| 1 |
<?php |
| 2 |
|
| 3 |
/** |
| 4 |
* +-----------------------------------------------------------------------+ |
| 5 |
* | This file is part of the Roundcube Webmail client | |
| 6 |
* | | |
| 7 |
* | Copyright (C) The Roundcube Dev Team | |
| 8 |
* | Copyright (C) Kolab Systems AG | |
| 9 |
* | | |
| 10 |
* | Licensed under the GNU General Public License version 3 or | |
| 11 |
* | any later version with exceptions for skins & plugins. | |
| 12 |
* | See the README file for a full license statement. | |
| 13 |
* | | |
| 14 |
* | PURPOSE: | |
| 15 |
* | MIME message parsing utilities | |
| 16 |
* +-----------------------------------------------------------------------+ |
| 17 |
* | Author: Thomas Bruederli <[email protected]> | |
| 18 |
* | Author: Aleksander Machniak <[email protected]> | |
| 19 |
* +-----------------------------------------------------------------------+ |
| 20 |
*/ |
| 21 |
|
| 22 |
namespace AcyMailing\Libraries\Imap2\Roundcube; |
| 23 |
|
| 24 |
use AcyMailing\Libraries\Imap2\rcube; |
| 25 |
use AcyMailing\Libraries\Imap2\rcube_mime_decode; |
| 26 |
|
| 27 |
/** |
| 28 |
* Class for parsing MIME messages |
| 29 |
* |
| 30 |
* @package Framework |
| 31 |
* @subpackage Storage |
| 32 |
*/ |
| 33 |
class Mime |
| 34 |
{ |
| 35 |
private static $default_charset; |
| 36 |
|
| 37 |
|
| 38 |
/** |
| 39 |
* Object constructor. |
| 40 |
*/ |
| 41 |
function __construct($default_charset = null) |
| 42 |
{ |
| 43 |
self::$default_charset = $default_charset; |
| 44 |
} |
| 45 |
|
| 46 |
/** |
| 47 |
* Returns message/object character set name |
| 48 |
* |
| 49 |
* @return string Character set name |
| 50 |
*/ |
| 51 |
public static function get_charset() |
| 52 |
{ |
| 53 |
if (self::$default_charset) { |
| 54 |
return self::$default_charset; |
| 55 |
} |
| 56 |
|
| 57 |
if ($charset = IMAP2_CHARSET) { |
| 58 |
return $charset; |
| 59 |
} |
| 60 |
|
| 61 |
return IMAP2_CHARSET; |
| 62 |
} |
| 63 |
|
| 64 |
/** |
| 65 |
* Parse the given raw message source and return a structure |
| 66 |
* of rcube_message_part objects. |
| 67 |
* |
| 68 |
* It makes use of the rcube_mime_decode library |
| 69 |
* |
| 70 |
* @param string $raw_body The message source |
| 71 |
* |
| 72 |
* @return object rcube_message_part The message structure |
| 73 |
*/ |
| 74 |
public static function parse_message($raw_body) |
| 75 |
{ |
| 76 |
$conf = [ |
| 77 |
'include_bodies' => true, |
| 78 |
'decode_bodies' => true, |
| 79 |
'decode_headers' => false, |
| 80 |
'default_charset' => self::get_charset(), |
| 81 |
]; |
| 82 |
|
| 83 |
$mime = new rcube_mime_decode($conf); |
| 84 |
|
| 85 |
return $mime->decode($raw_body); |
| 86 |
} |
| 87 |
|
| 88 |
/** |
| 89 |
* Split an address list into a structured array list |
| 90 |
* |
| 91 |
* @param string|array $input Input string (or list of strings) |
| 92 |
* @param int $max List only this number of addresses |
| 93 |
* @param boolean $decode Decode address strings |
| 94 |
* @param string $fallback Fallback charset if none specified |
| 95 |
* @param boolean $addronly Return flat array with e-mail addresses only |
| 96 |
* |
| 97 |
* @return array Indexed list of addresses |
| 98 |
*/ |
| 99 |
static function decode_address_list($input, $max = null, $decode = true, $fallback = null, $addronly = false) |
| 100 |
{ |
| 101 |
// A common case when the same header is used many times in a mail message |
| 102 |
if (is_array($input)) { |
| 103 |
$input = implode(', ', $input); |
| 104 |
} |
| 105 |
|
| 106 |
$a = self::parse_address_list($input, $decode, $fallback); |
| 107 |
$out = []; |
| 108 |
$j = 0; |
| 109 |
|
| 110 |
// Special chars as defined by RFC 822 need to in quoted string (or escaped). |
| 111 |
$special_chars = '[\(\)\<\>\\\.\[\]@,;:"]'; |
| 112 |
|
| 113 |
if (!is_array($a)) { |
| 114 |
return $out; |
| 115 |
} |
| 116 |
|
| 117 |
foreach ($a as $val) { |
| 118 |
$j++; |
| 119 |
$address = trim($val['address']); |
| 120 |
|
| 121 |
if ($addronly) { |
| 122 |
$out[$j] = $address; |
| 123 |
} else { |
| 124 |
$name = trim($val['name']); |
| 125 |
if ($name && $address && $name != $address) { |
| 126 |
$string = sprintf('%s <%s>', preg_match("/$special_chars/", $name) ? '"'.addcslashes($name, '"').'"' : $name, $address); |
| 127 |
} elseif ($address) { |
| 128 |
$string = $address; |
| 129 |
} elseif ($name) { |
| 130 |
$string = $name; |
| 131 |
} |
| 132 |
|
| 133 |
$out[$j] = ['name' => $name, 'mailto' => $address, 'string' => $string]; |
| 134 |
} |
| 135 |
|
| 136 |
if ($max && $j == $max) { |
| 137 |
break; |
| 138 |
} |
| 139 |
} |
| 140 |
|
| 141 |
return $out; |
| 142 |
} |
| 143 |
|
| 144 |
/** |
| 145 |
* Decode a message header value |
| 146 |
* |
| 147 |
* @param string $input Header value |
| 148 |
* @param string $fallback Fallback charset if none specified |
| 149 |
* |
| 150 |
* @return string Decoded string |
| 151 |
*/ |
| 152 |
public static function decode_header($input, $fallback = null) |
| 153 |
{ |
| 154 |
$str = self::decode_mime_string((string)$input, $fallback); |
| 155 |
|
| 156 |
return $str; |
| 157 |
} |
| 158 |
|
| 159 |
/** |
| 160 |
* Decode a mime-encoded string to internal charset |
| 161 |
* |
| 162 |
* @param string $input Header value |
| 163 |
* @param string $fallback Fallback charset if none specified |
| 164 |
* |
| 165 |
* @return string Decoded string |
| 166 |
*/ |
| 167 |
public static function decode_mime_string($input, $fallback = null) |
| 168 |
{ |
| 169 |
$default_charset = $fallback ? : self::get_charset(); |
| 170 |
|
| 171 |
// rfc: all line breaks or other characters not found |
| 172 |
// in the Base64 Alphabet must be ignored by decoding software |
| 173 |
// delete all blanks between MIME-lines, differently we can |
| 174 |
// receive unnecessary blanks and broken utf-8 symbols |
| 175 |
$input = preg_replace("/\?=\s+=\?/", '?==?', $input); |
| 176 |
|
| 177 |
// encoded-word regexp |
| 178 |
$re = '/=\?([^?]+)\?([BbQq])\?([^\n]*?)\?=/'; |
| 179 |
|
| 180 |
// Find all RFC2047's encoded words |
| 181 |
if (preg_match_all($re, $input, $matches, PREG_OFFSET_CAPTURE | PREG_SET_ORDER)) { |
| 182 |
// Initialize variables |
| 183 |
$tmp = []; |
| 184 |
$out = ''; |
| 185 |
$start = 0; |
| 186 |
|
| 187 |
foreach ($matches as $idx => $m) { |
| 188 |
$pos = $m[0][1]; |
| 189 |
$charset = $m[1][0]; |
| 190 |
$encoding = $m[2][0]; |
| 191 |
$text = $m[3][0]; |
| 192 |
$length = strlen($m[0][0]); |
| 193 |
|
| 194 |
// Append everything that is before the text to be decoded |
| 195 |
if ($start != $pos) { |
| 196 |
$substr = substr($input, $start, $pos - $start); |
| 197 |
$out .= Charset::convert($substr, $default_charset); |
| 198 |
$start = $pos; |
| 199 |
} |
| 200 |
$start += $length; |
| 201 |
|
| 202 |
// Per RFC2047, each string part "MUST represent an integral number |
| 203 |
// of characters . A multi-octet character may not be split across |
| 204 |
// adjacent encoded-words." However, some mailers break this, so we |
| 205 |
// try to handle characters spanned across parts anyway by iterating |
| 206 |
// through and aggregating sequential encoded parts with the same |
| 207 |
// character set and encoding, then perform the decoding on the |
| 208 |
// aggregation as a whole. |
| 209 |
|
| 210 |
$tmp[] = $text; |
| 211 |
if ($next_match = $matches[$idx + 1]) { |
| 212 |
if ($next_match[0][1] == $start |
| 213 |
&& $next_match[1][0] == $charset |
| 214 |
&& $next_match[2][0] == $encoding |
| 215 |
) { |
| 216 |
continue; |
| 217 |
} |
| 218 |
} |
| 219 |
|
| 220 |
$count = count($tmp); |
| 221 |
$text = ''; |
| 222 |
|
| 223 |
// Decode and join encoded-word's chunks |
| 224 |
if ($encoding == 'B' || $encoding == 'b') { |
| 225 |
$rest = ''; |
| 226 |
// base64 must be decoded a segment at a time. |
| 227 |
// However, there are broken implementations that continue |
| 228 |
// in the following word, we'll handle that (#6048) |
| 229 |
for ($i = 0 ; $i < $count ; $i++) { |
| 230 |
$chunk = $rest.$tmp[$i]; |
| 231 |
$length = strlen($chunk); |
| 232 |
if ($length % 4) { |
| 233 |
$length = floor($length / 4) * 4; |
| 234 |
$rest = substr($chunk, $length); |
| 235 |
$chunk = substr($chunk, 0, $length); |
| 236 |
} |
| 237 |
|
| 238 |
$text .= base64_decode($chunk); |
| 239 |
} |
| 240 |
} else { //if ($encoding == 'Q' || $encoding == 'q') { |
| 241 |
// quoted printable can be combined and processed at once |
| 242 |
for ($i = 0 ; $i < $count ; $i++) { |
| 243 |
$text .= $tmp[$i]; |
| 244 |
} |
| 245 |
|
| 246 |
$text = str_replace('_', ' ', $text); |
| 247 |
$text = quoted_printable_decode($text); |
| 248 |
} |
| 249 |
|
| 250 |
$out .= Charset::convert($text, $charset); |
| 251 |
$tmp = []; |
| 252 |
} |
| 253 |
|
| 254 |
// add the last part of the input string |
| 255 |
if ($start != strlen($input)) { |
| 256 |
$out .= Charset::convert(substr($input, $start), $default_charset); |
| 257 |
} |
| 258 |
|
| 259 |
// return the results |
| 260 |
return $out; |
| 261 |
} |
| 262 |
|
| 263 |
// no encoding information, use fallback |
| 264 |
return Charset::convert($input, $default_charset); |
| 265 |
} |
| 266 |
|
| 267 |
/** |
| 268 |
* Decode a mime part |
| 269 |
* |
| 270 |
* @param string $input Input string |
| 271 |
* @param string $encoding Part encoding |
| 272 |
* |
| 273 |
* @return string Decoded string |
| 274 |
*/ |
| 275 |
public static function decode($input, $encoding = '7bit') |
| 276 |
{ |
| 277 |
switch (strtolower($encoding)) { |
| 278 |
case 'quoted-printable': |
| 279 |
return quoted_printable_decode($input); |
| 280 |
case 'base64': |
| 281 |
return base64_decode($input); |
| 282 |
case 'x-uuencode': |
| 283 |
case 'x-uue': |
| 284 |
case 'uue': |
| 285 |
case 'uuencode': |
| 286 |
return convert_uudecode($input); |
| 287 |
case '7bit': |
| 288 |
default: |
| 289 |
return $input; |
| 290 |
} |
| 291 |
} |
| 292 |
|
| 293 |
/** |
| 294 |
* Split RFC822 header string into an associative array |
| 295 |
*/ |
| 296 |
public static function parse_headers($headers) |
| 297 |
{ |
| 298 |
$a_headers = []; |
| 299 |
$headers = preg_replace('/\r?\n(\t| )+/', ' ', $headers); |
| 300 |
$lines = explode("\n", $headers); |
| 301 |
$count = count($lines); |
| 302 |
|
| 303 |
for ($i = 0 ; $i < $count ; $i++) { |
| 304 |
if ($p = strpos($lines[$i], ': ')) { |
| 305 |
$field = strtolower(substr($lines[$i], 0, $p)); |
| 306 |
$value = trim(substr($lines[$i], $p + 1)); |
| 307 |
if (!empty($value)) { |
| 308 |
$a_headers[$field] = $value; |
| 309 |
} |
| 310 |
} |
| 311 |
} |
| 312 |
|
| 313 |
return $a_headers; |
| 314 |
} |
| 315 |
|
| 316 |
/** |
| 317 |
* E-mail address list parser |
| 318 |
*/ |
| 319 |
private static function parse_address_list($str, $decode = true, $fallback = null) |
| 320 |
{ |
| 321 |
// remove any newlines and carriage returns before |
| 322 |
$str = preg_replace('/\r?\n(\s|\t)?/', ' ', $str); |
| 323 |
|
| 324 |
// extract list items, remove comments |
| 325 |
$str = self::explode_header_string(',;', $str, true); |
| 326 |
$result = []; |
| 327 |
|
| 328 |
// simplified regexp, supporting quoted local part |
| 329 |
$email_rx = '(\S+|("\s*(?:[^"\f\n\r\t\v\b\s]+\s*)+"))@\S+'; |
| 330 |
|
| 331 |
foreach ($str as $key => $val) { |
| 332 |
$name = ''; |
| 333 |
$address = ''; |
| 334 |
$val = trim($val); |
| 335 |
|
| 336 |
if (preg_match('/(.*)<('.$email_rx.')>$/', $val, $m)) { |
| 337 |
$address = $m[2]; |
| 338 |
$name = trim($m[1]); |
| 339 |
} elseif (preg_match('/^('.$email_rx.')$/', $val, $m)) { |
| 340 |
$address = $m[1]; |
| 341 |
$name = ''; |
| 342 |
} // special case (#1489092) |
| 343 |
elseif (preg_match('/(\s*<MAILER-DAEMON>)$/', $val, $m)) { |
| 344 |
$address = 'MAILER-DAEMON'; |
| 345 |
$name = substr($val, 0, -strlen($m[1])); |
| 346 |
} elseif (preg_match('/('.$email_rx.')/', $val, $m)) { |
| 347 |
$name = $m[1]; |
| 348 |
} else { |
| 349 |
$name = $val; |
| 350 |
} |
| 351 |
|
| 352 |
// dequote and/or decode name |
| 353 |
if ($name) { |
| 354 |
if ($name[0] == '"' && $name[strlen($name) - 1] == '"') { |
| 355 |
$name = substr($name, 1, -1); |
| 356 |
$name = stripslashes($name); |
| 357 |
} |
| 358 |
if ($decode) { |
| 359 |
$name = self::decode_header($name, $fallback); |
| 360 |
// some clients encode addressee name with quotes around it |
| 361 |
if ($name[0] == '"' && $name[strlen($name) - 1] == '"') { |
| 362 |
$name = substr($name, 1, -1); |
| 363 |
} |
| 364 |
} |
| 365 |
} |
| 366 |
|
| 367 |
if (!$address && $name) { |
| 368 |
$address = $name; |
| 369 |
$name = ''; |
| 370 |
} |
| 371 |
|
| 372 |
if ($address) { |
| 373 |
$address = self::fix_email($address); |
| 374 |
$result[$key] = ['name' => $name, 'address' => $address]; |
| 375 |
} |
| 376 |
} |
| 377 |
|
| 378 |
return $result; |
| 379 |
} |
| 380 |
|
| 381 |
/** |
| 382 |
* Explodes header (e.g. address-list) string into array of strings |
| 383 |
* using specified separator characters with proper handling |
| 384 |
* of quoted-strings and comments (RFC2822) |
| 385 |
* |
| 386 |
* @param string $separator String containing separator characters |
| 387 |
* @param string $str Header string |
| 388 |
* @param bool $remove_comments Enable to remove comments |
| 389 |
* |
| 390 |
* @return array Header items |
| 391 |
*/ |
| 392 |
public static function explode_header_string($separator, $str, $remove_comments = false) |
| 393 |
{ |
| 394 |
$length = strlen($str); |
| 395 |
$result = []; |
| 396 |
$quoted = false; |
| 397 |
$comment = 0; |
| 398 |
$out = ''; |
| 399 |
|
| 400 |
for ($i = 0 ; $i < $length ; $i++) { |
| 401 |
// we're inside a quoted string |
| 402 |
if ($quoted) { |
| 403 |
if ($str[$i] == '"') { |
| 404 |
$quoted = false; |
| 405 |
} elseif ($str[$i] == "\\") { |
| 406 |
if ($comment <= 0) { |
| 407 |
$out .= "\\"; |
| 408 |
} |
| 409 |
$i++; |
| 410 |
} |
| 411 |
} // we are inside a comment string |
| 412 |
elseif ($comment > 0) { |
| 413 |
if ($str[$i] == ')') { |
| 414 |
$comment--; |
| 415 |
} elseif ($str[$i] == '(') { |
| 416 |
$comment++; |
| 417 |
} elseif ($str[$i] == "\\") { |
| 418 |
$i++; |
| 419 |
} |
| 420 |
continue; |
| 421 |
} // separator, add to result array |
| 422 |
elseif (strpos($separator, $str[$i]) !== false) { |
| 423 |
if ($out) { |
| 424 |
$result[] = $out; |
| 425 |
} |
| 426 |
$out = ''; |
| 427 |
continue; |
| 428 |
} // start of quoted string |
| 429 |
elseif ($str[$i] == '"') { |
| 430 |
$quoted = true; |
| 431 |
} // start of comment |
| 432 |
elseif ($remove_comments && $str[$i] == '(') { |
| 433 |
$comment++; |
| 434 |
} |
| 435 |
|
| 436 |
if ($comment <= 0) { |
| 437 |
$out .= $str[$i]; |
| 438 |
} |
| 439 |
} |
| 440 |
|
| 441 |
if ($out && $comment <= 0) { |
| 442 |
$result[] = $out; |
| 443 |
} |
| 444 |
|
| 445 |
return $result; |
| 446 |
} |
| 447 |
|
| 448 |
/** |
| 449 |
* Interpret a format=flowed message body according to RFC 2646 |
| 450 |
* |
| 451 |
* @param string $text Raw body formatted as flowed text |
| 452 |
* @param string $mark Mark each flowed line with specified character |
| 453 |
* @param boolean $delsp Remove the trailing space of each flowed line |
| 454 |
* |
| 455 |
* @return string Interpreted text with unwrapped lines and stuffed space removed |
| 456 |
*/ |
| 457 |
public static function unfold_flowed($text, $mark = null, $delsp = false) |
| 458 |
{ |
| 459 |
$text = preg_split('/\r?\n/', $text); |
| 460 |
$last = -1; |
| 461 |
$q_level = 0; |
| 462 |
$marks = []; |
| 463 |
|
| 464 |
foreach ($text as $idx => $line) { |
| 465 |
if ($q = strspn($line, '>')) { |
| 466 |
// remove quote chars |
| 467 |
$line = substr($line, $q); |
| 468 |
// remove (optional) space-staffing |
| 469 |
if ($line[0] === ' ') $line = substr($line, 1); |
| 470 |
|
| 471 |
// The same paragraph (We join current line with the previous one) when: |
| 472 |
// - the same level of quoting |
| 473 |
// - previous line was flowed |
| 474 |
// - previous line contains more than only one single space (and quote char(s)) |
| 475 |
if ($q == $q_level |
| 476 |
&& isset($text[$last]) && $text[$last][strlen($text[$last]) - 1] == ' ' |
| 477 |
&& !preg_match('/^>+ {0,1}$/', $text[$last]) |
| 478 |
) { |
| 479 |
if ($delsp) { |
| 480 |
$text[$last] = substr($text[$last], 0, -1); |
| 481 |
} |
| 482 |
$text[$last] .= $line; |
| 483 |
unset($text[$idx]); |
| 484 |
|
| 485 |
if ($mark) { |
| 486 |
$marks[$last] = true; |
| 487 |
} |
| 488 |
} else { |
| 489 |
$last = $idx; |
| 490 |
} |
| 491 |
} else { |
| 492 |
if ($line == '-- ') { |
| 493 |
$last = $idx; |
| 494 |
} else { |
| 495 |
// remove space-stuffing |
| 496 |
if ($line[0] === ' ') $line = substr($line, 1); |
| 497 |
|
| 498 |
if (isset($text[$last]) && $line && !$q_level |
| 499 |
&& $text[$last] != '-- ' |
| 500 |
&& $text[$last][strlen($text[$last]) - 1] == ' ' |
| 501 |
) { |
| 502 |
if ($delsp) { |
| 503 |
$text[$last] = substr($text[$last], 0, -1); |
| 504 |
} |
| 505 |
$text[$last] .= $line; |
| 506 |
unset($text[$idx]); |
| 507 |
|
| 508 |
if ($mark) { |
| 509 |
$marks[$last] = true; |
| 510 |
} |
| 511 |
} else { |
| 512 |
$text[$idx] = $line; |
| 513 |
$last = $idx; |
| 514 |
} |
| 515 |
} |
| 516 |
} |
| 517 |
$q_level = $q; |
| 518 |
} |
| 519 |
|
| 520 |
if (!empty($marks)) { |
| 521 |
foreach (array_keys($marks) as $mk) { |
| 522 |
$text[$mk] = $mark.$text[$mk]; |
| 523 |
} |
| 524 |
} |
| 525 |
|
| 526 |
return implode("\r\n", $text); |
| 527 |
} |
| 528 |
|
| 529 |
/** |
| 530 |
* Wrap the given text to comply with RFC 2646 |
| 531 |
* |
| 532 |
* @param string $text Text to wrap |
| 533 |
* @param int $length Length |
| 534 |
* @param string $charset Character encoding of $text |
| 535 |
* |
| 536 |
* @return string Wrapped text |
| 537 |
*/ |
| 538 |
public static function format_flowed($text, $length = 72, $charset = null) |
| 539 |
{ |
| 540 |
$text = preg_split('/\r?\n/', $text); |
| 541 |
|
| 542 |
foreach ($text as $idx => $line) { |
| 543 |
if ($line != '-- ') { |
| 544 |
if ($level = strspn($line, '>')) { |
| 545 |
// remove quote chars |
| 546 |
$line = substr($line, $level); |
| 547 |
// remove (optional) space-staffing and spaces before the line end |
| 548 |
$line = rtrim($line, ' '); |
| 549 |
if ($line[0] === ' ') $line = substr($line, 1); |
| 550 |
|
| 551 |
$prefix = str_repeat('>', $level).' '; |
| 552 |
$line = $prefix.self::wordwrap($line, $length - $level - 2, " \r\n$prefix", false, $charset); |
| 553 |
} elseif ($line) { |
| 554 |
$line = self::wordwrap(rtrim($line), $length - 2, " \r\n", false, $charset); |
| 555 |
// space-stuffing |
| 556 |
$line = preg_replace('/(^|\r\n)(From| |>)/', '\\1 \\2', $line); |
| 557 |
} |
| 558 |
|
| 559 |
$text[$idx] = $line; |
| 560 |
} |
| 561 |
} |
| 562 |
|
| 563 |
return implode("\r\n", $text); |
| 564 |
} |
| 565 |
|
| 566 |
/** |
| 567 |
* Improved wordwrap function with multibyte support. |
| 568 |
* The code is based on Zend_Text_MultiByte::wordWrap(). |
| 569 |
* |
| 570 |
* @param string $string Text to wrap |
| 571 |
* @param int $width Line width |
| 572 |
* @param string $break Line separator |
| 573 |
* @param bool $cut Enable to cut word |
| 574 |
* @param string $charset Charset of $string |
| 575 |
* @param bool $wrap_quoted When enabled quoted lines will not be wrapped |
| 576 |
* |
| 577 |
* @return string Text |
| 578 |
*/ |
| 579 |
public static function wordwrap($string, $width = 75, $break = "\n", $cut = false, $charset = null, $wrap_quoted = true) |
| 580 |
{ |
| 581 |
// Note: Never try to use iconv instead of mbstring functions here |
| 582 |
// Iconv's substr/strlen are 100x slower (#1489113) |
| 583 |
|
| 584 |
if ($charset && $charset != IMAP2_CHARSET) { |
| 585 |
mb_internal_encoding($charset); |
| 586 |
} |
| 587 |
|
| 588 |
// Convert \r\n to \n, this is our line-separator |
| 589 |
$string = str_replace("\r\n", "\n", $string); |
| 590 |
$separator = "\n"; // must be 1 character length |
| 591 |
$result = []; |
| 592 |
|
| 593 |
while (($stringLength = mb_strlen($string)) > 0) { |
| 594 |
$breakPos = mb_strpos($string, $separator, 0); |
| 595 |
|
| 596 |
// quoted line (do not wrap) |
| 597 |
if ($wrap_quoted && $string[0] == '>') { |
| 598 |
if ($breakPos === $stringLength - 1 || $breakPos === false) { |
| 599 |
$subString = $string; |
| 600 |
$cutLength = null; |
| 601 |
} else { |
| 602 |
$subString = mb_substr($string, 0, $breakPos); |
| 603 |
$cutLength = $breakPos + 1; |
| 604 |
} |
| 605 |
} // next line found and current line is shorter than the limit |
| 606 |
elseif ($breakPos !== false && $breakPos < $width) { |
| 607 |
if ($breakPos === $stringLength - 1) { |
| 608 |
$subString = $string; |
| 609 |
$cutLength = null; |
| 610 |
} else { |
| 611 |
$subString = mb_substr($string, 0, $breakPos); |
| 612 |
$cutLength = $breakPos + 1; |
| 613 |
} |
| 614 |
} else { |
| 615 |
$subString = mb_substr($string, 0, $width); |
| 616 |
|
| 617 |
// last line |
| 618 |
if ($breakPos === false && $subString === $string) { |
| 619 |
$cutLength = null; |
| 620 |
} else { |
| 621 |
$nextChar = mb_substr($string, $width, 1); |
| 622 |
|
| 623 |
if ($nextChar === ' ' || $nextChar === $separator) { |
| 624 |
$afterNextChar = mb_substr($string, $width + 1, 1); |
| 625 |
|
| 626 |
// Note: mb_substr() does never return False |
| 627 |
if ($afterNextChar === false || $afterNextChar === '') { |
| 628 |
$subString .= $nextChar; |
| 629 |
} |
| 630 |
|
| 631 |
$cutLength = mb_strlen($subString) + 1; |
| 632 |
} else { |
| 633 |
$spacePos = mb_strrpos($subString, ' ', 0); |
| 634 |
|
| 635 |
if ($spacePos !== false) { |
| 636 |
$subString = mb_substr($subString, 0, $spacePos); |
| 637 |
$cutLength = $spacePos + 1; |
| 638 |
} elseif ($cut === false) { |
| 639 |
$spacePos = mb_strpos($string, ' ', 0); |
| 640 |
|
| 641 |
if ($spacePos !== false && ($breakPos === false || $spacePos < $breakPos)) { |
| 642 |
$subString = mb_substr($string, 0, $spacePos); |
| 643 |
$cutLength = $spacePos + 1; |
| 644 |
} elseif ($breakPos === false) { |
| 645 |
$subString = $string; |
| 646 |
$cutLength = null; |
| 647 |
} else { |
| 648 |
$subString = mb_substr($string, 0, $breakPos); |
| 649 |
$cutLength = $breakPos + 1; |
| 650 |
} |
| 651 |
} else { |
| 652 |
$cutLength = $width; |
| 653 |
} |
| 654 |
} |
| 655 |
} |
| 656 |
} |
| 657 |
|
| 658 |
$result[] = $subString; |
| 659 |
|
| 660 |
if ($cutLength !== null) { |
| 661 |
$string = mb_substr($string, $cutLength, ($stringLength - $cutLength)); |
| 662 |
} else { |
| 663 |
break; |
| 664 |
} |
| 665 |
} |
| 666 |
|
| 667 |
if ($charset && $charset != IMAP2_CHARSET) { |
| 668 |
mb_internal_encoding(IMAP2_CHARSET); |
| 669 |
} |
| 670 |
|
| 671 |
return implode($break, $result); |
| 672 |
} |
| 673 |
|
| 674 |
/** |
| 675 |
* A method to guess the mime_type of an attachment. |
| 676 |
* |
| 677 |
* @param string $path Path to the file or file contents |
| 678 |
* @param string $name File name (with suffix) |
| 679 |
* @param string $failover Mime type supplied for failover |
| 680 |
* @param boolean $is_stream Set to True if $path contains file contents |
| 681 |
* @param boolean $skip_suffix Set to True if the config/mimetypes.php mappig should be ignored |
| 682 |
* |
| 683 |
* @return string |
| 684 |
* @author Till Klampaeckel <[email protected]> |
| 685 |
* @see http://de2.php.net/manual/en/ref.fileinfo.php |
| 686 |
* @see http://de2.php.net/mime_content_type |
| 687 |
*/ |
| 688 |
public static function file_content_type($path, $name, $failover = 'application/octet-stream', $is_stream = false, $skip_suffix = false) |
| 689 |
{ |
| 690 |
static $mime_ext = []; |
| 691 |
|
| 692 |
$mime_type = null; |
| 693 |
$config = rcube::get_instance()->config; |
| 694 |
|
| 695 |
if (!$skip_suffix && empty($mime_ext)) { |
| 696 |
foreach ($config->resolve_paths('mimetypes.php') as $fpath) { |
| 697 |
$mime_ext = array_merge($mime_ext, (array)@include($fpath)); |
| 698 |
} |
| 699 |
} |
| 700 |
|
| 701 |
// use file name suffix with hard-coded mime-type map |
| 702 |
if (!$skip_suffix && is_array($mime_ext) && $name) { |
| 703 |
if ($suffix = substr($name, strrpos($name, '.') + 1)) { |
| 704 |
$mime_type = $mime_ext[strtolower($suffix)]; |
| 705 |
} |
| 706 |
} |
| 707 |
|
| 708 |
// try fileinfo extension if available |
| 709 |
if (!$mime_type && function_exists('finfo_open')) { |
| 710 |
$mime_magic = $config->get('mime_magic'); |
| 711 |
// null as a 2nd argument should be the same as no argument |
| 712 |
// this however is not true on all systems/versions |
| 713 |
if ($mime_magic) { |
| 714 |
$finfo = finfo_open(FILEINFO_MIME, $mime_magic); |
| 715 |
} else { |
| 716 |
$finfo = finfo_open(FILEINFO_MIME); |
| 717 |
} |
| 718 |
|
| 719 |
if ($finfo) { |
| 720 |
$func = $is_stream ? 'finfo_buffer' : 'finfo_file'; |
| 721 |
$mime_type = $func($finfo, $path, FILEINFO_MIME_TYPE); |
| 722 |
finfo_close($finfo); |
| 723 |
} |
| 724 |
} |
| 725 |
|
| 726 |
// try PHP's mime_content_type |
| 727 |
if (!$mime_type && !$is_stream && function_exists('mime_content_type')) { |
| 728 |
$mime_type = @mime_content_type($path); |
| 729 |
} |
| 730 |
|
| 731 |
// fall back to user-submitted string |
| 732 |
if (!$mime_type) { |
| 733 |
$mime_type = $failover; |
| 734 |
} |
| 735 |
|
| 736 |
return $mime_type; |
| 737 |
} |
| 738 |
|
| 739 |
/** |
| 740 |
* Get mimetype => file extension mapping |
| 741 |
* |
| 742 |
* @param string Mime-Type to get extensions for |
| 743 |
* |
| 744 |
* @return array List of extensions matching the given mimetype or a hash array |
| 745 |
* with ext -> mimetype mappings if $mimetype is not given |
| 746 |
*/ |
| 747 |
public static function get_mime_extensions($mimetype = null) |
| 748 |
{ |
| 749 |
static $mime_types, $mime_extensions; |
| 750 |
|
| 751 |
// return cached data |
| 752 |
if (is_array($mime_types)) { |
| 753 |
return $mimetype ? $mime_types[$mimetype] : $mime_extensions; |
| 754 |
} |
| 755 |
|
| 756 |
// load mapping file |
| 757 |
$file_paths = []; |
| 758 |
|
| 759 |
if ($mime_types = rcube::get_instance()->config->get('mime_types')) { |
| 760 |
$file_paths[] = $mime_types; |
| 761 |
} |
| 762 |
|
| 763 |
// try common locations |
| 764 |
if (strtoupper(substr(PHP_OS, 0, 3)) == 'WIN') { |
| 765 |
$file_paths[] = 'C:/xampp/apache/conf/mime.types.'; |
| 766 |
} else { |
| 767 |
$file_paths[] = '/etc/mime.types'; |
| 768 |
$file_paths[] = '/etc/httpd/mime.types'; |
| 769 |
$file_paths[] = '/etc/httpd2/mime.types'; |
| 770 |
$file_paths[] = '/etc/apache/mime.types'; |
| 771 |
$file_paths[] = '/etc/apache2/mime.types'; |
| 772 |
$file_paths[] = '/etc/nginx/mime.types'; |
| 773 |
$file_paths[] = '/usr/local/etc/httpd/conf/mime.types'; |
| 774 |
$file_paths[] = '/usr/local/etc/apache/conf/mime.types'; |
| 775 |
$file_paths[] = '/usr/local/etc/apache24/mime.types'; |
| 776 |
} |
| 777 |
|
| 778 |
foreach ($file_paths as $fp) { |
| 779 |
if (@is_readable($fp)) { |
| 780 |
$lines = file($fp, FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES); |
| 781 |
break; |
| 782 |
} |
| 783 |
} |
| 784 |
|
| 785 |
$mime_types = $mime_extensions = []; |
| 786 |
$regex = "/([\w\+\-\.\/]+)\s+([\w\s]+)/i"; |
| 787 |
foreach ((array)$lines as $line) { |
| 788 |
// skip comments or mime types w/o any extensions |
| 789 |
if ($line[0] == '#' || !preg_match($regex, $line, $matches)) { |
| 790 |
continue; |
| 791 |
} |
| 792 |
|
| 793 |
$mime = $matches[1]; |
| 794 |
foreach (explode(' ', $matches[2]) as $ext) { |
| 795 |
$ext = trim($ext); |
| 796 |
$mime_types[$mime][] = $ext; |
| 797 |
$mime_extensions[$ext] = $mime; |
| 798 |
} |
| 799 |
} |
| 800 |
|
| 801 |
// fallback to some well-known types most important for daily emails |
| 802 |
if (empty($mime_types)) { |
| 803 |
foreach (rcube::get_instance()->config->resolve_paths('mimetypes.php') as $fpath) { |
| 804 |
$mime_extensions = array_merge($mime_extensions, (array)@include($fpath)); |
| 805 |
} |
| 806 |
|
| 807 |
foreach ($mime_extensions as $ext => $mime) { |
| 808 |
$mime_types[$mime][] = $ext; |
| 809 |
} |
| 810 |
} |
| 811 |
|
| 812 |
// Add some known aliases that aren't included by some mime.types (#1488891) |
| 813 |
// the order is important here so standard extensions have higher prio |
| 814 |
$aliases = [ |
| 815 |
'image/gif' => ['gif'], |
| 816 |
'image/png' => ['png'], |
| 817 |
'image/x-png' => ['png'], |
| 818 |
'image/jpeg' => ['jpg', 'jpeg', 'jpe'], |
| 819 |
'image/jpg' => ['jpg', 'jpeg', 'jpe'], |
| 820 |
'image/pjpeg' => ['jpg', 'jpeg', 'jpe'], |
| 821 |
'image/tiff' => ['tif'], |
| 822 |
'image/bmp' => ['bmp'], |
| 823 |
'image/x-ms-bmp' => ['bmp'], |
| 824 |
'message/rfc822' => ['eml'], |
| 825 |
'text/x-mail' => ['eml'], |
| 826 |
]; |
| 827 |
|
| 828 |
foreach ($aliases as $mime => $exts) { |
| 829 |
$mime_types[$mime] = array_unique(array_merge((array)$mime_types[$mime], $exts)); |
| 830 |
|
| 831 |
foreach ($exts as $ext) { |
| 832 |
if (!isset($mime_extensions[$ext])) { |
| 833 |
$mime_extensions[$ext] = $mime; |
| 834 |
} |
| 835 |
} |
| 836 |
} |
| 837 |
|
| 838 |
return $mimetype ? $mime_types[$mimetype] : $mime_extensions; |
| 839 |
} |
| 840 |
|
| 841 |
/** |
| 842 |
* Detect image type of the given binary data by checking magic numbers. |
| 843 |
* |
| 844 |
* @param string $data Binary file content |
| 845 |
* |
| 846 |
* @return string Detected mime-type or jpeg as fallback |
| 847 |
*/ |
| 848 |
public static function image_content_type($data) |
| 849 |
{ |
| 850 |
$type = 'jpeg'; |
| 851 |
if (preg_match('/^\x89\x50\x4E\x47/', $data)) { |
| 852 |
$type = 'png'; |
| 853 |
} elseif (preg_match('/^\x47\x49\x46\x38/', $data)) { |
| 854 |
$type = 'gif'; |
| 855 |
} elseif (preg_match('/^\x00\x00\x01\x00/', $data)) $type = 'ico'; |
| 856 |
|
| 857 |
// else if (preg_match('/^\xFF\xD8\xFF\xE0/', $data)) $type = 'jpeg'; |
| 858 |
|
| 859 |
return 'image/'.$type; |
| 860 |
} |
| 861 |
|
| 862 |
/** |
| 863 |
* Try to fix invalid email addresses |
| 864 |
*/ |
| 865 |
public static function fix_email($email) |
| 866 |
{ |
| 867 |
$parts = Utils::explode_quoted_string('@', $email); |
| 868 |
foreach ($parts as $idx => $part) { |
| 869 |
// remove redundant quoting (#1490040) |
| 870 |
if ($part[0] == '"' && preg_match('/^"([a-zA-Z0-9._+=-]+)"$/', $part, $m)) { |
| 871 |
$parts[$idx] = $m[1]; |
| 872 |
} |
| 873 |
} |
| 874 |
|
| 875 |
return implode('@', $parts); |
| 876 |
} |
| 877 |
} |
| 878 |
|
