) as $uri) {
- array_rename_keys($array[$uri], $parts, TRUE, $default);
- $offset = $array[$uri]['scheme'][1]; // Scheme's offset
-
- foreach(array_keys($array[$uri]) as $part) {
- // Remove offsets for each part
- $array[$uri][$part] = & $array[$uri][$part][0];
- }
- if ($normalize) {
- $array[$uri]['scheme'] = scheme_normalize($array[$uri]['scheme']);
- $array[$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']);
+if (! defined('LIB_DIR')) define('LIB_DIR', './');
+require(LIB_DIR . 'spam_pickup.php');
+require(LIB_DIR . 'spam_util.php');
- //$array[$uri]['uri'] = uri_array_implode($array[$uri]);
- if ($preserve_rawuri) $array[$uri]['rawuri'] = & $array[$uri][0];
- } else {
- $array[$uri]['uri'] = & $array[$uri][0]; // Raw
- }
- unset($array[$uri][0]); // Matched string itself
- if (! $preserve_chunk) {
- unset(
- $array[$uri]['scheme'],
- $array[$uri]['userinfo'],
- $array[$uri]['host'],
- $array[$uri]['port'],
- $array[$uri]['path'],
- $array[$uri]['file'],
- $array[$uri]['fragment']
- );
- }
+if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
- $array[$uri]['offset'] = $offset;
- $array[$uri]['area'] = 0;
- }
- return $array;
+// ---------------------
+// Regex
+
+// Rough implementation of globbing
+//
+// USAGE: $regex = '/^' . generate_glob_regex('*.txt', '/') . '$/i';
+//
+function generate_glob_regex($string = '', $divider = '/')
+{
+ static $from = array(
+ 1 => '*',
+ 11 => '?',
+ // 22 => '[', // Maybe cause regex compilation error (e.g. '[]')
+ // 23 => ']', //
+ );
+ static $mid = array(
+ 1 => '_AST_',
+ 11 => '_QUE_',
+ // 22 => '_RBR_',
+ // 23 => '_LBR_',
+ );
+ static $to = array(
+ 1 => '.*',
+ 11 => '.',
+ // 22 => '[',
+ // 23 => ']',
+ );
+
+ if (! is_string($string)) return '';
+
+ $string = str_replace($from, $mid, $string); // Hide
+ $string = preg_quote($string, $divider);
+ $string = str_replace($mid, $to, $string); // Unhide
+
+ return $string;
}
-// Preprocess: rawurldecode() and adding space(s) to detect/count some URIs _if possible_
-// NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
-// [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
-// [OK] http://victim.example.org/http://nasty.example.org
-function spam_uri_pickup_preprocess($string = '')
+// Generate host (FQDN, IPv4, ...) regex
+// 'localhost' : Matches with 'localhost' only
+// 'example.org' : Matches with 'example.org' only (See host_normalize() about 'www')
+// '.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
+// [TODO] '192.' : Matches with all IPv4 hosts started with '192.'
+// TODO: IPv4, CIDR?, IPv6
+function generate_host_regex($string = '', $divider = '/')
{
- if (is_string($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)
- );
+ if (! is_string($string)) return '';
+
+ if (mb_strpos($string, '.') === FALSE || is_ip($string)) {
+ // "localhost", IPv4, etc
+ return generate_glob_regex($string, $divider);
+ }
+
+ // FQDN or something
+ $part = explode('.', $string, 2);
+ if ($part[0] == '') {
+ // ".example.org"
+ $part[0] = '(?:.*\.)?';
+ } else if ($part[0] == '*') {
+ // "*.example.org"
+ $part[0] = '.*\.';
} else {
- return '';
+ // example.org, etc
+ return generate_glob_regex($string, $divider);
}
+
+ $part[1] = generate_glob_regex($part[1], $divider);
+
+ return implode('', $part);
}
-// TODO: Area selection (Check BBCode only, check anchor only, check ...)
-// Main function of spam-uri pickup
-function spam_uri_pickup($string = '')
+
+// ---------------------
+// Load
+
+// Load SPAM_INI_FILE and return parsed one
+function get_blocklist($list = '')
{
- $string = spam_uri_pickup_preprocess($string);
-
- $array = uri_pickup($string);
-
- // Area elevation for '(especially external)link' intension
- if (! empty($array)) {
- // Anchor tags by preg_match_all()
- // [OK] <a href="http://nasty.example.com">visit http://nasty.example.com/</a>
- // [OK] <a href=\'http://nasty.example.com/\' >discount foobar</a>
- // [NG] <a href="http://ng.example.com">visit http://ng.example.com _not_ended_
- // [NG] <a href= >Good site!</a> <a href= "#" >test</a>
- $areas = array();
- preg_match_all('#<a\b[^>]*href[^>]*>.*?</a\b[^>]*(>)#i',
- $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
- //var_dump(recursive_map('htmlspecialchars', $areas));
- foreach(array_keys($areas) as $area) {
- $areas[$area] = array(
- $areas[$area][0][1], // Area start (<a href>)
- $areas[$area][1][1], // Area end (</a>)
- );
- }
- area_measure($areas, $array);
-
- // phpBB's "BBCode" by preg_match_all()
- // [url]http://nasty.example.com/[/url]
- // [link]http://nasty.example.com/[/link]
- // [url=http://nasty.example.com]visit http://nasty.example.com/[/url]
- // [link http://nasty.example.com/]buy something[/link]
- // ?? [url=][/url]
- $areas = array();
- preg_match_all('#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#i',
- $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
- //var_dump(recursive_map('htmlspecialchars', $areas));
- foreach(array_keys($areas) as $area) {
- $areas[$area] = array(
- $areas[$area][0][1], // Area start ([url])
- $areas[$area][2][1], // Area end ([/url])
- );
+ static $regexes;
+
+ if ($list === NULL) {
+ $regexes = NULL; // Unset
+ return array();
+ }
+
+ if (! isset($regexes)) {
+ $regexes = array();
+ if (file_exists(SPAM_INI_FILE)) {
+ $blocklist = array();
+
+ include(SPAM_INI_FILE);
+ // $blocklist['list'] = array(
+ // //'goodhost' => FALSE;
+ // 'badhost' => TRUE;
+ // );
+ // $blocklist['badhost'] = array(
+ // '*.blogspot.com', // Blog services's subdomains (only)
+ // 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
+ // );
+
+ foreach(array(
+ 'pre',
+ 'list',
+ ) as $special) {
+
+ if (! isset($blocklist[$special])) continue;
+
+ $regexes[$special] = $blocklist[$special];
+
+ foreach(array_keys($blocklist[$special]) as $_list) {
+ if (! isset($blocklist[$_list])) continue;
+
+ foreach ($blocklist[$_list] as $key => $value) {
+ if (is_array($value)) {
+ $regexes[$_list][$key] = array();
+ foreach($value as $_key => $_value) {
+ get_blocklist_add($regexes[$_list][$key], $_key, $_value);
+ }
+ } else {
+ get_blocklist_add($regexes[$_list], $key, $value);
+ }
+ }
+
+ unset($blocklist[$_list]);
+ }
+ }
}
- area_measure($areas, $array);
-
- // Various Wiki syntax
- // [text_or_uri>text_or_uri]
- // [text_or_uri:text_or_uri]
- // [text_or_uri|text_or_uri]
- // [text_or_uri->text_or_uri]
- // [text_or_uri text_or_uri] // MediaWiki
- // MediaWiki: [http://nasty.example.com/ visit http://nasty.example.com/]
-
- // Remove 'offset's for area_measure()
- //foreach(array_keys($array) as $key)
- // unset($array[$key]['offset']);
}
- return $array;
+ if ($list === '') {
+ return $regexes; // ALL of
+ } else if (isset($regexes[$list])) {
+ return $regexes[$list]; // A part of
+ } else {
+ return array(); // Found nothing
+ }
}
-// $array['something'] => $array['wanted']
-function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '')
+// Subroutine of get_blocklist(): Add new regex to the $array
+function get_blocklist_add(& $array, $key = 0, $value = '*.example.org/path/to/file.html')
{
- if (! is_array($array) || ! is_array($keys))
- return FALSE;
-
- // Nondestructive test
- if (! $force)
- foreach(array_keys($keys) as $from)
- if (! isset($array[$from]))
- return FALSE;
-
- foreach($keys as $from => $to) {
- if ($from === $to) continue;
- if (! $force || isset($array[$from])) {
- $array[$to] = & $array[$from];
- unset($array[$from]);
- } else {
- $array[$to] = $default;
+ if (is_string($key)) {
+ $array[$key] = & $value; // Treat $value as a regex for FQDN(host)s
+ } else {
+ $regex = generate_host_regex($value, '#');
+ if (! empty($regex)) {
+ $array[$value] = '#^' . $regex . '$#i';
}
}
-
- return TRUE;
}
-// If in doubt, it's a little doubtful
-function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset')
+// Blocklist metrics: Separate $host, to $blocked and not blocked
+function blocklist_distiller(& $hosts, $keys = array('goodhost', 'badhost'), $asap = FALSE)
{
- if (! is_array($areas) || ! is_array($array)) return;
-
- $areas_keys = array_keys($areas);
- foreach(array_keys($array) as $u_index) {
- $offset = isset($array[$u_index][$o_key]) ?
- intval($array[$u_index][$o_key]) : 0;
- foreach($areas_keys as $a_index) {
- if (isset($array[$u_index][$a_key])) {
- $offset_s = intval($areas[$a_index][0]);
- $offset_e = intval($areas[$a_index][1]);
- // [Area => inside <= Area]
- if ($offset_s < $offset && $offset < $offset_e) {
- $array[$u_index][$a_key] += $belief;
+ if (! is_array($hosts)) $hosts = array($hosts);
+ if (! is_array($keys)) $keys = array($keys);
+
+ $list = get_blocklist('list');
+ $blocked = array();
+
+ foreach($keys as $key){
+ foreach (get_blocklist($key) as $label => $regex) {
+ if (is_array($regex)) {
+ foreach($regex as $_label => $_regex) {
+ $group = preg_grep($_regex, $hosts);
+ if ($group) {
+ $hosts = array_diff($hosts, $group);
+ $blocked[$key][$label][$_label] = $group;
+ if ($asap && $list[$key]) break;
+ }
+ }
+ } else {
+ $group = preg_grep($regex, $hosts);
+ if ($group) {
+ $hosts = array_diff($hosts, $group);
+ $blocked[$key][$label] = $group;
+ if ($asap && $list[$key]) break;
}
}
}
}
+
+ return $blocked;
}
// ---------------------
-// Part Two
-// Scheme normalization: Rename the schemes
-// snntp://example.org => nntps://example.org
-// NOTE: Keep the static list simple. See also port_normalize().
-function scheme_normalize($scheme = '')
+
+// Default (enabled) methods and thresholds (for content insertion)
+function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
{
- static $aliases = array(
- // alias => normalized
- 'pop' => 'pop3',
- 'news' => 'nntp',
- 'imap4' => 'imap',
- 'snntp' => 'nntps',
- 'snews' => 'nntps',
- 'spop3' => 'pop3s',
- 'pops' => 'pop3s',
+ $times = intval($times);
+ $t_area = intval($t_area);
+
+ $positive = array(
+ // Thresholds
+ 'quantity' => 8 * $times, // Allow N 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
+ 'area_bbcode' => $t_area, // Using [url] or [link] BBCode
+ //'uri_anchor' => $t_area, // URI inside <a href> HTML tag
+ //'uri_bbcode' => $t_area, // URI inside [url] or [link] BBCode
);
+ if ($rule) {
+ $bool = array(
+ // Rules
+ //'asap' => TRUE, // Quit or return As Soon As Possible
+ 'uniqhost' => TRUE, // Show uniq host (at block notification mail)
+ 'badhost' => TRUE, // Check badhost
+ );
+ } else {
+ $bool = array();
+ }
- $scheme = strtolower(trim($scheme));
- if (isset($aliases[$scheme])) $scheme = $aliases[$scheme];
+ // Remove non-$positive values
+ foreach (array_keys($positive) as $key) {
+ if ($positive[$key] < 0) unset($positive[$key]);
+ }
- return $scheme;
+ return $positive + $bool;
}
-// Port normalization: Suppress the (redundant) default port
-// HTTP://example.org:80/ => http://example.org/
-// HTTP://example.org:8080/ => http://example.org:8080/
-// HTTPS://example.org:443/ => https://example.org/
-function port_normalize($port, $scheme, $scheme_normalize = TRUE)
+// Simple/fast spam check
+function check_uri_spam($target = '', $method = array())
{
- // Schemes that users _maybe_ want to add protocol-handlers
- // to their web browsers. (and attackers _maybe_ want to use ...)
- // Reference: http://www.iana.org/assignments/port-numbers
- static $array = array(
- // scheme => default port
- 'ftp' => 21,
- 'ssh' => 22,
- 'telnet' => 23,
- 'smtp' => 25,
- 'tftp' => 69,
- 'gopher' => 70,
- 'finger' => 79,
- 'http' => 80,
- 'pop3' => 110,
- 'sftp' => 115,
- 'nntp' => 119,
- 'imap' => 143,
- 'irc' => 194,
- 'wais' => 210,
- 'https' => 443,
- 'nntps' => 563,
- 'rsync' => 873,
- 'ftps' => 990,
- 'telnets' => 992,
- 'imaps' => 993,
- 'ircs' => 994,
- 'pop3s' => 995,
- 'mysql' => 3306,
+ // Return value
+ $progress = array(
+ 'method' => array(
+ // Theme to do => Dummy, optional value, or optional array()
+ //'quantity' => 8,
+ //'uniqhost' => TRUE,
+ //'non_uniqhost'=> 3,
+ //'non_uniquri' => 3,
+ //'badhost' => TRUE,
+ //'area_anchor' => 0,
+ //'area_bbcode' => 0,
+ //'uri_anchor' => 0,
+ //'uri_bbcode' => 0,
+ ),
+ 'sum' => array(
+ // Theme => Volume found (int)
+ ),
+ 'is_spam' => array(
+ // Flag. If someting defined here,
+ // one or more spam will be included
+ // in this report
+ ),
+ 'blocked' => array(
+ // Hosts blocked
+ //'category' => array(
+ // 'host',
+ //)
+ ),
+ 'hosts' => array(
+ // Hosts not blocked
+ ),
);
- $port = trim($port);
- if ($port === '') return $port;
+ // ----------------------------------------
+ // Aliases
- if ($scheme_normalize) $scheme = scheme_normalize($scheme);
- if (isset($array[$scheme]) && $port == $array[$scheme])
- $port = ''; // Ignore the defaults
+ $sum = & $progress['sum'];
+ $is_spam = & $progress['is_spam'];
+ $progress['method'] = & $method; // Argument
+ $blocked = & $progress['blocked'];
+ $hosts = & $progress['hosts'];
+ $asap = isset($method['asap']);
- return $port;
-}
+ // ----------------------------------------
+ // Init
-// Path normalization
-// http://example.org => http://example.org/
-// http://example.org#hoge => http://example.org/#hoge
-// http://example.org/path/a/b/./c////./d => http://example.org/path/a/b/c/d
-// http://example.org/path/../../a/../back => http://example.org/back
-function path_normalize($path = '', $divider = '/', $addroot = TRUE)
-{
- if (! is_string($path) || $path == '') {
- $path = $addroot ? $divider : '';
- } else {
- $path = trim($path);
- $last = ($path[strlen($path) - 1] == $divider) ? $divider : '';
- $array = explode($divider, $path);
-
- // Remove paddings
- foreach(array_keys($array) as $key) {
- if ($array[$key] == '' || $array[$key] == '.')
- unset($array[$key]);
- }
- // Back-track
- $tmp = array();
- foreach($array as $value) {
- if ($value == '..') {
- array_pop($tmp);
- } else {
- array_push($tmp, $value);
+ if (! is_array($method) || empty($method)) {
+ $method = check_uri_spam_method();
+ }
+ foreach(array_keys($method) as $key) {
+ if (! isset($sum[$key])) $sum[$key] = 0;
+ }
+ if (! isset($sum['quantity'])) $sum['quantity'] = 0;
+
+ // ----------------------------------------
+ // Recurse
+
+ if (is_array($target)) {
+ foreach($target as $str) {
+ if (! is_string($str)) continue;
+
+ $_progress = check_uri_spam($str, $method); // Recurse
+
+ // Merge $sum
+ $_sum = & $_progress['sum'];
+ foreach (array_keys($_sum) as $key) {
+ if (! isset($sum[$key])) {
+ $sum[$key] = & $_sum[$key];
+ } else {
+ $sum[$key] += $_sum[$key];
+ }
+ }
+
+ // Merge $is_spam
+ $_is_spam = & $_progress['is_spam'];
+ foreach (array_keys($_is_spam) as $key) {
+ $is_spam[$key] = TRUE;
+ if ($asap) break;
}
+ if ($asap && $is_spam) break;
+
+ // Merge only
+ $blocked = array_merge_leaves($blocked, $_progress['blocked'], FALSE);
+ $hosts = array_merge_leaves($hosts, $_progress['hosts'], FALSE);
}
- $array = & $tmp;
- $path = $addroot ? $divider : '';
- if (! empty($array)) $path .= implode($divider, $array) . $last;
+ // Unique values
+ $blocked = array_unique_recursive($blocked);
+ $hosts = array_unique_recursive($hosts);
+
+ // Recount $sum['badhost']
+ $sum['badhost'] = array_count_leaves($blocked);
+
+ return $progress;
}
- return $path;
-}
+ // ----------------------------------------
+ // Area measure
-// An URI array => An URI (See uri_pickup())
-function uri_array_implode($uri = array())
-{
- if (empty($uri) || ! is_array($uri)) return NULL;
+ if (! $asap || ! $is_spam) {
- $tmp = array();
- if (isset($uri['scheme']) && $uri['scheme'] !== '') {
- $tmp[] = & $uri['scheme'];
- $tmp[] = '://';
+ // Method pickup
+ $_method = array();
+ foreach(array(
+ 'area_anchor', // There's HTML anchor tag
+ 'area_bbcode', // There's 'BBCode' linking tag
+ ) as $key) {
+ if (isset($method[$key])) $_method[$key] = TRUE;
+ }
+
+ if ($_method) {
+ $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
+ $_result = area_pickup($target, $_method + $_asap);
+ $_asap = NULL;
+ } else {
+ $_result = FALSE;
+ }
+
+ if ($_result) {
+ foreach(array_keys($_method) as $key) {
+ if (isset($_result[$key])) {
+ $sum[$key] = $_result[$key];
+ if (isset($method[$key]) && $sum[$key] > $method[$key]) {
+ $is_spam[$key] = TRUE;
+ }
+ }
+ }
+ }
+
+ unset($_asap, $_method, $_result);
}
- if (isset($uri['userinfo']) && $uri['userinfo'] !== '') {
- $tmp[] = & $uri['userinfo'];
- $tmp[] = '@';
+
+ // Return if ...
+ if ($asap && $is_spam) return $progress;
+
+ // ----------------------------------------
+ // URI: Pickup
+
+ $pickups = spam_uri_pickup($target, $method);
+
+
+ // Return if ...
+ if (empty($pickups)) return $progress;
+
+ // Normalize all
+ $pickups = uri_pickup_normalize($pickups);
+
+ // ----------------------------------------
+ // Pickup some part of URI
+
+ $hosts = array();
+ foreach ($pickups as $key => $pickup) {
+ $hosts[$key] = & $pickup['host'];
}
- if (isset($uri['host']) && $uri['host'] !== '') {
- $tmp[] = & $uri['host'];
+
+ // ----------------------------------------
+ // URI: Bad host <pre-filter> (Separate good/bad hosts from $hosts)
+
+ if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
+ $list = get_blocklist('pre');
+ $blocked = blocklist_distiller($hosts, array_keys($list), $asap);
+ foreach($list as $key => $type){
+ if (! $type) unset($blocked[$key]); // Ignore goodhost etc
+ }
+ unset($list);
+ if (! empty($blocked)) $is_spam['badhost'] = TRUE;
}
- if (isset($uri['port']) && $uri['port'] !== '') {
- $tmp[] = ':';
- $tmp[] = & $uri['port'];
+
+ // Return if ...
+ if ($asap && $is_spam) return $progress;
+
+ // Remove blocked from $pickups
+ foreach(array_keys($pickups) as $key) {
+ if (! isset($hosts[$key])) {
+ unset($pickups[$key]);
+ }
}
- if (isset($uri['path']) && $uri['path'] !== '') {
- $tmp[] = & $uri['path'];
+
+ // ----------------------------------------
+ // URI: Check quantity
+
+ $sum['quantity'] += count($pickups);
+ // URI quantity
+ if ((! $asap || ! $is_spam) && isset($method['quantity']) &&
+ $sum['quantity'] > $method['quantity']) {
+ $is_spam['quantity'] = TRUE;
}
- if (isset($uri['file']) && $uri['file'] !== '') {
- $tmp[] = & $uri['file'];
+
+ // ----------------------------------------
+ // URI: used inside HTML anchor tag pair
+
+ if ((! $asap || ! $is_spam) && isset($method['uri_anchor'])) {
+ $key = 'uri_anchor';
+ foreach($pickups as $pickup) {
+ if (isset($pickup['area'][$key])) {
+ $sum[$key] += $pickup['area'][$key];
+ if(isset($method[$key]) &&
+ $sum[$key] > $method[$key]) {
+ $is_spam[$key] = TRUE;
+ if ($asap && $is_spam) break;
+ }
+ if ($asap && $is_spam) break;
+ }
+ }
}
- if (isset($uri['fragment']) && $uri['fragment'] !== '') {
- $tmp[] = '#';
- $tmp[] = & $uri['fragment'];
+
+ // ----------------------------------------
+ // URI: used inside 'BBCode' pair
+
+ if ((! $asap || ! $is_spam) && isset($method['uri_bbcode'])) {
+ $key = 'uri_bbcode';
+ foreach($pickups as $pickup) {
+ if (isset($pickup['area'][$key])) {
+ $sum[$key] += $pickup['area'][$key];
+ if(isset($method[$key]) &&
+ $sum[$key] > $method[$key]) {
+ $is_spam[$key] = TRUE;
+ if ($asap && $is_spam) break;
+ }
+ if ($asap && $is_spam) break;
+ }
+ }
}
- return implode('', $tmp);
-}
+ // ----------------------------------------
+ // URI: Uniqueness (and removing non-uniques)
-// ---------------------
-// Part One : Checker
+ if ((! $asap || ! $is_spam) && isset($method['non_uniquri'])) {
+
+ $uris = array();
+ foreach (array_keys($pickups) as $key) {
+ $uris[$key] = uri_pickup_implode($pickups[$key]);
+ }
+ $count = count($uris);
+ $uris = array_unique($uris);
+ $sum['non_uniquri'] += $count - count($uris);
+ if ($sum['non_uniquri'] > $method['non_uniquri']) {
+ $is_spam['non_uniquri'] = TRUE;
+ }
+ if (! $asap || ! $is_spam) {
+ foreach (array_diff(array_keys($pickups),
+ array_keys($uris)) as $remove) {
+ unset($pickups[$remove]);
+ }
+ }
+ unset($uris);
+ }
+ // Return if ...
+ if ($asap && $is_spam) return $progress;
-// TODO: globbing for IP address or something
-// TODO: Ignore list
-// TODO: require_or_include_once(another file)
-function is_badhost($host = '')
-{
- static $blocklist_regex;
+ // ----------------------------------------
+ // Host: Uniqueness (uniq / non-uniq)
+
+ $hosts = array_unique($hosts);
+
+ if (isset($sum['uniqhost'])) $sum['uniqhost'] += count($hosts);
+ if ((! $asap || ! $is_spam) && isset($method['non_uniqhost'])) {
+ $sum['non_uniqhost'] = $sum['quantity'] - $sum['uniqhost'];
+ if ($sum['non_uniqhost'] > $method['non_uniqhost']) {
+ $is_spam['non_uniqhost'] = TRUE;
+ }
+ }
+
+ // Return if ...
+ if ($asap && $is_spam) return $progress;
- if (! isset($blocklist_regex)) {
- $blocklist = array(
- '.blogspot.com',
+ // ----------------------------------------
+ // URI: Bad host (Separate good/bad hosts from $hosts)
+
+ if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
+ $list = get_blocklist('list');
+ $blocked = array_merge_leaves(
+ $blocked,
+ blocklist_distiller($hosts, array_keys($list), $asap),
+ FALSE
);
- foreach ($blocklist as $part) {
- $blocklist_regex[] = '#\b' . preg_quote($part, '#') . '$#';
+ foreach($list as $key=>$type){
+ if (! $type) unset($blocked[$key]); // Ignore goodhost etc
}
+ unset($list);
+ if (! empty($blocked)) $is_spam['badhost'] = TRUE;
}
- foreach ($blocklist_regex as $regex) {
- if (preg_match($regex, $host)) {
- return TRUE;
+ // Return if ...
+ //if ($asap && $is_spam) return $progress;
+
+ // ----------------------------------------
+ // End
+
+ return $progress;
+}
+
+// ---------------------
+// Reporting
+
+// Summarize $progress (blocked only)
+function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
+{
+ if ($blockedonly) {
+ $tmp = array_keys($progress['is_spam']);
+ } else {
+ $tmp = array();
+ $method = & $progress['method'];
+ if (isset($progress['sum'])) {
+ foreach ($progress['sum'] as $key => $value) {
+ if (isset($method[$key]) && $value) {
+ $tmp[] = $key . '(' . $value . ')';
+ }
+ }
}
}
- return FALSE;
+ return implode(', ', $tmp);
}
-// TODO return TRUE or FALSE!
-// Simple/fast spam check
-function is_uri_spam($target = '')
+function summarize_detail_badhost($progress = array())
{
- static $blocklist = array(
- '.blogspot.com',
- );
- static $blocklist_regex;
+ if (! isset($progress['blocked']) || empty($progress['blocked'])) return '';
+
+ // Flat per group
+ $blocked = array();
+ foreach($progress['blocked'] as $list => $lvalue) {
+ foreach($lvalue as $group => $gvalue) {
+ $flat = implode(', ', array_flat_leaves($gvalue));
+ if ($flat === $group) {
+ $blocked[$list][] = $flat;
+ } else {
+ $blocked[$list][$group] = $flat;
+ }
+ }
+ }
- if (! isset($blocklist_regex)) {
- foreach ($blocklist as $part) {
- $blocklist_regex[] = '#\b' . preg_quote($part, '#') . '$#';
+ // Shrink per list
+ // From: 'A-1' => array('ie.to')
+ // To: 'A-1' => 'ie.to'
+ foreach($blocked as $list => $lvalue) {
+ if (is_array($lvalue) &&
+ count($lvalue) == 1 &&
+ is_numeric(key($lvalue))) {
+ $blocked[$list] = current($lvalue);
}
}
- $is_spam = FALSE;
- $urinum = 0;
+ return var_export_shrink($blocked, TRUE, TRUE);
+}
- if (is_array($target)) {
- foreach($target as $str) {
- // Recurse
- list($is_spam, $_urinum) = is_uri_spam($str);
- $urinum += $_urinum;
- if ($is_spam) break;
+function summarize_detail_newtral($progress = array())
+{
+ if (! isset($progress['hosts']) ||
+ ! is_array($progress['hosts']) ||
+ empty($progress['hosts'])) return '';
+
+ // Generate a responsible $trie
+ $trie = array();
+ foreach($progress['hosts'] as $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'
}
- } else {
- $pickups = spam_uri_pickup($target);
- $urinum += count($pickups);
- if (! empty($pickups)) {
- // Some users want to post some URLs, but ...
- if ($urinum > 8) {
- $is_spam = TRUE; // Too many!
- } else {
- foreach($pickups as $pickup) {
- if ($pickup['area'] < 0) {
- $is_spam = TRUE;
- break;
- }
- }
- }
+ $trie = array_merge_leaves($trie, array($resp => array($rest => NULL)), FALSE);
+ }
- foreach ($pickups as $pickup) {
- if (is_badhost($pickup['host'])) {
- $is_spam = TRUE;
- break;
+ // 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; // 'example.com'
+ } else {
+ $subs[] = $sub . '. '; // 'A.foo.bar. '
}
}
+ $result[] = ' \'' . $key . '\' => \'' . implode(', ', $subs) . '\',';
}
+ unset($trie[$key]);
}
-
- return array($is_spam, $urinum);
+ return
+ 'array (' . "\n" .
+ implode("\n", $result) . "\n" .
+ ')';
}
+
// ---------------------
+// Exit
-// Check User-Agent (not testing yet)
-function is_invalid_useragent($ua_name = '' /*, $ua_vars = ''*/ )
+// Freeing memories
+function spam_dispose()
+{
+ get_blocklist(NULL);
+ whois_responsibility(NULL);
+}
+
+// Common bahavior for blocking
+// NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
+function spam_exit($mode = '', $data = array())
{
- return $ua_name === '';
+ $exit = TRUE;
+
+ switch ($mode) {
+ case '':
+ echo("\n");
+ break;
+ case 'dump':
+ echo('<pre>' . "\n");
+ echo htmlsc(var_export($data, TRUE));
+ echo('</pre>' . "\n");
+ break;
+ };
+
+ if ($exit) exit; // Force exit
}
+
// ---------------------
+// Simple filtering
-// TODO: Separate check-part(s) and mail part
-// TODO: Multi-metrics (uri, host, user-agent, ...)
-// TODO: Mail to administrator with more measurement data?
+// TODO: Record them
// Simple/fast spam filter ($target: 'a string' or an array())
-function pkwk_spamfilter($action, $page, $target = array('title' => ''))
+function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $exitmode = '')
{
- $is_spam = FALSE;
+ $progress = check_uri_spam($target, $method);
- //$is_spam = is_invalid_useragent('NOTYET');
- if ($is_spam) {
- $action .= ' (Invalid User-Agent)';
+ if (empty($progress['is_spam'])) {
+ spam_dispose();
} else {
- list($is_spam) = is_uri_spam($target);
- }
- if ($is_spam) {
- // Mail to administrator(s)
- global $notify, $notify_subject;
- if ($notify) {
- $footer['ACTION'] = $action;
- $footer['PAGE'] = '[blocked] ' . $page;
- $footer['URI'] = get_script_uri() . '?' . rawurlencode($page);
- $footer['USER_AGENT'] = TRUE;
- $footer['REMOTE_ADDR'] = TRUE;
- pkwk_mail_notify($notify_subject, var_export($target, TRUE), $footer);
- unset($footer);
- }
- }
+// 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 = & $tmp;
- if ($is_spam) spam_exit();
+ pkwk_spamnotify($action, $page, $target, $progress, $method);
+ spam_exit($exitmode, $progress);
+ }
}
// ---------------------
+// PukiWiki original
-// Common bahavior for blocking
-// NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
-function spam_exit()
+// Mail to administrator(s)
+function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array(), $method = array())
{
- die("\n");
+ global $notify, $notify_subject;
+
+ if (! $notify) return;
+
+ $asap = isset($method['asap']);
+
+ $summary['ACTION'] = 'Blocked by: ' . summarize_spam_progress($progress, TRUE);
+ if (! $asap) {
+ $summary['METRICS'] = summarize_spam_progress($progress);
+ }
+
+ $tmp = summarize_detail_badhost($progress);
+ if ($tmp != '') $summary['DETAIL_BADHOST'] = $tmp;
+
+ $tmp = summarize_detail_newtral($progress);
+ if (! $asap && $tmp != '') $summary['DETAIL_NEUTRAL_HOST'] = $tmp;
+
+ $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, TRUE);
}
?>
OSDN Git Service
