<?php
-// $Id: spam.php,v 1.122 2007/03/04 03:59:37 henoheno Exp $
+// $Id: spam.php,v 1.197 2007/07/03 14:51:07 henoheno Exp $
// Copyright (C) 2006-2007 PukiWiki Developers Team
// License: GPL v2 or (at your option) any later version
//
//
// (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');
+if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php');
+if (! defined('DOMAIN_INI_FILE')) define('DOMAIN_INI_FILE', 'domain.ini.php');
// ---------------------
// Compat etc
}
}
+
// ---------------------
-// 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, $preserve_chunk = TRUE)
+// Utilities
+
+// Very roughly, shrink the lines of var_export()
+// NOTE: If the same data exists, it must be corrupted.
+function var_export_shrink($expression, $return = FALSE, $ignore_numeric_keys = FALSE)
{
- // 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
- '[a-z0-9.-]+' . // hostname(FQDN) : foo.example.org
- ')' .
- '(?::([0-9]*))?' . // 4: Port
- '((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info
- '([^\s<>"\'\[\]\#?]+)?' . // 6: File?
- '(?:\?([^\s<>"\'\[\]\#]+))?' . // 7: Query string
- '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 8: Fragment
- '#i',
- $string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE
- );
+ $result = var_export($expression, TRUE);
- // Shrink $array
- static $parts = array(
- 1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port',
- 5 => 'path', 6 => 'file', 7 => 'query', 8 => 'fragment'
+ $result = preg_replace(
+ // Remove a newline and spaces
+ '# => \n *array \(#', ' => array (',
+ $result
);
- $default = array('');
- foreach(array_keys($array) as $uri) {
- $_uri = & $array[$uri];
- array_rename_keys($_uri, $parts, TRUE, $default);
-
- $offset = $_uri['scheme'][1]; // Scheme's offset
- foreach(array_keys($_uri) as $part) {
- // Remove offsets for each part
- $_uri[$part] = & $_uri[$part][0];
- }
-
- 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'] = & $_uri[0]; // Raw
- }
- unset($_uri[0]); // Matched string itself
- if (! $preserve_chunk) {
- unset(
- $_uri['scheme'],
- $_uri['userinfo'],
- $_uri['host'],
- $_uri['port'],
- $_uri['path'],
- $_uri['file'],
- $_uri['query'],
- $_uri['fragment']
- );
- }
+ if ($ignore_numeric_keys) {
+ $result =preg_replace(
+ // Remove numeric keys
+ '#^( *)[0-9]+ => #m', '$1',
+ $result
+ );
+ }
- // Area offset for area_measure()
- $_uri['area']['offset'] = $offset;
+ if ($return) {
+ return $result;
+ } else {
+ echo $result;
+ return NULL;
}
+}
- return $array;
+// 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)));
}
-// Destructive normalize of URI array
-// NOTE: Give me the uri_pickup() result with chunks
-function uri_array_normalize(& $pickups, $preserve = TRUE)
+// ksort() by domain
+function ksort_by_domain(& $array)
{
- if (! is_array($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
+ $sort = array();
+ foreach(array_keys($array) as $key) {
+ $sort[delimiter_reverse($key)] = $key;
+ }
+ ksort($sort, SORT_STRING);
+ $result = array();
+ foreach($sort as $key) {
+ $result[$key] = & $array[$key];
}
+ $array = $result;
+}
- return $pickups;
+// Roughly strings(1) using PCRE
+// This function is useful to:
+// * Reduce the size of data, from removing unprintable binary data
+// * Detect _bare_strings_ from binary data
+// References:
+// http://www.freebsd.org/cgi/man.cgi?query=strings (Man-page of GNU strings)
+// http://www.pcre.org/pcre.txt
+// Note: mb_ereg_replace() is one of mbstring extension's functions
+// and need to init its encoding.
+function strings($binary = '', $min_len = 4, $ignore_space = FALSE, $multibyte = FALSE)
+{
+ // String only
+ $binary = (is_array($binary) || $binary === TRUE) ? '' : strval($binary);
+
+ $regex = $ignore_space ?
+ '[^[:graph:] \t\n]+' : // Remove "\0" etc, and readable spaces
+ '[^[:graph:][:space:]]+'; // Preserve readable spaces if possible
+
+ $binary = $multibyte ?
+ mb_ereg_replace($regex, "\n", $binary) :
+ preg_replace('/' . $regex . '/s', "\n", $binary);
+
+ if ($ignore_space) {
+ $binary = preg_replace(
+ array(
+ '/[ \t]{2,}/',
+ '/^[ \t]/m',
+ '/[ \t]$/m',
+ ),
+ array(
+ ' ',
+ '',
+ ''
+ ),
+ $binary);
+ }
+
+ if ($min_len > 1) {
+ // The last character seems "\n" or not
+ $br = (! empty($binary) && $binary[strlen($binary) - 1] == "\n") ? "\n" : '';
+
+ $min_len = min(1024, intval($min_len));
+ $regex = '/^.{' . $min_len . ',}/S';
+ $binary = implode("\n", preg_grep($regex, explode("\n", $binary))) . $br;
+ }
+
+ return $binary;
}
-// An URI array => An URI (See uri_pickup())
-function uri_array_implode($uri = array())
+
+// ---------------------
+// Utilities: Arrays
+
+// Count leaves (A leaf = value that is not an array, or an empty array)
+function array_count_leaves($array = array(), $count_empty = FALSE)
{
- if (empty($uri) || ! is_array($uri)) return NULL;
+ if (! is_array($array) || (empty($array) && $count_empty)) return 1;
- $tmp = array();
- if (isset($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'];
+ // Recurse
+ $count = 0;
+ foreach ($array as $part) {
+ $count += array_count_leaves($part, $count_empty);
}
- if (isset($uri['fragment']) && $uri['fragment'] !== '') {
- $tmp[] = '#';
- $tmp[] = & $uri['fragment'];
+ 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 implode('', $tmp);
+ return $unique ? array_values(array_unique($tmp)) : $tmp;
}
// $array['something'] => $array['wanted']
return TRUE;
}
-// ---------------------
-// Area pickup
-
-// Pickup all of markup areas
-function area_pickup($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_
- $regex = '#<a\b[^>]*\bhref\b[^>]*>.*?</a\b[^>]*(>)#i';
- if (isset($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;
- }
-
- // 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 = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#i';
- 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;
- }
-
- // 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] += $belief;
- }
- }
- }
- }
-}
-
-// ---------------------
-// Spam-uri pickup
-
-// Domain exposure callback (See spam_uri_pickup_preprocess())
-// http://victim.example.org/?foo+site:nasty.example.com+bar
-// => http://nasty.example.com/?refer=victim.example.org
-// NOTE: 'refer=' is not so good for (at this time).
-// Consider about using IP address of the victim, try to avoid that.
-function _preg_replace_callback_domain_exposure($matches = array())
-{
- $result = '';
-
- // Preserve the victim URI as a complicity or ...
- if (isset($matches[5])) {
- $result =
- $matches[1] . '://' . // scheme
- $matches[2] . '/' . // victim.example.org
- $matches[3]; // The rest of all (before victim)
- }
-
- // Flipped URI
- if (isset($matches[4])) {
- $result =
- $matches[1] . '://' . // scheme
- $matches[4] . // nasty.example.com
- '/?refer=' . strtolower($matches[2]) . // victim.example.org
- ' ' . $result;
- }
-
- return $result;
-}
-
-// Preprocess: rawurldecode() and adding space(s) and something
-// to detect/count some URIs _if possible_
-// NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:']
-// [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org
-// [OK] http://victim.example.org/http://nasty.example.org
-// TODO: link.toolbot.com, urlx.org
-function spam_uri_pickup_preprocess($string = '')
-{
- if (! is_string($string)) return '';
-
- $string = rawurldecode($string);
-
- // Domain exposure (See _preg_replace_callback_domain_exposure())
- $string = preg_replace_callback(
- array(
- '#(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',
- ),
- '_preg_replace_callback_domain_exposure',
- $string
- );
-
- // URI exposure (uriuri => uri uri)
- $string = preg_replace(
- array(
- '#(?<! )(?:https?|ftp):/#i',
- // '#[a-z][a-z0-9.+-]{1,8}://#i',
- // '#[a-z][a-z0-9.+-]{1,8}://#i'
- ),
- ' $0',
- $string
- );
-
- return $string;
-}
-
-// Main function of spam-uri pickup
-function spam_uri_pickup($string = '', $method = array())
-{
- if (! is_array($method) || empty($method)) {
- $method = check_uri_spam_method();
- }
-
- $string = spam_uri_pickup_preprocess($string);
-
- $array = uri_pickup($string);
-
- // Area elevation of URIs, for '(especially external)link' intension
- if (! empty($array)) {
- $_method = array();
- if (isset($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][$_key] = 0;
- }
- }
- foreach (array_keys($_method) as $_key) {
- if (isset($areas[$_key])) {
- area_measure($areas[$_key], $area_shadow, 1, $_key);
- }
- }
- }
- }
-
- // Remove 'offset's for area_measure()
- foreach(array_keys($array) as $key)
- unset($array[$key]['area']['offset']);
-
- return $array;
-}
-
-
-// ---------------------
-// Normalization
-
-// Scheme normalization: Renaming the schemes
-// snntp://example.org => nntps://example.org
-// NOTE: Keep the static lists simple. See also port_normalize().
-function scheme_normalize($scheme = '', $considerd_harmfull = TRUE)
-{
- // Abbreviations considerable they don't have link intension
- static $abbrevs = array(
- 'ttp' => 'http',
- 'ttps' => 'https',
- );
-
- // Alias => normalized
- static $aliases = array(
- 'pop' => 'pop3',
- 'news' => 'nntp',
- 'imap4' => 'imap',
- 'snntp' => 'nntps',
- 'snews' => 'nntps',
- 'spop3' => 'pop3s',
- 'pops' => 'pop3s',
- );
-
- $scheme = strtolower(trim($scheme));
- if (isset($abbrevs[$scheme])) {
- if ($considerd_harmfull) {
- $scheme = $abbrevs[$scheme];
- } else {
- $scheme = '';
- }
- }
- if (isset($aliases[$scheme])) $scheme = $aliases[$scheme];
-
- return $scheme;
-}
-
-// Port normalization: Suppress the (redundant) default port
-// HTTP://example.org:80/ => http://example.org/
-// HTTP://example.org:8080/ => http://example.org:8080/
-// HTTPS://example.org:443/ => https://example.org/
-function port_normalize($port, $scheme, $scheme_normalize = TRUE)
+// Remove redundant values from array()
+function array_unique_recursive($array = 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,
- );
-
- $port = trim($port);
- if ($port === '') return $port;
-
- if ($scheme_normalize) $scheme = scheme_normalize($scheme);
- if (isset($array[$scheme]) && $port == $array[$scheme])
- $port = ''; // Ignore the defaults
-
- return $port;
-}
+ if (! is_array($array)) return $array;
-// Path normalization
-// http://example.org => http://example.org/
-// http://example.org#hoge => http://example.org/#hoge
-// http://example.org/path/a/b/./c////./d => http://example.org/path/a/b/c/d
-// http://example.org/path/../../a/../back => http://example.org/back
-function path_normalize($path = '', $divider = '/', $addroot = TRUE)
-{
- if (! is_string($path) || $path == '')
- return $addroot ? $divider : '';
-
- $path = trim($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($tmp);
+ foreach($array as $key => $value){
+ if (is_array($value)) {
+ $array[$key] = array_unique_recursive($value);
} else {
- array_push($tmp, $value);
- }
- }
- $array = & $tmp;
-
- $path = $addroot ? $divider : '';
- if (! empty($array)) $path .= implode($divider, $array) . $last;
-
- return $path;
-}
-
-// DirectoryIndex normalize (Destructive and rough)
-function file_normalize($string = '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,
- 'default.htm' => TRUE,
- 'default.html' => TRUE,
- 'default.asp' => TRUE,
- 'default.aspx' => TRUE,
- );
-
- // 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 = & $string;
- }
-
- if (isset($array[strtolower($_string)])) {
- return '';
- } else {
- return $string;
- }
-}
-
-// Sort query-strings if possible (Destructive and rough)
-// [OK] &&&&f=d&b&d&c&a=0dd => a=0dd&b&c&d&f=d
-// [OK] nothing==&eg=dummy&eg=padding&eg=foobar => eg=foobar
-function query_normalize($string = '', $equal = FALSE, $equal_cutempty = TRUE)
-{
- $array = explode('&', $string);
-
- // Remove '&' paddings
- foreach(array_keys($array) as $key) {
- if ($array[$key] == '') {
- unset($array[$key]);
- }
- }
-
- // Consider '='-sepalated input and paddings
- if ($equal) {
- $equals = $not_equals = array();
- foreach ($array as $part) {
- if (strpos($part, '=') === FALSE) {
- $not_equals[] = $part;
+ if (isset($tmp[$value])) {
+ unset($array[$key]);
} else {
- list($key, $value) = explode('=', $part, 2);
- $value = ltrim($value, '=');
- if (! $equal_cutempty || $value != '') {
- $equals[$key] = $value;
- }
+ $tmp[$value] = TRUE;
}
}
-
- $array = & $not_equals;
- foreach ($equals as $key => $value) {
- $array[] = $key . '=' . $value;
- }
- unset($equals);
}
- natsort($array);
- return implode('&', $array);
+ return $array;
}
+
// ---------------------
// Part One : Checker
// 23 => ']',
);
- if (is_array($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 $string;
+}
+
+// Generate host (FQDN, IPv4, ...) regex
+// 'localhost' : Matches with 'localhost' only
+// 'example.org' : Matches with 'example.org' only (See host_normalize() about 'www')
+// '.example.org' : Matches with ALL FQDN ended with '.example.org'
+// '*.example.org' : Almost the same of '.example.org' except 'www.example.org'
+// '10.20.30.40' : Matches with IPv4 address '10.20.30.40' only
+// [TODO] '192.' : Matches with all IPv4 hosts started with '192.'
+// TODO: IPv4, CIDR?, IPv6
+function generate_host_regex($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 {
- $string = str_replace($from, $mid, $string); // Hide
- $string = preg_quote($string, $divider);
- $string = str_replace($mid, $to, $string); // Unhide
- return $string;
+ // 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('', $part);
+ }
+}
+
+// Rough hostname checker
+// [OK] 192.168.
+// TODO: Strict digit, 0x, CIDR, IPv6
+function is_ip($string = '')
+{
+ if (preg_match('/^' .
+ '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' .
+ '(?:[0-9]{1,3}\.){1,3}' . '$/',
+ $string)) {
+ return 4; // Seems IPv4(dot-decimal)
+ } else {
+ return 0; // Seems not IP
}
}
function get_blocklist($list = '')
{
- static $regexs;
+ static $regexes;
- if (! isset($regexs)) {
- $regexs = array();
+ if ($list === NULL) {
+ $regexes = NULL; // Unset
+ return array();
+ }
+
+ if (! isset($regexes)) {
+ $regexes = array();
if (file_exists(SPAM_INI_FILE)) {
$blocklist = array();
include(SPAM_INI_FILE);
// '*.blogspot.com', // Blog services's subdomains (only)
// 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#',
// );
- 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 $blocked;
}
+
+// ---------------------
+
+
// Default (enabled) methods and thresholds (for content insertion)
function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE)
{
// Simple/fast spam check
function check_uri_spam($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 (! isset($sum['quantity'])) $sum['quantity'] = 0;
+
if (is_array($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(), $count_empty_array = FALSE)
-{
- if (! is_array($array) || (empty($array) && $count_empty_array))
- return 1;
-
- // Recurse
- $result = 0;
- foreach ($array as $part) {
- $result += array_count_leaves($part, $count_empty_array);
- }
- return $result;
-}
-
// ---------------------
// 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 '';
+
+ // 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'
+ }
+ $trie = array_merge_recursive($trie, array($resp => array($rest => NULL)));
+ }
+
+ // 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;
+ } else {
+ $subs[] = $sub . '.' . $key;
+ }
+ }
+ $result[] = ' \'' . $key . '\' => \'' . implode(', ', $subs) . '\',';
+ }
+ unset($trie[$key]);
+ }
+ return
+ 'array (' . "\n" .
+ implode("\n", $result) . "\n" .
+ ')';
+}
+
+
+// Check responsibility-root of the FQDN
+// 'foo.bar.example.com' => 'example.com' (.com has the last whois for it)
+// 'foo.bar.example.au' => 'example.au' (.au has the last whois for it)
+// 'foo.bar.example.edu.au' => 'example.edu.au' (.edu.au has the last whois for it)
+// 'foo.bar.example.act.edu.au' => 'example.act.edu.au' (.act.edu.au has the last whois for it)
+function whois_responsibility($fqdn = 'foo.bar.example.com', $parent = FALSE, $implicit = TRUE)
+{
+ static $domain;
+
+ if ($fqdn === NULL) {
+ $domain = NULL; // Unset
+ return '';
+ }
+ if (! is_string($fqdn)) return '';
+
+ if (is_ip($fqdn)) return $fqdn;
+
+ if (! isset($domain)) {
+ $domain = array();
+ if (file_exists(DOMAIN_INI_FILE)) {
+ include(DOMAIN_INI_FILE); // Set
+ }
+ }
+
+ $result = array();
+ $dcursor = & $domain;
+ $array = array_reverse(explode('.', $fqdn));
+ $i = 0;
+ while(TRUE) {
+ if (! isset($array[$i])) break;
+ $acursor = $array[$i];
+ if (is_array($dcursor) && isset($dcursor[$acursor])) {
+ $result[] = & $array[$i];
+ $dcursor = & $dcursor[$acursor];
+ } else {
+ if (! $parent && isset($acursor)) {
+ $result[] = & $array[$i]; // Whois servers must know this subdomain
+ }
+ break;
+ }
+ ++$i;
+ }
+
+ // Implicit responsibility: Top-Level-Domains must not be yours
+ // 'bar.foo.something' => 'foo.something'
+ if ($implicit && count($result) == 1 && count($array) > 1) {
+ $result[] = & $array[1];
+ }
+
+ return $result ? implode('.', array_reverse($result)) : '';
+}
+
+
// ---------------------
// Exit
+// 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())
{
+ $exit = TRUE;
+
switch ($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
}
{
$progress = check_uri_spam($target, $method);
- if (! empty($progress['is_spam'])) {
- // Mail to administrator(s)
- pkwk_spamnotify($action, $page, $target, $progress, $method);
+ if (empty($progress['is_spam'])) {
+ spam_dispose();
+ } else {
+
+// 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 = & $tmp;
- // Exit
+ pkwk_spamnotify($action, $page, $target, $progress, $method);
spam_exit($exitmode, $progress);
}
}
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);