Phpdocumentor update - sed is your friend for these kinds of things ;)
[squirrelmail.git] / functions / mime.php
1 <?php
2
3 /**
4 * mime.php
5 *
6 * Copyright (c) 1999-2004 The SquirrelMail Project Team
7 * Licensed under the GNU GPL. For full terms see the file COPYING.
8 *
9 * This contains the functions necessary to detect and decode MIME
10 * messages.
11 *
12 * @version $Id$
13 * @package squirrelmail
14 */
15
16 /** The typical includes... */
17 require_once(SM_PATH . 'functions/imap.php');
18 require_once(SM_PATH . 'functions/attachment_common.php');
19
20 /* -------------------------------------------------------------------------- */
21 /* MIME DECODING */
22 /* -------------------------------------------------------------------------- */
23
24 /**
25 * Get the MIME structure
26 *
27 * This function gets the structure of a message and stores it in the "message" class.
28 * It will return this object for use with all relevant header information and
29 * fully parsed into the standard "message" object format.
30 */
31 function mime_structure ($bodystructure, $flags=array()) {
32
33 /* Isolate the body structure and remove beginning and end parenthesis. */
34 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
35 $read = trim(substr ($read, 0, -1));
36 $i = 0;
37 $msg = Message::parseStructure($read,$i);
38 if (!is_object($msg)) {
39 include_once(SM_PATH . 'functions/display_messages.php');
40 global $color, $mailbox;
41 /* removed urldecode because $_GET is auto urldecoded ??? */
42 displayPageHeader( $color, $mailbox );
43 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
44 $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
45 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
46 plain_error_message( $errormessage, $color );
47 echo '</body></html>';
48 exit;
49 }
50 if (count($flags)) {
51 foreach ($flags as $flag) {
52 $char = strtoupper($flag{1});
53 switch ($char) {
54 case 'S':
55 if (strtolower($flag) == '\\seen') {
56 $msg->is_seen = true;
57 }
58 break;
59 case 'A':
60 if (strtolower($flag) == '\\answered') {
61 $msg->is_answered = true;
62 }
63 break;
64 case 'D':
65 if (strtolower($flag) == '\\deleted') {
66 $msg->is_deleted = true;
67 }
68 break;
69 case 'F':
70 if (strtolower($flag) == '\\flagged') {
71 $msg->is_flagged = true;
72 }
73 break;
74 case 'M':
75 if (strtolower($flag) == '$mdnsent') {
76 $msg->is_mdnsent = true;
77 }
78 break;
79 default:
80 break;
81 }
82 }
83 }
84 // listEntities($msg);
85 return $msg;
86 }
87
88
89
90 /* This starts the parsing of a particular structure. It is called recursively,
91 * so it can be passed different structures. It returns an object of type
92 * $message.
93 * First, it checks to see if it is a multipart message. If it is, then it
94 * handles that as it sees is necessary. If it is just a regular entity,
95 * then it parses it and adds the necessary header information (by calling out
96 * to mime_get_elements()
97 */
98
99 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
100 /* Do a bit of error correction. If we couldn't find the entity id, just guess
101 * that it is the first one. That is usually the case anyway.
102 */
103
104 if (!$ent_id) {
105 $cmd = "FETCH $id BODY[]";
106 } else {
107 $cmd = "FETCH $id BODY[$ent_id]";
108 }
109
110 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
111
112 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
113 do {
114 $topline = trim(array_shift($data));
115 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
116
117 $wholemessage = implode('', $data);
118 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
119 $ret = substr($wholemessage, 0, $regs[1]);
120 /* There is some information in the content info header that could be important
121 * in order to parse html messages. Let's get them here.
122 */
123 // if ($ret{0} == '<') {
124 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
125 // }
126 } else if (ereg('"([^"]*)"', $topline, $regs)) {
127 $ret = $regs[1];
128 } else {
129 global $where, $what, $mailbox, $passed_id, $startMessage;
130 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
131 if (isset($where) && isset($what)) {
132 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
133 } else {
134 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
135 }
136 $par .= '&amp;response=' . urlencode($response) .
137 '&amp;message=' . urlencode($message) .
138 '&amp;topline=' . urlencode($topline);
139
140 echo '<tt><br>' .
141 '<table width="80%"><tr>' .
142 '<tr><td colspan=2>' .
143 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
144 '</td></tr>' .
145 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
146 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
147 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
148 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
149 "</table><BR></tt></font><hr>";
150
151 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
152 array_shift($data);
153 $wholemessage = implode('', $data);
154
155 $ret = $wholemessage;
156 }
157 return $ret;
158 }
159
160 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
161
162 /* Don't kill the connection if the browser is over a dialup
163 * and it would take over 30 seconds to download it.
164 * Dont call set_time_limit in safe mode.
165 */
166
167 if (!ini_get('safe_mode')) {
168 set_time_limit(0);
169 }
170 /* in case of base64 encoded attachments, do not buffer them.
171 Instead, echo the decoded attachment directly to screen */
172 if (strtolower($encoding) == 'base64') {
173 if (!$ent_id) {
174 $query = "FETCH $id BODY[]";
175 } else {
176 $query = "FETCH $id BODY[$ent_id]";
177 }
178 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode','php://stdout',true);
179 } else {
180 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
181 echo decodeBody($body, $encoding);
182 }
183
184 /*
185 TODO, use the same method for quoted printable.
186 However, I assume that quoted printable attachments aren't that large
187 so the performancegain / memory usage drop will be minimal.
188 If we decide to add that then we need to adapt sqimap_fread because
189 we need to split te result on \n and fread doesn't stop at \n. That
190 means we also should provide $results from sqimap_fread (by ref) to
191 te function and set $no_return to false. The $filter function for
192 quoted printable should handle unsetting of $results.
193 */
194 /*
195 TODO 2: find out how we write to the output stream php://stdout. fwrite
196 doesn't work because 'php://stdout isn't a stream.
197 */
198
199 return;
200 /*
201 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
202 $cnt = 0;
203 $continue = true;
204 $read = fgets ($imap_stream,8192);
205
206
207 // This could be bad -- if the section has sqimap_session_id() . ' OK'
208 // or similar, it will kill the download.
209 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
210 if (trim($read) == ')==') {
211 $read1 = $read;
212 $read = fgets ($imap_stream,4096);
213 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
214 return;
215 } else {
216 echo decodeBody($read1, $encoding) .
217 decodeBody($read, $encoding);
218 }
219 } else if ($cnt) {
220 echo decodeBody($read, $encoding);
221 }
222 $read = fgets ($imap_stream,4096);
223 $cnt++;
224 // break;
225 }
226 */
227 }
228
229 /* -[ END MIME DECODING ]----------------------------------------------------------- */
230
231 /* This is here for debugging purposes. It will print out a list
232 * of all the entity IDs that are in the $message object.
233 */
234 function listEntities ($message) {
235 if ($message) {
236 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
237 for ($i = 0; isset($message->entities[$i]); $i++) {
238 echo "$i : ";
239 $msg = listEntities($message->entities[$i]);
240
241 if ($msg) {
242 echo "return: ";
243 return $msg;
244 }
245 }
246 }
247 }
248
249 function getPriorityStr($priority) {
250 $priority_level = substr($priority,0,1);
251
252 switch($priority_level) {
253 /* Check for a higher then normal priority. */
254 case '1':
255 case '2':
256 $priority_string = _("High");
257 break;
258
259 /* Check for a lower then normal priority. */
260 case '4':
261 case '5':
262 $priority_string = _("Low");
263 break;
264
265 /* Check for a normal priority. */
266 case '3':
267 default:
268 $priority_level = '3';
269 $priority_string = _("Normal");
270 break;
271
272 }
273 return $priority_string;
274 }
275
276 /* returns a $message object for a particular entity id */
277 function getEntity ($message, $ent_id) {
278 return $message->getEntity($ent_id);
279 }
280
281 /* translateText
282 * Extracted from strings.php 23/03/2002
283 */
284
285 function translateText(&$body, $wrap_at, $charset) {
286 global $where, $what; /* from searching */
287 global $color; /* color theme */
288
289 require_once(SM_PATH . 'functions/url_parser.php');
290
291 $body_ary = explode("\n", $body);
292 for ($i=0; $i < count($body_ary); $i++) {
293 $line = $body_ary[$i];
294 if (strlen($line) - 2 >= $wrap_at) {
295 sqWordWrap($line, $wrap_at);
296 }
297 $line = charset_decode($charset, $line);
298 $line = str_replace("\t", ' ', $line);
299
300 parseUrl ($line);
301
302 $quotes = 0;
303 $pos = 0;
304 $j = strlen($line);
305
306 while ($pos < $j) {
307 if ($line[$pos] == ' ') {
308 $pos++;
309 } else if (strpos($line, '&gt;', $pos) === $pos) {
310 $pos += 4;
311 $quotes++;
312 } else {
313 break;
314 }
315 }
316
317 if ($quotes % 2) {
318 if (!isset($color[13])) {
319 $color[13] = '#800000';
320 }
321 $line = '<font color="' . $color[13] . '">' . $line . '</font>';
322 } elseif ($quotes) {
323 if (!isset($color[14])) {
324 $color[14] = '#FF0000';
325 }
326 $line = '<font color="' . $color[14] . '">' . $line . '</font>';
327 }
328
329 $body_ary[$i] = $line;
330 }
331 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
332 }
333
334 /* This returns a parsed string called $body. That string can then
335 * be displayed as the actual message in the HTML. It contains
336 * everything needed, including HTML Tags, Attachments at the
337 * bottom, etc.
338 */
339 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
340 /* This if statement checks for the entity to show as the
341 * primary message. To add more of them, just put them in the
342 * order that is their priority.
343 */
344 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
345 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
346 global $languages, $squirrelmail_language;
347
348 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
349 $view_unsafe_images = false;
350 }
351
352 $body = '';
353 $urlmailbox = urlencode($mailbox);
354 $body_message = getEntity($message, $ent_num);
355 if (($body_message->header->type0 == 'text') ||
356 ($body_message->header->type0 == 'rfc822')) {
357 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
358 $body = decodeBody($body, $body_message->header->encoding);
359
360 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
361 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
362 if (mb_detect_encoding($body) != 'ASCII') {
363 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
364 }
365 }
366 $hookResults = do_hook("message_body", $body);
367 $body = $hookResults[1];
368
369 /* If there are other types that shouldn't be formatted, add
370 * them here.
371 */
372
373 if ($body_message->header->type1 == 'html') {
374 if ($show_html_default <> 1) {
375 $entity_conv = array('&nbsp;' => ' ',
376 '<p>' => "\n",
377 '<P>' => "\n",
378 '<br>' => "\n",
379 '<BR>' => "\n",
380 '<br />' => "\n",
381 '<BR />' => "\n",
382 '&gt;' => '>',
383 '&lt;' => '<');
384 $body = strtr($body, $entity_conv);
385 $body = strip_tags($body);
386 $body = trim($body);
387 translateText($body, $wrap_at,
388 $body_message->header->getParameter('charset'));
389 } else {
390 $body = magicHTML($body, $id, $message, $mailbox);
391 }
392 } else {
393 translateText($body, $wrap_at,
394 $body_message->header->getParameter('charset'));
395 }
396 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
397 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
398 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
399 if (isset($passed_ent_id)) {
400 $link .= '&amp;passed_ent_id='.$passed_ent_id;
401 }
402 $body .= '<center><small><a href="download.php?absolute_dl=true&amp;' .
403 $link . '">' . _("Download this as a file") . '</a>';
404 if ($view_unsafe_images) {
405 $text = _("Hide Unsafe Images");
406 } else {
407 if (isset($has_unsafe_images) && $has_unsafe_images) {
408 $link .= '&amp;view_unsafe_images=1';
409 $text = _("View Unsafe Images");
410 } else {
411 $text = '';
412 }
413 }
414 if($text != '') {
415 $body .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
416 }
417 $body .= '</small></center><br>' . "\n";
418 }
419 return $body;
420 }
421
422
423 function formatAttachments($message, $exclude_id, $mailbox, $id) {
424 global $where, $what, $startMessage, $color;
425 static $ShownHTML = 0;
426
427 $att_ar = $message->getAttachments($exclude_id);
428
429 if (!count($att_ar)) return '';
430
431 $attachments = '';
432
433 $urlMailbox = urlencode($mailbox);
434
435 foreach ($att_ar as $att) {
436 $ent = $att->entity_id;
437 $header = $att->header;
438 $type0 = strtolower($header->type0);
439 $type1 = strtolower($header->type1);
440 $name = '';
441 $links['download link']['text'] = _("download");
442 $links['download link']['href'] = SM_PATH .
443 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
444 $ImageURL = '';
445 if ($type0 =='message' && $type1 == 'rfc822') {
446 $default_page = SM_PATH . 'src/read_body.php';
447 $rfc822_header = $att->rfc822_header;
448 $filename = $rfc822_header->subject;
449 if (trim( $filename ) == '') {
450 $filename = 'untitled-[' . $ent . ']' ;
451 }
452 $from_o = $rfc822_header->from;
453 if (is_object($from_o)) {
454 $from_name = decodeHeader($from_o->getAddress(false));
455 } else {
456 $from_name = _("Unknown sender");
457 }
458 $description = $from_name;
459 } else {
460 $default_page = SM_PATH . 'src/download.php';
461 if (is_object($header->disposition)) {
462 $filename = $header->disposition->getProperty('filename');
463 if (trim($filename) == '') {
464 $name = decodeHeader($header->disposition->getProperty('name'));
465 if (trim($name) == '') {
466 $name = $header->getParameter('name');
467 if(trim($name) == '') {
468 if (trim( $header->id ) == '') {
469 $filename = 'untitled-[' . $ent . ']' ;
470 } else {
471 $filename = 'cid: ' . $header->id;
472 }
473 } else {
474 $filename = $name;
475 }
476 } else {
477 $filename = $name;
478 }
479 }
480 } else {
481 $filename = $header->getParameter('name');
482 if (!trim($filename)) {
483 if (trim( $header->id ) == '') {
484 $filename = 'untitled-[' . $ent . ']' ;
485 } else {
486 $filename = 'cid: ' . $header->id;
487 }
488 }
489 }
490 if ($header->description) {
491 $description = decodeHeader($header->description);
492 } else {
493 $description = '';
494 }
495 }
496
497 $display_filename = $filename;
498 if (isset($passed_ent_id)) {
499 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
500 } else {
501 $passed_ent_id_link = '';
502 }
503 $defaultlink = $default_page . "?startMessage=$startMessage"
504 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
505 . '&amp;ent_id='.$ent.$passed_ent_id_link;
506 if ($where && $what) {
507 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
508 }
509 /* This executes the attachment hook with a specific MIME-type.
510 * If that doesn't have results, it tries if there's a rule
511 * for a more generic type.
512 */
513 $hookresults = do_hook("attachment $type0/$type1", $links,
514 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
515 $display_filename, $where, $what);
516 if(count($hookresults[1]) <= 1) {
517 $hookresults = do_hook("attachment $type0/*", $links,
518 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
519 $display_filename, $where, $what);
520 }
521
522 $links = $hookresults[1];
523 $defaultlink = $hookresults[6];
524
525 $attachments .= '<TR><TD>' .
526 '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
527 '<TD><SMALL><b>' . show_readable_size($header->size) .
528 '</b>&nbsp;&nbsp;</small></TD>' .
529 "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
530 '<TD><SMALL>';
531 $attachments .= '<b>' . $description . '</b>';
532 $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
533
534 $skipspaces = 1;
535 foreach ($links as $val) {
536 if ($skipspaces) {
537 $skipspaces = 0;
538 } else {
539 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
540 }
541 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
542 }
543 unset($links);
544 $attachments .= "</TD></TR>\n";
545 }
546 $attachmentadd = do_hook_function('attachments_bottom',$attachments);
547 if ($attachmentadd != '')
548 $attachments = $attachmentadd;
549 return $attachments;
550 }
551
552 function sqimap_base64_decode(&$string) {
553
554 // base64 enoded data goes in pairs of 4 bytes. To achieve on the
555 // fly decoding (to reduce memory usage) you have to check if the
556 // data has incomplete pairs
557
558 // remove the noise in order to check if the 4 bytes pairs are complete
559 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
560
561 $sStringRem = '';
562 $iMod = strlen($string) % 4;
563 if ($iMod) {
564 $sStringRem = substr($string,-$iMod);
565 // check if $sStringRem contains padding characters
566 if (substr($sStringRem,-1) != '=') {
567 $string = substr($string,0,-$iMod);
568 } else {
569 $sStringRem = '';
570 }
571 }
572 $string = base64_decode($string);
573 return $sStringRem;
574 }
575
576
577 /* This function decodes the body depending on the encoding type. */
578 function decodeBody($body, $encoding) {
579 global $show_html_default;
580
581 $body = str_replace("\r\n", "\n", $body);
582 $encoding = strtolower($encoding);
583
584 $encoding_handler = do_hook_function('decode_body', $encoding);
585
586
587 // plugins get first shot at decoding the body
588 //
589 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
590 $body = $encoding_handler('decode', $body);
591
592 } else if ($encoding == 'quoted-printable' ||
593 $encoding == 'quoted_printable') {
594 $body = quoted_printable_decode($body);
595
596 while (ereg("=\n", $body)) {
597 $body = ereg_replace ("=\n", '', $body);
598 }
599
600 } else if ($encoding == 'base64') {
601 $body = base64_decode($body);
602 }
603
604 // All other encodings are returned raw.
605 return $body;
606 }
607
608 /**
609 * Decodes headers
610 *
611 * This functions decode strings that is encoded according to
612 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
613 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
614 *
615 * @param string $string header string that has to be made readable
616 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
617 * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
618 * @param boolean $decide decide if string can be utfencoded. defaults to false
619 * @return string decoded header string
620 */
621 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
622 global $languages, $squirrelmail_language,$default_charset;
623 if (is_array($string)) {
624 $string = implode("\n", $string);
625 }
626
627 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
628 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
629 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
630 // Do we need to return at this point?
631 // return $string;
632 }
633 $i = 0;
634 $iLastMatch = -2;
635 $encoded = true;
636
637 $aString = explode(' ',$string);
638 $ret = '';
639 foreach ($aString as $chunk) {
640 if ($encoded && $chunk === '') {
641 continue;
642 } elseif ($chunk === '') {
643 $ret .= ' ';
644 continue;
645 }
646 $encoded = false;
647 /* if encoded words are not separated by a linear-space-white we still catch them */
648 $j = $i-1;
649
650 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
651 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
652 if ($iLastMatch !== $j) {
653 if ($htmlsave) {
654 $ret .= '&#32;';
655 } else {
656 $ret .= ' ';
657 }
658 }
659 $iLastMatch = $i;
660 $j = $i;
661 $ret .= $res[1];
662 $encoding = ucfirst($res[3]);
663
664 /* decide about valid decoding */
665 if ($decide && is_conversion_safe($res[2])) {
666 $utfencode=true;
667 $can_be_encoded=true;
668 } else {
669 $can_be_encoded=false;
670 }
671 switch ($encoding)
672 {
673 case 'B':
674 $replace = base64_decode($res[4]);
675 if ($can_be_encoded) {
676 /* convert string to different charset,
677 * if functions asks for it (usually in compose)
678 */
679 $ret .= charset_convert($res[2],$replace,$default_charset);
680 } else {
681 // convert string to html codes in order to display it
682 $ret .= charset_decode($res[2],$replace);
683 }
684 break;
685 case 'Q':
686 $replace = str_replace('_', ' ', $res[4]);
687 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
688 $replace);
689 if ($utfencode) {
690 if ($can_be_encoded) {
691 /* convert string to different charset,
692 * if functions asks for it (usually in compose)
693 */
694 $replace = charset_convert($res[2], $replace,$default_charset);
695 } else {
696 // convert string to html codes in order to display it
697 $replace = charset_decode($res[2], $replace);
698 }
699 } else {
700 if ($htmlsave) {
701 $replace = htmlspecialchars($replace);
702 }
703 }
704 $ret .= $replace;
705 break;
706 default:
707 break;
708 }
709 $chunk = $res[5];
710 $encoded = true;
711 }
712 if (!$encoded) {
713 if ($htmlsave) {
714 $ret .= '&#32;';
715 } else {
716 $ret .= ' ';
717 }
718 }
719
720 if (!$encoded && $htmlsave) {
721 $ret .= htmlspecialchars($chunk);
722 } else {
723 $ret .= $chunk;
724 }
725 ++$i;
726 }
727 /* remove the first added space */
728 if ($ret) {
729 if ($htmlsave) {
730 $ret = substr($ret,5);
731 } else {
732 $ret = substr($ret,1);
733 }
734 }
735
736 return $ret;
737 }
738
739 /**
740 * Encodes header as quoted-printable
741 *
742 * Encode a string according to RFC 1522 for use in headers if it
743 * contains 8-bit characters or anything that looks like it should
744 * be encoded.
745 *
746 * @param string $string header string, that has to be encoded
747 * @return string quoted-printable encoded string
748 */
749 function encodeHeader ($string) {
750 global $default_charset, $languages, $squirrelmail_language;
751
752 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
753 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
754 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
755 }
756
757 // Encode only if the string contains 8-bit characters or =?
758 $j = strlen($string);
759 $max_l = 75 - strlen($default_charset) - 7;
760 $aRet = array();
761 $ret = '';
762 $iEncStart = $enc_init = false;
763 $cur_l = $iOffset = 0;
764 for($i = 0; $i < $j; ++$i) {
765 switch($string{$i})
766 {
767 case '=':
768 case '<':
769 case '>':
770 case ',':
771 case '?':
772 case '_':
773 if ($iEncStart === false) {
774 $iEncStart = $i;
775 }
776 $cur_l+=3;
777 if ($cur_l > ($max_l-2)) {
778 /* if there is an stringpart that doesn't need encoding, add it */
779 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
780 $aRet[] = "=?$default_charset?Q?$ret?=";
781 $iOffset = $i;
782 $cur_l = 0;
783 $ret = '';
784 $iEncStart = false;
785 } else {
786 $ret .= sprintf("=%02X",ord($string{$i}));
787 }
788 break;
789 case '(':
790 case ')':
791 if ($iEncStart !== false) {
792 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
793 $aRet[] = "=?$default_charset?Q?$ret?=";
794 $iOffset = $i;
795 $cur_l = 0;
796 $ret = '';
797 $iEncStart = false;
798 }
799 break;
800 case ' ':
801 if ($iEncStart !== false) {
802 $cur_l++;
803 if ($cur_l > $max_l) {
804 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
805 $aRet[] = "=?$default_charset?Q?$ret?=";
806 $iOffset = $i;
807 $cur_l = 0;
808 $ret = '';
809 $iEncStart = false;
810 } else {
811 $ret .= '_';
812 }
813 }
814 break;
815 default:
816 $k = ord($string{$i});
817 if ($k > 126) {
818 if ($iEncStart === false) {
819 // do not start encoding in the middle of a string, also take the rest of the word.
820 $sLeadString = substr($string,0,$i);
821 $aLeadString = explode(' ',$sLeadString);
822 $sToBeEncoded = array_pop($aLeadString);
823 $iEncStart = $i - strlen($sToBeEncoded);
824 $ret .= $sToBeEncoded;
825 $cur_l += strlen($sToBeEncoded);
826 }
827 $cur_l += 3;
828 /* first we add the encoded string that reached it's max size */
829 if ($cur_l > ($max_l-2)) {
830 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
831 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
832 $cur_l = 3;
833 $ret = '';
834 $iOffset = $i;
835 $iEncStart = $i;
836 }
837 $enc_init = true;
838 $ret .= sprintf("=%02X", $k);
839 } else {
840 if ($iEncStart !== false) {
841 $cur_l++;
842 if ($cur_l > $max_l) {
843 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
844 $aRet[] = "=?$default_charset?Q?$ret?=";
845 $iEncStart = false;
846 $iOffset = $i;
847 $cur_l = 0;
848 $ret = '';
849 } else {
850 $ret .= $string{$i};
851 }
852 }
853 }
854 break;
855 }
856 }
857
858 if ($enc_init) {
859 if ($iEncStart !== false) {
860 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
861 $aRet[] = "=?$default_charset?Q?$ret?=";
862 } else {
863 $aRet[] = substr($string,$iOffset);
864 }
865 $string = implode('',$aRet);
866 }
867 return $string;
868 }
869
870 /* This function trys to locate the entity_id of a specific mime element */
871 function find_ent_id($id, $message) {
872 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
873 if ($message->entities[$i]->header->type0 == 'multipart') {
874 $ret = find_ent_id($id, $message->entities[$i]);
875 } else {
876 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
877 // if (sq_check_save_extension($message->entities[$i])) {
878 return $message->entities[$i]->entity_id;
879 // }
880 }
881 }
882 }
883 return $ret;
884 }
885
886 function sq_check_save_extension($message) {
887 $filename = $message->getFilename();
888 $ext = substr($filename, strrpos($filename,'.')+1);
889 $save_extensions = array('jpg','jpeg','gif','png','bmp');
890 return in_array($ext, $save_extensions);
891 }
892
893
894 /**
895 ** HTMLFILTER ROUTINES
896 */
897
898 /**
899 * This function is more or less a wrapper around stripslashes. Apparently
900 * Explorer is stupid enough to just remove the backslashes and then
901 * execute the content of the attribute as if nothing happened.
902 * Who does that?
903 *
904 * @param attvalue The value of the attribute
905 * @return attvalue The value of the attribute stripslashed.
906 */
907 function sq_unbackslash($attvalue){
908 /**
909 * Remove any backslashes. See if there are any first.
910 */
911
912 if (strstr($attvalue, '\\') !== false){
913 $attvalue = stripslashes($attvalue);
914 }
915 return $attvalue;
916 }
917
918 /**
919 * Kill any tabs, newlines, or carriage returns. Our friends the
920 * makers of the browser with 95% market value decided that it'd
921 * be funny to make "java[tab]script" be just as good as "javascript".
922 *
923 * @param attvalue The attribute value before extraneous spaces removed.
924 * @return attvalue The attribute value after extraneous spaces removed.
925 */
926 function sq_unspace($attvalue){
927 if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
928 $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
929 $attvalue);
930 }
931 return $attvalue;
932 }
933
934 /**
935 * This function returns the final tag out of the tag name, an array
936 * of attributes, and the type of the tag. This function is called by
937 * sq_sanitize internally.
938 *
939 * @param $tagname the name of the tag.
940 * @param $attary the array of attributes and their values
941 * @param $tagtype The type of the tag (see in comments).
942 * @return a string with the final tag representation.
943 */
944 function sq_tagprint($tagname, $attary, $tagtype){
945 $me = 'sq_tagprint';
946
947 if ($tagtype == 2){
948 $fulltag = '</' . $tagname . '>';
949 } else {
950 $fulltag = '<' . $tagname;
951 if (is_array($attary) && sizeof($attary)){
952 $atts = Array();
953 while (list($attname, $attvalue) = each($attary)){
954 array_push($atts, "$attname=$attvalue");
955 }
956 $fulltag .= ' ' . join(" ", $atts);
957 }
958 if ($tagtype == 3){
959 $fulltag .= ' /';
960 }
961 $fulltag .= '>';
962 }
963 return $fulltag;
964 }
965
966 /**
967 * A small helper function to use with array_walk. Modifies a by-ref
968 * value and makes it lowercase.
969 *
970 * @param $val a value passed by-ref.
971 * @return void since it modifies a by-ref value.
972 */
973 function sq_casenormalize(&$val){
974 $val = strtolower($val);
975 }
976
977 /**
978 * This function skips any whitespace from the current position within
979 * a string and to the next non-whitespace value.
980 *
981 * @param $body the string
982 * @param $offset the offset within the string where we should start
983 * looking for the next non-whitespace character.
984 * @return the location within the $body where the next
985 * non-whitespace char is located.
986 */
987 function sq_skipspace($body, $offset){
988 $me = 'sq_skipspace';
989 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
990 if (sizeof($matches{1})){
991 $count = strlen($matches{1});
992 $offset += $count;
993 }
994 return $offset;
995 }
996
997 /**
998 * This function looks for the next character within a string. It's
999 * really just a glorified "strpos", except it catches if failures
1000 * nicely.
1001 *
1002 * @param $body The string to look for needle in.
1003 * @param $offset Start looking from this position.
1004 * @param $needle The character/string to look for.
1005 * @return location of the next occurance of the needle, or
1006 * strlen($body) if needle wasn't found.
1007 */
1008 function sq_findnxstr($body, $offset, $needle){
1009 $me = 'sq_findnxstr';
1010 $pos = strpos($body, $needle, $offset);
1011 if ($pos === FALSE){
1012 $pos = strlen($body);
1013 }
1014 return $pos;
1015 }
1016
1017 /**
1018 * This function takes a PCRE-style regexp and tries to match it
1019 * within the string.
1020 *
1021 * @param $body The string to look for needle in.
1022 * @param $offset Start looking from here.
1023 * @param $reg A PCRE-style regex to match.
1024 * @return Returns a false if no matches found, or an array
1025 * with the following members:
1026 * - integer with the location of the match within $body
1027 * - string with whatever content between offset and the match
1028 * - string with whatever it is we matched
1029 */
1030 function sq_findnxreg($body, $offset, $reg){
1031 $me = 'sq_findnxreg';
1032 $matches = Array();
1033 $retarr = Array();
1034 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1035 if (!isset($matches{0}) || !$matches{0}){
1036 $retarr = false;
1037 } else {
1038 $retarr{0} = $offset + strlen($matches{1});
1039 $retarr{1} = $matches{1};
1040 $retarr{2} = $matches{2};
1041 }
1042 return $retarr;
1043 }
1044
1045 /**
1046 * This function looks for the next tag.
1047 *
1048 * @param $body String where to look for the next tag.
1049 * @param $offset Start looking from here.
1050 * @return false if no more tags exist in the body, or
1051 * an array with the following members:
1052 * - string with the name of the tag
1053 * - array with attributes and their values
1054 * - integer with tag type (1, 2, or 3)
1055 * - integer where the tag starts (starting "<")
1056 * - integer where the tag ends (ending ">")
1057 * first three members will be false, if the tag is invalid.
1058 */
1059 function sq_getnxtag($body, $offset){
1060 $me = 'sq_getnxtag';
1061 if ($offset > strlen($body)){
1062 return false;
1063 }
1064 $lt = sq_findnxstr($body, $offset, "<");
1065 if ($lt == strlen($body)){
1066 return false;
1067 }
1068 /**
1069 * We are here:
1070 * blah blah <tag attribute="value">
1071 * \---------^
1072 */
1073 $pos = sq_skipspace($body, $lt+1);
1074 if ($pos >= strlen($body)){
1075 return Array(false, false, false, $lt, strlen($body));
1076 }
1077 /**
1078 * There are 3 kinds of tags:
1079 * 1. Opening tag, e.g.:
1080 * <a href="blah">
1081 * 2. Closing tag, e.g.:
1082 * </a>
1083 * 3. XHTML-style content-less tag, e.g.:
1084 * <img src="blah"/>
1085 */
1086 $tagtype = false;
1087 switch (substr($body, $pos, 1)){
1088 case '/':
1089 $tagtype = 2;
1090 $pos++;
1091 break;
1092 case '!':
1093 /**
1094 * A comment or an SGML declaration.
1095 */
1096 if (substr($body, $pos+1, 2) == "--"){
1097 $gt = strpos($body, "-->", $pos);
1098 if ($gt === false){
1099 $gt = strlen($body);
1100 } else {
1101 $gt += 2;
1102 }
1103 return Array(false, false, false, $lt, $gt);
1104 } else {
1105 $gt = sq_findnxstr($body, $pos, ">");
1106 return Array(false, false, false, $lt, $gt);
1107 }
1108 break;
1109 default:
1110 /**
1111 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1112 * later.
1113 */
1114 $tagtype = 1;
1115 break;
1116 }
1117
1118 $tag_start = $pos;
1119 $tagname = '';
1120 /**
1121 * Look for next [\W-_], which will indicate the end of the tag name.
1122 */
1123 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1124 if ($regary == false){
1125 return Array(false, false, false, $lt, strlen($body));
1126 }
1127 list($pos, $tagname, $match) = $regary;
1128 $tagname = strtolower($tagname);
1129
1130 /**
1131 * $match can be either of these:
1132 * '>' indicating the end of the tag entirely.
1133 * '\s' indicating the end of the tag name.
1134 * '/' indicating that this is type-3 xhtml tag.
1135 *
1136 * Whatever else we find there indicates an invalid tag.
1137 */
1138 switch ($match){
1139 case '/':
1140 /**
1141 * This is an xhtml-style tag with a closing / at the
1142 * end, like so: <img src="blah"/>. Check if it's followed
1143 * by the closing bracket. If not, then this tag is invalid
1144 */
1145 if (substr($body, $pos, 2) == "/>"){
1146 $pos++;
1147 $tagtype = 3;
1148 } else {
1149 $gt = sq_findnxstr($body, $pos, ">");
1150 $retary = Array(false, false, false, $lt, $gt);
1151 return $retary;
1152 }
1153 case '>':
1154 return Array($tagname, false, $tagtype, $lt, $pos);
1155 break;
1156 default:
1157 /**
1158 * Check if it's whitespace
1159 */
1160 if (!preg_match('/\s/', $match)){
1161 /**
1162 * This is an invalid tag! Look for the next closing ">".
1163 */
1164 $gt = sq_findnxstr($body, $lt, ">");
1165 return Array(false, false, false, $lt, $gt);
1166 }
1167 break;
1168 }
1169
1170 /**
1171 * At this point we're here:
1172 * <tagname attribute='blah'>
1173 * \-------^
1174 *
1175 * At this point we loop in order to find all attributes.
1176 */
1177 $attname = '';
1178 $atttype = false;
1179 $attary = Array();
1180
1181 while ($pos <= strlen($body)){
1182 $pos = sq_skipspace($body, $pos);
1183 if ($pos == strlen($body)){
1184 /**
1185 * Non-closed tag.
1186 */
1187 return Array(false, false, false, $lt, $pos);
1188 }
1189 /**
1190 * See if we arrived at a ">" or "/>", which means that we reached
1191 * the end of the tag.
1192 */
1193 $matches = Array();
1194 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1195 /**
1196 * Yep. So we did.
1197 */
1198 $pos += strlen($matches{1});
1199 if ($matches{2} == "/>"){
1200 $tagtype = 3;
1201 $pos++;
1202 }
1203 return Array($tagname, $attary, $tagtype, $lt, $pos);
1204 }
1205
1206 /**
1207 * There are several types of attributes, with optional
1208 * [:space:] between members.
1209 * Type 1:
1210 * attrname[:space:]=[:space:]'CDATA'
1211 * Type 2:
1212 * attrname[:space:]=[:space:]"CDATA"
1213 * Type 3:
1214 * attr[:space:]=[:space:]CDATA
1215 * Type 4:
1216 * attrname
1217 *
1218 * We leave types 1 and 2 the same, type 3 we check for
1219 * '"' and convert to "&quot" if needed, then wrap in
1220 * double quotes. Type 4 we convert into:
1221 * attrname="yes".
1222 */
1223 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1224 if ($regary == false){
1225 /**
1226 * Looks like body ended before the end of tag.
1227 */
1228 return Array(false, false, false, $lt, strlen($body));
1229 }
1230 list($pos, $attname, $match) = $regary;
1231 $attname = strtolower($attname);
1232 /**
1233 * We arrived at the end of attribute name. Several things possible
1234 * here:
1235 * '>' means the end of the tag and this is attribute type 4
1236 * '/' if followed by '>' means the same thing as above
1237 * '\s' means a lot of things -- look what it's followed by.
1238 * anything else means the attribute is invalid.
1239 */
1240 switch($match){
1241 case '/':
1242 /**
1243 * This is an xhtml-style tag with a closing / at the
1244 * end, like so: <img src="blah"/>. Check if it's followed
1245 * by the closing bracket. If not, then this tag is invalid
1246 */
1247 if (substr($body, $pos, 2) == "/>"){
1248 $pos++;
1249 $tagtype = 3;
1250 } else {
1251 $gt = sq_findnxstr($body, $pos, ">");
1252 $retary = Array(false, false, false, $lt, $gt);
1253 return $retary;
1254 }
1255 case '>':
1256 $attary{$attname} = '"yes"';
1257 return Array($tagname, $attary, $tagtype, $lt, $pos);
1258 break;
1259 default:
1260 /**
1261 * Skip whitespace and see what we arrive at.
1262 */
1263 $pos = sq_skipspace($body, $pos);
1264 $char = substr($body, $pos, 1);
1265 /**
1266 * Two things are valid here:
1267 * '=' means this is attribute type 1 2 or 3.
1268 * \w means this was attribute type 4.
1269 * anything else we ignore and re-loop. End of tag and
1270 * invalid stuff will be caught by our checks at the beginning
1271 * of the loop.
1272 */
1273 if ($char == "="){
1274 $pos++;
1275 $pos = sq_skipspace($body, $pos);
1276 /**
1277 * Here are 3 possibilities:
1278 * "'" attribute type 1
1279 * '"' attribute type 2
1280 * everything else is the content of tag type 3
1281 */
1282 $quot = substr($body, $pos, 1);
1283 if ($quot == "'"){
1284 $regary = sq_findnxreg($body, $pos+1, "\'");
1285 if ($regary == false){
1286 return Array(false, false, false, $lt, strlen($body));
1287 }
1288 list($pos, $attval, $match) = $regary;
1289 $pos++;
1290 $attary{$attname} = "'" . $attval . "'";
1291 } else if ($quot == '"'){
1292 $regary = sq_findnxreg($body, $pos+1, '\"');
1293 if ($regary == false){
1294 return Array(false, false, false, $lt, strlen($body));
1295 }
1296 list($pos, $attval, $match) = $regary;
1297 $pos++;
1298 $attary{$attname} = '"' . $attval . '"';
1299 } else {
1300 /**
1301 * These are hateful. Look for \s, or >.
1302 */
1303 $regary = sq_findnxreg($body, $pos, "[\s>]");
1304 if ($regary == false){
1305 return Array(false, false, false, $lt, strlen($body));
1306 }
1307 list($pos, $attval, $match) = $regary;
1308 /**
1309 * If it's ">" it will be caught at the top.
1310 */
1311 $attval = preg_replace("/\"/s", "&quot;", $attval);
1312 $attary{$attname} = '"' . $attval . '"';
1313 }
1314 } else if (preg_match("|[\w/>]|", $char)) {
1315 /**
1316 * That was attribute type 4.
1317 */
1318 $attary{$attname} = '"yes"';
1319 } else {
1320 /**
1321 * An illegal character. Find next '>' and return.
1322 */
1323 $gt = sq_findnxstr($body, $pos, ">");
1324 return Array(false, false, false, $lt, $gt);
1325 }
1326 break;
1327 }
1328 }
1329 /**
1330 * The fact that we got here indicates that the tag end was never
1331 * found. Return invalid tag indication so it gets stripped.
1332 */
1333 return Array(false, false, false, $lt, strlen($body));
1334 }
1335
1336 /**
1337 * This function checks attribute values for entity-encoded values
1338 * and returns them translated into 8-bit strings so we can run
1339 * checks on them.
1340 *
1341 * @param $attvalue A string to run entity check against.
1342 * @return Translated value.
1343 */
1344
1345 function sq_deent($attvalue){
1346 $me = 'sq_deent';
1347 /**
1348 * See if we have to run the checks first. All entities must start
1349 * with "&".
1350 */
1351 if (strpos($attvalue, '&') === false){
1352 return $attvalue;
1353 }
1354 /**
1355 * Check named entities first.
1356 */
1357 $trans = get_html_translation_table(HTML_ENTITIES);
1358 /**
1359 * Leave &quot; in, as it can mess us up.
1360 */
1361 $trans = array_flip($trans);
1362 unset($trans{'&quot;'});
1363 while (list($ent, $val) = each($trans)){
1364 $attvalue = preg_replace('/' . $ent . '*/si', $val, $attvalue);
1365 }
1366 /**
1367 * Now translate numbered entities from 1 to 255 if needed.
1368 */
1369 if (strpos($attvalue, '#') !== false){
1370 $omit = Array(34, 39);
1371 for ($asc = 256; $asc >= 0; $asc--){
1372 if (!in_array($asc, $omit)){
1373 $chr = chr($asc);
1374 $octrule = '/\&#0*' . $asc . ';*/si';
1375 $hexrule = '/\&#x0*' . dechex($asc) . ';*/si';
1376 $attvalue = preg_replace($octrule, $chr, $attvalue);
1377 $attvalue = preg_replace($hexrule, $chr, $attvalue);
1378 }
1379 }
1380 }
1381 return $attvalue;
1382 }
1383
1384 /**
1385 * This function runs various checks against the attributes.
1386 *
1387 * @param $tagname String with the name of the tag.
1388 * @param $attary Array with all tag attributes.
1389 * @param $rm_attnames See description for sq_sanitize
1390 * @param $bad_attvals See description for sq_sanitize
1391 * @param $add_attr_to_tag See description for sq_sanitize
1392 * @param $message message object
1393 * @param $id message id
1394 * @return Array with modified attributes.
1395 */
1396 function sq_fixatts($tagname,
1397 $attary,
1398 $rm_attnames,
1399 $bad_attvals,
1400 $add_attr_to_tag,
1401 $message,
1402 $id,
1403 $mailbox
1404 ){
1405 $me = 'sq_fixatts';
1406 while (list($attname, $attvalue) = each($attary)){
1407 /**
1408 * See if this attribute should be removed.
1409 */
1410 foreach ($rm_attnames as $matchtag=>$matchattrs){
1411 if (preg_match($matchtag, $tagname)){
1412 foreach ($matchattrs as $matchattr){
1413 if (preg_match($matchattr, $attname)){
1414 unset($attary{$attname});
1415 continue;
1416 }
1417 }
1418 }
1419 }
1420 /**
1421 * Remove any backslashes, entities, and extraneous whitespace.
1422 */
1423 $attvalue = sq_unbackslash($attvalue);
1424 $attvalue = sq_deent($attvalue);
1425 $attvalue = sq_unspace($attvalue);
1426
1427 /**
1428 * Remove \r \n \t \0 " " "\\"
1429 */
1430 $attvalue = str_replace(Array("\r", "\n", "\t", "\0", " ", "\\"),
1431 Array('', '','','','',''), $attvalue);
1432
1433 /**
1434 * Now let's run checks on the attvalues.
1435 * I don't expect anyone to comprehend this. If you do,
1436 * get in touch with me so I can drive to where you live and
1437 * shake your hand personally. :)
1438 */
1439 foreach ($bad_attvals as $matchtag=>$matchattrs){
1440 if (preg_match($matchtag, $tagname)){
1441 foreach ($matchattrs as $matchattr=>$valary){
1442 if (preg_match($matchattr, $attname)){
1443 /**
1444 * There are two arrays in valary.
1445 * First is matches.
1446 * Second one is replacements
1447 */
1448 list($valmatch, $valrepl) = $valary;
1449 $newvalue =
1450 preg_replace($valmatch, $valrepl, $attvalue);
1451 if ($newvalue != $attvalue){
1452 $attary{$attname} = $newvalue;
1453 }
1454 }
1455 }
1456 }
1457 }
1458 /**
1459 * Turn cid: urls into http-friendly ones.
1460 */
1461 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1462 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1463 }
1464 }
1465 /**
1466 * See if we need to append any attributes to this tag.
1467 */
1468 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1469 if (preg_match($matchtag, $tagname)){
1470 $attary = array_merge($attary, $addattary);
1471 }
1472 }
1473 return $attary;
1474 }
1475
1476 /**
1477 * This function edits the style definition to make them friendly and
1478 * usable in squirrelmail.
1479 *
1480 * @param $message the message object
1481 * @param $id the message id
1482 * @param $content a string with whatever is between <style> and </style>
1483 * @param $mailbox the message mailbox
1484 * @return a string with edited content.
1485 */
1486 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1487 global $view_unsafe_images;
1488 $me = 'sq_fixstyle';
1489 $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1490 if ($ret == FALSE){
1491 return array(FALSE, strlen($body));
1492 }
1493 $newpos = $ret[0] + strlen($ret[2]);
1494 $content = $ret[1];
1495 /**
1496 * First look for general BODY style declaration, which would be
1497 * like so:
1498 * body {background: blah-blah}
1499 * and change it to .bodyclass so we can just assign it to a <div>
1500 */
1501 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1502 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1503 /**
1504 * Fix url('blah') declarations.
1505 */
1506 $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1507 "url(\\1$secremoveimg\\2)", $content);
1508 /**
1509 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1510 * is false.
1511 */
1512 if (!$view_unsafe_images){
1513 $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1514 "url(\\1$secremoveimg\\2)", $content);
1515 }
1516
1517 /**
1518 * Fix urls that refer to cid:
1519 */
1520 while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1521 $content, $matches)){
1522 $cidurl = $matches{1};
1523 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1524 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1525 "url($httpurl)", $content);
1526 }
1527
1528 /**
1529 * Fix stupid css declarations which lead to vulnerabilities
1530 * in IE.
1531 */
1532 $match = Array('/expression/i',
1533 '/behaviou*r/i',
1534 '/binding/i',
1535 '/include-source/i');
1536 $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1537 $content = preg_replace($match, $replace, $content);
1538 return array($content, $newpos);
1539 }
1540
1541 /**
1542 * This function converts cid: url's into the ones that can be viewed in
1543 * the browser.
1544 *
1545 * @param $message the message object
1546 * @param $id the message id
1547 * @param $cidurl the cid: url.
1548 * @param $mailbox the message mailbox
1549 * @return a string with a http-friendly url
1550 */
1551 function sq_cid2http($message, $id, $cidurl, $mailbox){
1552 /**
1553 * Get rid of quotes.
1554 */
1555 $quotchar = substr($cidurl, 0, 1);
1556 if ($quotchar == '"' || $quotchar == "'"){
1557 $cidurl = str_replace($quotchar, "", $cidurl);
1558 } else {
1559 $quotchar = '';
1560 }
1561 $cidurl = substr(trim($cidurl), 4);
1562 $linkurl = find_ent_id($cidurl, $message);
1563 /* in case of non-save cid links $httpurl should be replaced by a sort of
1564 unsave link image */
1565 $httpurl = '';
1566 if ($linkurl) {
1567 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1568 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1569 '&amp;ent_id=' . $linkurl . $quotchar;
1570 }
1571 return $httpurl;
1572 }
1573
1574 /**
1575 * This function changes the <body> tag into a <div> tag since we
1576 * can't really have a body-within-body.
1577 *
1578 * @param $attary an array of attributes and values of <body>
1579 * @param $mailbox mailbox we're currently reading (for cid2http)
1580 * @param $message current message (for cid2http)
1581 * @param $id current message id (for cid2http)
1582 * @return a modified array of attributes to be set for <div>
1583 */
1584 function sq_body2div($attary, $mailbox, $message, $id){
1585 $me = 'sq_body2div';
1586 $divattary = Array('class' => "'bodyclass'");
1587 $bgcolor = '#ffffff';
1588 $text = '#000000';
1589 $has_bgc_stl = $has_txt_stl = false;
1590 $styledef = '';
1591 if (is_array($attary) && sizeof($attary) > 0){
1592 foreach ($attary as $attname=>$attvalue){
1593 $quotchar = substr($attvalue, 0, 1);
1594 $attvalue = str_replace($quotchar, "", $attvalue);
1595 switch ($attname){
1596 case 'background':
1597 $attvalue = sq_cid2http($message, $id,
1598 $attvalue, $mailbox);
1599 $styledef .= "background-image: url('$attvalue'); ";
1600 break;
1601 case 'bgcolor':
1602 $has_bgc_stl = true;
1603 $styledef .= "background-color: $attvalue; ";
1604 break;
1605 case 'text':
1606 $has_txt_stl = true;
1607 $styledef .= "color: $attvalue; ";
1608 break;
1609 }
1610 }
1611 // Outlook defines a white bgcolor and no text color. This can lead to
1612 // white text on a white bg with certain themes.
1613 if ($has_bgc_stl && !$has_txt_stl) {
1614 $styledef .= "color: $text; ";
1615 }
1616 if (strlen($styledef) > 0){
1617 $divattary{"style"} = "\"$styledef\"";
1618 }
1619 }
1620 return $divattary;
1621 }
1622
1623 /**
1624 * This is the main function and the one you should actually be calling.
1625 * There are several variables you should be aware of an which need
1626 * special description.
1627 *
1628 * Since the description is quite lengthy, see it here:
1629 * http://www.mricon.com/html/phpfilter.html
1630 *
1631 * @param $body the string with HTML you wish to filter
1632 * @param $tag_list see description above
1633 * @param $rm_tags_with_content see description above
1634 * @param $self_closing_tags see description above
1635 * @param $force_tag_closing see description above
1636 * @param $rm_attnames see description above
1637 * @param $bad_attvals see description above
1638 * @param $add_attr_to_tag see description above
1639 * @param $message message object
1640 * @param $id message id
1641 * @return sanitized html safe to show on your pages.
1642 */
1643 function sq_sanitize($body,
1644 $tag_list,
1645 $rm_tags_with_content,
1646 $self_closing_tags,
1647 $force_tag_closing,
1648 $rm_attnames,
1649 $bad_attvals,
1650 $add_attr_to_tag,
1651 $message,
1652 $id,
1653 $mailbox
1654 ){
1655 $me = 'sq_sanitize';
1656 $rm_tags = array_shift($tag_list);
1657 /**
1658 * Normalize rm_tags and rm_tags_with_content.
1659 */
1660 @array_walk($tag_list, 'sq_casenormalize');
1661 @array_walk($rm_tags_with_content, 'sq_casenormalize');
1662 @array_walk($self_closing_tags, 'sq_casenormalize');
1663 /**
1664 * See if tag_list is of tags to remove or tags to allow.
1665 * false means remove these tags
1666 * true means allow these tags
1667 */
1668 $curpos = 0;
1669 $open_tags = Array();
1670 $trusted = "\n<!-- begin sanitized html -->\n";
1671 $skip_content = false;
1672 /**
1673 * Take care of netscape's stupid javascript entities like
1674 * &{alert('boo')};
1675 */
1676 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1677
1678 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1679 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1680 $free_content = substr($body, $curpos, $lt-$curpos);
1681 /**
1682 * Take care of <style>
1683 */
1684 if ($tagname == "style" && $tagtype == 1){
1685 list($free_content, $curpos) =
1686 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1687 if ($free_content != FALSE){
1688 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1689 $trusted .= $free_content;
1690 $trusted .= sq_tagprint($tagname, false, 2);
1691 }
1692 continue;
1693 }
1694 if ($skip_content == false){
1695 $trusted .= $free_content;
1696 }
1697 if ($tagname != FALSE){
1698 if ($tagtype == 2){
1699 if ($skip_content == $tagname){
1700 /**
1701 * Got to the end of tag we needed to remove.
1702 */
1703 $tagname = false;
1704 $skip_content = false;
1705 } else {
1706 if ($skip_content == false){
1707 if ($tagname == "body"){
1708 $tagname = "div";
1709 }
1710 if (isset($open_tags{$tagname}) &&
1711 $open_tags{$tagname} > 0){
1712 $open_tags{$tagname}--;
1713 } else {
1714 $tagname = false;
1715 }
1716 }
1717 }
1718 } else {
1719 /**
1720 * $rm_tags_with_content
1721 */
1722 if ($skip_content == false){
1723 /**
1724 * See if this is a self-closing type and change
1725 * tagtype appropriately.
1726 */
1727 if ($tagtype == 1
1728 && in_array($tagname, $self_closing_tags)){
1729 $tagtype = 3;
1730 }
1731 /**
1732 * See if we should skip this tag and any content
1733 * inside it.
1734 */
1735 if ($tagtype == 1 &&
1736 in_array($tagname, $rm_tags_with_content)){
1737 $skip_content = $tagname;
1738 } else {
1739 if (($rm_tags == false
1740 && in_array($tagname, $tag_list)) ||
1741 ($rm_tags == true &&
1742 !in_array($tagname, $tag_list))){
1743 $tagname = false;
1744 } else {
1745 /**
1746 * Convert body into div.
1747 */
1748 if ($tagname == "body"){
1749 $tagname = "div";
1750 $attary = sq_body2div($attary, $mailbox,
1751 $message, $id);
1752 }
1753 if ($tagtype == 1){
1754 if (isset($open_tags{$tagname})){
1755 $open_tags{$tagname}++;
1756 } else {
1757 $open_tags{$tagname}=1;
1758 }
1759 }
1760 /**
1761 * This is where we run other checks.
1762 */
1763 if (is_array($attary) && sizeof($attary) > 0){
1764 $attary = sq_fixatts($tagname,
1765 $attary,
1766 $rm_attnames,
1767 $bad_attvals,
1768 $add_attr_to_tag,
1769 $message,
1770 $id,
1771 $mailbox
1772 );
1773 }
1774 }
1775 }
1776 }
1777 }
1778 if ($tagname != false && $skip_content == false){
1779 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1780 }
1781 }
1782 $curpos = $gt+1;
1783 }
1784 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1785 if ($force_tag_closing == true){
1786 foreach ($open_tags as $tagname=>$opentimes){
1787 while ($opentimes > 0){
1788 $trusted .= '</' . $tagname . '>';
1789 $opentimes--;
1790 }
1791 }
1792 $trusted .= "\n";
1793 }
1794 $trusted .= "<!-- end sanitized html -->\n";
1795 return $trusted;
1796 }
1797
1798 /**
1799 * This is a wrapper function to call html sanitizing routines.
1800 *
1801 * @param $body the body of the message
1802 * @param $id the id of the message
1803 * @return a string with html safe to display in the browser.
1804 */
1805 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1806 global $attachment_common_show_images, $view_unsafe_images,
1807 $has_unsafe_images;
1808 /**
1809 * Don't display attached images in HTML mode.
1810 */
1811 $attachment_common_show_images = false;
1812 $tag_list = Array(
1813 false,
1814 "object",
1815 "meta",
1816 "html",
1817 "head",
1818 "base",
1819 "link",
1820 "frame",
1821 "iframe",
1822 "plaintext",
1823 "marquee"
1824 );
1825
1826 $rm_tags_with_content = Array(
1827 "script",
1828 "applet",
1829 "embed",
1830 "title",
1831 "frameset",
1832 "xml"
1833 );
1834
1835 $self_closing_tags = Array(
1836 "img",
1837 "br",
1838 "hr",
1839 "input"
1840 );
1841
1842 $force_tag_closing = true;
1843
1844 $rm_attnames = Array(
1845 "/.*/" =>
1846 Array(
1847 "/target/i",
1848 "/^on.*/i",
1849 "/^dynsrc/i",
1850 "/^data.*/i",
1851 "/^lowsrc.*/i"
1852 )
1853 );
1854
1855 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1856 $bad_attvals = Array(
1857 "/.*/" =>
1858 Array(
1859 "/^src|background/i" =>
1860 Array(
1861 Array(
1862 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1863 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1864 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1865 ),
1866 Array(
1867 "\\1$secremoveimg\\2",
1868 "\\1$secremoveimg\\2",
1869 "\\1$secremoveimg\\2",
1870 "\\1$secremoveimg\\2"
1871 )
1872 ),
1873 "/^href|action/i" =>
1874 Array(
1875 Array(
1876 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1877 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1878 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1879 ),
1880 Array(
1881 "\\1#\\1",
1882 "\\1#\\1",
1883 "\\1#\\1",
1884 "\\1#\\1"
1885 )
1886 ),
1887 "/^style/i" =>
1888 Array(
1889 Array(
1890 "/expression/i",
1891 "/binding/i",
1892 "/behaviou*r/i",
1893 "/include-source/i",
1894 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1895 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1896 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
1897 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
1898 ),
1899 Array(
1900 "idiocy",
1901 "idiocy",
1902 "idiocy",
1903 "idiocy",
1904 "url(\\1#\\1)",
1905 "url(\\1#\\1)",
1906 "url(\\1#\\1)",
1907 "url(\\1#\\1)",
1908 "\\1:url(\\2#\\3)"
1909 )
1910 )
1911 )
1912 );
1913 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1914 $view_unsafe_images = false;
1915 }
1916 if (!$view_unsafe_images){
1917 /**
1918 * Remove any references to http/https if view_unsafe_images set
1919 * to false.
1920 */
1921 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1922 '/^([\'\"])\s*https*:.*([\'\"])/si');
1923 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1924 "\\1$secremoveimg\\1");
1925 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1926 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1927 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1928 "url(\\1$secremoveimg\\1)");
1929 }
1930
1931 $add_attr_to_tag = Array(
1932 "/^a$/i" =>
1933 Array('target'=>'"_new"',
1934 'title'=>'"'._("This external link will open in a new window").'"'
1935 )
1936 );
1937 $trusted = sq_sanitize($body,
1938 $tag_list,
1939 $rm_tags_with_content,
1940 $self_closing_tags,
1941 $force_tag_closing,
1942 $rm_attnames,
1943 $bad_attvals,
1944 $add_attr_to_tag,
1945 $message,
1946 $id,
1947 $mailbox
1948 );
1949 if (preg_match("|$secremoveimg|i", $trusted)){
1950 $has_unsafe_images = true;
1951 }
1952 return $trusted;
1953 }
1954
1955 /**
1956 * function SendDownloadHeaders - send file to the browser
1957 *
1958 * Original Source: SM core src/download.php
1959 * moved here to make it available to other code, and separate
1960 * front end from back end functionality.
1961 *
1962 * @param string $type0 first half of mime type
1963 * @param string $type1 second half of mime type
1964 * @param string $filename filename to tell the browser for downloaded file
1965 * @param boolean $force whether to force the download dialog to pop
1966 * @param optional integer $filesize send the Content-Header and length to the browser
1967 * @return void
1968 */
1969 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
1970 global $languages, $squirrelmail_language;
1971 $isIE = $isIE6 = 0;
1972
1973 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
1974
1975 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
1976 strstr($HTTP_USER_AGENT, 'Opera') === false) {
1977 $isIE = 1;
1978 }
1979
1980 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
1981 strstr($HTTP_USER_AGENT, 'Opera') === false) {
1982 $isIE6 = 1;
1983 }
1984
1985 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
1986 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
1987 $filename =
1988 $languages[$squirrelmail_language]['XTRA_CODE']('downloadfilename', $filename, $HTTP_USER_AGENT);
1989 } else {
1990 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
1991 }
1992
1993 // A Pox on Microsoft and it's Internet Explorer!
1994 //
1995 // IE has lots of bugs with file downloads.
1996 // It also has problems with SSL. Both of these cause problems
1997 // for us in this function.
1998 //
1999 // See this article on Cache Control headers and SSL
2000 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2001 //
2002 // The best thing you can do for IE is to upgrade to the latest
2003 // version
2004 //set all the Cache Control Headers for IE
2005 if ($isIE) {
2006 header ("Pragma: public");
2007 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); # HTTP/1.1
2008 header ("Cache-Control: post-check=0, pre-check=0", false);
2009 header ("Cache-control: private");
2010
2011 //set the inline header for IE, we'll add the attachment header later if we need it
2012 header ("Content-Disposition: inline; filename=$filename");
2013 }
2014
2015 if (!$force) {
2016 // Try to show in browser window
2017 header ("Content-Disposition: inline; filename=\"$filename\"");
2018 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2019 } else {
2020 // Try to pop up the "save as" box
2021
2022 // IE makes this hard. It pops up 2 save boxes, or none.
2023 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2024 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2025 // But, according to Microsoft, it is "RFC compliant but doesn't
2026 // take into account some deviations that allowed within the
2027 // specification." Doesn't that mean RFC non-compliant?
2028 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2029
2030 // all browsers need the application/octet-stream header for this
2031 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2032
2033 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2034 // Do not have quotes around filename, but that applied to
2035 // "attachment"... does it apply to inline too?
2036 header ("Content-Disposition: attachment; filename=\"$filename\"");
2037
2038 if ($isIE && !$isIE6) {
2039 // This combination seems to work mostly. IE 5.5 SP 1 has
2040 // known issues (see the Microsoft Knowledge Base)
2041
2042 // This works for most types, but doesn't work with Word files
2043 header ("Content-Type: application/download; name=\"$filename\"");
2044
2045 // These are spares, just in case. :-)
2046 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2047 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2048 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2049 } else {
2050 // another application/octet-stream forces download for Netscape
2051 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2052 }
2053 }
2054
2055 //send the content-length header if the calling function provides it
2056 if ($filesize > 0) {
2057 header("Content-Length: $filesize");
2058 }
2059
2060 } // end fn SendDownloadHeaders
2061
2062 ?>