| commit | author | age | ||
| c321a9 | 1 | <?php |
| T | 2 | |
| 3 | /* | |
| 4 | +-----------------------------------------------------------------------+ | |
| 5 | | program/include/rcube_charset.php | | |
| 6 | | | | |
| 7 | | This file is part of the Roundcube Webmail client | | |
| 8 | | Copyright (C) 2005-2012, The Roundcube Dev Team | | |
| 9 | | Copyright (C) 2011-2012, Kolab Systems AG | | |
| 10 | | Copyright (C) 2000 Edmund Grimley Evans <edmundo@rano.org> | | |
| 7fe381 | 11 | | | |
| T | 12 | | Licensed under the GNU General Public License version 3 or | |
| 13 | | any later version with exceptions for skins & plugins. | | |
| 14 | | See the README file for a full license statement. | | |
| c321a9 | 15 | | | |
| T | 16 | | PURPOSE: | |
| 17 | | Provide charset conversion functionality | | |
| 18 | | | | |
| 19 | +-----------------------------------------------------------------------+ | |
| 20 | | Author: Thomas Bruederli <roundcube@gmail.com> | | |
| 21 | | Author: Aleksander Machniak <alec@alec.pl> | | |
| 22 | +-----------------------------------------------------------------------+ | |
| 23 | ||
| 24 | $Id$ | |
| 25 | ||
| 26 | */ | |
| 27 | ||
| 28 | /** | |
| 29 | * Character sets conversion functionality | |
| 30 | * | |
| 31 | * @package Core | |
| 32 | * @author Thomas Bruederli <roundcube@gmail.com> | |
| 33 | * @author Aleksander Machniak <alec@alec.pl> | |
| 34 | * @author Edmund Grimley Evans <edmundo@rano.org> | |
| 35 | */ | |
| 36 | class rcube_charset | |
| 37 | { | |
| 38 | // Aliases: some of them from HTML5 spec. | |
| 39 | static public $aliases = array( | |
| 40 | 'USASCII' => 'WINDOWS-1252', | |
| 41 | 'ANSIX31101983' => 'WINDOWS-1252', | |
| 42 | 'ANSIX341968' => 'WINDOWS-1252', | |
| 43 | 'UNKNOWN8BIT' => 'ISO-8859-15', | |
| 44 | 'UNKNOWN' => 'ISO-8859-15', | |
| 45 | 'USERDEFINED' => 'ISO-8859-15', | |
| 46 | 'KSC56011987' => 'EUC-KR', | |
| 1495ac | 47 | 'GB2312' => 'GBK', |
| A | 48 | 'GB231280' => 'GBK', |
| 49 | 'UNICODE' => 'UTF-8', | |
| 50 | 'UTF7IMAP' => 'UTF7-IMAP', | |
| 51 | 'TIS620' => 'WINDOWS-874', | |
| 52 | 'ISO88599' => 'WINDOWS-1254', | |
| 53 | 'ISO885911' => 'WINDOWS-874', | |
| 54 | 'MACROMAN' => 'MACINTOSH', | |
| c321a9 | 55 | '77' => 'MAC', |
| T | 56 | '128' => 'SHIFT-JIS', |
| 57 | '129' => 'CP949', | |
| 58 | '130' => 'CP1361', | |
| 59 | '134' => 'GBK', | |
| 60 | '136' => 'BIG5', | |
| 61 | '161' => 'WINDOWS-1253', | |
| 62 | '162' => 'WINDOWS-1254', | |
| 63 | '163' => 'WINDOWS-1258', | |
| 64 | '177' => 'WINDOWS-1255', | |
| 65 | '178' => 'WINDOWS-1256', | |
| 66 | '186' => 'WINDOWS-1257', | |
| 67 | '204' => 'WINDOWS-1251', | |
| 68 | '222' => 'WINDOWS-874', | |
| 69 | '238' => 'WINDOWS-1250', | |
| 70 | 'MS950' => 'CP950', | |
| 71 | 'WINDOWS949' => 'UHC', | |
| 72 | ); | |
| 73 | ||
| 74 | ||
| 75 | /** | |
| 76 | * Catch an error and throw an exception. | |
| 77 | * | |
| 78 | * @param int Level of the error | |
| 79 | * @param string Error message | |
| 80 | */ | |
| 66d215 | 81 | public static function error_handler($errno, $errstr) |
| c321a9 | 82 | { |
| T | 83 | throw new ErrorException($errstr, 0, $errno); |
| 84 | } | |
| 85 | ||
| 86 | ||
| 87 | /** | |
| 88 | * Parse and validate charset name string (see #1485758). | |
| 89 | * Sometimes charset string is malformed, there are also charset aliases | |
| 90 | * but we need strict names for charset conversion (specially utf8 class) | |
| 91 | * | |
| 92 | * @param string Input charset name | |
| 93 | * | |
| 94 | * @return string The validated charset name | |
| 95 | */ | |
| f5f9b4 | 96 | public static function parse_charset($input) |
| c321a9 | 97 | { |
| T | 98 | static $charsets = array(); |
| 99 | $charset = strtoupper($input); | |
| 100 | ||
| 101 | if (isset($charsets[$input])) { | |
| 102 | return $charsets[$input]; | |
| 103 | } | |
| 104 | ||
| 105 | $charset = preg_replace(array( | |
| 106 | '/^[^0-9A-Z]+/', // e.g. _ISO-8859-JP$SIO | |
| 107 | '/\$.*$/', // e.g. _ISO-8859-JP$SIO | |
| 108 | '/UNICODE-1-1-*/', // RFC1641/1642 | |
| 109 | '/^X-/', // X- prefix (e.g. X-ROMAN8 => ROMAN8) | |
| 110 | ), '', $charset); | |
| 111 | ||
| 112 | if ($charset == 'BINARY') { | |
| 113 | return $charsets[$input] = null; | |
| 114 | } | |
| 115 | ||
| 116 | // allow A-Z and 0-9 only | |
| 117 | $str = preg_replace('/[^A-Z0-9]/', '', $charset); | |
| 118 | ||
| 119 | if (isset(self::$aliases[$str])) { | |
| 120 | $result = self::$aliases[$str]; | |
| 121 | } | |
| 122 | // UTF | |
| 123 | else if (preg_match('/U[A-Z][A-Z](7|8|16|32)(BE|LE)*/', $str, $m)) { | |
| 124 | $result = 'UTF-' . $m[1] . $m[2]; | |
| 125 | } | |
| 126 | // ISO-8859 | |
| 127 | else if (preg_match('/ISO8859([0-9]{0,2})/', $str, $m)) { | |
| 128 | $iso = 'ISO-8859-' . ($m[1] ? $m[1] : 1); | |
| 129 | // some clients sends windows-1252 text as latin1, | |
| 130 | // it is safe to use windows-1252 for all latin1 | |
| 131 | $result = $iso == 'ISO-8859-1' ? 'WINDOWS-1252' : $iso; | |
| 132 | } | |
| 133 | // handle broken charset names e.g. WINDOWS-1250HTTP-EQUIVCONTENT-TYPE | |
| 134 | else if (preg_match('/(WIN|WINDOWS)([0-9]+)/', $str, $m)) { | |
| 135 | $result = 'WINDOWS-' . $m[2]; | |
| 136 | } | |
| 137 | // LATIN | |
| 138 | else if (preg_match('/LATIN(.*)/', $str, $m)) { | |
| 139 | $aliases = array('2' => 2, '3' => 3, '4' => 4, '5' => 9, '6' => 10, | |
| 140 | '7' => 13, '8' => 14, '9' => 15, '10' => 16, | |
| 141 | 'ARABIC' => 6, 'CYRILLIC' => 5, 'GREEK' => 7, 'GREEK1' => 7, 'HEBREW' => 8 | |
| 142 | ); | |
| 143 | ||
| 144 | // some clients sends windows-1252 text as latin1, | |
| 145 | // it is safe to use windows-1252 for all latin1 | |
| 146 | if ($m[1] == 1) { | |
| 147 | $result = 'WINDOWS-1252'; | |
| 148 | } | |
| 149 | // if iconv is not supported we need ISO labels, it's also safe for iconv | |
| 150 | else if (!empty($aliases[$m[1]])) { | |
| 151 | $result = 'ISO-8859-'.$aliases[$m[1]]; | |
| 152 | } | |
| 153 | // iconv requires convertion of e.g. LATIN-1 to LATIN1 | |
| 154 | else { | |
| 155 | $result = $str; | |
| 156 | } | |
| 157 | } | |
| 158 | else { | |
| 159 | $result = $charset; | |
| 160 | } | |
| 161 | ||
| 162 | $charsets[$input] = $result; | |
| 163 | ||
| 164 | return $result; | |
| 165 | } | |
| 166 | ||
| 167 | ||
| 168 | /** | |
| 169 | * Convert a string from one charset to another. | |
| 170 | * Uses mbstring and iconv functions if possible | |
| 171 | * | |
| 172 | * @param string Input string | |
| 173 | * @param string Suspected charset of the input string | |
| 174 | * @param string Target charset to convert to; defaults to RCMAIL_CHARSET | |
| 175 | * | |
| 176 | * @return string Converted string | |
| 177 | */ | |
| 178 | public static function convert($str, $from, $to = null) | |
| 179 | { | |
| 180 | static $iconv_options = null; | |
| 181 | static $mbstring_list = null; | |
| 182 | static $conv = null; | |
| 183 | ||
| f5f9b4 | 184 | $to = empty($to) ? strtoupper(RCMAIL_CHARSET) : self::parse_charset($to); |
| A | 185 | $from = self::parse_charset($from); |
| c321a9 | 186 | |
| T | 187 | if ($from == $to || empty($str) || empty($from)) { |
| 188 | return $str; | |
| 189 | } | |
| 190 | ||
| f5f9b4 | 191 | if ($iconv_options === null) { |
| A | 192 | if (function_exists('iconv')) { |
| c321a9 | 193 | // ignore characters not available in output charset |
| T | 194 | $iconv_options = '//IGNORE'; |
| 195 | if (iconv('', $iconv_options, '') === false) { | |
| 196 | // iconv implementation does not support options | |
| 197 | $iconv_options = ''; | |
| 198 | } | |
| 199 | } | |
| f5f9b4 | 200 | } |
| c321a9 | 201 | |
| f5f9b4 | 202 | // convert charset using iconv module |
| A | 203 | if ($iconv_options !== null && $from != 'UTF7-IMAP' && $to != 'UTF7-IMAP') { |
| c321a9 | 204 | // throw an exception if iconv reports an illegal character in input |
| T | 205 | // it means that input string has been truncated |
| 206 | set_error_handler(array('rcube_charset', 'error_handler'), E_NOTICE); | |
| 207 | try { | |
| 208 | $_iconv = iconv($from, $to . $iconv_options, $str); | |
| 209 | } catch (ErrorException $e) { | |
| 210 | $_iconv = false; | |
| 211 | } | |
| 212 | restore_error_handler(); | |
| 213 | ||
| 214 | if ($_iconv !== false) { | |
| 215 | return $_iconv; | |
| 216 | } | |
| 217 | } | |
| 218 | ||
| f5f9b4 | 219 | if ($mbstring_list === null) { |
| A | 220 | if (extension_loaded('mbstring')) { |
| c321a9 | 221 | $mbstring_list = mb_list_encodings(); |
| T | 222 | $mbstring_list = array_map('strtoupper', $mbstring_list); |
| 223 | } | |
| f5f9b4 | 224 | } |
| A | 225 | |
| 226 | // convert charset using mbstring module | |
| 227 | if ($mbstring_list !== null) { | |
| 228 | $aliases['WINDOWS-1257'] = 'ISO-8859-13'; | |
| c321a9 | 229 | |
| T | 230 | $mb_from = $aliases[$from] ? $aliases[$from] : $from; |
| 231 | $mb_to = $aliases[$to] ? $aliases[$to] : $to; | |
| 232 | ||
| 233 | // return if encoding found, string matches encoding and convert succeeded | |
| 234 | if (in_array($mb_from, $mbstring_list) && in_array($mb_to, $mbstring_list)) { | |
| 235 | if (mb_check_encoding($str, $mb_from) && ($out = mb_convert_encoding($str, $mb_to, $mb_from))) { | |
| 236 | return $out; | |
| 237 | } | |
| 238 | } | |
| 239 | } | |
| 240 | ||
| 241 | // convert charset using bundled classes/functions | |
| 242 | if ($to == 'UTF-8') { | |
| 243 | if ($from == 'UTF7-IMAP') { | |
| 244 | if ($_str = self::utf7imap_to_utf8($str)) { | |
| 245 | return $_str; | |
| 246 | } | |
| 247 | } | |
| 248 | else if ($from == 'UTF-7') { | |
| 249 | if ($_str = self::utf7_to_utf8($str)) { | |
| 250 | return $_str; | |
| 251 | } | |
| 252 | } | |
| 253 | else if ($from == 'ISO-8859-1' && function_exists('utf8_encode')) { | |
| 254 | return utf8_encode($str); | |
| 255 | } | |
| 256 | else if (class_exists('utf8')) { | |
| 257 | if (!$conv) { | |
| 258 | $conv = new utf8($from); | |
| 259 | } | |
| 260 | else { | |
| 261 | $conv->loadCharset($from); | |
| 262 | } | |
| 263 | ||
| 264 | if ($_str = $conv->strToUtf8($str)) { | |
| 265 | return $_str; | |
| 266 | } | |
| 267 | } | |
| 268 | } | |
| 269 | ||
| 270 | // encode string for output | |
| 271 | if ($from == 'UTF-8') { | |
| 272 | // @TODO: we need a function for UTF-7 (RFC2152) conversion | |
| 273 | if ($to == 'UTF7-IMAP' || $to == 'UTF-7') { | |
| 20efa5 | 274 | if ($_str = self::utf8_to_utf7imap($str)) { |
| c321a9 | 275 | return $_str; |
| T | 276 | } |
| 277 | } | |
| 278 | else if ($to == 'ISO-8859-1' && function_exists('utf8_decode')) { | |
| 279 | return utf8_decode($str); | |
| 280 | } | |
| 281 | else if (class_exists('utf8')) { | |
| 282 | if (!$conv) { | |
| 283 | $conv = new utf8($to); | |
| 284 | } | |
| 285 | else { | |
| 286 | $conv->loadCharset($from); | |
| 287 | } | |
| 288 | ||
| 289 | if ($_str = $conv->strToUtf8($str)) { | |
| 290 | return $_str; | |
| 291 | } | |
| 292 | } | |
| 293 | } | |
| 294 | ||
| 295 | // return original string | |
| 296 | return $str; | |
| 297 | } | |
| 298 | ||
| 299 | ||
| 300 | /** | |
| 301 | * Converts string from standard UTF-7 (RFC 2152) to UTF-8. | |
| 302 | * | |
| 303 | * @param string Input string (UTF-7) | |
| 304 | * | |
| 305 | * @return string Converted string (UTF-8) | |
| 306 | */ | |
| 307 | public static function utf7_to_utf8($str) | |
| 308 | { | |
| 309 | $Index_64 = array( | |
| 310 | 0,0,0,0, 0,0,0,0, 0,0,0,0, 0,0,0,0, | |
| 311 | 0,0,0,0, 0,0,0,0, 0,0,0,0, 0,0,0,0, | |
| 312 | 0,0,0,0, 0,0,0,0, 0,0,0,1, 0,0,0,0, | |
| 313 | 1,1,1,1, 1,1,1,1, 1,1,0,0, 0,0,0,0, | |
| 314 | 0,1,1,1, 1,1,1,1, 1,1,1,1, 1,1,1,1, | |
| 315 | 1,1,1,1, 1,1,1,1, 1,1,1,0, 0,0,0,0, | |
| 316 | 0,1,1,1, 1,1,1,1, 1,1,1,1, 1,1,1,1, | |
| 317 | 1,1,1,1, 1,1,1,1, 1,1,1,0, 0,0,0,0, | |
| 318 | ); | |
| 319 | ||
| 320 | $u7len = strlen($str); | |
| 321 | $str = strval($str); | |
| 322 | $res = ''; | |
| 323 | ||
| 324 | for ($i=0; $u7len > 0; $i++, $u7len--) { | |
| 325 | $u7 = $str[$i]; | |
| 326 | if ($u7 == '+') { | |
| 327 | $i++; | |
| 328 | $u7len--; | |
| 329 | $ch = ''; | |
| 330 | ||
| 331 | for (; $u7len > 0; $i++, $u7len--) { | |
| 332 | $u7 = $str[$i]; | |
| 333 | ||
| 334 | if (!$Index_64[ord($u7)]) { | |
| 335 | break; | |
| 336 | } | |
| 337 | ||
| 1495ac | 338 | $ch .= $u7; |
| c321a9 | 339 | } |
| T | 340 | |
| 341 | if ($ch == '') { | |
| 342 | if ($u7 == '-') { | |
| 343 | $res .= '+'; | |
| 344 | } | |
| 345 | ||
| 346 | continue; | |
| 347 | } | |
| 348 | ||
| 349 | $res .= self::utf16_to_utf8(base64_decode($ch)); | |
| 350 | } | |
| 351 | else { | |
| 352 | $res .= $u7; | |
| 353 | } | |
| 354 | } | |
| 355 | ||
| 356 | return $res; | |
| 357 | } | |
| 358 | ||
| 359 | ||
| 360 | /** | |
| 361 | * Converts string from UTF-16 to UTF-8 (helper for utf-7 to utf-8 conversion) | |
| 362 | * | |
| 363 | * @param string Input string | |
| 364 | * | |
| 365 | * @return string The converted string | |
| 366 | */ | |
| 367 | public static function utf16_to_utf8($str) | |
| 368 | { | |
| 369 | $len = strlen($str); | |
| 370 | $dec = ''; | |
| 371 | ||
| 372 | for ($i = 0; $i < $len; $i += 2) { | |
| 373 | $c = ord($str[$i]) << 8 | ord($str[$i + 1]); | |
| 374 | if ($c >= 0x0001 && $c <= 0x007F) { | |
| 375 | $dec .= chr($c); | |
| 376 | } | |
| 377 | else if ($c > 0x07FF) { | |
| 378 | $dec .= chr(0xE0 | (($c >> 12) & 0x0F)); | |
| 379 | $dec .= chr(0x80 | (($c >> 6) & 0x3F)); | |
| 380 | $dec .= chr(0x80 | (($c >> 0) & 0x3F)); | |
| 381 | } | |
| 382 | else { | |
| 383 | $dec .= chr(0xC0 | (($c >> 6) & 0x1F)); | |
| 384 | $dec .= chr(0x80 | (($c >> 0) & 0x3F)); | |
| 385 | } | |
| 386 | } | |
| 387 | ||
| 388 | return $dec; | |
| 389 | } | |
| 390 | ||
| 391 | ||
| 392 | /** | |
| 393 | * Convert the data ($str) from RFC 2060's UTF-7 to UTF-8. | |
| 394 | * If input data is invalid, return the original input string. | |
| 395 | * RFC 2060 obviously intends the encoding to be unique (see | |
| 396 | * point 5 in section 5.1.3), so we reject any non-canonical | |
| 397 | * form, such as &ACY- (instead of &-) or &AMA-&AMA- (instead | |
| 398 | * of &AMAAwA-). | |
| 399 | * | |
| 400 | * Translated from C to PHP by Thomas Bruederli <roundcube@gmail.com> | |
| 401 | * | |
| 402 | * @param string $str Input string (UTF7-IMAP) | |
| 403 | * | |
| 404 | * @return string Output string (UTF-8) | |
| 405 | */ | |
| 406 | public static function utf7imap_to_utf8($str) | |
| 407 | { | |
| 408 | $Index_64 = array( | |
| 409 | -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1, | |
| 410 | -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1, | |
| 411 | -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,62, 63,-1,-1,-1, | |
| 412 | 52,53,54,55, 56,57,58,59, 60,61,-1,-1, -1,-1,-1,-1, | |
| 413 | -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10, 11,12,13,14, | |
| 414 | 15,16,17,18, 19,20,21,22, 23,24,25,-1, -1,-1,-1,-1, | |
| 415 | -1,26,27,28, 29,30,31,32, 33,34,35,36, 37,38,39,40, | |
| 416 | 41,42,43,44, 45,46,47,48, 49,50,51,-1, -1,-1,-1,-1 | |
| 417 | ); | |
| 418 | ||
| 419 | $u7len = strlen($str); | |
| 420 | $str = strval($str); | |
| 421 | $p = ''; | |
| 422 | $err = ''; | |
| 423 | ||
| 424 | for ($i=0; $u7len > 0; $i++, $u7len--) { | |
| 425 | $u7 = $str[$i]; | |
| 426 | if ($u7 == '&') { | |
| 427 | $i++; | |
| 428 | $u7len--; | |
| 429 | $u7 = $str[$i]; | |
| 430 | ||
| 431 | if ($u7len && $u7 == '-') { | |
| 432 | $p .= '&'; | |
| 433 | continue; | |
| 434 | } | |
| 435 | ||
| 436 | $ch = 0; | |
| 437 | $k = 10; | |
| 438 | for (; $u7len > 0; $i++, $u7len--) { | |
| 439 | $u7 = $str[$i]; | |
| 440 | ||
| 441 | if ((ord($u7) & 0x80) || ($b = $Index_64[ord($u7)]) == -1) { | |
| 442 | break; | |
| 443 | } | |
| 444 | ||
| 445 | if ($k > 0) { | |
| 446 | $ch |= $b << $k; | |
| 447 | $k -= 6; | |
| 448 | } | |
| 449 | else { | |
| 450 | $ch |= $b >> (-$k); | |
| 451 | if ($ch < 0x80) { | |
| 452 | // Printable US-ASCII | |
| 453 | if (0x20 <= $ch && $ch < 0x7f) { | |
| 454 | return $err; | |
| 455 | } | |
| 456 | $p .= chr($ch); | |
| 457 | } | |
| 458 | else if ($ch < 0x800) { | |
| 459 | $p .= chr(0xc0 | ($ch >> 6)); | |
| 460 | $p .= chr(0x80 | ($ch & 0x3f)); | |
| 461 | } | |
| 462 | else { | |
| 463 | $p .= chr(0xe0 | ($ch >> 12)); | |
| 464 | $p .= chr(0x80 | (($ch >> 6) & 0x3f)); | |
| 465 | $p .= chr(0x80 | ($ch & 0x3f)); | |
| 466 | } | |
| 467 | ||
| 468 | $ch = ($b << (16 + $k)) & 0xffff; | |
| 469 | $k += 10; | |
| 470 | } | |
| 471 | } | |
| 472 | ||
| 473 | // Non-zero or too many extra bits | |
| 474 | if ($ch || $k < 6) { | |
| 475 | return $err; | |
| 476 | } | |
| 477 | ||
| 478 | // BASE64 not properly terminated | |
| 479 | if (!$u7len || $u7 != '-') { | |
| 480 | return $err; | |
| 481 | } | |
| 482 | ||
| 483 | // Adjacent BASE64 sections | |
| 484 | if ($u7len > 2 && $str[$i+1] == '&' && $str[$i+2] != '-') { | |
| 485 | return $err; | |
| 486 | } | |
| 487 | } | |
| 488 | // Not printable US-ASCII | |
| 489 | else if (ord($u7) < 0x20 || ord($u7) >= 0x7f) { | |
| 490 | return $err; | |
| 491 | } | |
| 492 | else { | |
| 493 | $p .= $u7; | |
| 494 | } | |
| 495 | } | |
| 496 | ||
| 497 | return $p; | |
| 498 | } | |
| 499 | ||
| 500 | ||
| 501 | /** | |
| 502 | * Convert the data ($str) from UTF-8 to RFC 2060's UTF-7. | |
| 503 | * Unicode characters above U+FFFF are replaced by U+FFFE. | |
| 504 | * If input data is invalid, return an empty string. | |
| 505 | * | |
| 506 | * Translated from C to PHP by Thomas Bruederli <roundcube@gmail.com> | |
| 507 | * | |
| 508 | * @param string $str Input string (UTF-8) | |
| 509 | * | |
| 510 | * @return string Output string (UTF7-IMAP) | |
| 511 | */ | |
| 512 | public static function utf8_to_utf7imap($str) | |
| 513 | { | |
| 514 | $B64Chars = array( | |
| 515 | 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', | |
| 516 | 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'a', 'b', 'c', 'd', | |
| 517 | 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', | |
| 518 | 't', 'u', 'v', 'w', 'x', 'y', 'z', '0', '1', '2', '3', '4', '5', '6', '7', | |
| 519 | '8', '9', '+', ',' | |
| 520 | ); | |
| 521 | ||
| 522 | $u8len = strlen($str); | |
| 523 | $base64 = 0; | |
| 524 | $i = 0; | |
| 525 | $p = ''; | |
| 526 | $err = ''; | |
| 527 | ||
| 528 | while ($u8len) { | |
| 529 | $u8 = $str[$i]; | |
| 530 | $c = ord($u8); | |
| 531 | ||
| 532 | if ($c < 0x80) { | |
| 533 | $ch = $c; | |
| 534 | $n = 0; | |
| 535 | } | |
| 536 | else if ($c < 0xc2) { | |
| 537 | return $err; | |
| 538 | } | |
| 539 | else if ($c < 0xe0) { | |
| 540 | $ch = $c & 0x1f; | |
| 541 | $n = 1; | |
| 542 | } | |
| 543 | else if ($c < 0xf0) { | |
| 544 | $ch = $c & 0x0f; | |
| 545 | $n = 2; | |
| 546 | } | |
| 547 | else if ($c < 0xf8) { | |
| 548 | $ch = $c & 0x07; | |
| 549 | $n = 3; | |
| 550 | } | |
| 551 | else if ($c < 0xfc) { | |
| 552 | $ch = $c & 0x03; | |
| 553 | $n = 4; | |
| 554 | } | |
| 555 | else if ($c < 0xfe) { | |
| 556 | $ch = $c & 0x01; | |
| 557 | $n = 5; | |
| 558 | } | |
| 559 | else { | |
| 560 | return $err; | |
| 561 | } | |
| 562 | ||
| 563 | $i++; | |
| 564 | $u8len--; | |
| 565 | ||
| 566 | if ($n > $u8len) { | |
| 567 | return $err; | |
| 568 | } | |
| 569 | ||
| 570 | for ($j=0; $j < $n; $j++) { | |
| 571 | $o = ord($str[$i+$j]); | |
| 572 | if (($o & 0xc0) != 0x80) { | |
| 573 | return $err; | |
| 574 | } | |
| 575 | $ch = ($ch << 6) | ($o & 0x3f); | |
| 576 | } | |
| 577 | ||
| 578 | if ($n > 1 && !($ch >> ($n * 5 + 1))) { | |
| 579 | return $err; | |
| 580 | } | |
| 581 | ||
| 582 | $i += $n; | |
| 583 | $u8len -= $n; | |
| 584 | ||
| 585 | if ($ch < 0x20 || $ch >= 0x7f) { | |
| 586 | if (!$base64) { | |
| 587 | $p .= '&'; | |
| 588 | $base64 = 1; | |
| 589 | $b = 0; | |
| 590 | $k = 10; | |
| 591 | } | |
| 592 | if ($ch & ~0xffff) { | |
| 593 | $ch = 0xfffe; | |
| 594 | } | |
| 595 | ||
| 596 | $p .= $B64Chars[($b | $ch >> $k)]; | |
| 597 | $k -= 6; | |
| 598 | for (; $k >= 0; $k -= 6) { | |
| 599 | $p .= $B64Chars[(($ch >> $k) & 0x3f)]; | |
| 600 | } | |
| 601 | ||
| 602 | $b = ($ch << (-$k)) & 0x3f; | |
| 603 | $k += 16; | |
| 604 | } | |
| 605 | else { | |
| 606 | if ($base64) { | |
| 607 | if ($k > 10) { | |
| 608 | $p .= $B64Chars[$b]; | |
| 609 | } | |
| 610 | $p .= '-'; | |
| 611 | $base64 = 0; | |
| 612 | } | |
| 613 | ||
| 614 | $p .= chr($ch); | |
| 615 | if (chr($ch) == '&') { | |
| 616 | $p .= '-'; | |
| 617 | } | |
| 618 | } | |
| 619 | } | |
| 620 | ||
| 621 | if ($base64) { | |
| 622 | if ($k > 10) { | |
| 623 | $p .= $B64Chars[$b]; | |
| 624 | } | |
| 625 | $p .= '-'; | |
| 626 | } | |
| 627 | ||
| 628 | return $p; | |
| 629 | } | |
| 630 | ||
| 631 | ||
| 632 | /** | |
| 633 | * A method to guess character set of a string. | |
| 634 | * | |
| 635 | * @param string $string String. | |
| 636 | * @param string $failover Default result for failover. | |
| 637 | * | |
| 638 | * @return string Charset name | |
| 639 | */ | |
| 640 | public static function detect($string, $failover='') | |
| 641 | { | |
| 642 | if (!function_exists('mb_detect_encoding')) { | |
| 643 | return $failover; | |
| 644 | } | |
| 645 | ||
| 646 | // FIXME: the order is important, because sometimes | |
| 647 | // iso string is detected as euc-jp and etc. | |
| 648 | $enc = array( | |
| 649 | 'UTF-8', 'SJIS', 'BIG5', 'GB2312', | |
| 650 | 'ISO-8859-1', 'ISO-8859-2', 'ISO-8859-3', 'ISO-8859-4', | |
| 651 | 'ISO-8859-5', 'ISO-8859-6', 'ISO-8859-7', 'ISO-8859-8', 'ISO-8859-9', | |
| 652 | 'ISO-8859-10', 'ISO-8859-13', 'ISO-8859-14', 'ISO-8859-15', 'ISO-8859-16', | |
| 653 | 'WINDOWS-1252', 'WINDOWS-1251', 'EUC-JP', 'EUC-TW', 'KOI8-R', | |
| 654 | 'ISO-2022-KR', 'ISO-2022-JP' | |
| 655 | ); | |
| 656 | ||
| 657 | $result = mb_detect_encoding($string, join(',', $enc)); | |
| 658 | ||
| 659 | return $result ? $result : $failover; | |
| 660 | } | |
| 661 | ||
| 662 | ||
| 663 | /** | |
| 664 | * Removes non-unicode characters from input. | |
| 665 | * | |
| 666 | * @param mixed $input String or array. | |
| 667 | * | |
| 668 | * @return mixed String or array | |
| 669 | */ | |
| 670 | public static function clean($input) | |
| 671 | { | |
| 672 | // handle input of type array | |
| 673 | if (is_array($input)) { | |
| 674 | foreach ($input as $idx => $val) { | |
| 675 | $input[$idx] = self::clean($val); | |
| 676 | } | |
| 677 | return $input; | |
| 678 | } | |
| 679 | ||
| 680 | if (!is_string($input) || $input == '') { | |
| 681 | return $input; | |
| 682 | } | |
| 683 | ||
| 684 | // iconv/mbstring are much faster (especially with long strings) | |
| 685 | if (function_exists('mb_convert_encoding')) { | |
| 686 | if (($res = mb_convert_encoding($input, 'UTF-8', 'UTF-8')) !== false) { | |
| 687 | return $res; | |
| 688 | } | |
| 689 | } | |
| 690 | ||
| 691 | if (function_exists('iconv')) { | |
| 692 | if (($res = @iconv('UTF-8', 'UTF-8//IGNORE', $input)) !== false) { | |
| 693 | return $res; | |
| 694 | } | |
| 695 | } | |
| 696 | ||
| 697 | $seq = ''; | |
| 698 | $out = ''; | |
| 699 | $regexp = '/^('. | |
| 700 | // '[\x00-\x7F]'. // UTF8-1 | |
| 701 | '|[\xC2-\xDF][\x80-\xBF]'. // UTF8-2 | |
| 702 | '|\xE0[\xA0-\xBF][\x80-\xBF]'. // UTF8-3 | |
| 703 | '|[\xE1-\xEC][\x80-\xBF][\x80-\xBF]'. // UTF8-3 | |
| 704 | '|\xED[\x80-\x9F][\x80-\xBF]'. // UTF8-3 | |
| 705 | '|[\xEE-\xEF][\x80-\xBF][\x80-\xBF]'. // UTF8-3 | |
| 706 | '|\xF0[\x90-\xBF][\x80-\xBF][\x80-\xBF]'. // UTF8-4 | |
| 707 | '|[\xF1-\xF3][\x80-\xBF][\x80-\xBF][\x80-\xBF]'.// UTF8-4 | |
| 708 | '|\xF4[\x80-\x8F][\x80-\xBF][\x80-\xBF]'. // UTF8-4 | |
| 709 | ')$/'; | |
| 710 | ||
| 711 | for ($i = 0, $len = strlen($input); $i < $len; $i++) { | |
| 712 | $chr = $input[$i]; | |
| 713 | $ord = ord($chr); | |
| 714 | ||
| 715 | // 1-byte character | |
| 716 | if ($ord <= 0x7F) { | |
| 717 | if ($seq) { | |
| 718 | $out .= preg_match($regexp, $seq) ? $seq : ''; | |
| 719 | } | |
| 720 | $seq = ''; | |
| 721 | $out .= $chr; | |
| 722 | // first (or second) byte of multibyte sequence | |
| 723 | } | |
| 724 | else if ($ord >= 0xC0) { | |
| 725 | if (strlen($seq) > 1) { | |
| 1495ac | 726 | $out .= preg_match($regexp, $seq) ? $seq : ''; |
| c321a9 | 727 | $seq = ''; |
| T | 728 | } |
| 729 | else if ($seq && ord($seq) < 0xC0) { | |
| 730 | $seq = ''; | |
| 731 | } | |
| 732 | $seq .= $chr; | |
| 733 | // next byte of multibyte sequence | |
| 734 | } | |
| 735 | else if ($seq) { | |
| 736 | $seq .= $chr; | |
| 737 | } | |
| 738 | } | |
| 739 | ||
| 740 | if ($seq) { | |
| 741 | $out .= preg_match($regexp, $seq) ? $seq : ''; | |
| 742 | } | |
| 743 | ||
| 744 | return $out; | |
| 745 | } | |
| 746 | ||
| 747 | } | |
