Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
97.57% |
241 / 247 |
|
89.66% |
26 / 29 |
CRAP | |
0.00% |
0 / 1 |
| IPUtils | |
97.57% |
241 / 247 |
|
89.66% |
26 / 29 |
114 | |
0.00% |
0 / 1 |
| isIPAddress | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isIPv6 | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isIPv4 | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isValid | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
2 | |||
| isValidIPv4 | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isValidIPv6 | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isValidIPv4Range | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isValidIPv6Range | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isValidRange | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
2 | |||
| sanitizeIP | |
100.00% |
30 / 30 |
|
100.00% |
1 / 1 |
9 | |||
| prettifyIP | |
100.00% |
23 / 23 |
|
100.00% |
1 / 1 |
8 | |||
| splitHostAndPort | |
100.00% |
17 / 17 |
|
100.00% |
1 / 1 |
8 | |||
| combineHostAndPort | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
4 | |||
| formatHex | |
100.00% |
3 / 3 |
|
100.00% |
1 / 1 |
2 | |||
| hexToOctet | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
2 | |||
| hexToQuad | |
100.00% |
7 / 7 |
|
100.00% |
1 / 1 |
3 | |||
| isPublic | |
100.00% |
14 / 14 |
|
100.00% |
1 / 1 |
2 | |||
| toHex | |
100.00% |
10 / 10 |
|
100.00% |
1 / 1 |
6 | |||
| IPv6ToRawHex | |
85.71% |
6 / 7 |
|
0.00% |
0 / 1 |
3.03 | |||
| parseCIDR | |
93.75% |
15 / 16 |
|
0.00% |
0 / 1 |
9.02 | |||
| parseRange | |
100.00% |
22 / 22 |
|
100.00% |
1 / 1 |
12 | |||
| parseCIDR6 | |
71.43% |
10 / 14 |
|
0.00% |
0 / 1 |
8.14 | |||
| parseRange6 | |
100.00% |
21 / 21 |
|
100.00% |
1 / 1 |
7 | |||
| isInRange | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
2 | |||
| isInRanges | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
3 | |||
| canonicalize | |
100.00% |
11 / 11 |
|
100.00% |
1 / 1 |
5 | |||
| sanitizeRange | |
100.00% |
6 / 6 |
|
100.00% |
1 / 1 |
2 | |||
| getSubnet | |
100.00% |
8 / 8 |
|
100.00% |
1 / 1 |
3 | |||
| getIPsInRange | |
100.00% |
11 / 11 |
|
100.00% |
1 / 1 |
6 | |||
| 1 | <?php |
| 2 | /** |
| 3 | * This program is free software; you can redistribute it and/or modify |
| 4 | * it under the terms of the GNU General Public License as published by |
| 5 | * the Free Software Foundation; either version 2 of the License, or |
| 6 | * (at your option) any later version. |
| 7 | * |
| 8 | * This program is distributed in the hope that it will be useful, |
| 9 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 10 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 11 | * GNU General Public License for more details. |
| 12 | * |
| 13 | * You should have received a copy of the GNU General Public License along |
| 14 | * with this program; if not, write to the Free Software Foundation, Inc., |
| 15 | * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. |
| 16 | * http://www.gnu.org/copyleft/gpl.html |
| 17 | * |
| 18 | * @file |
| 19 | * @author Antoine Musso "<hashar at free dot fr>" |
| 20 | */ |
| 21 | |
| 22 | namespace Wikimedia; |
| 23 | |
| 24 | use InvalidArgumentException; |
| 25 | |
| 26 | /** |
| 27 | * Play with IP addresses and IP ranges. |
| 28 | */ |
| 29 | class IPUtils { |
| 30 | |
| 31 | /** |
| 32 | * An IPv4 address is made of 4 bytes from x00 to xFF which is d0 to d255 |
| 33 | */ |
| 34 | public const RE_IP_BYTE = '(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|0?[0-9]?[0-9])'; |
| 35 | |
| 36 | private const RE_IP_ADD = self::RE_IP_BYTE . '\.' . self::RE_IP_BYTE . '\.' |
| 37 | . self::RE_IP_BYTE . '\.' . self::RE_IP_BYTE; |
| 38 | /** |
| 39 | * An IPv4 range is an IP address and a prefix (d0 to d32) |
| 40 | */ |
| 41 | private const RE_IP_PREFIX = '(3[0-2]|[12][0-9]|[0-9])'; |
| 42 | |
| 43 | private const RE_IP_RANGE = '(' . self::RE_IP_ADD . '\/' . self::RE_IP_PREFIX . '|' |
| 44 | . self::RE_IP_ADD . ' ?\- ?' . self::RE_IP_ADD . ')'; |
| 45 | |
| 46 | /** |
| 47 | * An IPv6 address is made up of 8 words (each x0000 to xFFFF). |
| 48 | * However, the "::" abbreviation can be used on consecutive x0000 words. |
| 49 | */ |
| 50 | private const RE_IPV6_WORD = '([0-9A-Fa-f]{1,4})'; |
| 51 | /** |
| 52 | * An IPv6 range is an IP address and a prefix (d0 to d128) |
| 53 | */ |
| 54 | private const RE_IPV6_PREFIX = '(12[0-8]|1[01][0-9]|[1-9][0-9]|[0-9])'; |
| 55 | |
| 56 | private const RE_IPV6_ADD = |
| 57 | '(?:' . |
| 58 | // starts with "::" (including "::") |
| 59 | ':(?::|(?::' . self::RE_IPV6_WORD . '){1,7})' . |
| 60 | '|' . |
| 61 | // ends with "::" (except "::") |
| 62 | self::RE_IPV6_WORD . '(?::' . self::RE_IPV6_WORD . '){0,6}::' . |
| 63 | '|' . |
| 64 | // contains one "::" in the middle (the ^ makes the test fail if none found) |
| 65 | self::RE_IPV6_WORD . '(?::((?(-1)|:))?' . self::RE_IPV6_WORD . '){1,6}(?(-2)|^)' . |
| 66 | '|' . |
| 67 | // contains no "::" |
| 68 | self::RE_IPV6_WORD . '(?::' . self::RE_IPV6_WORD . '){7}' . |
| 69 | ')'; |
| 70 | /** |
| 71 | * An IPv6 range is an IP address and a prefix (d1 to d128) |
| 72 | */ |
| 73 | private const RE_IPV6_RANGE = '(' . self::RE_IPV6_ADD . '\/' . self::RE_IPV6_PREFIX |
| 74 | . '|' . self::RE_IPV6_ADD . ' ?\- ?' . self::RE_IPV6_ADD . ')'; |
| 75 | /** |
| 76 | * For IPv6 canonicalization (NOT for strict validation; these are quite lax!) |
| 77 | */ |
| 78 | public const RE_IPV6_GAP = ':(?:0+:)*(?::(?:0+:)*)?'; |
| 79 | /** @private */ |
| 80 | private const RE_IPV6_V4_PREFIX = '0*' . self::RE_IPV6_GAP . '(?:ffff:)?'; |
| 81 | |
| 82 | /** |
| 83 | * This might be useful for regexps used elsewhere, matches any IPv4 or IPv6 address or network |
| 84 | */ |
| 85 | private const RE_IP_ADDRESS_STRING = |
| 86 | '(?:' . |
| 87 | // IPv4 |
| 88 | self::RE_IP_ADD . '(?:\/' . self::RE_IP_PREFIX . ')?' . |
| 89 | '|' . |
| 90 | // IPv6 |
| 91 | self::RE_IPV6_ADD . '(?:\/' . self::RE_IPV6_PREFIX . ')?' . |
| 92 | ')'; |
| 93 | |
| 94 | /** |
| 95 | * Maximum number of IP addresses that can be retrieved from a given range. |
| 96 | */ |
| 97 | private const MAXIMUM_IPS_FROM_RANGE = 2 ** 16; |
| 98 | |
| 99 | /** |
| 100 | * Determine if a string is as valid IP address or network (CIDR prefix). |
| 101 | * SIIT IPv4-translated addresses are rejected. |
| 102 | * @note canonicalize() tries to convert translated addresses to IPv4. |
| 103 | * |
| 104 | * @param string $ip Possible IP address |
| 105 | * @return bool |
| 106 | */ |
| 107 | public static function isIPAddress( $ip ) { |
| 108 | return (bool)preg_match( '/^' . self::RE_IP_ADDRESS_STRING . '$/', $ip ); |
| 109 | } |
| 110 | |
| 111 | /** |
| 112 | * Given a string, determine if it as valid IP in IPv6 only. |
| 113 | * @note Unlike isValid(), this looks for networks too. |
| 114 | * |
| 115 | * @param string $ip Possible IP address |
| 116 | * @return bool |
| 117 | */ |
| 118 | public static function isIPv6( $ip ) { |
| 119 | return (bool)preg_match( '/^' . self::RE_IPV6_ADD . '(?:\/' . self::RE_IPV6_PREFIX . ')?$/', $ip ); |
| 120 | } |
| 121 | |
| 122 | /** |
| 123 | * Given a string, determine if it as valid IP in IPv4 only. |
| 124 | * @note Unlike isValid(), this looks for networks too. |
| 125 | * |
| 126 | * @param string $ip Possible IP address |
| 127 | * @return bool |
| 128 | */ |
| 129 | public static function isIPv4( $ip ) { |
| 130 | return (bool)preg_match( '/^' . self::RE_IP_ADD . '(?:\/' . self::RE_IP_PREFIX . ')?$/', $ip ); |
| 131 | } |
| 132 | |
| 133 | /** |
| 134 | * Validate an IP address. Ranges are NOT considered valid. |
| 135 | * SIIT IPv4-translated addresses are rejected. |
| 136 | * @note canonicalize() tries to convert translated addresses to IPv4. |
| 137 | * |
| 138 | * @param string $ip |
| 139 | * @return bool True if it is valid |
| 140 | */ |
| 141 | public static function isValid( $ip ) { |
| 142 | // Test IPv4 before IPv6 as it's more common. |
| 143 | return self::isValidIPv4( $ip ) || self::isValidIPv6( $ip ); |
| 144 | } |
| 145 | |
| 146 | /** |
| 147 | * Validate an IPv4 address. Ranges are NOT considered valid. |
| 148 | * |
| 149 | * @param string $ip |
| 150 | * @return bool True if it is valid |
| 151 | */ |
| 152 | public static function isValidIPv4( $ip ) { |
| 153 | return (bool)preg_match( '/^' . self::RE_IP_ADD . '$/', $ip ); |
| 154 | } |
| 155 | |
| 156 | /** |
| 157 | * Validate an IPv6 address. Ranges are NOT considered valid. |
| 158 | * SIIT IPv4-translated addresses are rejected. |
| 159 | * @note canonicalize() tries to convert translated addresses to IPv4. |
| 160 | * |
| 161 | * @param string $ip |
| 162 | * @return bool True if it is valid |
| 163 | */ |
| 164 | public static function isValidIPv6( $ip ) { |
| 165 | return (bool)preg_match( '/^' . self::RE_IPV6_ADD . '$/', $ip ); |
| 166 | } |
| 167 | |
| 168 | /** |
| 169 | * Validate an IPv4 range (valid IPv4 address with a valid CIDR prefix or explicit range). |
| 170 | * |
| 171 | * @param string $ipRange |
| 172 | * @return bool True if input is valid |
| 173 | */ |
| 174 | private static function isValidIPv4Range( $ipRange ) { |
| 175 | return (bool)preg_match( '/^' . self::RE_IP_RANGE . '$/', $ipRange ); |
| 176 | } |
| 177 | |
| 178 | /** |
| 179 | * Validate an IPv6 range (valid IPv6 address with a valid CIDR prefix or explicit range). |
| 180 | * |
| 181 | * @param string $ipRange |
| 182 | * @return bool True if input is valid |
| 183 | */ |
| 184 | private static function isValidIPv6Range( $ipRange ) { |
| 185 | return (bool)preg_match( '/^' . self::RE_IPV6_RANGE . '$/', $ipRange ); |
| 186 | } |
| 187 | |
| 188 | /** |
| 189 | * Validate an IP range (valid in either IPv4 OR IPv6; given with valid CIDR prefix or in explicit notation). |
| 190 | * SIIT IPv4-translated addresses are rejected. |
| 191 | * @note canonicalize() tries to convert translated addresses to IPv4. |
| 192 | * |
| 193 | * @param string $ipRange |
| 194 | * @return bool True if it is valid |
| 195 | */ |
| 196 | public static function isValidRange( $ipRange ) { |
| 197 | // Test IPv4 before IPv6 as it's more common. |
| 198 | return self::isValidIPv4Range( $ipRange ) || self::isValidIPv6Range( $ipRange ); |
| 199 | } |
| 200 | |
| 201 | /** |
| 202 | * Convert an IP into a verbose, uppercase, normalized form. |
| 203 | * Both IPv4 and IPv6 addresses are trimmed. Additionally, |
| 204 | * IPv6 addresses in octet notation are expanded to 8 words; |
| 205 | * IPv4 addresses have leading zeros, in each octet, removed. |
| 206 | * |
| 207 | * @param string $ip IP address in quad or octet form (CIDR or not). |
| 208 | * @return string|null |
| 209 | */ |
| 210 | public static function sanitizeIP( $ip ) { |
| 211 | $ip = trim( $ip ); |
| 212 | if ( $ip === '' ) { |
| 213 | return null; |
| 214 | } |
| 215 | // If not an IP, just return trimmed value, since sanitizeIP() is called |
| 216 | // in a number of contexts where usernames are supplied as input. |
| 217 | if ( !self::isIPAddress( $ip ) ) { |
| 218 | return $ip; |
| 219 | } |
| 220 | if ( self::isIPv4( $ip ) ) { |
| 221 | // Remove leading 0's from octet representation of IPv4 address |
| 222 | return preg_replace( '!(?:^|(?<=\.))0+(?=[1-9]|0[./]|0$)!', '', $ip ); |
| 223 | } |
| 224 | // Remove any whitespaces, convert to upper case |
| 225 | $ip = strtoupper( $ip ); |
| 226 | // Expand zero abbreviations |
| 227 | $abbrevPos = strpos( $ip, '::' ); |
| 228 | if ( $abbrevPos !== false ) { |
| 229 | // We know this is valid IPv6. Find the last index of the |
| 230 | // address before any CIDR number (e.g. "a:b:c::/24"). |
| 231 | $CIDRStart = strpos( $ip, "/" ); |
| 232 | $addressEnd = ( $CIDRStart !== false ) |
| 233 | ? $CIDRStart - 1 |
| 234 | : strlen( $ip ) - 1; |
| 235 | // If the '::' is at the beginning... |
| 236 | if ( $abbrevPos === 0 ) { |
| 237 | $repeat = '0:'; |
| 238 | // for the address '::' |
| 239 | $extra = $ip === '::' ? '0' : ''; |
| 240 | // 7+2 (due to '::') |
| 241 | $pad = 9; |
| 242 | // If the '::' is at the end... |
| 243 | } elseif ( $abbrevPos === $addressEnd - 1 ) { |
| 244 | $repeat = ':0'; |
| 245 | $extra = ''; |
| 246 | // 7+2 (due to '::') |
| 247 | $pad = 9; |
| 248 | // If the '::' is in the middle... |
| 249 | } else { |
| 250 | $repeat = ':0'; |
| 251 | $extra = ':'; |
| 252 | // 6+2 (due to '::') |
| 253 | $pad = 8; |
| 254 | } |
| 255 | $ip = str_replace( '::', |
| 256 | str_repeat( $repeat, $pad - substr_count( $ip, ':' ) ) . $extra, |
| 257 | $ip |
| 258 | ); |
| 259 | } |
| 260 | // Remove leading zeros from each bloc as needed |
| 261 | return preg_replace( '/(^|:)0+(' . self::RE_IPV6_WORD . ')/', '$1$2', $ip ); |
| 262 | } |
| 263 | |
| 264 | /** |
| 265 | * Prettify an IP for display to end users. |
| 266 | * This will make it more compact and lower-case. |
| 267 | * |
| 268 | * @param string $ip |
| 269 | * @return string|null |
| 270 | */ |
| 271 | public static function prettifyIP( $ip ) { |
| 272 | // normalize (removes '::') |
| 273 | $ip = self::sanitizeIP( $ip ); |
| 274 | if ( $ip === null ) { |
| 275 | return null; |
| 276 | } |
| 277 | if ( self::isIPv6( $ip ) ) { |
| 278 | // Split IP into an address and a CIDR |
| 279 | if ( strpos( $ip, '/' ) !== false ) { |
| 280 | [ $ip, $cidr ] = explode( '/', $ip, 2 ); |
| 281 | } else { |
| 282 | [ $ip, $cidr ] = [ $ip, '' ]; |
| 283 | } |
| 284 | // Get the largest slice of words with multiple zeros |
| 285 | $offset = 0; |
| 286 | $longest = $longestPos = false; |
| 287 | while ( preg_match( |
| 288 | '!(?:^|:)0(?::0)+(?:$|:)!', $ip, $m, PREG_OFFSET_CAPTURE, $offset |
| 289 | ) ) { |
| 290 | // full match |
| 291 | [ $match, $pos ] = $m[0]; |
| 292 | if ( strlen( (string)$match ) > strlen( (string)$longest ) ) { |
| 293 | $longest = $match; |
| 294 | $longestPos = $pos; |
| 295 | } |
| 296 | // advance |
| 297 | $offset = $pos + strlen( $match ); |
| 298 | } |
| 299 | if ( $longest !== false ) { |
| 300 | // Replace this portion of the string with the '::' abbreviation |
| 301 | $ip = substr_replace( $ip, '::', $longestPos, strlen( $longest ) ); |
| 302 | } |
| 303 | // Add any CIDR back on |
| 304 | if ( $cidr !== '' ) { |
| 305 | $ip = "{$ip}/{$cidr}"; |
| 306 | } |
| 307 | // Convert to lower case to make it more readable |
| 308 | $ip = strtolower( $ip ); |
| 309 | } |
| 310 | |
| 311 | return $ip; |
| 312 | } |
| 313 | |
| 314 | /** |
| 315 | * Given a host/port string, like one might find in the host part of a URL |
| 316 | * per RFC 2732, split the hostname part and the port part and return an |
| 317 | * array with an element for each. If there is no port part, the array will |
| 318 | * have false in place of the port. If the string was invalid in some way, |
| 319 | * false is returned. |
| 320 | * |
| 321 | * This was easy with IPv4 and was generally done in an ad-hoc way, but |
| 322 | * with IPv6 it's somewhat more complicated due to the need to parse the |
| 323 | * square brackets and colons. |
| 324 | * |
| 325 | * A bare IPv6 address is accepted despite the lack of square brackets. |
| 326 | * |
| 327 | * @param string $both The string with the host (or IPv4/IPv6 address) and port |
| 328 | * @return array|false Array normally, false on certain failures |
| 329 | */ |
| 330 | public static function splitHostAndPort( $both ) { |
| 331 | if ( substr( $both, 0, 1 ) === '[' ) { |
| 332 | if ( preg_match( '/^\[(' . self::RE_IPV6_ADD . ')\](?::(?P<port>\d+))?$/', $both, $m ) ) { |
| 333 | if ( isset( $m['port'] ) ) { |
| 334 | return [ $m[1], intval( $m['port'] ) ]; |
| 335 | } |
| 336 | |
| 337 | return [ $m[1], false ]; |
| 338 | } |
| 339 | |
| 340 | // Square bracket found but no IPv6 |
| 341 | return false; |
| 342 | } |
| 343 | $numColons = substr_count( $both, ':' ); |
| 344 | if ( $numColons >= 2 ) { |
| 345 | // Is it a bare IPv6 address? |
| 346 | if ( preg_match( '/^' . self::RE_IPV6_ADD . '$/', $both ) ) { |
| 347 | return [ $both, false ]; |
| 348 | } |
| 349 | |
| 350 | // Not valid IPv6, but too many colons for anything else |
| 351 | return false; |
| 352 | } |
| 353 | if ( $numColons >= 1 ) { |
| 354 | // Host:port? |
| 355 | $bits = explode( ':', $both ); |
| 356 | if ( preg_match( '/^\d+/', $bits[1] ) ) { |
| 357 | return [ $bits[0], intval( $bits[1] ) ]; |
| 358 | } |
| 359 | |
| 360 | // Not a valid port |
| 361 | return false; |
| 362 | } |
| 363 | |
| 364 | // Plain hostname |
| 365 | return [ $both, false ]; |
| 366 | } |
| 367 | |
| 368 | /** |
| 369 | * Given a host name and a port, combine them into host/port string like |
| 370 | * you might find in a URL. If the host contains a colon, wrap it in square |
| 371 | * brackets like in RFC 2732. If the port matches the default port, omit |
| 372 | * the port specification |
| 373 | * |
| 374 | * @param string $host |
| 375 | * @param int $port |
| 376 | * @param bool|int $defaultPort |
| 377 | * @return string |
| 378 | */ |
| 379 | public static function combineHostAndPort( $host, $port, $defaultPort = false ) { |
| 380 | if ( strpos( $host, ':' ) !== false ) { |
| 381 | $host = "[$host]"; |
| 382 | } |
| 383 | if ( $defaultPort !== false && $port === $defaultPort ) { |
| 384 | return $host; |
| 385 | } |
| 386 | |
| 387 | return "$host:$port"; |
| 388 | } |
| 389 | |
| 390 | /** |
| 391 | * Convert an IPv4 or IPv6 hexadecimal representation back to readable format |
| 392 | * |
| 393 | * @param string $hex Number, with "v6-" prefix if it is IPv6 |
| 394 | * @return string Quad-dotted (IPv4) or octet notation (IPv6) |
| 395 | */ |
| 396 | public static function formatHex( $hex ) { |
| 397 | if ( substr( $hex, 0, 3 ) === 'v6-' ) { |
| 398 | // IPv6 |
| 399 | return self::hexToOctet( substr( $hex, 3 ) ); |
| 400 | } |
| 401 | |
| 402 | // IPv4 |
| 403 | return self::hexToQuad( $hex ); |
| 404 | } |
| 405 | |
| 406 | /** |
| 407 | * Converts a hexadecimal number to an IPv6 address in octet notation |
| 408 | * |
| 409 | * @param string $ip_hex Pure hex (no v6- prefix) |
| 410 | * @return string (of format a:b:c:d:e:f:g:h) |
| 411 | */ |
| 412 | public static function hexToOctet( $ip_hex ) { |
| 413 | // Pad hex to 32 chars (128 bits) |
| 414 | $ip_hex = str_pad( strtoupper( $ip_hex ), 32, '0', STR_PAD_LEFT ); |
| 415 | // Separate into 8 words |
| 416 | $ip_oct = substr( $ip_hex, 0, 4 ); |
| 417 | for ( $n = 1; $n < 8; $n++ ) { |
| 418 | $ip_oct .= ':' . substr( $ip_hex, 4 * $n, 4 ); |
| 419 | } |
| 420 | // NO leading zeroes |
| 421 | return preg_replace( '/(^|:)0+(' . self::RE_IPV6_WORD . ')/', '$1$2', $ip_oct ); |
| 422 | } |
| 423 | |
| 424 | /** |
| 425 | * Converts a hexadecimal number to an IPv4 address in quad-dotted notation |
| 426 | * |
| 427 | * @param string $ip_hex Pure hex |
| 428 | * @return string (of format a.b.c.d) |
| 429 | */ |
| 430 | public static function hexToQuad( $ip_hex ) { |
| 431 | // Pad hex to 8 chars (32 bits) |
| 432 | $ip_hex = str_pad( strtoupper( $ip_hex ), 8, '0', STR_PAD_LEFT ); |
| 433 | // Separate into four quads |
| 434 | $s = ''; |
| 435 | for ( $i = 0; $i < 4; $i++ ) { |
| 436 | if ( $s !== '' ) { |
| 437 | $s .= '.'; |
| 438 | } |
| 439 | $s .= base_convert( substr( $ip_hex, $i * 2, 2 ), 16, 10 ); |
| 440 | } |
| 441 | |
| 442 | return $s; |
| 443 | } |
| 444 | |
| 445 | /** |
| 446 | * Determine if an IP address really is an IP address, and if it is public, |
| 447 | * i.e. not RFC 1918 or similar |
| 448 | * |
| 449 | * @param string $ip |
| 450 | * @return bool |
| 451 | */ |
| 452 | public static function isPublic( $ip ) { |
| 453 | static $privateSet = null; |
| 454 | if ( !$privateSet ) { |
| 455 | $privateSet = new IPSet( [ |
| 456 | // RFC 1918 (private) |
| 457 | '10.0.0.0/8', |
| 458 | // RFC 1918 (private) |
| 459 | '172.16.0.0/12', |
| 460 | // RFC 1918 (private) |
| 461 | '192.168.0.0/16', |
| 462 | // this network |
| 463 | '0.0.0.0/8', |
| 464 | // loopback |
| 465 | '127.0.0.0/8', |
| 466 | // RFC 4193 (local) |
| 467 | 'fc00::/7', |
| 468 | // loopback |
| 469 | '0:0:0:0:0:0:0:1', |
| 470 | // link-local |
| 471 | '169.254.0.0/16', |
| 472 | // link-local |
| 473 | 'fe80::/10', |
| 474 | ] ); |
| 475 | } |
| 476 | return !$privateSet->match( $ip ); |
| 477 | } |
| 478 | |
| 479 | /** |
| 480 | * Return a zero-padded upper case hexadecimal representation of an IP address. |
| 481 | * |
| 482 | * Hexadecimal addresses are used because they can easily be extended to |
| 483 | * IPv6 support. To separate the ranges, the return value from this |
| 484 | * function for an IPv6 address will be prefixed with "v6-", a non- |
| 485 | * hexadecimal string which sorts after the IPv4 addresses. |
| 486 | * |
| 487 | * @param string $ip Quad dotted/octet IP address. |
| 488 | * @return string|bool False on failure |
| 489 | */ |
| 490 | public static function toHex( $ip ) { |
| 491 | if ( self::isIPv6( $ip ) ) { |
| 492 | $n = 'v6-' . self::IPv6ToRawHex( $ip ); |
| 493 | } elseif ( self::isIPv4( $ip ) ) { |
| 494 | // T62035/T97897: An IP with leading 0's fails in ip2long sometimes (e.g. *.08), |
| 495 | // also double/triple 0 needs to be changed to just a single 0 for ip2long. |
| 496 | $ip = self::sanitizeIP( $ip ); |
| 497 | $n = ip2long( $ip ); |
| 498 | if ( $n < 0 ) { |
| 499 | // We don't run code coverage on a 32-bit OS or Windows, so this will never be exercised |
| 500 | // @codeCoverageIgnoreStart |
| 501 | $n += 2 ** 32; |
| 502 | // On 32-bit platforms (and on Windows), 2^32 does not fit into an int, |
| 503 | // so $n becomes a float. We convert it to string instead. |
| 504 | if ( is_float( $n ) ) { |
| 505 | $n = (string)$n; |
| 506 | } |
| 507 | // @codeCoverageIgnoreEnd |
| 508 | } |
| 509 | if ( $n !== false ) { |
| 510 | // Floating points can handle the conversion; faster than \Wikimedia\base_convert() |
| 511 | $n = strtoupper( str_pad( base_convert( $n, 10, 16 ), 8, '0', STR_PAD_LEFT ) ); |
| 512 | } |
| 513 | } else { |
| 514 | $n = false; |
| 515 | } |
| 516 | |
| 517 | return $n; |
| 518 | } |
| 519 | |
| 520 | /** |
| 521 | * Given an IPv6 address in octet notation, returns a pure hex string. |
| 522 | * |
| 523 | * @param string $ip Octet ipv6 IP address. |
| 524 | * @return string|bool Pure hex (uppercase); false on failure |
| 525 | */ |
| 526 | private static function IPv6ToRawHex( $ip ) { |
| 527 | $ip = self::sanitizeIP( $ip ); |
| 528 | if ( !$ip ) { |
| 529 | return false; |
| 530 | } |
| 531 | $r_ip = ''; |
| 532 | foreach ( explode( ':', $ip ) as $v ) { |
| 533 | $r_ip .= str_pad( $v, 4, '0', STR_PAD_LEFT ); |
| 534 | } |
| 535 | |
| 536 | return $r_ip; |
| 537 | } |
| 538 | |
| 539 | /** |
| 540 | * Convert a network specification in CIDR notation |
| 541 | * to an integer network and a number of bits |
| 542 | * |
| 543 | * @param string $range IP with CIDR prefix |
| 544 | * @return array [int|string, int] |
| 545 | */ |
| 546 | public static function parseCIDR( $range ) { |
| 547 | if ( self::isIPv6( $range ) ) { |
| 548 | return self::parseCIDR6( $range ); |
| 549 | } |
| 550 | $parts = explode( '/', $range, 2 ); |
| 551 | if ( count( $parts ) !== 2 ) { |
| 552 | return [ false, false ]; |
| 553 | } |
| 554 | [ $network, $bits ] = $parts; |
| 555 | $network = ip2long( $network ); |
| 556 | if ( $network !== false && is_numeric( $bits ) && $bits >= 0 && $bits <= 32 ) { |
| 557 | '@phan-var int $bits'; |
| 558 | if ( $bits === 0 ) { |
| 559 | $network = 0; |
| 560 | } else { |
| 561 | $network &= ~( ( 1 << ( 32 - (int)$bits ) ) - 1 ); |
| 562 | } |
| 563 | // Convert to unsigned |
| 564 | if ( $network < 0 ) { |
| 565 | $network += 2 ** 32; |
| 566 | } |
| 567 | } else { |
| 568 | $network = false; |
| 569 | $bits = false; |
| 570 | } |
| 571 | |
| 572 | return [ $network, $bits ]; |
| 573 | } |
| 574 | |
| 575 | /** |
| 576 | * Given a string range in a number of formats, |
| 577 | * return the start and end of the range in hexadecimal. |
| 578 | * |
| 579 | * Formats are: |
| 580 | * 1.2.3.4/24 CIDR |
| 581 | * 1.2.3.4 - 1.2.3.5 Explicit range |
| 582 | * 1.2.3.4 Single IP |
| 583 | * |
| 584 | * 2001:0db8:85a3::7344/96 CIDR |
| 585 | * 2001:0db8:85a3::7344 - 2001:0db8:85a3::7344 Explicit range |
| 586 | * 2001:0db8:85a3::7344 Single IP |
| 587 | * @param string $range IP range |
| 588 | * @return array{string,string}|array{false,false} If the start or end of the range |
| 589 | * is invalid, then array `[false, false]` is returned |
| 590 | */ |
| 591 | public static function parseRange( $range ) { |
| 592 | // CIDR notation |
| 593 | if ( strpos( $range, '/' ) !== false ) { |
| 594 | if ( self::isIPv6( $range ) ) { |
| 595 | return self::parseRange6( $range ); |
| 596 | } |
| 597 | [ $network, $bits ] = self::parseCIDR( $range ); |
| 598 | if ( $network === false ) { |
| 599 | $start = $end = false; |
| 600 | } else { |
| 601 | $start = sprintf( '%08X', $network ); |
| 602 | $end = sprintf( '%08X', $network + 2 ** ( 32 - $bits ) - 1 ); |
| 603 | } |
| 604 | // Explicit range |
| 605 | } elseif ( strpos( $range, '-' ) !== false ) { |
| 606 | [ $start, $end ] = array_map( 'trim', explode( '-', $range, 2 ) ); |
| 607 | if ( self::isIPv6( $start ) && self::isIPv6( $end ) ) { |
| 608 | return self::parseRange6( $range ); |
| 609 | } |
| 610 | if ( self::isIPv4( $start ) && self::isIPv4( $end ) ) { |
| 611 | $start = self::toHex( $start ); |
| 612 | $end = self::toHex( $end ); |
| 613 | if ( $start > $end ) { |
| 614 | $start = $end = false; |
| 615 | } |
| 616 | } else { |
| 617 | $start = $end = false; |
| 618 | } |
| 619 | } else { |
| 620 | // Single IP |
| 621 | $start = $end = self::toHex( $range ); |
| 622 | } |
| 623 | if ( $start === false || $end === false ) { |
| 624 | return [ false, false ]; |
| 625 | } |
| 626 | |
| 627 | return [ $start, $end ]; |
| 628 | } |
| 629 | |
| 630 | /** |
| 631 | * Convert a network specification in IPv6 CIDR notation to an |
| 632 | * integer network and a number of bits |
| 633 | * |
| 634 | * @param string $range |
| 635 | * |
| 636 | * @return array{string,int}|array{false,false} |
| 637 | */ |
| 638 | private static function parseCIDR6( $range ) { |
| 639 | // Explode into <expanded IP,range> |
| 640 | $parts = explode( '/', self::sanitizeIP( $range ), 2 ); |
| 641 | if ( count( $parts ) !== 2 ) { |
| 642 | return [ false, false ]; |
| 643 | } |
| 644 | [ $network, $bits ] = $parts; |
| 645 | $network = self::IPv6ToRawHex( $network ); |
| 646 | if ( $network !== false && is_numeric( $bits ) && $bits >= 0 && $bits <= 128 ) { |
| 647 | '@phan-var int $bits'; |
| 648 | if ( $bits === 0 ) { |
| 649 | $network = "0"; |
| 650 | } else { |
| 651 | // Native 32 bit functions WONT work here!!! |
| 652 | // Convert to a padded binary number |
| 653 | $network = \Wikimedia\base_convert( $network, 16, 2, 128 ); |
| 654 | // Truncate the last (128-$bits) bits and replace them with zeros |
| 655 | $network = str_pad( substr( $network, 0, (int)$bits ), 128, '0', STR_PAD_RIGHT ); |
| 656 | // Convert back to an integer |
| 657 | $network = \Wikimedia\base_convert( $network, 2, 10 ); |
| 658 | } |
| 659 | } else { |
| 660 | $network = false; |
| 661 | $bits = false; |
| 662 | } |
| 663 | |
| 664 | return [ $network, (int)$bits ]; |
| 665 | } |
| 666 | |
| 667 | /** |
| 668 | * Given a string range in a number of formats, return the |
| 669 | * start and end of the range in hexadecimal. For IPv6. |
| 670 | * |
| 671 | * Formats are: |
| 672 | * 2001:0db8:85a3::7344/96 CIDR |
| 673 | * 2001:0db8:85a3::7344 - 2001:0db8:85a3::7344 Explicit range |
| 674 | * 2001:0db8:85a3::7344/96 Single IP |
| 675 | * |
| 676 | * @param string $range |
| 677 | * |
| 678 | * @return array [string, string]|array [false, false] If the start or end of the range |
| 679 | * is invalid, then array [false, false] is returned |
| 680 | */ |
| 681 | private static function parseRange6( $range ) { |
| 682 | // Expand any IPv6 IP |
| 683 | $range = self::sanitizeIP( $range ); |
| 684 | |
| 685 | $start = false; |
| 686 | $end = false; |
| 687 | |
| 688 | // CIDR notation... |
| 689 | if ( strpos( $range, '/' ) !== false ) { |
| 690 | [ $network, $bits ] = self::parseCIDR6( $range ); |
| 691 | if ( $network !== false ) { |
| 692 | $start = \Wikimedia\base_convert( $network, 10, 16, 32, false ); |
| 693 | // Turn network to binary (again) |
| 694 | $end = \Wikimedia\base_convert( $network, 10, 2, 128 ); |
| 695 | // Truncate the last (128-$bits) bits and replace them with ones |
| 696 | $end = str_pad( substr( $end, 0, $bits ), 128, '1', STR_PAD_RIGHT ); |
| 697 | // Convert to hex |
| 698 | $end = \Wikimedia\base_convert( $end, 2, 16, 32, false ); |
| 699 | // see toHex() comment |
| 700 | $start = "v6-$start"; |
| 701 | $end = "v6-$end"; |
| 702 | } |
| 703 | // Explicit range notation... |
| 704 | } elseif ( strpos( $range, '-' ) !== false ) { |
| 705 | [ $start, $end ] = array_map( 'trim', explode( '-', $range, 2 ) ); |
| 706 | $start = self::toHex( $start ); |
| 707 | $end = self::toHex( $end ); |
| 708 | if ( $start > $end ) { |
| 709 | $start = $end = false; |
| 710 | } |
| 711 | } |
| 712 | |
| 713 | if ( $start === false || $end === false ) { |
| 714 | return [ false, false ]; |
| 715 | } |
| 716 | |
| 717 | return [ $start, $end ]; |
| 718 | } |
| 719 | |
| 720 | /** |
| 721 | * Determine if a given IPv4/IPv6 address is in a given CIDR network |
| 722 | * |
| 723 | * @param string $addr The address to check against the given range. |
| 724 | * @param string $range The range to check the given address against. |
| 725 | * @return bool Whether or not the given address is in the given range. |
| 726 | * |
| 727 | * @note This can return unexpected results for invalid arguments! |
| 728 | * Make sure you pass a valid IP address and IP range. |
| 729 | */ |
| 730 | public static function isInRange( $addr, $range ) { |
| 731 | $hexIP = self::toHex( $addr ); |
| 732 | [ $start, $end ] = self::parseRange( $range ); |
| 733 | |
| 734 | return strcmp( $hexIP, $start ) >= 0 && |
| 735 | strcmp( $hexIP, $end ) <= 0; |
| 736 | } |
| 737 | |
| 738 | /** |
| 739 | * Determines if an IP address is a list of CIDR a.b.c.d/n ranges. |
| 740 | * |
| 741 | * @param string $ip the IP to check |
| 742 | * @param array $ranges the IP ranges, each element a range |
| 743 | * |
| 744 | * @return bool true if the specified adress belongs to the specified range; otherwise, false. |
| 745 | */ |
| 746 | public static function isInRanges( $ip, $ranges ) { |
| 747 | foreach ( $ranges as $range ) { |
| 748 | if ( self::isInRange( $ip, $range ) ) { |
| 749 | return true; |
| 750 | } |
| 751 | } |
| 752 | return false; |
| 753 | } |
| 754 | |
| 755 | /** |
| 756 | * Convert some unusual representations of IPv4 addresses to their |
| 757 | * canonical dotted quad representation. |
| 758 | * |
| 759 | * This currently only checks a few IPV4-to-IPv6 related cases. More |
| 760 | * unusual representations may be added later. |
| 761 | * |
| 762 | * @param string $addr Something that might be an IP address |
| 763 | * @return string|null Valid IP address or null |
| 764 | * @return-taint none |
| 765 | */ |
| 766 | public static function canonicalize( $addr ) { |
| 767 | // remove zone info (T37738) |
| 768 | $addr = preg_replace( '/\%.*/', '', $addr ); |
| 769 | |
| 770 | // If it's already a valid IPv4 address, nothing to do |
| 771 | if ( self::isValidIPv4( $addr ) ) { |
| 772 | return $addr; |
| 773 | } |
| 774 | |
| 775 | // https://en.wikipedia.org/wiki/IPv6#IPv4-mapped_IPv6_addresses |
| 776 | // Turn mapped addresses from: |
| 777 | // ::ce:ffff:1.2.3.4 to 1.2.3.4 (IPv4-mapped IPv6 addresses) |
| 778 | // ::1.2.3.4 to 1.2.3.4 (IPv4-compatible IPv6 address) |
| 779 | // IPv4-compatible IPv6 addresses are now deprecated https://tools.ietf.org/html/rfc4291#section-2.5.5.1 |
| 780 | if ( preg_match( '/^' . self::RE_IPV6_V4_PREFIX . '(' . self::RE_IP_ADD . ')$/i', $addr, $m ) ) { |
| 781 | return $m[1]; |
| 782 | } |
| 783 | |
| 784 | // Converts :ffff:1F to 255.255.0.31 |
| 785 | // Is this actually used/needed? |
| 786 | if ( preg_match( '/^' . self::RE_IPV6_V4_PREFIX . self::RE_IPV6_WORD . |
| 787 | ':' . self::RE_IPV6_WORD . '$/i', $addr, $m ) |
| 788 | ) { |
| 789 | return long2ip( ( hexdec( $m[1] ) << 16 ) + hexdec( $m[2] ) ); |
| 790 | } |
| 791 | |
| 792 | // It's a valid IPv6 address that we haven't canonicalized, so return it |
| 793 | if ( self::isValidIPv6( $addr ) ) { |
| 794 | return $addr; |
| 795 | } |
| 796 | |
| 797 | // Not a valid IP address |
| 798 | return null; |
| 799 | } |
| 800 | |
| 801 | /** |
| 802 | * Gets rid of unneeded numbers in quad-dotted/octet IP strings |
| 803 | * For example, 127.111.113.151/24 -> 127.111.113.0/24 |
| 804 | * @param string $range IP address to normalize |
| 805 | * @return string |
| 806 | */ |
| 807 | public static function sanitizeRange( $range ) { |
| 808 | [ , $bits ] = self::parseCIDR( $range ); |
| 809 | [ $start, ] = self::parseRange( $range ); |
| 810 | $start = self::formatHex( $start ); |
| 811 | if ( $bits === false ) { |
| 812 | // wasn't actually a range |
| 813 | return $start; |
| 814 | } |
| 815 | |
| 816 | return "$start/$bits"; |
| 817 | } |
| 818 | |
| 819 | /** |
| 820 | * Returns the subnet of a given IP |
| 821 | * |
| 822 | * @param string $ip |
| 823 | * @return string|false |
| 824 | */ |
| 825 | public static function getSubnet( $ip ) { |
| 826 | $matches = []; |
| 827 | $subnet = false; |
| 828 | if ( self::isIPv6( $ip ) ) { |
| 829 | $parts = self::parseRange( "$ip/64" ); |
| 830 | $subnet = $parts[0]; |
| 831 | } elseif ( preg_match( '/^' . self::RE_IP_ADD . '$/', $ip, $matches ) ) { |
| 832 | // IPv4 |
| 833 | $subnet = "{$matches[1]}.{$matches[2]}.{$matches[3]}"; |
| 834 | } |
| 835 | return $subnet; |
| 836 | } |
| 837 | |
| 838 | /** |
| 839 | * Return all the addresses in a given range |
| 840 | * |
| 841 | * This currently does not support IPv6 ranges and is limited to /16 block (65535 addresses). |
| 842 | * |
| 843 | * @param string $range IP ranges to get the IPs within |
| 844 | * @return string[] Array of addresses in the range |
| 845 | * @throws InvalidArgumentException If input uses IPv6 |
| 846 | * @throws InvalidArgumentException If input range is too large |
| 847 | */ |
| 848 | public static function getIPsInRange( $range ): array { |
| 849 | // No IPv6 for now. |
| 850 | if ( self::isValidIPv6( $range ) || self::isValidIPv6Range( $range ) ) { |
| 851 | throw new InvalidArgumentException( 'Cannot retrieve addresses for IPv6 range: ' . $range ); |
| 852 | } |
| 853 | |
| 854 | [ $start, $end ] = self::parseRange( $range ); |
| 855 | if ( $start === false || $start === $end ) { |
| 856 | throw new InvalidArgumentException( 'Invalid range given: ' . $range ); |
| 857 | } |
| 858 | |
| 859 | if ( hexdec( $end ) - hexdec( $start ) > self::MAXIMUM_IPS_FROM_RANGE ) { |
| 860 | throw new InvalidArgumentException( "Range {$range} is too large, it contains more than " |
| 861 | . self::MAXIMUM_IPS_FROM_RANGE . ' addresses' ); |
| 862 | } |
| 863 | |
| 864 | $start = ip2long( self::formatHex( $start ) ); |
| 865 | $end = ip2long( self::formatHex( $end ) ); |
| 866 | |
| 867 | return array_map( 'long2ip', range( $start, $end ) ); |
| 868 | } |
| 869 | } |