lib/classes/ip_utils.php

   1 <?php
   2 // This file is part of Moodle - http://moodle.org/
   3 //
   4 // Moodle is free software: you can redistribute it and/or modify
   5 // it under the terms of the GNU General Public License as published by
   6 // the Free Software Foundation, either version 3 of the License, or
   7 // (at your option) any later version.
   8 //
   9 // Moodle is distributed in the hope that it will be useful,
  10 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 // GNU General Public License for more details.
  13 //
  14 // You should have received a copy of the GNU General Public License
  15 // along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
  16
  17 /**
  18  * Contains a simple class providing some useful internet protocol-related functions.
  19  *
  20  * @package   core
  21  * @copyright 2016 Jake Dallimore
  22  * @license   http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
  23  * @author    Jake Dallimore <jrhdallimore@gmail.com>
  24  */
  25
  26 namespace core;
  27
  28 defined('MOODLE_INTERNAL') || exit();
  29
  30 /**
  31  * Static helper class providing some useful internet-protocol-related functions.
  32  *
  33  * @package   core
  34  * @copyright 2016 Jake Dallimore
  35  * @license   http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
  36  * @author    Jake Dallimore <jrhdallimore@gmail.com>
  37  */
  38 final class ip_utils {
  39     /**
  40      * Syntax checking for domain names, including fully qualified domain names.
  41      *
  42      * This function does not verify the existence of the domain name. It only verifies syntactic correctness.
  43      * This is based on RFC1034/1035 and does not provide support for validation of internationalised domain names (IDNs).
  44      * All IDNs must be prior-converted to their ascii-compatible encoding before being passed to this function.
  45      *
  46      * @param string $domainname the input string to check.
  47      * @return bool true if the string has valid syntax, false otherwise.
  48      */
  49     public static function is_domain_name($domainname) {
  50         if (!is_string($domainname)) {
  51             return false;
  52         }
  53         // Usually the trailing dot (null label) is omitted, but is valid if supplied. We'll just remove it and validate as normal.
  54         $domainname = rtrim($domainname, '.');
  55
  56         // The entire name cannot exceed 253 ascii characters (255 octets, less the leading label-length byte and null label byte).
  57         if (strlen($domainname) > 253) {
  58             return false;
  59         }
  60         // Tertiary domain labels can have 63 octets max, and must not have begin or end with a hyphen.
  61         // The TLD label cannot begin with a number, but otherwise, is only loosely restricted here (TLD list is not checked).
  62         $domaintertiary = '([a-zA-Z0-9](([a-zA-Z0-9-]{0,61})[a-zA-Z0-9])?\.)*';
  63         $domaintoplevel = '([a-zA-Z](([a-zA-Z0-9-]*)[a-zA-Z0-9])?)';
  64         $address = '(' . $domaintertiary .  $domaintoplevel . ')';
  65         $regexp = '#^' . $address . '$#i'; // Case insensitive matching.
  66         return preg_match($regexp, $domainname, $match) == true; // False for error, 0 for no match - we treat the same.
  67     }
  68
  69     /**
  70      * Checks whether the input string is a valid wildcard domain matching pattern.
  71      *
  72      * A domain matching pattern is essentially a domain name with a single, leading wildcard (*) label, and at least one other
  73      * label. The wildcard label is considered to match at least one label at or above (to the left of) its position in the string,
  74      * but will not match the trailing domain (everything to its right).
  75      *
  76      * The string must be dot-separated, and the whole pattern must follow the domain name syntax rules defined in RFC1034/1035.
  77      * Namely, the character type (ascii), total-length (253) and label-length (63) restrictions. This function only confirms
  78      * syntactic correctness. It does not check for the existence of the domain/subdomains.
  79      *
  80      * For example, the string '*.example.com' is a pattern deemed to match any direct subdomain of
  81      * example.com (such as test.example.com), any higher level subdomains (e.g. another.test.example.com) but will not match
  82      * the 'example.com' domain itself.
  83      *
  84      * @param string $pattern the string to check.
  85      * @return bool true if the input string is a valid domain wildcard matching pattern, false otherwise.
  86      */
  87     public static function is_domain_matching_pattern($pattern) {
  88         if (!is_string($pattern)) {
  89             return false;
  90         }
  91         // Usually the trailing dot (null label) is omitted, but is valid if supplied. We'll just remove it and validate as normal.
  92         $pattern = rtrim($pattern, '.');
  93
  94         // The entire name cannot exceed 253 ascii characters (255 octets, less the leading label-length byte and null label byte).
  95         if (strlen($pattern) > 253) {
  96             return false;
  97         }
  98         // A valid pattern must left-positioned wildcard symbol (*).
  99         // Tertiary domain labels can have 63 octets max, and must not have begin or end with a hyphen.
 100         // The TLD label cannot begin with a number, but otherwise, is only loosely restricted here (TLD list is not checked).
 101         $wildcard = '((\*)\.){1}';
 102         $domaintertiary = '([a-zA-Z0-9](([a-zA-Z0-9-]{0,61})[a-zA-Z0-9])?\.)*';
 103         $domaintoplevel = '([a-zA-Z](([a-zA-Z0-9-]*)[a-zA-Z0-9])?)';
 104         $address = '(' . $wildcard . $domaintertiary .  $domaintoplevel . ')';
 105         $regexp = '#^' . $address . '$#i'; // Case insensitive matching.
 106         return preg_match($regexp, $pattern, $match) == true; // False for error, 0 for no match - we treat the same.
 107     }
 108
 109     /**
 110      * Syntax validation for IP addresses, supporting both IPv4 and Ipv6 formats.
 111      *
 112      * @param string $address the address to check.
 113      * @return bool true if the address is a valid IPv4 of IPv6 address, false otherwise.
 114      */
 115     public static function is_ip_address($address) {
 116         return filter_var($address, FILTER_VALIDATE_IP) !== false;
 117     }
 118
 119     /**
 120      * Syntax validation for IPv4 addresses.
 121      *
 122      * @param string $address the address to check.
 123      * @return bool true if the address is a valid IPv4 address, false otherwise.
 124      */
 125     public static function is_ipv4_address($address) {
 126         return filter_var($address, FILTER_VALIDATE_IP, FILTER_FLAG_IPV4) !== false;
 127     }
 128
 129     /**
 130      * Syntax checking for IPv4 address ranges.
 131      * Supports CIDR notation and last-group ranges.
 132      * Eg. 127.0.0.0/24 or 127.0.0.80-255
 133      *
 134      * @param string $addressrange the address range to check.
 135      * @return bool true if the string is a valid range representation, false otherwise.
 136      */
 137     public static function is_ipv4_range($addressrange) {
 138         // Check CIDR notation.
 139         if (preg_match('#^(\d{1,3})\.(\d{1,3})\.(\d{1,3})\.(\d{1,3})\/(\d{1,2})$#', $addressrange, $match)) {
 140             $address = "{$match[1]}.{$match[2]}.{$match[3]}.{$match[4]}";
 141             return self::is_ipv4_address($address) && $match[5] <= 32;
 142         }
 143         // Check last-group notation.
 144         if (preg_match('#^(\d{1,3})\.(\d{1,3})\.(\d{1,3})\.(\d{1,3})-(\d{1,3})$#', $addressrange, $match)) {
 145             $address = "{$match[1]}.{$match[2]}.{$match[3]}.{$match[4]}";
 146             return self::is_ipv4_address($address) && $match[5] <= 255 && $match[5] >= $match[4];
 147         }
 148         return false;
 149     }
 150
 151     /**
 152      * Syntax validation for IPv6 addresses.
 153      * This function does not check whether the address is assigned, only its syntactical correctness.
 154      *
 155      * @param string $address the address to check.
 156      * @return bool true if the address is a valid IPv6 address, false otherwise.
 157      */
 158     public static function is_ipv6_address($address) {
 159         return filter_var($address, FILTER_VALIDATE_IP, FILTER_FLAG_IPV6) !== false;
 160     }
 161
 162     /**
 163      * Syntax validation for IPv6 address ranges.
 164      * Supports CIDR notation and last-group ranges.
 165      * Eg. fe80::d98c/64 or fe80::d98c-ffee
 166      *
 167      * @param string $addressrange the IPv6 address range to check.
 168      * @return bool true if the string is a valid range representation, false otherwise.
 169      */
 170     public static function is_ipv6_range($addressrange) {
 171         // Check CIDR notation.
 172         $ipv6parts = explode('/', $addressrange);
 173         if (count($ipv6parts) == 2) {
 174             $range = (int)$ipv6parts[1];
 175             return self::is_ipv6_address($ipv6parts[0]) && (string)$range === $ipv6parts[1] && $range >= 0 && $range <= 128;
 176         }
 177         // Check last-group notation.
 178         $ipv6parts = explode('-', $addressrange);
 179         if (count($ipv6parts) == 2) {
 180             $addressparts = explode(':', $ipv6parts[0]);
 181             $rangestart = $addressparts[count($addressparts) - 1];
 182             $rangeend = $ipv6parts[1];
 183             return self::is_ipv6_address($ipv6parts[0]) && ctype_xdigit($rangestart) && ctype_xdigit($rangeend)
 184             && strlen($rangeend) <= 4 && strlen($rangestart) <= 4 && hexdec($rangeend) >= hexdec($rangestart);
 185         }
 186         return false;
 187     }
 188
 189     /**
 190      * Checks the domain name against a list of allowed domains. The list of allowed domains is may use
 191      * wildcards that match {@link is_domain_matching_pattern()}.
 192      *
 193      * @param  string $domain Domain address
 194      * @param  array $alloweddomains An array of allowed domains.
 195      * @return boolean True if the domain matches one of the entries in the allowed domains list.
 196      */
 197     public static function is_domain_in_allowed_list($domain, $alloweddomains) {
 198
 199         if (!self::is_domain_name($domain)) {
 200             return false;
 201         }
 202
 203         foreach ($alloweddomains as $alloweddomain) {
 204             if (strpos($alloweddomain, '*') !== false) {
 205                 if (!self::is_domain_matching_pattern($alloweddomain)) {
 206                     continue;
 207                 }
 208                 // Use of wildcard for possible subdomains.
 209                 $escapeperiods = str_replace('.', '\.', $alloweddomain);
 210                 $replacewildcard = str_replace('*', '.*', $escapeperiods);
 211                 $ultimatepattern = '/' . $replacewildcard . '$/';
 212                 if (preg_match($ultimatepattern, $domain)) {
 213                     return true;
 214                 }
 215             } else {
 216                 if (!self::is_domain_name($alloweddomain)) {
 217                     continue;
 218                 }
 219                 // Strict domain setting.
 220                 if ($domain === $alloweddomain) {
 221                     return true;
 222                 }
 223             }
 224         }
 225         return false;
 226     }
 227 }