<?php
-// $Id: spam.php,v 1.120 2007/02/21 14:19:58 henoheno Exp $
+// $Id: spam.php,v 1.165 2007/05/20 02:22:56 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): $method['uri_XXX'] related feature
if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
}
}
+// ----
+
+// Very roughly, shrink the lines of var_export()
+// NOTE: If the same data exists, it must be corrupted.
+function var_export_shrink($expression, $return = FALSE, $ignore_numeric_keys = FALSE)
+{
+ $result =preg_replace(
+ // Remove a newline and spaces
+ '# => \n *array \(#', ' => array (',
+ var_export($expression, TRUE)
+ );
+
+ if ($ignore_numeric_keys) {
+ $result =preg_replace(
+ // Remove numeric keys
+ '#^( *)[0-9]+ => #m', '$1',
+ $result
+ );
+ }
+
+ if ($return) {
+ return $result;
+ } else {
+ echo $result;
+ return NULL;
+ }
+}
+
+// Remove redundant values from array()
+function array_unique_recursive($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;
+}
+
+// Renumber all numeric keys from 0
+function array_renumber_numeric_keys(& $array)
+{
+ if (! is_array($array)) return $array;
+
+ $count = -1;
+ $tmp = array();
+ foreach($array as $key => $value){
+ if (is_array($value)) array_renumber_numeric_keys($array[$key]); // Recurse
+ if (is_numeric($key)) $tmp[$key] = ++$count;
+ }
+ array_rename_keys($array, $tmp);
+
+ 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;
+}
+
+// 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, $string)));
+}
+
+
// ---------------------
// URI pickup
// [OK] http://nasty.example.org:80/foo/xxx#nasty_string/bar
// [OK] ftp://nasty.example.org:80/dfsdfs
// [OK] ftp://cnn.example.com&story=breaking_news@10.0.0.1/top_story.htm (from RFC3986)
-function uri_pickup($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
- '[a-z0-9.-]+' . // hostname(FQDN) : foo.example.org
+ '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' . // IPv4(dot-decimal): 001.22.3.44
+ '[a-z0-9][a-z0-9.-]+[a-z0-9]' . // hostname(FQDN) : foo.example.org
')' .
'(?::([0-9]*))?' . // 4: Port
'((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
$string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE
);
- // Shrink $array
+ // Format the $array
static $parts = array(
1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port',
5 => 'path', 6 => 'file', 7 => 'query', 8 => 'fragment'
foreach(array_keys($array) as $uri) {
$_uri = & $array[$uri];
array_rename_keys($_uri, $parts, TRUE, $default);
-
- $offset = $_uri['scheme'][1]; // Scheme's offset
+ $offset = $_uri['scheme'][1]; // Scheme's offset = URI's offset
foreach(array_keys($_uri) as $part) {
- // Remove offsets for each part
- $_uri[$part] = & $_uri[$part][0];
+ $_uri[$part] = & $_uri[$part][0]; // Remove offsets
}
+ }
- if ($normalize) {
- $_uri['scheme'] = scheme_normalize($_uri['scheme']);
- if ($_uri['scheme'] === '') {
- unset($array[$uri]);
- continue;
- }
- $_uri['host'] = strtolower($_uri['host']);
- $_uri['port'] = port_normalize($_uri['port'], $_uri['scheme'], FALSE);
- $_uri['path'] = path_normalize($_uri['path']);
- if ($preserve_rawuri) $_uri['rawuri'] = & $_uri[0];
-
- // DEBUG
- //$_uri['uri'] = uri_array_implode($_uri);
- } else {
- $_uri['uri'] = & $_uri[0]; // Raw
+ foreach(array_keys($array) as $uri) {
+ $_uri = & $array[$uri];
+ if ($_uri['scheme'] === '') {
+ unset($array[$uri]); // Considererd harmless
+ continue;
}
unset($_uri[0]); // Matched string itself
- if (! $preserve_chunk) {
- unset(
- $_uri['scheme'],
- $_uri['userinfo'],
- $_uri['host'],
- $_uri['port'],
- $_uri['path'],
- $_uri['file'],
- $_uri['query'],
- $_uri['fragment']
- );
- }
-
- // Area offset for area_measure()
- $_uri['area']['offset'] = $offset;
+ $_uri['area']['offset'] = $offset; // Area offset for area_measure()
}
return $array;
}
-// Destructive normalize of URI array
-// NOTE: Give me the uri_pickup() result with chunks
-function uri_array_normalize(& $pickups, $preserve = TRUE)
+// 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 $pickups;
- foreach (array_keys($pickups) as $key) {
- $_key = & $pickups[$key];
- $_key['path'] = isset($_key['path']) ? strtolower($_key['path']) : '';
- $_key['file'] = isset($_key['file']) ? file_normalize($_key['file']) : '';
- $_key['query'] = isset($_key['query']) ? query_normalize(strtolower($_key['query']), TRUE) : '';
- $_key['fragment'] = (isset($_key['fragment']) && $preserve) ?
- strtolower($_key['fragment']) : ''; // Just ignore
+ 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 $pickups;
}
// An URI array => An URI (See uri_pickup())
-function uri_array_implode($uri = array())
+// 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;
// [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[^>]*(>)#i';
+ $regex = '#<a\b[^>]*\bhref\b[^>]*>.*?</a\b[^>]*(>)#is';
if (isset($method['area_anchor'])) {
$areas = array();
$count = isset($method['asap']) ?
// [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 = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#i';
+ $regex = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#is';
if (isset($method['area_bbcode'])) {
$areas = array();
$count = isset($method['asap']) ?
// 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(
+ '#h?ttp://' .
+ '(' .
+ '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 (gate-big5)
+ // http://victim.example.org/gate/big5/nasty.example.org/path
+ // => http://nasty.example.org/?refer=victim.example.org and original
+ $string = preg_replace(
+ '#h?ttp://' .
+ '(' .
+ 'big5.51job.com' . '|' .
+ 'big5.china.com' . '|' .
+ 'big5.xinhuanet.com' . '|' .
+ ')' .
+ '/gate/big5' .
+ '/([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' .
+ '#i', // nasty.example.org
+ 'http://$2/?refer=$1 $0', // Preserve $0 or remove?
+ $string
+ );
+
// Domain exposure (See _preg_replace_callback_domain_exposure())
$string = preg_replace_callback(
array(
'_preg_replace_callback_domain_exposure',
$string
);
-
-
// URI exposure (uriuri => uri uri)
$string = preg_replace(
return $string;
}
-// Main function of spam-uri pickup
+// Main function of spam-uri pickup,
+// A wrapper function of uri_pickup()
function spam_uri_pickup($string = '', $method = array())
{
if (! is_array($method) || empty($method)) {
// Scheme normalization: Renaming the schemes
// snntp://example.org => nntps://example.org
// NOTE: Keep the static lists simple. See also port_normalize().
-function scheme_normalize($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 == '')
- return $addroot ? $divider : '';
+ if (! is_string($divider)) return is_string($path) ? $path : '';
+
+ if ($add_root) {
+ $first_div = & $divider;
+ } else {
+ $first_div = '';
+ }
+ if (! is_string($path) || $path == '') return $first_div;
+
+ if (strpos($path, $divider, strlen($path) - strlen($divider)) === FALSE) {
+ $last_div = '';
+ } else {
+ $last_div = & $divider;
+ }
- $path = trim($path);
- $last = ($path[strlen($path) - 1] == $divider) ? $divider : '';
$array = explode($divider, $path);
- // Remove paddings
+ // Remove paddings ('//' and '/./')
foreach(array_keys($array) as $key) {
- if ($array[$key] == '' || $array[$key] == '.')
+ if ($array[$key] == '' || $array[$key] == '.') {
unset($array[$key]);
+ }
}
- // Back-track
+
+ // Remove back-tracks ('/../')
$tmp = array();
foreach($array as $value) {
if ($value == '..') {
}
$array = & $tmp;
- $path = $addroot ? $divider : '';
- if (! empty($array)) $path .= implode($divider, $array) . $last;
-
- return $path;
+ if (empty($array)) {
+ return $first_div;
+ } else {
+ return $first_div . implode($divider, $array) . $last_div;
+ }
}
// DirectoryIndex normalize (Destructive and rough)
-function file_normalize($string = 'index.html.en')
+// TODO: sample.en.ja.html.gz => sample.html
+function file_normalize($file = 'index.html.en')
{
- static $array = array(
- 'index' => TRUE, // Some system can omit the suffix
- 'index.htm' => TRUE,
- 'index.html' => TRUE,
- 'index.shtml' => TRUE,
- 'index.jsp' => TRUE,
- 'index.php' => TRUE,
- 'index.php3' => TRUE,
- 'index.php4' => TRUE,
- //'index.pl' => TRUE,
- //'index.py' => TRUE,
- //'index.rb' => TRUE,
- 'index.cgi' => TRUE,
+ static $simple_defaults = array(
'default.htm' => TRUE,
'default.html' => TRUE,
'default.asp' => TRUE,
'default.aspx' => TRUE,
+ 'index' => TRUE, // Some system can omit the suffix
);
- // Content-negothiation filter:
- // Roughly removing ISO 639 -like
- // 2-letter suffixes (See RFC3066)
- $matches = array();
- if (preg_match('/(.*)\.[a-z][a-z](?:-[a-z][a-z])?$/i', $string, $matches)) {
- $_string = $matches[1];
- } else {
- $_string = & $string;
- }
+ static $content_suffix = array(
+ // index.xxx, sample.xxx
+ 'htm' => TRUE,
+ 'html' => TRUE,
+ 'shtml' => TRUE,
+ 'jsp' => TRUE,
+ 'php' => TRUE,
+ 'php3' => TRUE,
+ 'php4' => TRUE,
+ 'pl' => TRUE,
+ 'py' => TRUE,
+ 'rb' => TRUE,
+ 'cgi' => TRUE,
+ 'xml' => TRUE,
+ );
- if (isset($array[strtolower($_string)])) {
- return '';
- } else {
- return $string;
+ 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[$last_key]);
+ }
+ }
+ // Cut language and charset suffixes
+ foreach($suffixes as $key => $value){
+ if (isset($language_suffix[$value]) || isset($charset_suffix[$value])) {
+ unset($suffixes[$key]);
+ }
}
+ if (empty($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 = FALSE, $equal_cutempty = TRUE)
+function query_normalize($string = '', $equal = TRUE, $equal_cutempty = TRUE, $stortolower = TRUE)
{
+ if (! is_string($string)) return '';
+ if ($stortolower) $string = strtolower($string);
+
$array = explode('&', $string);
// Remove '&' paddings
// ---------------------
// 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 $string;
+ return 0; // Seems not IP
+ }
+}
+
+// Generate host (FQDN, IPv4, ...) regex
+// 'localhost' : Matches with 'localhost' only
+// 'example.org' : Matches with 'example.org' only (See host_normalize() about 'www')
+// '.example.org' : Matches with ALL FQDN ended with '.example.org'
+// '*.example.org' : Almost the same of '.example.org' except 'www.example.org'
+// '10.20.30.40' : Matches with IPv4 address '10.20.30.40' only
+// [TODO] '192.' : Matches with all IPv4 hosts started with '192.'
+// TODO: IPv4, CIDR?, IPv6
+function generate_host_regex($string = '', $divider = '/')
+{
+ if (! is_string($string)) return '';
+
+ 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('', $part);
}
}
function get_blocklist($list = '')
{
- static $regexs;
+ static $f_dispose = FALSE, $regexes;
- if (! isset($regexs)) {
- $regexs = array();
+ if ($list === NULL) {
+ $f_dispose = TRUE;
+ $regexes = NULL; // Unset
+ return array();
+ }
+
+ if (! isset($regexes)) {
+ if ($f_dispose === TRUE) die(__FUNCTION__ . '(): Memory already disposed');
+
+ $regexes = array();
if (file_exists(SPAM_INI_FILE)) {
$blocklist = array();
include(SPAM_INI_FILE);
// '*.blogspot.com', // Blog services's subdomains (only)
// 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
// );
- foreach(array('goodhost', 'badhost') as $_list) {
- if (! isset($blocklist[$list])) continue;
+ if (isset($blocklist['list'])) {
+ $regexes['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 => $value) {
- if (is_string($key)) {
- $regexs[$_list][$key] = $value;
+ if (is_array($value)) {
+ $regexes[$_list][$key] = array();
+ foreach($value as $_key => $_value) {
+ get_blocklist_add($regexes[$_list][$key], $_key, $_value);
+ }
} else {
- $regexs[$_list][$value] =
- '/^(?:www\.)?' . generate_glob_regex($value, '/') . '$/i';
+ get_blocklist_add($regexes[$_list], $key, $value);
}
}
+ unset($blocklist[$_list]);
}
}
}
- if ($list == '') {
- return $regexs; // ALL
- } else if (isset($regexs[$list])) {
- return $regexs[$list];
- } else {
+ if ($list === '') {
+ return $regexes; // ALL
+ } else if (isset($regexes[$list])) {
+ return $regexes[$list];
+ } else {
return array();
}
}
-function is_badhost($hosts = array(), $asap = TRUE, & $remains)
+// Subroutine of get_blocklist()
+function get_blocklist_add(& $array, $key = 0, $value = '*.example.org')
{
- $result = array();
- if (! is_array($hosts)) $hosts = array($hosts);
- foreach(array_keys($hosts) as $key) {
- if (! is_string($hosts[$key])) unset($hosts[$key]);
- }
- if (empty($hosts)) return $result;
-
- foreach (get_blocklist('goodhost') as $regex) {
- $hosts = preg_grep_invert($regex, $hosts);
+ if (is_string($key)) {
+ $array[$key] = & $value; // Treat $value as a regex
+ } else {
+ $array[$value] = '/^' . generate_host_regex($value, '/') . '$/i';
}
- if (empty($hosts)) return $result;
+}
- $tmp = array();
- foreach (get_blocklist('badhost') as $label => $regex) {
- $result[$label] = preg_grep($regex, $hosts);
- if (empty($result[$label])) {
- unset($result[$label]);
- } else {
- $hosts = array_diff($hosts, $result[$label]);
- if ($asap) 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;
+ }
+ }
}
}
- $remains = $hosts;
-
- return $result;
+ return $blocked;
}
// Default (enabled) methods and thresholds (for content insertion)
// Simple/fast spam check
function check_uri_spam($target = '', $method = array())
{
- if (! is_array($method) || empty($method)) {
- $method = check_uri_spam_method();
- }
+ // 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(
- 'quantity' => 0,
- 'uniqhost' => 0,
- 'non_uniqhost'=> 0,
- 'non_uniquri' => 0,
- 'badhost' => 0,
- 'area_anchor' => 0,
- 'area_bbcode' => 0,
- 'uri_anchor' => 0,
- 'uri_bbcode' => 0,
+ // 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
),
- 'is_spam' => array(),
- 'method' => & $method,
- 'remains' => array(),
);
+
+ // Aliases
$sum = & $progress['sum'];
$is_spam = & $progress['is_spam'];
- $remains = & $progress['remains'];
+ $progress['method'] = & $method; // Argument
+ $blocked = & $progress['blocked'];
+ $hosts = & $progress['hosts'];
$asap = isset($method['asap']);
- // Recurse
+ // 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($target)) {
foreach($target as $str) {
- // Recurse
- $_progress = check_uri_spam($str, $method);
- $_sum = & $_progress['sum'];
- $_is_spam = & $_progress['is_spam'];
- $_remains = & $_progress['remains'];
+ if (! is_string($str)) continue;
+
+ $_progress = check_uri_spam($str, $method); // Recurse
+
+ // Merge $sum
+ $_sum = & $_progress['sum'];
foreach (array_keys($_sum) as $key) {
- $sum[$key] += $_sum[$key];
- }
- foreach (array_keys($_is_spam) as $key) {
- if (is_array($_is_spam[$key])) {
- // Marge keys (badhost)
- foreach(array_keys($_is_spam[$key]) as $_key) {
- if (! isset($is_spam[$key][$_key])) {
- $is_spam[$key][$_key] = $_is_spam[$key][$_key];
- } else {
- $is_spam[$key][$_key] += $_is_spam[$key][$_key];
- }
- }
+ if (! isset($sum[$key])) {
+ $sum[$key] = & $_sum[$key];
} else {
- $is_spam[$key] = TRUE;
+ $sum[$key] += $_sum[$key];
}
}
- foreach ($_remains as $key=>$value) {
- foreach ($value as $_key=>$_value) {
- if (is_int($_key)) {
- $remains[$key][] = $_value;
- } else {
- $remains[$key][$_key] = $_value;
- }
- }
+
+ // 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
+ $blocked = array_merge_recursive($blocked, $_progress['blocked']);
+ $hosts = array_merge_recursive($hosts, $_progress['hosts']);
}
+
+ // Unique values
+ $blocked = array_unique_recursive($blocked);
+ $hosts = array_unique_recursive($hosts);
+
+ // Recount $sum['badhost']
+ $sum['badhost'] = array_count_leaves($blocked);
+
return $progress;
}
if ($asap && $is_spam) return $progress;
// URI: Pickup
- $pickups = spam_uri_pickup($target, $method);
- //$remains['uri_pickup'] = & $pickups;
+ $pickups = uri_pickup_normalize(spam_uri_pickup($target, $method));
// Return if ...
if (empty($pickups)) return $progress;
// URI: Uniqueness (and removing non-uniques)
if ((! $asap || ! $is_spam) && isset($method['non_uniquri'])) {
- // Destructive normalize of URIs
- uri_array_normalize($pickups);
-
$uris = array();
foreach (array_keys($pickups) as $key) {
- $uris[$key] = uri_array_implode($pickups[$key]);
+ $uris[$key] = uri_pickup_implode($pickups[$key]);
}
$count = count($uris);
$uris = array_unique($uris);
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;
- // URI: Bad host
+ // URI: Bad host (Separate good/bad hosts from $hosts)
if ((! $asap || ! $is_spam) && isset($method['badhost'])) {
- $__remains = array();
- if ($asap) {
- $badhost = is_badhost($hosts, $asap, $__remains);
- } else {
- $badhost = is_badhost($hosts, $asap, $__remains);
- if ($__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) {
- $is_spam['badhost'][$keys] =
- array_count_leaves($badhost[$keys]);
- }
- unset($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 $progress;
}
-// Count leaves
-function array_count_leaves($array = array(), $count_empty_array = FALSE)
+// Count leaves (A leaf = value that is not an array, or an empty array)
+function array_count_leaves($array = array(), $count_empty = FALSE)
{
- if (! is_array($array) || (empty($array) && $count_empty_array))
- return 1;
+ if (! is_array($array) || (empty($array) && $count_empty)) return 1;
// Recurse
- $result = 0;
+ $count = 0;
foreach ($array as $part) {
- $result += array_count_leaves($part, $count_empty_array);
+ $count += array_count_leaves($part, $count_empty);
}
- return $result;
+ return $count;
}
+// An array-leaves to a flat array
+function array_flat_leaves($array, $unique = TRUE)
+{
+ if (! is_array($array)) return $array;
+
+ $tmp = array();
+ foreach(array_keys($array) as $key) {
+ if (is_array($array[$key])) {
+ // Recurse
+ foreach(array_flat_leaves($array[$key]) as $_value) {
+ $tmp[] = $_value;
+ }
+ } else {
+ $tmp[] = & $array[$key];
+ }
+ }
+
+ return $unique ? array_values(array_unique($tmp)) : $tmp;
+}
+
+// An array() to an array leaf
+function array_leaf($array = array('A', 'B', 'C.D'), $stem = FALSE, $edge = array())
+{
+ $leaf = array();
+ $tmp = & $leaf;
+ foreach($array as $arg) {
+ if (! is_string($arg) && ! is_int($arg)) continue;
+ $tmp[$arg] = array();
+ $parent = & $tmp;
+ $tmp = & $tmp[$arg];
+ }
+ if ($stem) {
+ $parent[key($parent)] = & $edge;
+ } else {
+ $parent = key($parent);
+ }
+
+ return $leaf; // array('A' => array('B' => 'C.D'))
+}
+
+
// ---------------------
// Reporting
-// TODO: Don't show unused $method!
// Summarize $progress (blocked only)
function summarize_spam_progress($progress = array(), $blockedonly = FALSE)
{
$method = & $progress['method'];
if (isset($progress['sum'])) {
foreach ($progress['sum'] as $key => $value) {
- if (isset($method[$key])) {
+ if (isset($method[$key]) && $value) {
$tmp[] = $key . '(' . $value . ')';
}
}
return implode(', ', $tmp);
}
+function summarize_detail_badhost($progress = array())
+{
+ 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] = $flat;
+ }
+ }
+ }
+
+ // Shrink per list
+ // From: 'A-1' => array('ie.to')
+ // To: 'A-1' => 'ie.to'
+ foreach($blocked as $list => $lvalue) {
+ if (is_array($lvalue) &&
+ count($lvalue) == 1 &&
+ is_numeric(key($lvalue))) {
+ $blocked[$list] = current($lvalue);
+ }
+ }
+
+ return var_export_shrink($blocked, TRUE, TRUE);
+}
+
+function summarize_detail_newtral($progress = array())
+{
+ if (! isset($progress['hosts']) ||
+ ! is_array($progress['hosts']) ||
+ empty($progress['hosts'])) return '';
+
+ $result = '';
+ if (FALSE) {
+ // Sort by domain
+ $tmp = array();
+ foreach($progress['hosts'] as $value) {
+ $tmp[delimiter_reverse($value)] = $value;
+ }
+ ksort($tmp, SORT_STRING);
+ $result = count($tmp) . ' (' .implode(', ', $tmp) . ')';
+ } else {
+ $tmp = array();
+ foreach($progress['hosts'] as $value) {
+ $tmp = array_merge_recursive(
+ $tmp,
+ array_leaf(explode('.', delimiter_reverse($value) . '.'), TRUE, $value)
+ );
+ }
+ ksort($tmp, SORT_STRING);
+
+ separate_and_joinkey_leaves($tmp, '.', TRUE, TRUE);
+ separate_and_joinkey_leaves($tmp, '.', TRUE, FALSE);
+ separate_and_joinkey_leaves($tmp, '.', TRUE, FALSE);
+ //separate_and_joinkey_leaves($tmp, '.', TRUE, FALSE);
+
+ foreach($tmp as $key => $value) {
+ if (is_array($value)) {
+ ksort($tmp[$key]);
+ // $tmp[$key] = implode(', ', array_flat_leaves($value));
+ }
+ }
+
+ $result = var_export_shrink($tmp, TRUE, TRUE);
+ }
+
+ return $result;
+}
+
+function separate_and_joinkey_leaves(
+ & $array, // array('A' => array('B' => 'C.D')),
+ $delim = '.', $reversejoin = FALSE, $allowmulti = FALSE)
+{
+ if (! is_array($array)) return $array;
+
+ $result = array();
+ foreach(array_keys($array) as $key) {
+ if (! is_array($array[$key]) || (! $allowmulti && count($array[$key]) > 1)) {
+ $result[$key] = & $array[$key]; // Do nothing
+ } else {
+ foreach(array_keys($array[$key]) as $_key) {
+ $joinkey = $reversejoin ?
+ $_key . $delim . $key :
+ $key . $delim . $_key;
+ $result[$joinkey] = & $array[$key][$_key];
+ }
+ }
+ }
+
+ $array = & $result;
+
+ return $result; // array('A.B' => 'C.D')
+}
+
+
// ---------------------
// Exit
// NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked'
function spam_exit($mode = '', $data = array())
{
+ // Dispose
+ get_blocklist(NULL);
+
+ $exit = TRUE;
switch ($mode) {
- case '': echo("\n"); break;
+ case '':
+ echo("\n");
+ break;
case 'dump':
echo('<pre>' . "\n");
echo htmlspecialchars(var_export($data, TRUE));
break;
};
- // Force exit
- exit;
+ if ($exit) exit; // Force exit
}
if (! $asap) {
$summary['METRICS'] = summarize_spam_progress($progress);
}
- if (isset($progress['is_spam']['badhost'])) {
- $badhost = array();
- foreach($progress['is_spam']['badhost'] as $glob=>$number) {
- $badhost[] = $glob . '(' . $number . ')';
- }
- $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']))
- ) .
- ')';
- }
+
+ $tmp = summarize_detail_badhost($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);