<?php
-// $Id: spam.php,v 1.53 2006/12/03 02:47:35 henoheno Exp $
-// Copyright (C) 2006 PukiWiki Developers Team
+// $Id: spam.php,v 1.152 2007/05/05 07:28:33 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): $method['uri_XXX'] related feature
if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
+// ---------------------
+// Compat etc
+
+// (PHP 4 >= 4.2.0): var_export(): mail-reporting and dump related
+if (! function_exists('var_export')) {
+ function var_export() {
+ return 'var_export() is not found on this server' . "\n";
+ }
+}
+
+// (PHP 4 >= 4.2.0): preg_grep() enables invert option
+function preg_grep_invert($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 $input;
+ }
+ }
+}
+
+// Remove redundant values from array()
+function array_unique_recursive($array = array())
+{
+ if (! is_array($array)) return $array;
+
+ $tmp = array();
+ foreach($array as $key => $value){
+ if (is_array($value)) {
+ $array[$key] = array_unique_recursive($value);
+ } else {
+ if (isset($tmp[$value])) {
+ unset($array[$key]);
+ } else {
+ $tmp[$value] = TRUE;
+ }
+ }
+ }
+
+ return $array;
+}
+
+// Roughly strings(1) using PCRE
+// This function is useful to:
+// * Reduce the size of data, from removing unprintable binary data
+// * Detect _bare_strings_ from binary data
+// References:
+// http://www.freebsd.org/cgi/man.cgi?query=strings (Man-page of GNU strings)
+// http://www.pcre.org/pcre.txt
+function strings($binary = '', $min_len = 4, $ignore_space = FALSE)
+{
+ if ($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) {
+ $min_len = min(1024, intval($min_len));
+ $binary =
+ implode("\n",
+ preg_grep('/^.{' . $min_len . ',}/S',
+ explode("\n", $binary)
+ )
+ );
+ }
+
+ return $binary;
+}
+
+
+// ---------------------
+// URI pickup
+
// 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, $preserve_chunk = TRUE)
+function uri_pickup($string = '')
{
+ if (! is_string($string)) return array();
+
// 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
+ '#(\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
- '([^\s<>"\'\[\]\#]+)?' . // 6: File and query string
- '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 7: Fragment
+ '([^\s<>"\'\[\]\#?]+)?' . // 6: File?
+ '(?:\?([^\s<>"\'\[\]\#]+))?' . // 7: Query string
+ '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 8: Fragment
'#i',
$string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE
);
- //var_dump(recursive_map('htmlspecialchars', $array));
- // Shrink $array
+ // Format the $array
static $parts = array(
1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port',
- 5 => 'path', 6 => 'file', 7 => 'fragment'
+ 5 => 'path', 6 => 'file', 7 => 'query', 8 => '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
+ $_uri = & $array[$uri];
+ array_rename_keys($_uri, $parts, TRUE, $default);
+ $offset = $_uri['scheme'][1]; // Scheme's offset = URI's offset
+ foreach(array_keys($_uri) as $part) {
+ $_uri[$part] = & $_uri[$part][0]; // Remove offsets
+ }
+ }
- foreach(array_keys($array[$uri]) as $part) {
- // Remove offsets for each part
- $array[$uri][$part] = & $array[$uri][$part][0];
+ foreach(array_keys($array) as $uri) {
+ $_uri = & $array[$uri];
+ if ($_uri['scheme'] === '') {
+ unset($array[$uri]); // Considererd harmless
+ continue;
}
+ unset($_uri[0]); // Matched string itself
+ $_uri['area']['offset'] = $offset; // Area offset for area_measure()
+ }
- if ($normalize) {
- $array[$uri]['scheme'] = scheme_normalize($array[$uri]['scheme']);
- if ($array[$uri]['scheme'] === '') {
- 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']);
- 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']
+ return $array;
+}
+
+// Normalize an array of URI arrays
+// NOTE: Give me the uri_pickup() results
+function uri_pickup_normalize(& $pickups, $destructive = TRUE)
+{
+ if (! is_array($pickups)) return $pickups;
+
+ if ($destructive) {
+ foreach (array_keys($pickups) as $key) {
+ $_key = & $pickups[$key];
+ $_key['scheme'] = isset($_key['scheme']) ? scheme_normalize($_key['scheme']) : '';
+ $_key['host'] = isset($_key['host']) ? host_normalize($_key['host']) : '';
+ $_key['port'] = isset($_key['port']) ? port_normalize($_key['port'], $_key['scheme'], FALSE) : '';
+ $_key['path'] = isset($_key['path']) ? strtolower(path_normalize($_key['path'])) : '';
+ $_key['file'] = isset($_key['file']) ? file_normalize($_key['file']) : '';
+ $_key['query'] = isset($_key['query']) ? query_normalize($_key['query']) : '';
+ $_key['fragment'] = isset($_key['fragment']) ? strtolower($_key['fragment']) : '';
+ }
+ } else {
+ foreach (array_keys($pickups) as $key) {
+ $_key = & $pickups[$key];
+ $_key['scheme'] = isset($_key['scheme']) ? scheme_normalize($_key['scheme']) : '';
+ $_key['host'] = isset($_key['host']) ? strtolower($_key['host']) : '';
+ $_key['port'] = isset($_key['port']) ? port_normalize($_key['port'], $_key['scheme'], FALSE) : '';
+ $_key['path'] = isset($_key['path']) ? path_normalize($_key['path']) : '';
+ }
+ }
+
+ return $pickups;
+}
+
+// An URI array => An URI (See uri_pickup())
+// USAGE:
+// $pickups = uri_pickup('a string include some URIs');
+// $uris = array();
+// foreach (array_keys($pickups) as $key) {
+// $uris[$key] = uri_pickup_implode($pickups[$key]);
+// }
+function uri_pickup_implode($uri = array())
+{
+ if (empty($uri) || ! is_array($uri)) return NULL;
+
+ $tmp = array();
+ if (isset($uri['scheme']) && $uri['scheme'] !== '') {
+ $tmp[] = & $uri['scheme'];
+ $tmp[] = '://';
+ }
+ if (isset($uri['userinfo']) && $uri['userinfo'] !== '') {
+ $tmp[] = & $uri['userinfo'];
+ $tmp[] = '@';
+ }
+ if (isset($uri['host']) && $uri['host'] !== '') {
+ $tmp[] = & $uri['host'];
+ }
+ if (isset($uri['port']) && $uri['port'] !== '') {
+ $tmp[] = ':';
+ $tmp[] = & $uri['port'];
+ }
+ if (isset($uri['path']) && $uri['path'] !== '') {
+ $tmp[] = & $uri['path'];
+ }
+ if (isset($uri['file']) && $uri['file'] !== '') {
+ $tmp[] = & $uri['file'];
+ }
+ if (isset($uri['query']) && $uri['query'] !== '') {
+ $tmp[] = '?';
+ $tmp[] = & $uri['query'];
+ }
+ if (isset($uri['fragment']) && $uri['fragment'] !== '') {
+ $tmp[] = '#';
+ $tmp[] = & $uri['fragment'];
+ }
+
+ return implode('', $tmp);
+}
+
+// $array['something'] => $array['wanted']
+function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '')
+{
+ if (! is_array($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] = $default;
+ }
+ }
+
+ return TRUE;
+}
+
+// ---------------------
+// Area pickup
+
+// Pickup all of markup areas
+function area_pickup($string = '', $method = array())
+{
+ $area = array();
+ if (empty($method)) return $area;
+
+ // Anchor tag pair by preg_match and preg_match_all()
+ // [OK] <a href></a>
+ // [OK] <a href= >Good site!</a>
+ // [OK] <a href= "#" >test</a>
+ // [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_
+ $regex = '#<a\b[^>]*\bhref\b[^>]*>.*?</a\b[^>]*(>)#is';
+ if (isset($method['area_anchor'])) {
+ $areas = array();
+ $count = isset($method['asap']) ?
+ preg_match($regex, $string) :
+ preg_match_all($regex, $string, $areas);
+ if (! empty($count)) $area['area_anchor'] = $count;
+ }
+ if (isset($method['uri_anchor'])) {
+ $areas = array();
+ preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
+ foreach(array_keys($areas) as $_area) {
+ $areas[$_area] = array(
+ $areas[$_area][0][1], // Area start (<a href>)
+ $areas[$_area][1][1], // Area end (</a>)
);
}
+ if (! empty($areas)) $area['uri_anchor'] = $areas;
+ }
- $array[$uri]['area']['offset'] = $offset;
+ // phpBB's "BBCode" pair by preg_match and preg_match_all()
+ // [OK] [url][/url]
+ // [OK] [url]http://nasty.example.com/[/url]
+ // [OK] [link]http://nasty.example.com/[/link]
+ // [OK] [url=http://nasty.example.com]visit http://nasty.example.com/[/url]
+ // [OK] [link http://nasty.example.com/]buy something[/link]
+ $regex = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#is';
+ if (isset($method['area_bbcode'])) {
+ $areas = array();
+ $count = isset($method['asap']) ?
+ preg_match($regex, $string) :
+ preg_match_all($regex, $string, $areas, PREG_SET_ORDER);
+ if (! empty($count)) $area['area_bbcode'] = $count;
+ }
+ if (isset($method['uri_bbcode'])) {
+ $areas = array();
+ preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE);
+ foreach(array_keys($areas) as $_area) {
+ $areas[$_area] = array(
+ $areas[$_area][0][1], // Area start ([url])
+ $areas[$_area][2][1], // Area end ([/url])
+ );
+ }
+ if (! empty($areas)) $area['uri_bbcode'] = $areas;
}
- return $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/]
+
+ return $area;
}
+// If in doubt, it's a little doubtful
+// if (Area => inside <= Area) $brief += -1
+function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset')
+{
+ if (! is_array($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] += $belief;
+ }
+ }
+ }
+ }
+}
+
+// ---------------------
+// Spam-uri pickup
+
// Domain exposure callback (See spam_uri_pickup_preprocess())
// http://victim.example.org/?foo+site:nasty.example.com+bar
// => http://nasty.example.com/?refer=victim.example.org
// Preprocess: rawurldecode() and adding space(s) and something
// to detect/count some URIs _if possible_
// NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
+// [OK] http://victim.example.org/?site:nasty.example.org
+// [OK] http://victim.example.org/nasty.example.org
// [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
// [OK] http://victim.example.org/http://nasty.example.org
function spam_uri_pickup_preprocess($string = '')
$string = rawurldecode($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://' .
+ '(' .
+ 'ime\.nu' . '|' . // 2ch.net
+ 'ime\.st' . '|' . // 2ch.net
+ 'link\.toolbot\.com' . '|' .
+ 'urlx\.org' .
+ ')' .
+ '/([a-z0-9.%_-]+\.[a-z0-9.%_-]+)#i', // nasty.example.org
+ 'http://$2/?refer=$1 $0', // Preserve $0 or remove?
+ $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
+ '#(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 $string;
}
-// Main function of spam-uri pickup
-function spam_uri_pickup($string = '', $area = array())
+// Main function of spam-uri pickup,
+// A wrapper function of uri_pickup()
+function spam_uri_pickup($string = '', $method = array())
{
- if (! is_array($area) || empty($area)) {
- $area = array(
- 'anchor' => TRUE,
- 'bbcode' => TRUE,
- );
+ if (! is_array($method) || empty($method)) {
+ $method = check_uri_spam_method();
}
$string = spam_uri_pickup_preprocess($string);
$array = uri_pickup($string);
- // Area elevation for '(especially external)link' intension
+ // Area elevation of URIs, for '(especially external)link' intension
if (! empty($array)) {
-
- $area_shadow = array();
- foreach(array_keys($array) as $key){
- $area_shadow[$key] = & $array[$key]['area'];
- $area_shadow[$key]['anchor'] = 0;
- $area_shadow[$key]['bbcode'] = 0;
- }
-
- // 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>
- if (isset($area['anchor'])) {
- $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>)
- );
+ $_method = array();
+ if (isset($method['uri_anchor'])) $_method['uri_anchor'] = & $method['uri_anchor'];
+ if (isset($method['uri_bbcode'])) $_method['uri_bbcode'] = & $method['uri_bbcode'];
+ $areas = area_pickup($string, $_method, TRUE);
+ if (! empty($areas)) {
+ $area_shadow = array();
+ foreach (array_keys($array) as $key) {
+ $area_shadow[$key] = & $array[$key]['area'];
+ foreach (array_keys($_method) as $_key) {
+ $area_shadow[$key][$_key] = 0;
+ }
}
- area_measure($areas, $area_shadow, 1, 'anchor');
- }
-
- // phpBB's "BBCode" by preg_match_all()
- // [url]http://nasty.example.com/[/url]
- // [link]http://nasty.example.com/[/link]
- // [url=http://nasty.example.com]visit http://nasty.example.com/[/url]
- // [link http://nasty.example.com/]buy something[/link]
- // ?? [url=][/url]
- if (isset($area['bbcode'])) {
- $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])
- );
+ foreach (array_keys($_method) as $_key) {
+ if (isset($areas[$_key])) {
+ area_measure($areas[$_key], $area_shadow, 1, $_key);
+ }
}
- area_measure($areas, $area_shadow, 1, 'bbcode');
- }
-
- // Various Wiki syntax
- // [text_or_uri>text_or_uri]
- // [text_or_uri:text_or_uri]
- // [text_or_uri|text_or_uri]
- // [text_or_uri->text_or_uri]
- // [text_or_uri text_or_uri] // MediaWiki
- // MediaWiki: [http://nasty.example.com/ visit http://nasty.example.com/]
-
- // Remove 'offset's for area_measure()
- foreach(array_keys($array) as $key)
- unset($array[$key]['area']['offset']);
- }
-
- return $array;
-}
-
-// $array['something'] => $array['wanted']
-function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '')
-{
- if (! is_array($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] = $default;
}
}
- return TRUE;
-}
-
-// If in doubt, it's a little doubtful
-function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset')
-{
- if (! is_array($areas) || ! is_array($array)) return;
+ // Remove 'offset's for area_measure()
+ foreach(array_keys($array) as $key)
+ unset($array[$key]['area']['offset']);
- $areas_keys = array_keys($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;
- }
- }
- }
- }
+ return $array;
}
// ---------------------
-// Part Two
+// Normalization
// Scheme normalization: Renaming the schemes
// snntp://example.org => nntps://example.org
// NOTE: Keep the static lists simple. See also port_normalize().
-function scheme_normalize($scheme = '', $considerd_harmfull = TRUE)
+function scheme_normalize($scheme = '', $abbrevs_harmfull = TRUE)
{
- // Abbreviations considerable they don't have link intension
+ // Abbreviations they have no intention of link
static $abbrevs = array(
'ttp' => 'http',
'ttps' => 'https',
);
- // Alias => normalized
+ // Aliases => normalized ones
static $aliases = array(
'pop' => 'pop3',
'news' => 'nntp',
'pops' => 'pop3s',
);
- $scheme = strtolower(trim($scheme));
+ if (! is_string($scheme)) return '';
+
+ $scheme = strtolower($scheme);
if (isset($abbrevs[$scheme])) {
- if ($considerd_harmfull) {
- $scheme = $abbrevs[$scheme];
- } else {
- $scheme = '';
- }
+ $scheme = $abbrevs_harmfull ? $abbrevs[$scheme] : '';
+ }
+ if (isset($aliases[$scheme])) {
+ $scheme = $aliases[$scheme];
}
- if (isset($aliases[$scheme])) $scheme = $aliases[$scheme];
return $scheme;
}
+// Hostname normlization (Destructive)
+// www.foo => www.foo ('foo' seems TLD)
+// www.foo.bar => foo.bar
+// www.10.20 => www.10.20 (Invalid hostname)
+// NOTE:
+// 'www' is mostly used as traditional hostname of WWW server.
+// 'www.foo.bar' may be identical with 'foo.bar'.
+function host_normalize($host = '')
+{
+ if (! is_string($host)) return '';
+
+ $host = strtolower($host);
+ $matches = array();
+ if (preg_match('/^www\.(.+\.[a-z]+)$/', $host, $matches)) {
+ return $matches[1];
+ } else {
+ return $host;
+ }
+}
+
// Port normalization: Suppress the (redundant) default port
// HTTP://example.org:80/ => http://example.org/
// HTTP://example.org:8080/ => http://example.org:8080/
// HTTPS://example.org:443/ => https://example.org/
-function port_normalize($port, $scheme, $scheme_normalize = TRUE)
+function port_normalize($port, $scheme, $scheme_normalize = FALSE)
{
// Schemes that users _maybe_ want to add protocol-handlers
// to their web browsers. (and attackers _maybe_ want to use ...)
'mysql' => 3306,
);
- $port = trim($port);
- if ($port === '') return $port;
+ // intval() converts '0-1' to '0', so preg_match() rejects these invalid ones
+ if (! is_numeric($port) || $port < 0 || preg_match('/[^0-9]/i', $port))
+ return '';
+ $port = intval($port);
if ($scheme_normalize) $scheme = scheme_normalize($scheme);
if (isset($array[$scheme]) && $port == $array[$scheme])
$port = ''; // Ignore the defaults
// http://example.org#hoge => http://example.org/#hoge
// http://example.org/path/a/b/./c////./d => http://example.org/path/a/b/c/d
// http://example.org/path/../../a/../back => http://example.org/back
-function path_normalize($path = '', $divider = '/', $addroot = TRUE)
+function path_normalize($path = '', $divider = '/', $add_root = TRUE)
{
- if (! is_string($path) || $path == '') {
- $path = $addroot ? $divider : '';
+ if (! is_string($divider)) return is_string($path) ? $path : '';
+
+ if ($add_root) {
+ $first_div = & $divider;
} else {
- $path = trim($path);
- $last = ($path[strlen($path) - 1] == $divider) ? $divider : '';
- $array = explode($divider, $path);
+ $first_div = '';
+ }
+ if (! is_string($path) || $path == '') return $first_div;
- // Remove paddings
- foreach(array_keys($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, $value);
- }
+ if (strpos($path, $divider, strlen($path) - strlen($divider)) === FALSE) {
+ $last_div = '';
+ } else {
+ $last_div = & $divider;
+ }
+
+ $array = explode($divider, $path);
+
+ // Remove paddings ('//' and '/./')
+ foreach(array_keys($array) as $key) {
+ if ($array[$key] == '' || $array[$key] == '.') {
+ unset($array[$key]);
}
- $array = & $tmp;
+ }
- $path = $addroot ? $divider : '';
- if (! empty($array)) $path .= implode($divider, $array) . $last;
+ // Remove back-tracks ('/../')
+ $tmp = array();
+ foreach($array as $value) {
+ if ($value == '..') {
+ array_pop($tmp);
+ } else {
+ array_push($tmp, $value);
+ }
}
+ $array = & $tmp;
- return $path;
+ if (empty($array)) {
+ return $first_div;
+ } else {
+ return $first_div . implode($divider, $array) . $last_div;
+ }
}
-// An URI array => An URI (See uri_pickup())
-function uri_array_implode($uri = array())
+// DirectoryIndex normalize (Destructive and rough)
+// TODO: sample.en.ja.html.gz => sample.html
+function file_normalize($file = 'index.html.en')
{
- if (empty($uri) || ! is_array($uri)) return NULL;
-
- $tmp = array();
- if (isset($uri['scheme']) && $uri['scheme'] !== '') {
- $tmp[] = & $uri['scheme'];
- $tmp[] = '://';
- }
- if (isset($uri['userinfo']) && $uri['userinfo'] !== '') {
- $tmp[] = & $uri['userinfo'];
- $tmp[] = '@';
- }
- if (isset($uri['host']) && $uri['host'] !== '') {
- $tmp[] = & $uri['host'];
- }
- if (isset($uri['port']) && $uri['port'] !== '') {
- $tmp[] = ':';
- $tmp[] = & $uri['port'];
+ static $simple_defaults = array(
+ 'default.htm' => TRUE,
+ 'default.html' => TRUE,
+ 'default.asp' => TRUE,
+ 'default.aspx' => TRUE,
+ 'index' => TRUE, // Some system can omit the suffix
+ );
+
+ static $content_suffix = array(
+ // index.xxx, sample.xxx
+ 'htm' => TRUE,
+ 'html' => TRUE,
+ 'shtml' => TRUE,
+ 'jsp' => TRUE,
+ 'php' => TRUE,
+ 'php3' => TRUE,
+ 'php4' => TRUE,
+ 'pl' => TRUE,
+ 'py' => TRUE,
+ 'rb' => TRUE,
+ 'cgi' => TRUE,
+ 'xml' => TRUE,
+ );
+
+ static $language_suffix = array(
+ // Reference: Apache 2.0.59 'AddLanguage' default
+ 'ca' => TRUE,
+ 'cs' => TRUE, // cs
+ 'cz' => TRUE, // cs
+ 'de' => TRUE,
+ 'dk' => TRUE, // da
+ 'el' => TRUE,
+ 'en' => TRUE,
+ 'eo' => TRUE,
+ 'es' => TRUE,
+ 'et' => TRUE,
+ 'fr' => TRUE,
+ 'he' => TRUE,
+ 'hr' => TRUE,
+ 'it' => TRUE,
+ 'ja' => TRUE,
+ 'ko' => TRUE,
+ 'ltz' => TRUE,
+ 'nl' => TRUE,
+ 'nn' => TRUE,
+ 'no' => TRUE,
+ 'po' => TRUE,
+ 'pt' => TRUE,
+ 'pt-br' => TRUE,
+ 'ru' => TRUE,
+ 'sv' => TRUE,
+ 'zh-cn' => TRUE,
+ 'zh-tw' => TRUE,
+
+ // Reference: Apache 2.0.59 default 'index.html' variants
+ 'ee' => TRUE,
+ 'lb' => TRUE,
+ 'var' => TRUE,
+ );
+
+ static $charset_suffix = array(
+ // Reference: Apache 2.0.59 'AddCharset' default
+ 'iso8859-1' => TRUE, // ISO-8859-1
+ 'latin1' => TRUE, // ISO-8859-1
+ 'iso8859-2' => TRUE, // ISO-8859-2
+ 'latin2' => TRUE, // ISO-8859-2
+ 'cen' => TRUE, // ISO-8859-2
+ 'iso8859-3' => TRUE, // ISO-8859-3
+ 'latin3' => TRUE, // ISO-8859-3
+ 'iso8859-4' => TRUE, // ISO-8859-4
+ 'latin4' => TRUE, // ISO-8859-4
+ 'iso8859-5' => TRUE, // ISO-8859-5
+ 'latin5' => TRUE, // ISO-8859-5
+ 'cyr' => TRUE, // ISO-8859-5
+ 'iso-ru' => TRUE, // ISO-8859-5
+ 'iso8859-6' => TRUE, // ISO-8859-6
+ 'latin6' => TRUE, // ISO-8859-6
+ 'arb' => TRUE, // ISO-8859-6
+ 'iso8859-7' => TRUE, // ISO-8859-7
+ 'latin7' => TRUE, // ISO-8859-7
+ 'grk' => TRUE, // ISO-8859-7
+ 'iso8859-8' => TRUE, // ISO-8859-8
+ 'latin8' => TRUE, // ISO-8859-8
+ 'heb' => TRUE, // ISO-8859-8
+ 'iso8859-9' => TRUE, // ISO-8859-9
+ 'latin9' => TRUE, // ISO-8859-9
+ 'trk' => TRUE, // ISO-8859-9
+ 'iso2022-jp'=> TRUE, // ISO-2022-JP
+ 'jis' => TRUE, // ISO-2022-JP
+ 'iso2022-kr'=> TRUE, // ISO-2022-KR
+ 'kis' => TRUE, // ISO-2022-KR
+ 'iso2022-cn'=> TRUE, // ISO-2022-CN
+ 'cis' => TRUE, // ISO-2022-CN
+ 'big5' => TRUE,
+ 'cp-1251' => TRUE, // ru, WINDOWS-1251
+ 'win-1251' => TRUE, // ru, WINDOWS-1251
+ 'cp866' => TRUE, // ru
+ 'koi8-r' => TRUE, // ru, KOI8-r
+ 'koi8-ru' => TRUE, // ru, KOI8-r
+ 'koi8-uk' => TRUE, // ru, KOI8-ru
+ 'ua' => TRUE, // ru, KOI8-ru
+ 'ucs2' => TRUE, // ru, ISO-10646-UCS-2
+ 'ucs4' => TRUE, // ru, ISO-10646-UCS-4
+ 'utf8' => TRUE,
+
+ // Reference: Apache 2.0.59 default 'index.html' variants
+ 'euc-kr' => TRUE,
+ 'gb2312' => TRUE,
+ );
+
+ // May uncompress by web browsers on the fly
+ // Must be at the last of the filename
+ // Reference: Apache 2.0.59 'AddEncoding'
+ static $encoding_suffix = array(
+ 'z' => TRUE,
+ 'gz' => TRUE,
+ );
+
+ if (! is_string($file)) return '';
+ $_file = strtolower($file);
+ if (isset($simple_defaults[$_file])) return '';
+
+
+ // Roughly removing language/character-set/encoding suffixes
+ // References:
+ // * Apache 2 document about 'Content-negotiaton', 'mod_mime' and 'mod_negotiation'
+ // http://httpd.apache.org/docs/2.0/content-negotiation.html
+ // http://httpd.apache.org/docs/2.0/mod/mod_mime.html
+ // http://httpd.apache.org/docs/2.0/mod/mod_negotiation.html
+ // * http://www.iana.org/assignments/character-sets
+ // * RFC3066: Tags for the Identification of Languages
+ // http://www.ietf.org/rfc/rfc3066.txt
+ // * ISO 639: codes of 'language names'
+ $suffixes = explode('.', $_file);
+ $body = array_shift($suffixes);
+ if ($suffixes) {
+ // Remove the last .gz/.z
+ $last_key = end(array_keys($suffixes));
+ if (isset($encoding_suffix[$suffixes[$last_key]])) {
+ unset($suffixes[$last_key]);
+ }
}
- if (isset($uri['path']) && $uri['path'] !== '') {
- $tmp[] = & $uri['path'];
+ // Cut language and charset suffixes
+ foreach($suffixes as $key => $value){
+ if (isset($language_suffix[$value]) || isset($charset_suffix[$value])) {
+ unset($suffixes[$key]);
+ }
}
- if (isset($uri['file']) && $uri['file'] !== '') {
- $tmp[] = & $uri['file'];
+ if (empty($suffixes)) return $body;
+
+ // Index.xxx
+ $count = count($suffixes);
+ reset($suffixes);
+ $current = current($suffixes);
+ if ($body == 'index' && $count == 1 && isset($content_suffix[$current])) return '';
+
+ return $file;
+}
+
+// Sort query-strings if possible (Destructive and rough)
+// [OK] &&&&f=d&b&d&c&a=0dd => a=0dd&b&c&d&f=d
+// [OK] nothing==&eg=dummy&eg=padding&eg=foobar => eg=foobar
+function query_normalize($string = '', $equal = TRUE, $equal_cutempty = TRUE, $stortolower = TRUE)
+{
+ if (! is_string($string)) return '';
+ if ($stortolower) $string = strtolower($string);
+
+ $array = explode('&', $string);
+
+ // Remove '&' paddings
+ foreach(array_keys($array) as $key) {
+ if ($array[$key] == '') {
+ unset($array[$key]);
+ }
}
- if (isset($uri['fragment']) && $uri['fragment'] !== '') {
- $tmp[] = '#';
- $tmp[] = & $uri['fragment'];
+
+ // Consider '='-sepalated input and paddings
+ if ($equal) {
+ $equals = $not_equals = array();
+ foreach ($array as $part) {
+ if (strpos($part, '=') === FALSE) {
+ $not_equals[] = $part;
+ } else {
+ list($key, $value) = explode('=', $part, 2);
+ $value = ltrim($value, '=');
+ if (! $equal_cutempty || $value != '') {
+ $equals[$key] = $value;
+ }
+ }
+ }
+
+ $array = & $not_equals;
+ foreach ($equals as $key => $value) {
+ $array[] = $key . '=' . $value;
+ }
+ unset($equals);
}
- return implode('', $tmp);
+ natsort($array);
+ return implode('&', $array);
}
// ---------------------
// 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), $divider)
- )
- ) .
- ')';
+ if (! is_string($string)) return '';
+
+ $string = str_replace($from, $mid, $string); // Hide
+ $string = preg_quote($string, $divider);
+ $string = str_replace($mid, $to, $string); // Unhide
+
+ return $string;
+}
+
+// Rough hostname checker
+// [OK] 192.168.
+// TODO: Strict digit, 0x, CIDR, IPv6
+function is_ip($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 = '', $asap = TRUE)
+// Generate host (FQDN, IPv4, ...) regex
+// 'localhost' : Matches with 'localhost' only
+// 'example.org' : Matches with 'example.org' only (See host_normalize() about 'www')
+// '.example.org' : Matches with ALL FQDN ended with '.example.org'
+// '*.example.org' : Almost the same of '.example.org' except 'www.example.org'
+// '10.20.30.40' : Matches with IPv4 address '10.20.30.40' only
+// [TODO] '192.' : Matches with all IPv4 hosts started with '192.'
+// TODO: IPv4, CIDR?, IPv6
+function generate_host_regex($string = '', $divider = '/')
{
- static $regex;
+ if (! is_string($string)) return '';
- if (! isset($regex)) {
- $regex = array();
- $regex['badhost'] = array();
+ if (mb_strpos($string, '.') === FALSE)
+ return generate_glob_regex($string, $divider);
- // Sample
- if (TRUE) {
- $blocklist['badhost'] = array(
- //'*', // Deny all uri
- //'10.20.*.*', // 10.20.example.com also matches
- //'*.blogspot.com', // Blog services subdomains
- //array('blogspot.com', '*.blogspot.com')
- );
- foreach ($blocklist['badhost'] as $part) {
- $regex['badhost'][] = '/^' . generate_glob_regex($part) . '$/i';
- }
+ $result = '';
+ if (is_ip($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) {
- $regex['badhost'][] = '/^' . generate_glob_regex($part) . '$/i';
+ include(SPAM_INI_FILE);
+ // $blocklist['badhost'] = array(
+ // '*.blogspot.com', // Blog services's subdomains (only)
+ // 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
+ // );
+ if (isset($blocklist['list'])) {
+ $regexs['list'] = & $blocklist['list'];
+ } else {
+ // Default
+ $blocklist['list'] = array(
+ 'goodhost' => FALSE,
+ 'badhost' => TRUE,
+ );
+ }
+ foreach(array_keys($blocklist['list']) 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 => $_value) {
+ get_blocklist_add($regexs[$_list][$key], $_key, $_value);
+ }
+ } else {
+ get_blocklist_add($regexs[$_list], $key, $value);
+ }
+ }
+ unset($blocklist[$_list]);
}
}
}
- //var_dump($regex);
- $result = 0;
- if (! is_array($hosts)) $hosts = array($hosts);
+ if ($list == '') {
+ return $regexs; // ALL
+ } else if (isset($regexs[$list])) {
+ return $regexs[$list];
+ } else {
+ return array();
+ }
+}
- foreach($hosts as $host) {
- if (! is_string($host)) $host = '';
+// Subroutine of get_blocklist()
+function get_blocklist_add(& $array, $key = 0, $value = '*.example.org')
+{
+ if (is_string($key)) {
+ $array[$key] = & $value; // Treat $value as a regex
+ } else {
+ $array[$value] = '/^' . generate_host_regex($value, '/') . '$/i';
+ }
+}
- // badhost
- foreach ($regex['badhost'] as $_regex) {
- if (preg_match($_regex, $host)) {
- ++$result;
- if ($asap) {
- return $result;
- } else {
- break;
+// Blocklist metrics: Separate $host, to $blocked and not blocked
+function blocklist_distiller(& $hosts, $keys = array('goodhost', 'badhost'), $asap = FALSE)
+{
+ if (! is_array($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 $result;
+ 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) : $blocked;
}
-// Default (enabled) methods and thresholds
+
+// Default (enabled) methods and thresholds (for content insertion)
function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
{
$times = intval($times);
$t_area = intval($t_area);
- // Thresholds
- $method = array(
- 'quantity' => 8 * $times, // Allow N URIs
- 'non_uniq' => 3 * $times, // Allow N duped (and normalized) URIs
- );
-
- // Areas
- $area = array(
- //'total' => $t_area, // Allow N areas total, enabled below
- 'anchor' => $t_area, // Inside <a href> HTML tag
- 'bbcode' => $t_area, // Inside [url] or [link] BBCode
- );
-
- // Rules
- $rules = array(
- 'badhost' => TRUE, // Check badhost
+ $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' => $t_area, // URI inside [url] or [link] BBCode
);
-
- // Remove unused
- foreach (array_keys($method) as $key) {
- if ($method[$key] < 0) unset($method[$key]);
+ if ($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 {
+ $bool = array();
}
- foreach (array_keys($area) as $key) {
- if ($area[$key] < 0) unset($area[$key]);
+
+ // Remove non-$positive values
+ foreach (array_keys($positive) as $key) {
+ if ($positive[$key] < 0) unset($positive[$key]);
}
- $area = empty($area) ? array() : array('area' => $area);
- $rules = $rule ? $rules : array();
- return $method + $area + $rules;
+ return $positive + $bool;
}
-
-// TODO: Simplify. !empty(['_action']) just means $is_spam
// Simple/fast spam check
-function check_uri_spam($target = '', $method = array(), $asap = TRUE)
+function check_uri_spam($target = '', $method = array())
{
- $is_spam = FALSE;
+ // Return value
$progress = array(
- 'quantity' => 0,
- 'area' => array(
- 'total' => 0,
- 'anchor' => 0,
- 'bbcode' => 0,
- ),
- 'non_uniq' => 0,
- 'uniqhost' => 0,
- 'badhost' => 0,
- '_action' => 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
+ ),
+ );
+
+ // Aliases
+ $sum = & $progress['sum'];
+ $is_spam = & $progress['is_spam'];
+ $progress['method'] = & $method; // Argument
+ $blocked = & $progress['blocked'];
+ $hosts = & $progress['hosts'];
+ $asap = isset($method['asap']);
+ // Init
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 (is_array($target)) {
- // Recurse
foreach($target as $str) {
- list($_is_spam, $_progress) = check_uri_spam($str, $method, $asap);
- $is_spam = $is_spam || $_is_spam;
- $progress['quantity'] += $_progress['quantity'];
- $progress['area']['total'] += $_progress['area']['total'];
- $progress['area']['anchor'] += $_progress['area']['anchor'];
- $progress['area']['bbcode'] += $_progress['area']['bbcode'];
- $progress['non_uniq'] += $_progress['non_uniq'];
- $progress['uniqhost'] += $_progress['uniqhost'];
- $progress['badhost'] += $_progress['badhost'];
- foreach($_progress['_action'] as $key => $value) {
- if (is_array($value)) {
- foreach(array_keys($value) as $_key) {
- $progress['_action'][$key][$_key] = TRUE;
- }
+ if (! is_string($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 {
- $progress['_action'][$key] = TRUE;
+ $sum[$key] += $_sum[$key];
}
}
- if ($is_spam && $asap) break;
+
+ // 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 $blocked
+ $blocked = array_merge_leaves($blocked, $_progress['blocked']);
+
+ // Merge $hosts
+ $hosts = array_merge_leaves($hosts, $_progress['hosts']);
}
- } else {
- $pickups = spam_uri_pickup($target);
- if (! empty($pickups)) {
- $progress['quantity'] += count($pickups);
-
- // URI quantity
- if ((! $is_spam || ! $asap) && isset($method['quantity']) &&
- $progress['quantity'] > $method['quantity']) {
- $is_spam = TRUE;
- $progress['_action']['quantity'] = TRUE;
+
+ // Recount $sum['badhost']
+ $sum['badhost'] = array_count_leaves($blocked);
+
+ return $progress;
+ }
+
+ // Area: There's HTML anchor tag
+ if ((! $asap || ! $is_spam) && isset($method['area_anchor'])) {
+ $key = 'area_anchor';
+ $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
+ $result = area_pickup($target, array($key => TRUE) + $_asap);
+ if ($result) {
+ $sum[$key] = $result[$key];
+ if (isset($method[$key]) && $sum[$key] > $method[$key]) {
+ $is_spam[$key] = TRUE;
}
- //var_dump($method['quantity'], $is_spam);
-
- // Using invalid area
- if ((! $is_spam || ! $asap) && isset($method['area'])) {
- foreach($pickups as $pickup) {
- foreach ($pickup['area'] as $key => $value) {
- if ($key == 'offset') continue;
- $progress['area']['total'] += $value;
- $progress['area'][$key] += $value;
- if (isset($method['area']['total']) &&
- $progress['area']['total'] > $method['area']['total']) {
- $is_spam = TRUE;
- $progress['_action']['area']['total'] = TRUE;
- if ($is_spam && $asap) break;
- }
- if(isset($method['area'][$key]) &&
- $progress['area'][$key] > $method['area'][$key]) {
- $is_spam = TRUE;
- $progress['_action']['area'][$key] = TRUE;
- if ($is_spam && $asap) break;
- }
- }
- if ($is_spam && $asap) break;
- }
+ }
+ }
+
+ // Area: There's 'BBCode' linking tag
+ if ((! $asap || ! $is_spam) && isset($method['area_bbcode'])) {
+ $key = 'area_bbcode';
+ $_asap = isset($method['asap']) ? array('asap' => TRUE) : array();
+ $result = area_pickup($target, array($key => TRUE) + $_asap);
+ if ($result) {
+ $sum[$key] = $result[$key];
+ if (isset($method[$key]) && $sum[$key] > $method[$key]) {
+ $is_spam[$key] = TRUE;
}
- //var_dump($method['area'], $is_spam);
+ }
+ }
- // URI uniqueness (and removing non-uniques)
- if ((! $is_spam || ! $asap) && isset($method['non_uniq'])) {
- $uris = array();
- foreach ($pickups as $key => $pickup) {
- $uris[$key] = uri_array_implode($pickup);
- }
- $count = count($uris);
- $uris = array_unique($uris);
- $progress['non_uniq'] += $count - count($uris);
- if ($progress['non_uniq'] > $method['non_uniq']) {
- $is_spam = TRUE;
- $progress['_action']['non_uniq'] = TRUE;
- }
- if (! $asap || ! $is_spam) {
- foreach (array_diff(array_keys($pickups),
- array_keys($uris)) as $remove) {
- unset($pickups[$remove]);
- }
+ // Return if ...
+ if ($asap && $is_spam) return $progress;
+
+ // URI: Pickup
+ $pickups = uri_pickup_normalize(spam_uri_pickup($target, $method));
+
+ // Return if ...
+ if (empty($pickups)) return $progress;
+
+ // URI: Check quantity
+ $sum['quantity'] += count($pickups);
+ // URI quantity
+ if ((! $asap || ! $is_spam) && isset($method['quantity']) &&
+ $sum['quantity'] > $method['quantity']) {
+ $is_spam['quantity'] = TRUE;
+ }
+
+ // 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 && $is_spam) break;
}
- unset($uris);
+ if ($asap && $is_spam) break;
}
- //var_dump($method['non_uniq'], $is_spam);
+ }
+ }
- // Unique host
- $hosts = array();
- foreach ($pickups as $pickup) {
- $hosts[] = & $pickup['host'];
- }
- $hosts = array_unique($hosts);
- $progress['uniqhost'] += count($hosts);
- //var_dump($method['uniqhost'], $is_spam);
-
- // Bad host
- if ((! $is_spam || ! $asap) && isset($method['badhost'])) {
- $count = is_badhost($hosts, $asap);
- $progress['badhost'] += $count;
- if ($count !== 0) {
- $progress['_action']['badhost'] = TRUE;
- $is_spam = TRUE;
+ // 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;
+ }
+ }
+ }
+
+ // URI: Uniqueness (and removing non-uniques)
+ 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[$remove]);
}
- //var_dump($method['badhost'], $is_spam);
}
+ unset($uris);
+ }
+
+ // Return if ...
+ if ($asap && $is_spam) return $progress;
+
+ // Host: Uniqueness (uniq / non-uniq)
+ foreach ($pickups as $pickup) $hosts[] = & $pickup['host'];
+ $hosts = array_unique($hosts);
+ $sum['uniqhost'] += count($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;
+
+ // URI: Bad host (Separate good/bad hosts from $hosts)
+ if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
+
+ // is_badhost()
+ $list = get_blocklist('list');
+ $blocked = blocklist_distiller($hosts, array_keys($list), $asap);
+ foreach($list as $key=>$type){
+ if (! $type) unset($blocked[$key]); // Ignore goodhost etc
+ }
+ unset($list);
+
+ if (! empty($blocked)) $is_spam['badhost'] = TRUE;
}
- return array($is_spam, $progress);
+ return $progress;
}
+// Count leaves (A leaf = value that is not an array, or an empty array)
+function array_count_leaves($array = array(), $count_empty = FALSE)
+{
+ if (! is_array($array) || (empty($array) && $count_empty)) return 1;
+
+ // Recurse
+ $count = 0;
+ foreach ($array as $part) {
+ $count += array_count_leaves($part, $count_empty);
+ }
+ return $count;
+}
+
+// Merge two leaves' value
+function array_merge_leaves(& $array1, & $array2, $unique_values = TRUE)
+{
+ // All NUMERIC keys are always renumbered from 0
+ $array = array_merge_recursive($array1, $array2);
+
+ // Redundant values (and keys) are vanished
+ if ($unique_values) $array = array_unique_recursive($array);
+
+ return $array;
+}
+
+
// ---------------------
// Reporting
-// TODO: Show all
+// TODO: Don't show unused $method!
// Summarize $progress (blocked only)
-function summarize_check_uri_spam_progress($progress = array(), $shownum = TRUE)
+function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
{
- //$list = check_uri_spam_method();
-
- $tmp = array();
- foreach (array_keys($progress['_action']) as $_action) {
- if (is_array($progress['_action'][$_action])) {
- foreach (array_keys($progress['_action'][$_action]) as $_area) {
- $tmp[] = $_action . '=>' . $_area .
- ($shownum ? '(' . $progress[$_action][$_area] . ')' : '');
+ if ($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 . ')';
+ }
}
- } else {
- $tmp[] = $_action .
- ($shownum ? '(' . $progress[$_action] . ')' : '');
}
}
return implode(', ', $tmp);
}
+// Very roughly, shrink the lines of var_export()
+// NOTE: If the same data exists, it must be corrupted.
+function var_export_shrink($expression, $return = FALSE)
+{
+ $result =preg_replace(
+ // Remove a newline and spaces
+ '# => \n *array \(#', ' => array (',
+ var_export($expression, TRUE)
+ );
+
+ if ($return) {
+ return $result;
+ } else {
+ echo $result;
+ return NULL;
+ }
+}
+
+function summarize_detail_badhost($progress = array())
+{
+ if (! isset($progress['is_spam']['badhost'])) return '';
+
+ return var_export_shrink($progress['blocked'], TRUE);
+}
+
+// Reverse $string with specified delimiter
+function delimiter_reverse($string = 'foo.bar.example.com', $from_delim = '.', $to_delim = '.')
+{
+ if (! is_string($string) || ! is_string($from_delim) || ! is_string($to_delim))
+ return $string;
+
+ // com.example.bar.foo
+ return implode($to_delim, array_reverse(explode($from_delim, $string)));
+}
+
+function summarize_detail_newtral($progress = array())
+{
+ if (! isset($progress['hosts']) ||
+ ! is_array($progress['hosts']) ||
+ empty($progress['hosts'])) return '';
+
+ // Sort by domain
+ $tmp = array();
+ foreach($progress['hosts'] as $value) {
+ $tmp[delimiter_reverse($value)] = $value;
+ }
+ ksort($tmp);
+
+ return count($tmp) . ' (' .implode(', ', $tmp) . ')';
+}
+
+
// ---------------------
// Exit
// Common bahavior for blocking
// NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
-function spam_exit()
+function spam_exit($mode = '', $data = array())
{
- die("\n");
+ switch ($mode) {
+ case '': echo("\n"); break;
+ case 'dump':
+ echo('<pre>' . "\n");
+ echo htmlspecialchars(var_export($data, TRUE));
+ echo('</pre>' . "\n");
+ break;
+ };
+
+ // Force exit
+ exit;
}
// ---------------------
-//
+// Simple filtering
+// TODO: Record them
// Simple/fast spam filter ($target: 'a string' or an array())
-function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $asap = FALSE)
+function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $exitmode = '')
{
- global $notify;
-
- list($is_spam, $progress) = check_uri_spam($target, $method, $asap);
+ $progress = check_uri_spam($target, $method);
- if ($is_spam) {
+ if (! empty($progress['is_spam'])) {
// Mail to administrator(s)
- if ($notify) pkwk_spamnotify($action, $page, $target, $progress);
- // End
- spam_exit();
+ pkwk_spamnotify($action, $page, $target, $progress, $method);
+
+ // Exit
+ spam_exit($exitmode, $progress);
}
}
// PukiWiki original
// Mail to administrator(s)
-function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array())
+function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array(), $method = array())
{
- global $notify_subject;
+ 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;
- $footer['ACTION'] = 'Blocked by: ' . summarize_check_uri_spam_progress($progress);
- $footer['COMMENT'] = $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);
+ $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);
}
?>