Applied bugfixes to htmlfilter code from stable.
[squirrelmail.git] / functions / mime.php
1 <?php
2
3 /**
4 * mime.php
5 *
6 * Copyright (c) 1999-2003 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 * $Id$
13 */
14
15 require_once(SM_PATH . 'functions/imap.php');
16 require_once(SM_PATH . 'functions/attachment_common.php');
17
18 /* -------------------------------------------------------------------------- */
19 /* MIME DECODING */
20 /* -------------------------------------------------------------------------- */
21
22 /* This function gets the structure of a message and stores it in the "message" class.
23 * It will return this object for use with all relevant header information and
24 * fully parsed into the standard "message" object format.
25 */
26
27 function mime_structure ($bodystructure, $flags=array()) {
28
29 /* Isolate the body structure and remove beginning and end parenthesis. */
30 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
31 $read = trim(substr ($read, 0, -1));
32 $i = 0;
33 $msg = Message::parseStructure($read,$i);
34 if (!is_object($msg)) {
35 include_once(SM_PATH . 'functions/display_messages.php');
36 global $color, $mailbox;
37 /* removed urldecode because $_GET is auto urldecoded ??? */
38 displayPageHeader( $color, $mailbox );
39 echo "<BODY TEXT=\"$color[8]\" BGCOLOR=\"$color[4]\" LINK=\"$color[7]\" VLINK=\"$color[7]\" ALINK=\"$color[7]\">\n\n" .
40 '<CENTER>';
41 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
42 $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
43 $errormessage .= '<table><tr><td>' . htmlspecialchars($read) . '</td></tr></table>';
44 plain_error_message( $errormessage, $color );
45 echo '</body></html>';
46 exit;
47 }
48 if (count($flags)) {
49 foreach ($flags as $flag) {
50 $char = strtoupper($flag{1});
51 switch ($char) {
52 case 'S':
53 if (strtolower($flag) == '\\seen') {
54 $msg->is_seen = true;
55 }
56 break;
57 case 'A':
58 if (strtolower($flag) == '\\answered') {
59 $msg->is_answered = true;
60 }
61 break;
62 case 'D':
63 if (strtolower($flag) == '\\deleted') {
64 $msg->is_deleted = true;
65 }
66 break;
67 case 'F':
68 if (strtolower($flag) == '\\flagged') {
69 $msg->is_flagged = true;
70 }
71 break;
72 case 'M':
73 if (strtolower($flag) == '$mdnsent') {
74 $msg->is_mdnsent = true;
75 }
76 break;
77 default:
78 break;
79 }
80 }
81 }
82 // listEntities($msg);
83 return $msg;
84 }
85
86
87
88 /* This starts the parsing of a particular structure. It is called recursively,
89 * so it can be passed different structures. It returns an object of type
90 * $message.
91 * First, it checks to see if it is a multipart message. If it is, then it
92 * handles that as it sees is necessary. If it is just a regular entity,
93 * then it parses it and adds the necessary header information (by calling out
94 * to mime_get_elements()
95 */
96
97 function mime_fetch_body($imap_stream, $id, $ent_id=1) {
98 global $uid_support;
99 /* Do a bit of error correction. If we couldn't find the entity id, just guess
100 * that it is the first one. That is usually the case anyway.
101 */
102
103 if (!$ent_id) {
104 $cmd = "FETCH $id BODY[]";
105 } else {
106 $cmd = "FETCH $id BODY[$ent_id]";
107 }
108
109 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
110 do {
111 $topline = trim(array_shift($data));
112 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
113
114 $wholemessage = implode('', $data);
115 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
116 $ret = substr($wholemessage, 0, $regs[1]);
117 /* There is some information in the content info header that could be important
118 * in order to parse html messages. Let's get them here.
119 */
120 // if ($ret{0} == '<') {
121 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
122 // }
123 } else if (ereg('"([^"]*)"', $topline, $regs)) {
124 $ret = $regs[1];
125 } else {
126 global $where, $what, $mailbox, $passed_id, $startMessage;
127 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
128 if (isset($where) && isset($what)) {
129 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
130 } else {
131 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
132 }
133 $par .= '&amp;response=' . urlencode($response) .
134 '&amp;message=' . urlencode($message) .
135 '&amp;topline=' . urlencode($topline);
136
137 echo '<tt><br>' .
138 '<table width="80%"><tr>' .
139 '<tr><td colspan=2>' .
140 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
141 '</td></tr>' .
142 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
143 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
144 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
145 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
146 "</table><BR></tt></font><hr>";
147
148 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
149 array_shift($data);
150 $wholemessage = implode('', $data);
151
152 $ret = $wholemessage;
153 }
154 return $ret;
155 }
156
157 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
158 global $uid_support;
159
160 /* Don't kill the connection if the browser is over a dialup
161 * and it would take over 30 seconds to download it.
162 * DonĀ“t call set_time_limit in safe mode.
163 */
164
165 if (!ini_get('safe_mode')) {
166 set_time_limit(0);
167 }
168 /* in case of base64 encoded attachments, do not buffer them.
169 Instead, echo the decoded attachment directly to screen */
170 if (strtolower($encoding) == 'base64') {
171 if (!$ent_id) {
172 $query = "FETCH $id BODY[]";
173 } else {
174 $query = "FETCH $id BODY[$ent_id]";
175 }
176 sqimap_run_command($imap_stream,$query,true,$response,$message,$uid_support,'sqimap_base64_decode','php://stdout',true);
177 } else {
178 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
179 echo decodeBody($body, $encoding);
180 }
181
182 /*
183 TODO, use the same method for quoted printable.
184 However, I assume that quoted printable attachments aren't that large
185 so the performancegain / memory usage drop will be minimal.
186 If we decide to add that then we need to adapt sqimap_fread because
187 we need to split te result on \n and fread doesn't stop at \n. That
188 means we also should provide $results from sqimap_fread (by ref) to
189 te function and set $no_return to false. The $filter function for
190 quoted printable should handle unsetting of $results.
191 */
192 /*
193 TODO 2: find out how we write to the output stream php://stdout. fwrite
194 doesn't work because 'php://stdout isn't a stream.
195 */
196
197 return;
198 /*
199 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
200 $cnt = 0;
201 $continue = true;
202 $read = fgets ($imap_stream,8192);
203
204
205 // This could be bad -- if the section has sqimap_session_id() . ' OK'
206 // or similar, it will kill the download.
207 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
208 if (trim($read) == ')==') {
209 $read1 = $read;
210 $read = fgets ($imap_stream,4096);
211 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
212 return;
213 } else {
214 echo decodeBody($read1, $encoding) .
215 decodeBody($read, $encoding);
216 }
217 } else if ($cnt) {
218 echo decodeBody($read, $encoding);
219 }
220 $read = fgets ($imap_stream,4096);
221 $cnt++;
222 // break;
223 }
224 */
225 }
226
227 /* -[ END MIME DECODING ]----------------------------------------------------------- */
228
229 /* This is here for debugging purposes. It will print out a list
230 * of all the entity IDs that are in the $message object.
231 */
232 function listEntities ($message) {
233 if ($message) {
234 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
235 for ($i = 0; isset($message->entities[$i]); $i++) {
236 echo "$i : ";
237 $msg = listEntities($message->entities[$i]);
238
239 if ($msg) {
240 echo "return: ";
241 return $msg;
242 }
243 }
244 }
245 }
246
247 function getPriorityStr($priority) {
248 $priority_level = substr($priority,0,1);
249
250 switch($priority_level) {
251 /* Check for a higher then normal priority. */
252 case '1':
253 case '2':
254 $priority_string = _("High");
255 break;
256
257 /* Check for a lower then normal priority. */
258 case '4':
259 case '5':
260 $priority_string = _("Low");
261 break;
262
263 /* Check for a normal priority. */
264 case '3':
265 default:
266 $priority_level = '3';
267 $priority_string = _("Normal");
268 break;
269
270 }
271 return $priority_string;
272 }
273
274 /* returns a $message object for a particular entity id */
275 function getEntity ($message, $ent_id) {
276 return $message->getEntity($ent_id);
277 }
278
279 /* translateText
280 * Extracted from strings.php 23/03/2002
281 */
282
283 function translateText(&$body, $wrap_at, $charset) {
284 global $where, $what; /* from searching */
285 global $color; /* color theme */
286
287 require_once(SM_PATH . 'functions/url_parser.php');
288
289 $body_ary = explode("\n", $body);
290 for ($i=0; $i < count($body_ary); $i++) {
291 $line = $body_ary[$i];
292 if (strlen($line) - 2 >= $wrap_at) {
293 sqWordWrap($line, $wrap_at);
294 }
295 $line = charset_decode($charset, $line);
296 $line = str_replace("\t", ' ', $line);
297
298 parseUrl ($line);
299
300 $quotes = 0;
301 $pos = 0;
302 $j = strlen($line);
303
304 while ($pos < $j) {
305 if ($line[$pos] == ' ') {
306 $pos++;
307 } else if (strpos($line, '&gt;', $pos) === $pos) {
308 $pos += 4;
309 $quotes++;
310 } else {
311 break;
312 }
313 }
314
315 if ($quotes % 2) {
316 if (!isset($color[13])) {
317 $color[13] = '#800000';
318 }
319 $line = '<font color="' . $color[13] . '">' . $line . '</font>';
320 } elseif ($quotes) {
321 if (!isset($color[14])) {
322 $color[14] = '#FF0000';
323 }
324 $line = '<font color="' . $color[14] . '">' . $line . '</font>';
325 }
326
327 $body_ary[$i] = $line;
328 }
329 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
330 }
331
332 /* This returns a parsed string called $body. That string can then
333 * be displayed as the actual message in the HTML. It contains
334 * everything needed, including HTML Tags, Attachments at the
335 * bottom, etc.
336 */
337 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
338 /* This if statement checks for the entity to show as the
339 * primary message. To add more of them, just put them in the
340 * order that is their priority.
341 */
342 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
343 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
344 global $languages, $squirrelmail_language;
345
346 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
347 $view_unsafe_images = false;
348 }
349
350 $body = '';
351 $urlmailbox = urlencode($mailbox);
352 $body_message = getEntity($message, $ent_num);
353 if (($body_message->header->type0 == 'text') ||
354 ($body_message->header->type0 == 'rfc822')) {
355 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
356 $body = decodeBody($body, $body_message->header->encoding);
357
358 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
359 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
360 if (mb_detect_encoding($body) != 'ASCII') {
361 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
362 }
363 }
364 $hookResults = do_hook("message_body", $body);
365 $body = $hookResults[1];
366
367 /* If there are other types that shouldn't be formatted, add
368 * them here.
369 */
370
371 if ($body_message->header->type1 == 'html') {
372 if ($show_html_default <> 1) {
373 $entity_conv = array('&nbsp;' => ' ',
374 '<p>' => "\n",
375 '<br>' => "\n",
376 '<P>' => "\n",
377 '<BR>' => "\n",
378 '&gt;' => '>',
379 '&lt;' => '<');
380 $body = strtr($body, $entity_conv);
381 $body = strip_tags($body);
382 $body = trim($body);
383 translateText($body, $wrap_at,
384 $body_message->header->getParameter('charset'));
385 } else {
386 $body = magicHTML($body, $id, $message, $mailbox);
387 }
388 } else {
389 translateText($body, $wrap_at,
390 $body_message->header->getParameter('charset'));
391 }
392 $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
393 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
394 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
395 if (isset($passed_ent_id)) {
396 $link .= '&amp;passed_ent_id='.$passed_ent_id;
397 }
398 if ($view_unsafe_images) {
399 $text = _("Hide Unsafe Images");
400 } else {
401 if (isset($has_unsafe_images) && $has_unsafe_images) {
402 $link .= '&amp;view_unsafe_images=1';
403 $text = _("View Unsafe Images");
404 } else {
405 $text = '';
406 }
407 }
408 $body .= '<center><small><a href="'.$link.'">'.$text.
409 '</a></small></center><br>' . "\n";
410 }
411 return $body;
412 }
413
414
415 function formatAttachments($message, $exclude_id, $mailbox, $id) {
416 global $where, $what, $startMessage, $color;
417 static $ShownHTML = 0;
418
419 $att_ar = $message->getAttachments($exclude_id);
420
421 if (!count($att_ar)) return '';
422
423 $attachments = '';
424
425 $urlMailbox = urlencode($mailbox);
426
427 foreach ($att_ar as $att) {
428 $ent = $att->entity_id;
429 $header = $att->header;
430 $type0 = strtolower($header->type0);
431 $type1 = strtolower($header->type1);
432 $name = '';
433 $links['download link']['text'] = _("download");
434 $links['download link']['href'] = SM_PATH .
435 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
436 $ImageURL = '';
437 if ($type0 =='message' && $type1 == 'rfc822') {
438 $default_page = SM_PATH . 'src/read_body.php';
439 $rfc822_header = $att->rfc822_header;
440 $filename = $rfc822_header->subject;
441 if (trim( $filename ) == '') {
442 $filename = 'untitled-[' . $ent . ']' ;
443 }
444 $from_o = $rfc822_header->from;
445 if (is_object($from_o)) {
446 $from_name = $from_o->getAddress(false);
447 } else {
448 $from_name = _("Unknown sender");
449 }
450 $from_name = decodeHeader(($from_name));
451 $description = $from_name;
452 } else {
453 $default_page = SM_PATH . 'src/download.php';
454 if (is_object($header->disposition)) {
455 $filename = $header->disposition->getProperty('filename');
456 if (trim($filename) == '') {
457 $name = decodeHeader($header->disposition->getProperty('name'));
458 if (trim($name) == '') {
459 $name = $header->getParameter('name');
460 if(trim($name) == '') {
461 if (trim( $header->id ) == '') {
462 $filename = 'untitled-[' . $ent . ']' ;
463 } else {
464 $filename = 'cid: ' . $header->id;
465 }
466 } else {
467 $filename = $name;
468 }
469 } else {
470 $filename = $name;
471 }
472 }
473 } else {
474 $filename = $header->getParameter('name');
475 if (!trim($filename)) {
476 if (trim( $header->id ) == '') {
477 $filename = 'untitled-[' . $ent . ']' ;
478 } else {
479 $filename = 'cid: ' . $header->id;
480 }
481 }
482 }
483 if ($header->description) {
484 $description = decodeHeader($header->description);
485 } else {
486 $description = '';
487 }
488 }
489
490 $display_filename = $filename;
491 if (isset($passed_ent_id)) {
492 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
493 } else {
494 $passed_ent_id_link = '';
495 }
496 $defaultlink = $default_page . "?startMessage=$startMessage"
497 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
498 . '&amp;ent_id='.$ent.$passed_ent_id_link;
499 if ($where && $what) {
500 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
501 }
502 /* This executes the attachment hook with a specific MIME-type.
503 * If that doesn't have results, it tries if there's a rule
504 * for a more generic type.
505 */
506 $hookresults = do_hook("attachment $type0/$type1", $links,
507 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
508 $display_filename, $where, $what);
509 if(count($hookresults[1]) <= 1) {
510 $hookresults = do_hook("attachment $type0/*", $links,
511 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
512 $display_filename, $where, $what);
513 }
514
515 $links = $hookresults[1];
516 $defaultlink = $hookresults[6];
517
518 $attachments .= '<TR><TD>' .
519 '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
520 '<TD><SMALL><b>' . show_readable_size($header->size) .
521 '</b>&nbsp;&nbsp;</small></TD>' .
522 "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
523 '<TD><SMALL>';
524 $attachments .= '<b>' . $description . '</b>';
525 $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
526
527 $skipspaces = 1;
528 foreach ($links as $val) {
529 if ($skipspaces) {
530 $skipspaces = 0;
531 } else {
532 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
533 }
534 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
535 }
536 unset($links);
537 $attachments .= "</TD></TR>\n";
538 }
539 return $attachments;
540 }
541
542 function sqimap_base64_decode(&$string) {
543 $string = str_replace("\r\n", "\n", $string);
544 $string = base64_decode($string);
545 }
546
547 /* This function decodes the body depending on the encoding type. */
548 function decodeBody($body, $encoding) {
549 global $show_html_default;
550
551 $body = str_replace("\r\n", "\n", $body);
552 $encoding = strtolower($encoding);
553
554 if ($encoding == 'quoted-printable' ||
555 $encoding == 'quoted_printable') {
556 $body = quoted_printable_decode($body);
557
558 while (ereg("=\n", $body)) {
559 $body = ereg_replace ("=\n", '', $body);
560 }
561
562 } else if ($encoding == 'base64') {
563 $body = base64_decode($body);
564 }
565
566 // All other encodings are returned raw.
567 return $body;
568 }
569
570 /*
571 * This functions decode strings that is encoded according to
572 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
573 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
574 */
575 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
576 global $languages, $squirrelmail_language;
577 if (is_array($string)) {
578 $string = implode("\n", $string);
579 }
580
581 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
582 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
583 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
584 // Do we need to return at this point?
585 // return $string;
586 }
587 $i = 0;
588 $iLastMatch = -2;
589 $encoded = false;
590
591 $aString = explode(' ',$string);
592 $ret = '';
593 foreach ($aString as $chunk) {
594 if ($encoded && !$chunk) {
595 continue;
596 } elseif (!$chunk) {
597 $ret .= ' ';
598 continue;
599 }
600 $encoded = false;
601 /* if encoded words are not separated by a linear-space-white we still catch them */
602 $j = $i-1;
603 // if ($chunk{0} === '=') { /* performance, saves an unnessecarry preg call */
604 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
605 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
606 if ($iLastMatch !== $j) {
607 if ($htmlsave) {
608 $ret .= '&nbsp;';
609 } else {
610 $ret .= ' ';
611 }
612 }
613 $iLastMatch = $i;
614 $j = $i;
615 $ret .= $res[1];
616 $encoding = ucfirst($res[3]);
617 switch ($encoding)
618 {
619 case 'B':
620 $replace = base64_decode($res[4]);
621 $ret .= charset_decode($res[2],$replace);
622 break;
623 case 'Q':
624 $replace = str_replace('_', ' ', $res[4]);
625 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
626 $replace);
627 /* Only encode into entities by default. Some places
628 * don't need the encoding, like the compose form.
629 */
630 if ($utfencode) {
631 $replace = charset_decode($res[2], $replace);
632 } else {
633 if ($htmlsave) {
634 $replace = htmlspecialchars($replace);
635 }
636 }
637 $ret .= $replace;
638 break;
639 default:
640 break;
641 }
642 $chunk = $res[5];
643 $encoded = true;
644 }
645 // }
646 if (!$encoded) {
647 if ($htmlsave) {
648 $ret .= '&nbsp;';
649 } else {
650 $ret .= ' ';
651 }
652 }
653
654 if (!$encoded && $htmlsave) {
655 $ret .= htmlspecialchars($chunk);
656 } else {
657 $ret .= $chunk;
658 }
659 ++$i;
660 }
661 /* remove the first added space */
662 if ($ret) {
663 if ($htmlsave) {
664 $ret = substr($ret,6);
665 } else {
666 $ret = substr($ret,1);
667 }
668 }
669
670 return $ret;
671 }
672
673 /*
674 * Encode a string according to RFC 1522 for use in headers if it
675 * contains 8-bit characters or anything that looks like it should
676 * be encoded.
677 */
678 function encodeHeader ($string) {
679 global $default_charset, $languages, $squirrelmail_language;
680
681 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
682 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
683 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
684 }
685 if (strtolower($default_charset) == 'iso-8859-1') {
686 $string = str_replace("\240",' ',$string);
687 }
688
689 // Encode only if the string contains 8-bit characters or =?
690 $j = strlen($string);
691 $max_l = 75 - strlen($default_charset) - 7;
692 $aRet = array();
693 $ret = '';
694 $iEncStart = $enc_init = false;
695 $cur_l = $iOffset = 0;
696 for($i = 0; $i < $j; ++$i) {
697 switch($string{$i})
698 {
699 case '=':
700 case '<':
701 case '>':
702 case ',':
703 case '?':
704 case '_':
705 if ($iEncStart === false) {
706 $iEncStart = $i;
707 }
708 $cur_l+=3;
709 if ($cur_l > ($max_l-2)) {
710 /* if there is an stringpart that doesn't need encoding, add it */
711 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
712 $aRet[] = "=?$default_charset?Q?$ret?=";
713 $iOffset = $i;
714 $cur_l = 0;
715 $ret = '';
716 $iEncStart = false;
717 } else {
718 $ret .= sprintf("=%02X",ord($string{$i}));
719 }
720 break;
721 case '(':
722 case ')':
723 if ($iEncStart !== false) {
724 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
725 $aRet[] = "=?$default_charset?Q?$ret?=";
726 $iOffset = $i;
727 $cur_l = 0;
728 $ret = '';
729 $iEncStart = false;
730 }
731 break;
732 case ' ':
733 if ($iEncStart !== false) {
734 $cur_l++;
735 if ($cur_l > $max_l) {
736 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
737 $aRet[] = "=?$default_charset?Q?$ret?=";
738 $iOffset = $i;
739 $cur_l = 0;
740 $ret = '';
741 $iEncStart = false;
742 } else {
743 $ret .= '_';
744 }
745 }
746 break;
747 default:
748 $k = ord($string{$i});
749 if ($k > 126) {
750 if ($iEncStart === false) {
751 // do not start encoding in the middle of a string, also take the rest of the word.
752 $sLeadString = substr($string,0,$i);
753 $aLeadString = explode(' ',$sLeadString);
754 $sToBeEncoded = array_pop($aLeadString);
755 $iEncStart = $i - strlen($sToBeEncoded);
756 $ret .= $sToBeEncoded;
757 $cur_l += strlen($sToBeEncoded);
758 }
759 $cur_l += 3;
760 /* first we add the encoded string that reached it's max size */
761 if ($cur_l > ($max_l-2)) {
762 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
763 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
764 $cur_l = 3;
765 $ret = '';
766 $iOffset = $i;
767 $iEncStart = $i;
768 }
769 $enc_init = true;
770 $ret .= sprintf("=%02X", $k);
771 } else {
772 if ($iEncStart !== false) {
773 $cur_l++;
774 if ($cur_l > $max_l) {
775 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
776 $aRet[] = "=?$default_charset?Q?$ret?=";
777 $iEncStart = false;
778 $iOffset = $i;
779 $cur_l = 0;
780 $ret = '';
781 } else {
782 $ret .= $string{$i};
783 }
784 }
785 }
786 break;
787 }
788 }
789
790 if ($enc_init) {
791 if ($iEncStart !== false) {
792 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
793 $aRet[] = "=?$default_charset?Q?$ret?=";
794 } else {
795 $aRet[] = substr($string,$iOffset);
796 }
797 $string = implode('',$aRet);
798 }
799 return $string;
800 }
801
802 /* This function trys to locate the entity_id of a specific mime element */
803 function find_ent_id($id, $message) {
804 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
805 if ($message->entities[$i]->header->type0 == 'multipart') {
806 $ret = find_ent_id($id, $message->entities[$i]);
807 } else {
808 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
809 // if (sq_check_save_extension($message->entities[$i])) {
810 return $message->entities[$i]->entity_id;
811 // }
812 }
813 }
814 }
815 return $ret;
816 }
817
818 function sq_check_save_extension($message) {
819 $filename = $message->getFilename();
820 $ext = substr($filename, strrpos($filename,'.')+1);
821 $save_extensions = array('jpg','jpeg','gif','png','bmp');
822 return in_array($ext, $save_extensions);
823 }
824
825
826 /**
827 ** HTMLFILTER ROUTINES
828 */
829
830 /**
831 * This function returns the final tag out of the tag name, an array
832 * of attributes, and the type of the tag. This function is called by
833 * sq_sanitize internally.
834 *
835 * @param $tagname the name of the tag.
836 * @param $attary the array of attributes and their values
837 * @param $tagtype The type of the tag (see in comments).
838 * @return a string with the final tag representation.
839 */
840 function sq_tagprint($tagname, $attary, $tagtype){
841 $me = 'sq_tagprint';
842
843 if ($tagtype == 2){
844 $fulltag = '</' . $tagname . '>';
845 } else {
846 $fulltag = '<' . $tagname;
847 if (is_array($attary) && sizeof($attary)){
848 $atts = Array();
849 while (list($attname, $attvalue) = each($attary)){
850 array_push($atts, "$attname=$attvalue");
851 }
852 $fulltag .= ' ' . join(" ", $atts);
853 }
854 if ($tagtype == 3){
855 $fulltag .= ' /';
856 }
857 $fulltag .= '>';
858 }
859 return $fulltag;
860 }
861
862 /**
863 * A small helper function to use with array_walk. Modifies a by-ref
864 * value and makes it lowercase.
865 *
866 * @param $val a value passed by-ref.
867 * @return void since it modifies a by-ref value.
868 */
869 function sq_casenormalize(&$val){
870 $val = strtolower($val);
871 }
872
873 /**
874 * This function skips any whitespace from the current position within
875 * a string and to the next non-whitespace value.
876 *
877 * @param $body the string
878 * @param $offset the offset within the string where we should start
879 * looking for the next non-whitespace character.
880 * @return the location within the $body where the next
881 * non-whitespace char is located.
882 */
883 function sq_skipspace($body, $offset){
884 $me = 'sq_skipspace';
885 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
886 if (sizeof($matches{1})){
887 $count = strlen($matches{1});
888 $offset += $count;
889 }
890 return $offset;
891 }
892
893 /**
894 * This function looks for the next character within a string. It's
895 * really just a glorified "strpos", except it catches if failures
896 * nicely.
897 *
898 * @param $body The string to look for needle in.
899 * @param $offset Start looking from this position.
900 * @param $needle The character/string to look for.
901 * @return location of the next occurance of the needle, or
902 * strlen($body) if needle wasn't found.
903 */
904 function sq_findnxstr($body, $offset, $needle){
905 $me = 'sq_findnxstr';
906 $pos = strpos($body, $needle, $offset);
907 if ($pos === FALSE){
908 $pos = strlen($body);
909 }
910 return $pos;
911 }
912
913 /**
914 * This function takes a PCRE-style regexp and tries to match it
915 * within the string.
916 *
917 * @param $body The string to look for needle in.
918 * @param $offset Start looking from here.
919 * @param $reg A PCRE-style regex to match.
920 * @return Returns a false if no matches found, or an array
921 * with the following members:
922 * - integer with the location of the match within $body
923 * - string with whatever content between offset and the match
924 * - string with whatever it is we matched
925 */
926 function sq_findnxreg($body, $offset, $reg){
927 $me = 'sq_findnxreg';
928 $matches = Array();
929 $retarr = Array();
930 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
931 if (!isset($matches{0}) || !$matches{0}){
932 $retarr = false;
933 } else {
934 $retarr{0} = $offset + strlen($matches{1});
935 $retarr{1} = $matches{1};
936 $retarr{2} = $matches{2};
937 }
938 return $retarr;
939 }
940
941 /**
942 * This function looks for the next tag.
943 *
944 * @param $body String where to look for the next tag.
945 * @param $offset Start looking from here.
946 * @return false if no more tags exist in the body, or
947 * an array with the following members:
948 * - string with the name of the tag
949 * - array with attributes and their values
950 * - integer with tag type (1, 2, or 3)
951 * - integer where the tag starts (starting "<")
952 * - integer where the tag ends (ending ">")
953 * first three members will be false, if the tag is invalid.
954 */
955 function sq_getnxtag($body, $offset){
956 $me = 'sq_getnxtag';
957 if ($offset > strlen($body)){
958 return false;
959 }
960 $lt = sq_findnxstr($body, $offset, "<");
961 if ($lt == strlen($body)){
962 return false;
963 }
964 /**
965 * We are here:
966 * blah blah <tag attribute="value">
967 * \---------^
968 */
969 $pos = sq_skipspace($body, $lt+1);
970 if ($pos >= strlen($body)){
971 return Array(false, false, false, $lt, strlen($body));
972 }
973 /**
974 * There are 3 kinds of tags:
975 * 1. Opening tag, e.g.:
976 * <a href="blah">
977 * 2. Closing tag, e.g.:
978 * </a>
979 * 3. XHTML-style content-less tag, e.g.:
980 * <img src="blah"/>
981 */
982 $tagtype = false;
983 switch (substr($body, $pos, 1)){
984 case '/':
985 $tagtype = 2;
986 $pos++;
987 break;
988 case '!':
989 /**
990 * A comment or an SGML declaration.
991 */
992 if (substr($body, $pos+1, 2) == "--"){
993 $gt = strpos($body, "-->", $pos);
994 if ($gt === false){
995 $gt = strlen($body);
996 } else {
997 $gt += 2;
998 }
999 return Array(false, false, false, $lt, $gt);
1000 } else {
1001 $gt = sq_findnxstr($body, $pos, ">");
1002 return Array(false, false, false, $lt, $gt);
1003 }
1004 break;
1005 default:
1006 /**
1007 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1008 * later.
1009 */
1010 $tagtype = 1;
1011 break;
1012 }
1013
1014 $tag_start = $pos;
1015 $tagname = '';
1016 /**
1017 * Look for next [\W-_], which will indicate the end of the tag name.
1018 */
1019 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1020 if ($regary == false){
1021 return Array(false, false, false, $lt, strlen($body));
1022 }
1023 list($pos, $tagname, $match) = $regary;
1024 $tagname = strtolower($tagname);
1025
1026 /**
1027 * $match can be either of these:
1028 * '>' indicating the end of the tag entirely.
1029 * '\s' indicating the end of the tag name.
1030 * '/' indicating that this is type-3 xhtml tag.
1031 *
1032 * Whatever else we find there indicates an invalid tag.
1033 */
1034 switch ($match){
1035 case '/':
1036 /**
1037 * This is an xhtml-style tag with a closing / at the
1038 * end, like so: <img src="blah"/>. Check if it's followed
1039 * by the closing bracket. If not, then this tag is invalid
1040 */
1041 if (substr($body, $pos, 2) == "/>"){
1042 $pos++;
1043 $tagtype = 3;
1044 } else {
1045 $gt = sq_findnxstr($body, $pos, ">");
1046 $retary = Array(false, false, false, $lt, $gt);
1047 return $retary;
1048 }
1049 case '>':
1050 return Array($tagname, false, $tagtype, $lt, $pos);
1051 break;
1052 default:
1053 /**
1054 * Check if it's whitespace
1055 */
1056 if (!preg_match('/\s/', $match)){
1057 /**
1058 * This is an invalid tag! Look for the next closing ">".
1059 */
1060 $gt = sq_findnxstr($body, $lt, ">");
1061 return Array(false, false, false, $lt, $gt);
1062 }
1063 break;
1064 }
1065
1066 /**
1067 * At this point we're here:
1068 * <tagname attribute='blah'>
1069 * \-------^
1070 *
1071 * At this point we loop in order to find all attributes.
1072 */
1073 $attname = '';
1074 $atttype = false;
1075 $attary = Array();
1076
1077 while ($pos <= strlen($body)){
1078 $pos = sq_skipspace($body, $pos);
1079 if ($pos == strlen($body)){
1080 /**
1081 * Non-closed tag.
1082 */
1083 return Array(false, false, false, $lt, $pos);
1084 }
1085 /**
1086 * See if we arrived at a ">" or "/>", which means that we reached
1087 * the end of the tag.
1088 */
1089 $matches = Array();
1090 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1091 /**
1092 * Yep. So we did.
1093 */
1094 $pos += strlen($matches{1});
1095 if ($matches{2} == "/>"){
1096 $tagtype = 3;
1097 $pos++;
1098 }
1099 return Array($tagname, $attary, $tagtype, $lt, $pos);
1100 }
1101
1102 /**
1103 * There are several types of attributes, with optional
1104 * [:space:] between members.
1105 * Type 1:
1106 * attrname[:space:]=[:space:]'CDATA'
1107 * Type 2:
1108 * attrname[:space:]=[:space:]"CDATA"
1109 * Type 3:
1110 * attr[:space:]=[:space:]CDATA
1111 * Type 4:
1112 * attrname
1113 *
1114 * We leave types 1 and 2 the same, type 3 we check for
1115 * '"' and convert to "&quot" if needed, then wrap in
1116 * double quotes. Type 4 we convert into:
1117 * attrname="yes".
1118 */
1119 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1120 if ($regary == false){
1121 /**
1122 * Looks like body ended before the end of tag.
1123 */
1124 return Array(false, false, false, $lt, strlen($body));
1125 }
1126 list($pos, $attname, $match) = $regary;
1127 $attname = strtolower($attname);
1128 /**
1129 * We arrived at the end of attribute name. Several things possible
1130 * here:
1131 * '>' means the end of the tag and this is attribute type 4
1132 * '/' if followed by '>' means the same thing as above
1133 * '\s' means a lot of things -- look what it's followed by.
1134 * anything else means the attribute is invalid.
1135 */
1136 switch($match){
1137 case '/':
1138 /**
1139 * This is an xhtml-style tag with a closing / at the
1140 * end, like so: <img src="blah"/>. Check if it's followed
1141 * by the closing bracket. If not, then this tag is invalid
1142 */
1143 if (substr($body, $pos, 2) == "/>"){
1144 $pos++;
1145 $tagtype = 3;
1146 } else {
1147 $gt = sq_findnxstr($body, $pos, ">");
1148 $retary = Array(false, false, false, $lt, $gt);
1149 return $retary;
1150 }
1151 case '>':
1152 $attary{$attname} = '"yes"';
1153 return Array($tagname, $attary, $tagtype, $lt, $pos);
1154 break;
1155 default:
1156 /**
1157 * Skip whitespace and see what we arrive at.
1158 */
1159 $pos = sq_skipspace($body, $pos);
1160 $char = substr($body, $pos, 1);
1161 /**
1162 * Two things are valid here:
1163 * '=' means this is attribute type 1 2 or 3.
1164 * \w means this was attribute type 4.
1165 * anything else we ignore and re-loop. End of tag and
1166 * invalid stuff will be caught by our checks at the beginning
1167 * of the loop.
1168 */
1169 if ($char == "="){
1170 $pos++;
1171 $pos = sq_skipspace($body, $pos);
1172 /**
1173 * Here are 3 possibilities:
1174 * "'" attribute type 1
1175 * '"' attribute type 2
1176 * everything else is the content of tag type 3
1177 */
1178 $quot = substr($body, $pos, 1);
1179 if ($quot == "'"){
1180 $regary = sq_findnxreg($body, $pos+1, "\'");
1181 if ($regary == false){
1182 return Array(false, false, false, $lt, strlen($body));
1183 }
1184 list($pos, $attval, $match) = $regary;
1185 $pos++;
1186 $attary{$attname} = "'" . $attval . "'";
1187 } else if ($quot == '"'){
1188 $regary = sq_findnxreg($body, $pos+1, '\"');
1189 if ($regary == false){
1190 return Array(false, false, false, $lt, strlen($body));
1191 }
1192 list($pos, $attval, $match) = $regary;
1193 $pos++;
1194 $attary{$attname} = '"' . $attval . '"';
1195 } else {
1196 /**
1197 * These are hateful. Look for \s, or >.
1198 */
1199 $regary = sq_findnxreg($body, $pos, "[\s>]");
1200 if ($regary == false){
1201 return Array(false, false, false, $lt, strlen($body));
1202 }
1203 list($pos, $attval, $match) = $regary;
1204 /**
1205 * If it's ">" it will be caught at the top.
1206 */
1207 $attval = preg_replace("/\"/s", "&quot;", $attval);
1208 $attary{$attname} = '"' . $attval . '"';
1209 }
1210 } else if (preg_match("|[\w/>]|", $char)) {
1211 /**
1212 * That was attribute type 4.
1213 */
1214 $attary{$attname} = '"yes"';
1215 } else {
1216 /**
1217 * An illegal character. Find next '>' and return.
1218 */
1219 $gt = sq_findnxstr($body, $pos, ">");
1220 return Array(false, false, false, $lt, $gt);
1221 }
1222 break;
1223 }
1224 }
1225 /**
1226 * The fact that we got here indicates that the tag end was never
1227 * found. Return invalid tag indication so it gets stripped.
1228 */
1229 return Array(false, false, false, $lt, strlen($body));
1230 }
1231
1232 /**
1233 * This function checks attribute values for entity-encoded values
1234 * and returns them translated into 8-bit strings so we can run
1235 * checks on them.
1236 *
1237 * @param $attvalue A string to run entity check against.
1238 * @return Translated value.
1239 */
1240 function sq_deent($attvalue){
1241 $me = 'sq_deent';
1242 /**
1243 * See if we have to run the checks first. All entities must start
1244 * with "&".
1245 */
1246 if (strpos($attvalue, "&") === false){
1247 return $attvalue;
1248 }
1249 /**
1250 * Check named entities first.
1251 */
1252 $trans = get_html_translation_table(HTML_ENTITIES);
1253 /**
1254 * Leave &quot; in, as it can mess us up.
1255 */
1256 $trans = array_flip($trans);
1257 unset($trans{"&quot;"});
1258 while (list($ent, $val) = each($trans)){
1259 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1260 }
1261 /**
1262 * Now translate numbered entities from 1 to 255 if needed.
1263 */
1264 if (strpos($attvalue, "#") !== false){
1265 $omit = Array(34, 39);
1266 for ($asc=1; $asc<256; $asc++){
1267 if (!in_array($asc, $omit)){
1268 $chr = chr($asc);
1269 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1270 $attvalue);
1271 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1272 "$chr\\1", $attvalue);
1273 }
1274 }
1275 }
1276 return $attvalue;
1277 }
1278
1279 /**
1280 * This function runs various checks against the attributes.
1281 *
1282 * @param $tagname String with the name of the tag.
1283 * @param $attary Array with all tag attributes.
1284 * @param $rm_attnames See description for sq_sanitize
1285 * @param $bad_attvals See description for sq_sanitize
1286 * @param $add_attr_to_tag See description for sq_sanitize
1287 * @param $message message object
1288 * @param $id message id
1289 * @return Array with modified attributes.
1290 */
1291 function sq_fixatts($tagname,
1292 $attary,
1293 $rm_attnames,
1294 $bad_attvals,
1295 $add_attr_to_tag,
1296 $message,
1297 $id,
1298 $mailbox
1299 ){
1300 $me = 'sq_fixatts';
1301 while (list($attname, $attvalue) = each($attary)){
1302 /**
1303 * See if this attribute should be removed.
1304 */
1305 foreach ($rm_attnames as $matchtag=>$matchattrs){
1306 if (preg_match($matchtag, $tagname)){
1307 foreach ($matchattrs as $matchattr){
1308 if (preg_match($matchattr, $attname)){
1309 unset($attary{$attname});
1310 continue;
1311 }
1312 }
1313 }
1314 }
1315 /**
1316 * Remove any entities.
1317 */
1318 $attvalue = sq_deent($attvalue);
1319
1320 /**
1321 * Now let's run checks on the attvalues.
1322 * I don't expect anyone to comprehend this. If you do,
1323 * get in touch with me so I can drive to where you live and
1324 * shake your hand personally. :)
1325 */
1326 foreach ($bad_attvals as $matchtag=>$matchattrs){
1327 if (preg_match($matchtag, $tagname)){
1328 foreach ($matchattrs as $matchattr=>$valary){
1329 if (preg_match($matchattr, $attname)){
1330 /**
1331 * There are two arrays in valary.
1332 * First is matches.
1333 * Second one is replacements
1334 */
1335 list($valmatch, $valrepl) = $valary;
1336 $newvalue =
1337 preg_replace($valmatch, $valrepl, $attvalue);
1338 if ($newvalue != $attvalue){
1339 $attary{$attname} = $newvalue;
1340 }
1341 }
1342 }
1343 }
1344 }
1345 /**
1346 * Turn cid: urls into http-friendly ones.
1347 */
1348 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1349 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1350 }
1351 }
1352 /**
1353 * See if we need to append any attributes to this tag.
1354 */
1355 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1356 if (preg_match($matchtag, $tagname)){
1357 $attary = array_merge($attary, $addattary);
1358 }
1359 }
1360 return $attary;
1361 }
1362
1363 /**
1364 * This function edits the style definition to make them friendly and
1365 * usable in squirrelmail.
1366 *
1367 * @param $message the message object
1368 * @param $id the message id
1369 * @param $content a string with whatever is between <style> and </style>
1370 * @return a string with edited content.
1371 */
1372 function sq_fixstyle($body, $pos, $message, $id){
1373 global $view_unsafe_images;
1374 $me = 'sq_fixstyle';
1375 $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1376 if ($ret == FALSE){
1377 return array(FALSE, strlen($body));
1378 }
1379 $newpos = $ret[0] + strlen($ret[2]);
1380 $content = $ret[1];
1381 /**
1382 * First look for general BODY style declaration, which would be
1383 * like so:
1384 * body {background: blah-blah}
1385 * and change it to .bodyclass so we can just assign it to a <div>
1386 */
1387 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1388 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1389 /**
1390 * Fix url('blah') declarations.
1391 */
1392 $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1393 "url(\\1$secremoveimg\\2)", $content);
1394 /**
1395 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1396 * is false.
1397 */
1398 if (!$view_unsafe_images){
1399 $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1400 "url(\\1$secremoveimg\\2)", $content);
1401 }
1402
1403 /**
1404 * Fix urls that refer to cid:
1405 */
1406 while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1407 $content, $matches)){
1408 $cidurl = $matches{1};
1409 $httpurl = sq_cid2http($message, $id, $cidurl);
1410 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1411 "url($httpurl)", $content);
1412 }
1413
1414 /**
1415 * Fix stupid css declarations which lead to vulnerabilities
1416 * in IE.
1417 */
1418 $match = Array('/expression/i',
1419 '/behaviou*r/i',
1420 '/binding/i');
1421 $replace = Array('idiocy', 'idiocy', 'idiocy');
1422 $content = preg_replace($match, $replace, $content);
1423 return array($content, $newpos);
1424 }
1425
1426 /**
1427 * This function converts cid: url's into the ones that can be viewed in
1428 * the browser.
1429 *
1430 * @param $message the message object
1431 * @param $id the message id
1432 * @param $cidurl the cid: url.
1433 * @return a string with a http-friendly url
1434 */
1435 function sq_cid2http($message, $id, $cidurl, $mailbox){
1436 /**
1437 * Get rid of quotes.
1438 */
1439 $quotchar = substr($cidurl, 0, 1);
1440 $cidurl = str_replace($quotchar, "", $cidurl);
1441 $cidurl = substr(trim($cidurl), 4);
1442 $linkurl = find_ent_id($cidurl, $message);
1443 /* in case of non-save cid links $httpurl should be replaced by a sort of
1444 unsave link image */
1445 $httpurl = '';
1446 if ($linkurl) {
1447 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1448 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1449 '&amp;ent_id=' . $linkurl . $quotchar;
1450 }
1451 return $httpurl;
1452 }
1453
1454 /**
1455 * This function changes the <body> tag into a <div> tag since we
1456 * can't really have a body-within-body.
1457 *
1458 * @param $attary an array of attributes and values of <body>
1459 * @return a modified array of attributes to be set for <div>
1460 */
1461 function sq_body2div($attary){
1462 $me = 'sq_body2div';
1463 $divattary = Array('class' => "'bodyclass'");
1464 $bgcolor = '#ffffff';
1465 $text = '#000000';
1466 $styledef = '';
1467 if (is_array($attary) && sizeof($attary) > 0){
1468 foreach ($attary as $attname=>$attvalue){
1469 $quotchar = substr($attvalue, 0, 1);
1470 $attvalue = str_replace($quotchar, "", $attvalue);
1471 switch ($attname){
1472 case 'background':
1473 $styledef .= "background-image: url('$attvalue'); ";
1474 break;
1475 case 'bgcolor':
1476 $styledef .= "background-color: $attvalue; ";
1477 break;
1478 case 'text':
1479 $styledef .= "color: $attvalue; ";
1480 break;
1481 }
1482 }
1483 if (strlen($styledef) > 0){
1484 $divattary{"style"} = "\"$styledef\"";
1485 }
1486 }
1487 return $divattary;
1488 }
1489
1490 /**
1491 * This is the main function and the one you should actually be calling.
1492 * There are several variables you should be aware of an which need
1493 * special description.
1494 *
1495 * Since the description is quite lengthy, see it here:
1496 * http://www.mricon.com/html/phpfilter.html
1497 *
1498 * @param $body the string with HTML you wish to filter
1499 * @param $tag_list see description above
1500 * @param $rm_tags_with_content see description above
1501 * @param $self_closing_tags see description above
1502 * @param $force_tag_closing see description above
1503 * @param $rm_attnames see description above
1504 * @param $bad_attvals see description above
1505 * @param $add_attr_to_tag see description above
1506 * @param $message message object
1507 * @param $id message id
1508 * @return sanitized html safe to show on your pages.
1509 */
1510 function sq_sanitize($body,
1511 $tag_list,
1512 $rm_tags_with_content,
1513 $self_closing_tags,
1514 $force_tag_closing,
1515 $rm_attnames,
1516 $bad_attvals,
1517 $add_attr_to_tag,
1518 $message,
1519 $id,
1520 $mailbox
1521 ){
1522 $me = 'sq_sanitize';
1523 $rm_tags = array_shift($tag_list);
1524 /**
1525 * Normalize rm_tags and rm_tags_with_content.
1526 */
1527 @array_walk($tag_list, 'sq_casenormalize');
1528 @array_walk($rm_tags_with_content, 'sq_casenormalize');
1529 @array_walk($self_closing_tags, 'sq_casenormalize');
1530 /**
1531 * See if tag_list is of tags to remove or tags to allow.
1532 * false means remove these tags
1533 * true means allow these tags
1534 */
1535 $curpos = 0;
1536 $open_tags = Array();
1537 $trusted = "<!-- begin sanitized html -->\n";
1538 $skip_content = false;
1539 /**
1540 * Take care of netscape's stupid javascript entities like
1541 * &{alert('boo')};
1542 */
1543 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1544
1545 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1546 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1547 $free_content = substr($body, $curpos, $lt-$curpos);
1548 /**
1549 * Take care of <style>
1550 */
1551 if ($tagname == "style" && $tagtype == 1){
1552 list($free_content, $curpos) =
1553 sq_fixstyle($body, $gt+1, $message, $id);
1554 if ($free_content != FALSE){
1555 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1556 $trusted .= $free_content;
1557 $trusted .= sq_tagprint($tagname, false, 2);
1558 }
1559 continue;
1560 }
1561 if ($skip_content == false){
1562 $trusted .= $free_content;
1563 }
1564 if ($tagname != FALSE){
1565 if ($tagtype == 2){
1566 if ($skip_content == $tagname){
1567 /**
1568 * Got to the end of tag we needed to remove.
1569 */
1570 $tagname = false;
1571 $skip_content = false;
1572 } else {
1573 if ($skip_content == false){
1574 if ($tagname == "body"){
1575 $tagname = "div";
1576 } else {
1577 if (isset($open_tags{$tagname}) &&
1578 $open_tags{$tagname} > 0){
1579 $open_tags{$tagname}--;
1580 } else {
1581 $tagname = false;
1582 }
1583 }
1584 }
1585 }
1586 } else {
1587 /**
1588 * $rm_tags_with_content
1589 */
1590 if ($skip_content == false){
1591 /**
1592 * See if this is a self-closing type and change
1593 * tagtype appropriately.
1594 */
1595 if ($tagtype == 1
1596 && in_array($tagname, $self_closing_tags)){
1597 $tagtype=3;
1598 }
1599 /**
1600 * See if we should skip this tag and any content
1601 * inside it.
1602 */
1603 if ($tagtype == 1 &&
1604 in_array($tagname, $rm_tags_with_content)){
1605 $skip_content = $tagname;
1606 } else {
1607 if (($rm_tags == false
1608 && in_array($tagname, $tag_list)) ||
1609 ($rm_tags == true &&
1610 !in_array($tagname, $tag_list))){
1611 $tagname = false;
1612 } else {
1613 if ($tagtype == 1){
1614 if (isset($open_tags{$tagname})){
1615 $open_tags{$tagname}++;
1616 } else {
1617 $open_tags{$tagname}=1;
1618 }
1619 }
1620 /**
1621 * This is where we run other checks.
1622 */
1623 if (is_array($attary) && sizeof($attary) > 0){
1624 $attary = sq_fixatts($tagname,
1625 $attary,
1626 $rm_attnames,
1627 $bad_attvals,
1628 $add_attr_to_tag,
1629 $message,
1630 $id,
1631 $mailbox
1632 );
1633 }
1634 /**
1635 * Convert body into div.
1636 */
1637 if ($tagname == "body"){
1638 $tagname = "div";
1639 $attary = sq_body2div($attary, $message, $id);
1640 }
1641 }
1642 }
1643 }
1644 }
1645 if ($tagname != false && $skip_content == false){
1646 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1647 }
1648 }
1649 $curpos = $gt+1;
1650 }
1651 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1652 if ($force_tag_closing == true){
1653 foreach ($open_tags as $tagname=>$opentimes){
1654 while ($opentimes > 0){
1655 $trusted .= '</' . $tagname . '>';
1656 $opentimes--;
1657 }
1658 }
1659 $trusted .= "\n";
1660 }
1661 $trusted .= "<!-- end sanitized html -->\n";
1662 return $trusted;
1663 }
1664
1665 /**
1666 * This is a wrapper function to call html sanitizing routines.
1667 *
1668 * @param $body the body of the message
1669 * @param $id the id of the message
1670 * @return a string with html safe to display in the browser.
1671 */
1672 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1673 global $attachment_common_show_images, $view_unsafe_images,
1674 $has_unsafe_images;
1675 /**
1676 * Don't display attached images in HTML mode.
1677 */
1678 $attachment_common_show_images = false;
1679 $tag_list = Array(
1680 false,
1681 "object",
1682 "meta",
1683 "html",
1684 "head",
1685 "base",
1686 "link",
1687 "frame",
1688 "iframe"
1689 );
1690
1691 $rm_tags_with_content = Array(
1692 "script",
1693 "applet",
1694 "embed",
1695 "title"
1696 );
1697
1698 $self_closing_tags = Array(
1699 "img",
1700 "br",
1701 "hr",
1702 "input"
1703 );
1704
1705 $force_tag_closing = false;
1706
1707 $rm_attnames = Array(
1708 "/.*/" =>
1709 Array(
1710 "/target/i",
1711 "/^on.*/i",
1712 "/^dynsrc/i",
1713 "/^data.*/i",
1714 "/^lowsrc.*/i"
1715 )
1716 );
1717
1718 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1719 $bad_attvals = Array(
1720 "/.*/" =>
1721 Array(
1722 "/^src|background/i" =>
1723 Array(
1724 Array(
1725 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1726 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1727 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1728 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1729 ),
1730 Array(
1731 "\\1$secremoveimg\\2",
1732 "\\1$secremoveimg\\2",
1733 "\\1$secremoveimg\\2",
1734 "\\1$secremoveimg\\2"
1735 )
1736 ),
1737 "/^href|action/i" =>
1738 Array(
1739 Array(
1740 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1741 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1742 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1743 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1744 ),
1745 Array(
1746 "\\1#\\2",
1747 "\\1#\\2",
1748 "\\1#\\2",
1749 "\\1#\\2"
1750 )
1751 ),
1752 "/^style/i" =>
1753 Array(
1754 Array(
1755 "/expression/i",
1756 "/binding/i",
1757 "/behaviou*r/i",
1758 "|url\s*\(\s*([\'\"])\s*\.\./.*([\'\"])\s*\)|si",
1759 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1760 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1761 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si"
1762 ),
1763 Array(
1764 "idiocy",
1765 "idiocy",
1766 "idiocy",
1767 "url(\\1#\\2)",
1768 "url(\\1#\\2)",
1769 "url(\\1#\\2)",
1770 "url(\\1#\\2)"
1771 )
1772 )
1773 )
1774 );
1775 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1776 $view_unsafe_images = false;
1777 }
1778 if (!$view_unsafe_images){
1779 /**
1780 * Remove any references to http/https if view_unsafe_images set
1781 * to false.
1782 */
1783 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1784 '/^([\'\"])\s*https*:.*([\'\"])/si');
1785 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1786 "\\1$secremoveimg\\2");
1787 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1788 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1789 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1790 "url(\\1$secremoveimg\\2)");
1791 }
1792
1793 $add_attr_to_tag = Array(
1794 "/^a$/i" => Array('target'=>'"_new"')
1795 );
1796 $trusted = sq_sanitize($body,
1797 $tag_list,
1798 $rm_tags_with_content,
1799 $self_closing_tags,
1800 $force_tag_closing,
1801 $rm_attnames,
1802 $bad_attvals,
1803 $add_attr_to_tag,
1804 $message,
1805 $id,
1806 $mailbox
1807 );
1808 if (preg_match("|$secremoveimg|i", $trusted)){
1809 $has_unsafe_images = true;
1810 }
1811 return $trusted;
1812 }
1813
1814 ?>