| commit | author | age | ||
| 1c4f23 | 1 | <?php |
| A | 2 | |
| 3 | /** | |
| 4 | +-----------------------------------------------------------------------+ | |
| 5 | | program/include/rcube_mime.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 | | Licensed under the GNU GPL | | |
| 11 | | | | |
| 12 | | PURPOSE: | | |
| 13 | | MIME message parsing utilities | | |
| 14 | | | | |
| 15 | +-----------------------------------------------------------------------+ | |
| 16 | | Author: Thomas Bruederli <roundcube@gmail.com> | | |
| 17 | | Author: Aleksander Machniak <alec@alec.pl> | | |
| 18 | +-----------------------------------------------------------------------+ | |
| 19 | ||
| 20 | $Id$ | |
| 21 | ||
| 22 | */ | |
| 23 | ||
| 24 | ||
| 25 | /** | |
| 26 | * Class for parsing MIME messages | |
| 27 | * | |
| 28 | * @package Mail | |
| 29 | * @author Thomas Bruederli <roundcube@gmail.com> | |
| 30 | * @author Aleksander Machniak <alec@alec.pl> | |
| 31 | */ | |
| 32 | class rcube_mime | |
| 33 | { | |
| 34 | private static $default_charset = RCMAIL_CHARSET; | |
| 35 | ||
| 36 | ||
| 37 | /** | |
| 38 | * Object constructor. | |
| 39 | */ | |
| 40 | function __construct($default_charset = null) | |
| 41 | { | |
| 42 | if ($default_charset) { | |
| 43 | self::$default_charset = $default_charset; | |
| 44 | } | |
| 45 | else { | |
| 46 | self::$default_charset = rcmail::get_instance()->config->get('default_charset', RCMAIL_CHARSET); | |
| 47 | } | |
| 48 | } | |
| 49 | ||
| 50 | ||
| 51 | /** | |
| 52 | * Split an address list into a structured array list | |
| 53 | * | |
| 54 | * @param string $input Input string | |
| 55 | * @param int $max List only this number of addresses | |
| 56 | * @param boolean $decode Decode address strings | |
| 57 | * @param string $fallback Fallback charset if none specified | |
| 58 | * | |
| 59 | * @return array Indexed list of addresses | |
| 60 | */ | |
| 61 | static function decode_address_list($input, $max = null, $decode = true, $fallback = null) | |
| 62 | { | |
| 63 | $a = self::parse_address_list($input, $decode, $fallback); | |
| 64 | $out = array(); | |
| 65 | $j = 0; | |
| 66 | ||
| 67 | // Special chars as defined by RFC 822 need to in quoted string (or escaped). | |
| 68 | $special_chars = '[\(\)\<\>\\\.\[\]@,;:"]'; | |
| 69 | ||
| 70 | if (!is_array($a)) | |
| 71 | return $out; | |
| 72 | ||
| 73 | foreach ($a as $val) { | |
| 74 | $j++; | |
| 75 | $address = trim($val['address']); | |
| 76 | $name = trim($val['name']); | |
| 77 | ||
| 78 | if ($name && $address && $name != $address) | |
| 79 | $string = sprintf('%s <%s>', preg_match("/$special_chars/", $name) ? '"'.addcslashes($name, '"').'"' : $name, $address); | |
| 80 | else if ($address) | |
| 81 | $string = $address; | |
| 82 | else if ($name) | |
| 83 | $string = $name; | |
| 84 | ||
| 85 | $out[$j] = array( | |
| 86 | 'name' => $name, | |
| 87 | 'mailto' => $address, | |
| 88 | 'string' => $string | |
| 89 | ); | |
| 90 | ||
| 91 | if ($max && $j==$max) | |
| 92 | break; | |
| 93 | } | |
| 94 | ||
| 95 | return $out; | |
| 96 | } | |
| 97 | ||
| 98 | ||
| 99 | /** | |
| 100 | * Decode a message header value | |
| 101 | * | |
| 102 | * @param string $input Header value | |
| 103 | * @param string $fallback Fallback charset if none specified | |
| 104 | * | |
| 105 | * @return string Decoded string | |
| 106 | */ | |
| 107 | public static function decode_header($input, $fallback = null) | |
| 108 | { | |
| 109 | $str = self::decode_mime_string((string)$input, $fallback); | |
| 110 | ||
| 111 | return $str; | |
| 112 | } | |
| 113 | ||
| 114 | ||
| 115 | /** | |
| 116 | * Decode a mime-encoded string to internal charset | |
| 117 | * | |
| 118 | * @param string $input Header value | |
| 119 | * @param string $fallback Fallback charset if none specified | |
| 120 | * | |
| 121 | * @return string Decoded string | |
| 122 | */ | |
| 123 | public static function decode_mime_string($input, $fallback = null) | |
| 124 | { | |
| 125 | $default_charset = !empty($fallback) ? $fallback : self::$default_charset; | |
| 126 | ||
| 127 | // rfc: all line breaks or other characters not found | |
| 128 | // in the Base64 Alphabet must be ignored by decoding software | |
| 129 | // delete all blanks between MIME-lines, differently we can | |
| 130 | // receive unnecessary blanks and broken utf-8 symbols | |
| 131 | $input = preg_replace("/\?=\s+=\?/", '?==?', $input); | |
| 132 | ||
| 133 | // encoded-word regexp | |
| 134 | $re = '/=\?([^?]+)\?([BbQq])\?([^\n]*?)\?=/'; | |
| 135 | ||
| 136 | // Find all RFC2047's encoded words | |
| 137 | if (preg_match_all($re, $input, $matches, PREG_OFFSET_CAPTURE | PREG_SET_ORDER)) { | |
| 138 | // Initialize variables | |
| 139 | $tmp = array(); | |
| 140 | $out = ''; | |
| 141 | $start = 0; | |
| 142 | ||
| 143 | foreach ($matches as $idx => $m) { | |
| 144 | $pos = $m[0][1]; | |
| 145 | $charset = $m[1][0]; | |
| 146 | $encoding = $m[2][0]; | |
| 147 | $text = $m[3][0]; | |
| 148 | $length = strlen($m[0][0]); | |
| 149 | ||
| 150 | // Append everything that is before the text to be decoded | |
| 151 | if ($start != $pos) { | |
| 152 | $substr = substr($input, $start, $pos-$start); | |
| 153 | $out .= rcube_charset_convert($substr, $default_charset); | |
| 154 | $start = $pos; | |
| 155 | } | |
| 156 | $start += $length; | |
| 157 | ||
| 158 | // Per RFC2047, each string part "MUST represent an integral number | |
| 159 | // of characters . A multi-octet character may not be split across | |
| 160 | // adjacent encoded-words." However, some mailers break this, so we | |
| 161 | // try to handle characters spanned across parts anyway by iterating | |
| 162 | // through and aggregating sequential encoded parts with the same | |
| 163 | // character set and encoding, then perform the decoding on the | |
| 164 | // aggregation as a whole. | |
| 165 | ||
| 166 | $tmp[] = $text; | |
| 167 | if ($next_match = $matches[$idx+1]) { | |
| 168 | if ($next_match[0][1] == $start | |
| 169 | && $next_match[1][0] == $charset | |
| 170 | && $next_match[2][0] == $encoding | |
| 171 | ) { | |
| 172 | continue; | |
| 173 | } | |
| 174 | } | |
| 175 | ||
| 176 | $count = count($tmp); | |
| 177 | $text = ''; | |
| 178 | ||
| 179 | // Decode and join encoded-word's chunks | |
| 180 | if ($encoding == 'B' || $encoding == 'b') { | |
| 181 | // base64 must be decoded a segment at a time | |
| 182 | for ($i=0; $i<$count; $i++) | |
| 183 | $text .= base64_decode($tmp[$i]); | |
| 184 | } | |
| 185 | else { //if ($encoding == 'Q' || $encoding == 'q') { | |
| 186 | // quoted printable can be combined and processed at once | |
| 187 | for ($i=0; $i<$count; $i++) | |
| 188 | $text .= $tmp[$i]; | |
| 189 | ||
| 190 | $text = str_replace('_', ' ', $text); | |
| 191 | $text = quoted_printable_decode($text); | |
| 192 | } | |
| 193 | ||
| 194 | $out .= rcube_charset_convert($text, $charset); | |
| 195 | $tmp = array(); | |
| 196 | } | |
| 197 | ||
| 198 | // add the last part of the input string | |
| 199 | if ($start != strlen($input)) { | |
| 200 | $out .= rcube_charset_convert(substr($input, $start), $default_charset); | |
| 201 | } | |
| 202 | ||
| 203 | // return the results | |
| 204 | return $out; | |
| 205 | } | |
| 206 | ||
| 207 | // no encoding information, use fallback | |
| 208 | return rcube_charset_convert($input, $default_charset); | |
| 209 | } | |
| 210 | ||
| 211 | ||
| 212 | /** | |
| 213 | * Decode a mime part | |
| 214 | * | |
| 215 | * @param string $input Input string | |
| 216 | * @param string $encoding Part encoding | |
| 217 | * @return string Decoded string | |
| 218 | */ | |
| 219 | public static function decode($input, $encoding = '7bit') | |
| 220 | { | |
| 221 | switch (strtolower($encoding)) { | |
| 222 | case 'quoted-printable': | |
| 223 | return quoted_printable_decode($input); | |
| 224 | case 'base64': | |
| 225 | return base64_decode($input); | |
| 226 | case 'x-uuencode': | |
| 227 | case 'x-uue': | |
| 228 | case 'uue': | |
| 229 | case 'uuencode': | |
| 230 | return convert_uudecode($input); | |
| 231 | case '7bit': | |
| 232 | default: | |
| 233 | return $input; | |
| 234 | } | |
| 235 | } | |
| 236 | ||
| 237 | ||
| 238 | /** | |
| 239 | * Split RFC822 header string into an associative array | |
| 240 | * @access private | |
| 241 | */ | |
| 242 | public static function parse_headers($headers) | |
| 243 | { | |
| 244 | $a_headers = array(); | |
| 245 | $headers = preg_replace('/\r?\n(\t| )+/', ' ', $headers); | |
| 246 | $lines = explode("\n", $headers); | |
| 247 | $c = count($lines); | |
| 248 | ||
| 249 | for ($i=0; $i<$c; $i++) { | |
| 250 | if ($p = strpos($lines[$i], ': ')) { | |
| 251 | $field = strtolower(substr($lines[$i], 0, $p)); | |
| 252 | $value = trim(substr($lines[$i], $p+1)); | |
| 253 | if (!empty($value)) | |
| 254 | $a_headers[$field] = $value; | |
| 255 | } | |
| 256 | } | |
| 257 | ||
| 258 | return $a_headers; | |
| 259 | } | |
| 260 | ||
| 261 | ||
| 262 | /** | |
| 263 | * @access private | |
| 264 | */ | |
| 265 | private static function parse_address_list($str, $decode = true, $fallback = null) | |
| 266 | { | |
| 267 | // remove any newlines and carriage returns before | |
| 268 | $str = preg_replace('/\r?\n(\s|\t)?/', ' ', $str); | |
| 269 | ||
| 270 | // extract list items, remove comments | |
| 271 | $str = self::explode_header_string(',;', $str, true); | |
| 272 | $result = array(); | |
| 273 | ||
| 274 | // simplified regexp, supporting quoted local part | |
| 275 | $email_rx = '(\S+|("\s*(?:[^"\f\n\r\t\v\b\s]+\s*)+"))@\S+'; | |
| 276 | ||
| 277 | foreach ($str as $key => $val) { | |
| 278 | $name = ''; | |
| 279 | $address = ''; | |
| 280 | $val = trim($val); | |
| 281 | ||
| 282 | if (preg_match('/(.*)<('.$email_rx.')>$/', $val, $m)) { | |
| 283 | $address = $m[2]; | |
| 284 | $name = trim($m[1]); | |
| 285 | } | |
| 286 | else if (preg_match('/^('.$email_rx.')$/', $val, $m)) { | |
| 287 | $address = $m[1]; | |
| 288 | $name = ''; | |
| 289 | } | |
| 290 | else { | |
| 291 | $name = $val; | |
| 292 | } | |
| 293 | ||
| 294 | // dequote and/or decode name | |
| 295 | if ($name) { | |
| 296 | if ($name[0] == '"' && $name[strlen($name)-1] == '"') { | |
| 297 | $name = substr($name, 1, -1); | |
| 298 | $name = stripslashes($name); | |
| 299 | } | |
| 300 | if ($decode) { | |
| 301 | $name = self::decode_header($name, $fallback); | |
| 302 | } | |
| 303 | } | |
| 304 | ||
| 305 | if (!$address && $name) { | |
| 306 | $address = $name; | |
| 307 | } | |
| 308 | ||
| 309 | if ($address) { | |
| 310 | $result[$key] = array('name' => $name, 'address' => $address); | |
| 311 | } | |
| 312 | } | |
| 313 | ||
| 314 | return $result; | |
| 315 | } | |
| 316 | ||
| 317 | ||
| 318 | /** | |
| 319 | * Explodes header (e.g. address-list) string into array of strings | |
| 320 | * using specified separator characters with proper handling | |
| 321 | * of quoted-strings and comments (RFC2822) | |
| 322 | * | |
| 323 | * @param string $separator String containing separator characters | |
| 324 | * @param string $str Header string | |
| 325 | * @param bool $remove_comments Enable to remove comments | |
| 326 | * | |
| 327 | * @return array Header items | |
| 328 | */ | |
| 329 | public static function explode_header_string($separator, $str, $remove_comments = false) | |
| 330 | { | |
| 331 | $length = strlen($str); | |
| 332 | $result = array(); | |
| 333 | $quoted = false; | |
| 334 | $comment = 0; | |
| 335 | $out = ''; | |
| 336 | ||
| 337 | for ($i=0; $i<$length; $i++) { | |
| 338 | // we're inside a quoted string | |
| 339 | if ($quoted) { | |
| 340 | if ($str[$i] == '"') { | |
| 341 | $quoted = false; | |
| 342 | } | |
| 343 | else if ($str[$i] == "\\") { | |
| 344 | if ($comment <= 0) { | |
| 345 | $out .= "\\"; | |
| 346 | } | |
| 347 | $i++; | |
| 348 | } | |
| 349 | } | |
| 350 | // we are inside a comment string | |
| 351 | else if ($comment > 0) { | |
| 352 | if ($str[$i] == ')') { | |
| 353 | $comment--; | |
| 354 | } | |
| 355 | else if ($str[$i] == '(') { | |
| 356 | $comment++; | |
| 357 | } | |
| 358 | else if ($str[$i] == "\\") { | |
| 359 | $i++; | |
| 360 | } | |
| 361 | continue; | |
| 362 | } | |
| 363 | // separator, add to result array | |
| 364 | else if (strpos($separator, $str[$i]) !== false) { | |
| 365 | if ($out) { | |
| 366 | $result[] = $out; | |
| 367 | } | |
| 368 | $out = ''; | |
| 369 | continue; | |
| 370 | } | |
| 371 | // start of quoted string | |
| 372 | else if ($str[$i] == '"') { | |
| 373 | $quoted = true; | |
| 374 | } | |
| 375 | // start of comment | |
| 376 | else if ($remove_comments && $str[$i] == '(') { | |
| 377 | $comment++; | |
| 378 | } | |
| 379 | ||
| 380 | if ($comment <= 0) { | |
| 381 | $out .= $str[$i]; | |
| 382 | } | |
| 383 | } | |
| 384 | ||
| 385 | if ($out && $comment <= 0) { | |
| 386 | $result[] = $out; | |
| 387 | } | |
| 388 | ||
| 389 | return $result; | |
| 390 | } | |
| 391 | ||
| 392 | ||
| 393 | /** | |
| 394 | * Interpret a format=flowed message body according to RFC 2646 | |
| 395 | * | |
| 396 | * @param string $text Raw body formatted as flowed text | |
| 397 | * | |
| 398 | * @return string Interpreted text with unwrapped lines and stuffed space removed | |
| 399 | */ | |
| 400 | public static function unfold_flowed($text) | |
| 401 | { | |
| 402 | $text = preg_split('/\r?\n/', $text); | |
| 403 | $last = -1; | |
| 404 | $q_level = 0; | |
| 405 | ||
| 406 | foreach ($text as $idx => $line) { | |
| 407 | if ($line[0] == '>' && preg_match('/^(>+\s*)/', $line, $regs)) { | |
| 408 | $q = strlen(str_replace(' ', '', $regs[0])); | |
| 409 | $line = substr($line, strlen($regs[0])); | |
| 410 | ||
| 411 | if ($q == $q_level && $line | |
| 412 | && isset($text[$last]) | |
| 413 | && $text[$last][strlen($text[$last])-1] == ' ' | |
| 414 | ) { | |
| 415 | $text[$last] .= $line; | |
| 416 | unset($text[$idx]); | |
| 417 | } | |
| 418 | else { | |
| 419 | $last = $idx; | |
| 420 | } | |
| 421 | } | |
| 422 | else { | |
| 423 | $q = 0; | |
| 424 | if ($line == '-- ') { | |
| 425 | $last = $idx; | |
| 426 | } | |
| 427 | else { | |
| 428 | // remove space-stuffing | |
| 429 | $line = preg_replace('/^\s/', '', $line); | |
| 430 | ||
| 431 | if (isset($text[$last]) && $line | |
| 432 | && $text[$last] != '-- ' | |
| 433 | && $text[$last][strlen($text[$last])-1] == ' ' | |
| 434 | ) { | |
| 435 | $text[$last] .= $line; | |
| 436 | unset($text[$idx]); | |
| 437 | } | |
| 438 | else { | |
| 439 | $text[$idx] = $line; | |
| 440 | $last = $idx; | |
| 441 | } | |
| 442 | } | |
| 443 | } | |
| 444 | $q_level = $q; | |
| 445 | } | |
| 446 | ||
| 447 | return implode("\r\n", $text); | |
| 448 | } | |
| 449 | ||
| 450 | ||
| 451 | /** | |
| 452 | * Wrap the given text to comply with RFC 2646 | |
| 453 | * | |
| 454 | * @param string $text Text to wrap | |
| 455 | * @param int $length Length | |
| 456 | * | |
| 457 | * @return string Wrapped text | |
| 458 | */ | |
| 459 | public static function format_flowed($text, $length = 72) | |
| 460 | { | |
| 461 | $text = preg_split('/\r?\n/', $text); | |
| 462 | ||
| 463 | foreach ($text as $idx => $line) { | |
| 464 | if ($line != '-- ') { | |
| 465 | if ($line[0] == '>' && preg_match('/^(>+)/', $line, $regs)) { | |
| 466 | $prefix = $regs[0]; | |
| 467 | $level = strlen($prefix); | |
| 468 | $line = rtrim(substr($line, $level)); | |
| 469 | $line = $prefix . rc_wordwrap($line, $length - $level - 2, " \r\n$prefix "); | |
| 470 | } | |
| 471 | else if ($line) { | |
| 472 | $line = rc_wordwrap(rtrim($line), $length - 2, " \r\n"); | |
| 473 | // space-stuffing | |
| 474 | $line = preg_replace('/(^|\r\n)(From| |>)/', '\\1 \\2', $line); | |
| 475 | } | |
| 476 | ||
| 477 | $text[$idx] = $line; | |
| 478 | } | |
| 479 | } | |
| 480 | ||
| 481 | return implode("\r\n", $text); | |
| 482 | } | |
| 483 | ||
| 484 | } | |
