From 7a42173a16fcf7c53023aee4610ac42523e86226 Mon Sep 17 00:00:00 2001 From: Aleksander Machniak <alec@alec.pl> Date: Thu, 13 Aug 2015 03:04:19 -0400 Subject: [PATCH] Simplify rcube_utils::check_ip() --- program/lib/Roundcube/rcube_utils.php | 301 +++++++++++++++++++++++++++++++++---------------- 1 files changed, 202 insertions(+), 99 deletions(-) diff --git a/program/lib/Roundcube/rcube_utils.php b/program/lib/Roundcube/rcube_utils.php index db41a6e..924b0d6 100644 --- a/program/lib/Roundcube/rcube_utils.php +++ b/program/lib/Roundcube/rcube_utils.php @@ -1,6 +1,6 @@ <?php -/* +/** +-----------------------------------------------------------------------+ | This file is part of the Roundcube Webmail client | | Copyright (C) 2008-2012, The Roundcube Dev Team | @@ -103,13 +103,14 @@ } foreach ($domain_array as $part) { - if (!preg_match('/^(([A-Za-z0-9][A-Za-z0-9-]{0,61}[A-Za-z0-9])|([A-Za-z0-9]))$/', $part)) { + if (!preg_match('/^((xn--)?([A-Za-z0-9][A-Za-z0-9-]{0,61}[A-Za-z0-9])|([A-Za-z0-9]))$/', $part)) { return false; } } // last domain part - if (preg_match('/[^a-zA-Z]/', array_pop($domain_array))) { + $last_part = array_pop($domain_array); + if (strpos($last_part, 'xn--') !== 0 && preg_match('/[^a-zA-Z]/', $last_part)) { return false; } @@ -117,17 +118,6 @@ if (!$dns_check || !$rcube->config->get('email_dns_check')) { return true; - } - - if (strtoupper(substr(PHP_OS, 0, 3)) == 'WIN' && version_compare(PHP_VERSION, '5.3.0', '<')) { - $lookup = array(); - @exec("nslookup -type=MX " . escapeshellarg($domain_part) . " 2>&1", $lookup); - foreach ($lookup as $line) { - if (strpos($line, 'MX preference')) { - return true; - } - } - return false; } // find MX record(s) @@ -144,7 +134,6 @@ return false; } - /** * Validates IPv4 or IPv6 address * @@ -154,41 +143,8 @@ */ public static function check_ip($ip) { - // IPv6, but there's no build-in IPv6 support - if (strpos($ip, ':') !== false && !defined('AF_INET6')) { - $parts = explode(':', $ip); - $count = count($parts); - - if ($count > 8 || $count < 2) { - return false; - } - - foreach ($parts as $idx => $part) { - $length = strlen($part); - if (!$length) { - // there can be only one :: - if ($found_empty) { - return false; - } - $found_empty = true; - } - // last part can be an IPv4 address - else if ($idx == $count - 1) { - if (!preg_match('/^[0-9a-f]{1,4}$/i', $part)) { - return @inet_pton($part) !== false; - } - } - else if (!preg_match('/^[0-9a-f]{1,4}$/i', $part)) { - return false; - } - } - - return true; - } - - return @inet_pton($ip) !== false; + return filter_var($ip, FILTER_VALIDATE_IP) !== false; } - /** * Check whether the HTTP referer matches the current request @@ -201,7 +157,6 @@ $referer = parse_url(self::request_header('Referer')); return $referer['host'] == self::request_header('Host') && $referer['path'] == $uri['path']; } - /** * Replacing specials characters to a specific encoding type @@ -266,8 +221,8 @@ $js_rep_table["'"] = "\\'"; $js_rep_table["\\"] = "\\\\"; // Unicode line and paragraph separators (#1486310) - $js_rep_table[chr(hexdec(E2)).chr(hexdec(80)).chr(hexdec(A8))] = '
'; - $js_rep_table[chr(hexdec(E2)).chr(hexdec(80)).chr(hexdec(A9))] = '
'; + $js_rep_table[chr(hexdec('E2')).chr(hexdec('80')).chr(hexdec('A8'))] = '
'; + $js_rep_table[chr(hexdec('E2')).chr(hexdec('80')).chr(hexdec('A9'))] = '
'; } // encode for javascript use @@ -292,7 +247,6 @@ // no encoding given -> return original string return $str; } - /** * Read input value and convert it for internal use @@ -333,7 +287,6 @@ return self::parse_input_value($value, $allow_html, $charset); } - /** * Parse/validate input value. See self::get_input_value() @@ -385,7 +338,6 @@ return $value; } - /** * Convert array of request parameters (prefixed with _) * to a regular array with non-prefixed keys. @@ -411,7 +363,6 @@ return $out; } - /** * Convert the given string into a valid HTML identifier * Same functionality as done in app.js with rcube_webmail.html_identifier() @@ -425,7 +376,6 @@ return asciiwords($str, true, '_'); } } - /** * Replace all css definitions with #container [def] @@ -503,7 +453,6 @@ return $source; } - /** * Generate CSS classes from mimetype and filename extension * @@ -534,7 +483,6 @@ return join(" ", $classes); } - /** * Decode escaped entities used by known XSS exploits. * See http://downloads.securityfocus.com/vulnerabilities/exploits/26800.eml for examples @@ -553,7 +501,6 @@ return $out; } - /** * preg_replace_callback callback for xss_entity_decode * @@ -565,7 +512,6 @@ { return chr(hexdec($matches[1])); } - /** * Check if we can process not exceeding memory_limit @@ -582,7 +528,6 @@ return $mem_limit > 0 && $memory + $need > $mem_limit ? false : true; } - /** * Check if working in SSL mode * @@ -593,24 +538,23 @@ */ public static function https_check($port=null, $use_https=true) { - global $RCMAIL; - if (!empty($_SERVER['HTTPS']) && strtolower($_SERVER['HTTPS']) != 'off') { return true; } - if (!empty($_SERVER['HTTP_X_FORWARDED_PROTO']) && strtolower($_SERVER['HTTP_X_FORWARDED_PROTO']) == 'https') { + if (!empty($_SERVER['HTTP_X_FORWARDED_PROTO']) + && strtolower($_SERVER['HTTP_X_FORWARDED_PROTO']) == 'https' + && in_array($_SERVER['REMOTE_ADDR'], rcube::get_instance()->config->get('proxy_whitelist', array()))) { return true; } if ($port && $_SERVER['SERVER_PORT'] == $port) { return true; } - if ($use_https && isset($RCMAIL) && $RCMAIL->config->get('use_https')) { + if ($use_https && rcube::get_instance()->config->get('use_https')) { return true; } return false; } - /** * Replaces hostname variables. @@ -622,6 +566,10 @@ */ public static function parse_host($name, $host = '') { + if (!is_string($name)) { + return $name; + } + // %n - host $n = preg_replace('/:\d+$/', '', $_SERVER['SERVER_NAME']); // %t - host name without first part, e.g. %n=mail.domain.tld, %t=domain.tld @@ -642,10 +590,8 @@ } } - $name = str_replace(array('%n', '%t', '%d', '%h', '%z', '%s'), array($n, $t, $d, $h, $z, $s[2]), $name); - return $name; + return str_replace(array('%n', '%t', '%d', '%h', '%z', '%s'), array($n, $t, $d, $h, $z, $s[2]), $name); } - /** * Returns remote IP address and forwarded addresses if found @@ -672,7 +618,6 @@ return $address; } - /** * Returns the real remote IP address * @@ -680,9 +625,26 @@ */ public static function remote_addr() { - foreach (array('HTTP_X_FORWARDED_FOR','HTTP_X_REAL_IP','REMOTE_ADDR') as $prop) { - if (!empty($_SERVER[$prop])) - return $_SERVER[$prop]; + // Check if any of the headers are set first to improve performance + if (!empty($_SERVER['HTTP_X_FORWARDED_FOR']) || !empty($_SERVER['HTTP_X_REAL_IP'])) { + $proxy_whitelist = rcube::get_instance()->config->get('proxy_whitelist', array()); + if (in_array($_SERVER['REMOTE_ADDR'], $proxy_whitelist)) { + if (!empty($_SERVER['HTTP_X_FORWARDED_FOR'])) { + foreach(array_reverse(explode(',', $_SERVER['HTTP_X_FORWARDED_FOR'])) as $forwarded_ip) { + if (!in_array($forwarded_ip, $proxy_whitelist)) { + return $forwarded_ip; + } + } + } + + if (!empty($_SERVER['HTTP_X_REAL_IP'])) { + return $_SERVER['HTTP_X_REAL_IP']; + } + } + } + + if (!empty($_SERVER['REMOTE_ADDR'])) { + return $_SERVER['REMOTE_ADDR']; } return ''; @@ -737,17 +699,18 @@ return $result; } - /** * Improved equivalent to strtotime() * * @param string $date Date string + * @param object DateTimeZone to use for DateTime object * * @return int Unix timestamp */ - public static function strtotime($date) + public static function strtotime($date, $timezone = null) { $date = self::clean_datestr($date); + $tzname = $timezone ? ' ' . $timezone->getName() : ''; // unix timestamp if (is_numeric($date)) { @@ -756,7 +719,7 @@ // if date parsing fails, we have a date in non-rfc format. // remove token from the end and try again - while ((($ts = @strtotime($date)) === false) || ($ts < 0)) { + while ((($ts = @strtotime($date . $tzname)) === false) || ($ts < 0)) { $d = explode(' ', $date); array_pop($d); if (!$d) { @@ -772,10 +735,11 @@ * Date parsing function that turns the given value into a DateTime object * * @param string $date Date string + * @param object DateTimeZone to use for DateTime object * * @return object DateTime instance or false on failure */ - public static function anytodatetime($date) + public static function anytodatetime($date, $timezone = null) { if (is_object($date) && is_a($date, 'DateTime')) { return $date; @@ -787,7 +751,7 @@ // try to parse string with DateTime first if (!empty($date)) { try { - $dt = new DateTime($date); + $dt = $timezone ? new DateTime($date, $timezone) : new DateTime($date); } catch (Exception $e) { // ignore @@ -795,9 +759,12 @@ } // try our advanced strtotime() method - if (!$dt && ($timestamp = self::strtotime($date))) { + if (!$dt && ($timestamp = self::strtotime($date, $timezone))) { try { $dt = new DateTime("@".$timestamp); + if ($timezone) { + $dt->setTimezone($timezone); + } } catch (Exception $e) { // ignore @@ -862,7 +829,6 @@ return self::idn_convert($str, true); } - /* * Idn_to_ascii wrapper. * Intl/Idn modules version of this function doesn't work with e-mail address @@ -872,8 +838,7 @@ return self::idn_convert($str, false); } - - public static function idn_convert($input, $is_utf=false) + public static function idn_convert($input, $is_utf = false) { if ($at = strpos($input, '@')) { $user = substr($input, 0, $at); @@ -896,26 +861,34 @@ * Split the given string into word tokens * * @param string Input to tokenize + * @param integer Minimum length of a single token * @return array List of tokens */ - public static function tokenize_string($str) + public static function tokenize_string($str, $minlen = 2) { - return explode(" ", preg_replace( - array('/[\s;\/+-]+/i', '/(\d)[-.\s]+(\d)/', '/\s\w{1,3}\s/u'), - array(' ', '\\1\\2', ' '), - $str)); + $expr = array('/[\s;,"\'\/+-]+/ui', '/(\d)[-.\s]+(\d)/u'); + $repl = array(' ', '\\1\\2'); + + if ($minlen > 1) { + $minlen--; + $expr[] = "/(^|\s+)\w{1,$minlen}(\s+|$)/u"; + $repl[] = ' '; + } + + return array_filter(explode(" ", preg_replace($expr, $repl, $str))); } /** * Normalize the given string for fulltext search. - * Currently only optimized for Latin-1 characters; to be extended + * Currently only optimized for ISO-8859-1 and ISO-8859-2 characters; to be extended * * @param string Input string (UTF-8) * @param boolean True to return list of words as array + * @param integer Minimum length of tokens * * @return mixed Normalized string or a list of normalized tokens */ - public static function normalize_string($str, $as_array = false) + public static function normalize_string($str, $as_array = false, $minlen = 2) { // replace 4-byte unicode characters with '?' character, // these are not supported in default utf-8 charset on mysql, @@ -927,20 +900,59 @@ . ')/', '?', $str); // split by words - $arr = self::tokenize_string($str); + $arr = self::tokenize_string($str, $minlen); + + // detect character set + if (utf8_encode(utf8_decode($str)) == $str) { + // ISO-8859-1 (or ASCII) + preg_match_all('/./u', 'äâàåáãæçéêëèïîìíñöôòøõóüûùúýÿ', $keys); + preg_match_all('/./', 'aaaaaaaceeeeiiiinoooooouuuuyy', $values); + + $mapping = array_combine($keys[0], $values[0]); + $mapping = array_merge($mapping, array('ß' => 'ss', 'ae' => 'a', 'oe' => 'o', 'ue' => 'u')); + } + else if (rcube_charset::convert(rcube_charset::convert($str, 'UTF-8', 'ISO-8859-2'), 'ISO-8859-2', 'UTF-8') == $str) { + // ISO-8859-2 + preg_match_all('/./u', 'ąáâäćçčéęëěíîłľĺńňóôöŕřśšşťţůúűüźžżý', $keys); + preg_match_all('/./', 'aaaaccceeeeiilllnnooorrsssttuuuuzzzy', $values); + + $mapping = array_combine($keys[0], $values[0]); + $mapping = array_merge($mapping, array('ß' => 'ss', 'ae' => 'a', 'oe' => 'o', 'ue' => 'u')); + } foreach ($arr as $i => $part) { - if (utf8_encode(utf8_decode($part)) == $part) { // is latin-1 ? - $arr[$i] = utf8_encode(strtr(strtolower(strtr(utf8_decode($part), - 'ÇçäâàåéêëèïîìÅÉöôòüûùÿøØáíóúñÑÁÂÀãÃÊËÈÍÎÏÓÔõÕÚÛÙýÝ', - 'ccaaaaeeeeiiiaeooouuuyooaiounnaaaaaeeeiiioooouuuyy')), - array('ß' => 'ss', 'ae' => 'a', 'oe' => 'o', 'ue' => 'u'))); + $part = mb_strtolower($part); + + if (!empty($mapping)) { + $part = strtr($part, $mapping); } - else - $arr[$i] = mb_strtolower($part); + + $arr[$i] = $part; } return $as_array ? $arr : join(" ", $arr); + } + + /** + * Compare two strings for matching words (order not relevant) + * + * @param string Haystack + * @param string Needle + * @return boolen True if match, False otherwise + */ + public static function words_match($haystack, $needle) + { + $a_needle = self::tokenize_string($needle, 1); + $haystack = join(" ", self::tokenize_string($haystack, 1)); + + $hits = 0; + foreach ($a_needle as $w) { + if (stripos($haystack, $w) !== false) { + $hits++; + } + } + + return $hits >= count($a_needle); } /** @@ -1019,7 +1031,6 @@ } } - /** * Find out if the string content means true or false * @@ -1034,4 +1045,96 @@ return !in_array($str, array('false', '0', 'no', 'off', 'nein', ''), true); } + /** + * OS-dependent absolute path detection + */ + public static function is_absolute_path($path) + { + if (strtoupper(substr(PHP_OS, 0, 3)) == 'WIN') { + return (bool) preg_match('!^[a-z]:[\\\\/]!i', $path); + } + else { + return $path[0] == '/'; + } + } + + /** + * Resolve relative URL + * + * @param string $url Relative URL + * + * @return string Absolute URL + */ + public static function resolve_url($url) + { + // prepend protocol://hostname:port + if (!preg_match('|^https?://|', $url)) { + $schema = 'http'; + $default_port = 80; + + if (self::https_check()) { + $schema = 'https'; + $default_port = 443; + } + + $prefix = $schema . '://' . preg_replace('/:\d+$/', '', $_SERVER['HTTP_HOST']); + if ($_SERVER['SERVER_PORT'] != $default_port) { + $prefix .= ':' . $_SERVER['SERVER_PORT']; + } + + $url = $prefix . ($url[0] == '/' ? '' : '/') . $url; + } + + return $url; + } + + /** + * Generate a ramdom string + * + * @param int $length String length + * @param bool $raw Return RAW data instead of hex + * + * @return string The generated random string + */ + public static function random_bytes($length, $raw = false) + { + $rlen = $raw ? $length : ceil($length / 2); + $random = openssl_random_pseudo_bytes($rlen); + + if ($raw) { + return $random; + } + + $random = bin2hex($random); + + // if the length wasn't even... + if ($length < strlen($random)) { + $random = substr($random, 0, $length); + } + + return $random; + } + + /** + * Format current date according to specified format. + * This method supports microseconds (u). + * + * @param string $format Date format (default: 'd-M-Y H:i:s O') + * + * @return string Formatted date + */ + public static function date_format($format = null) + { + if (empty($format)) { + $format = 'd-M-Y H:i:s O'; + } + + if (strpos($format, 'u') !== false + && ($date = date_create_from_format('U.u.e', microtime(true) . '.' . date_default_timezone_get())) + ) { + return $date->format($format); + } + + return date($format); + } } -- Gitblit v1.9.1