4ff2542a88f499503fa98b5d14e2619c153b4ba7
[squirrelmail.git] / functions / mime.php
1 <?php
2
3 /**
4 * mime.php
5 *
6 * Copyright (c) 1999-2002 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 $msg =& new Message();
33 $res = $msg->parseStructure($read);
34 $msg = $res[0];
35 if (!is_object($msg)) {
36 include_once(SM_PATH . 'functions/display_messages.php');
37 global $color, $mailbox;
38 displayPageHeader( $color, urldecode($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 $msg->setEnt('0');
49 if (count($flags)) {
50 foreach ($flags as $flag) {
51 $char = strtoupper($flag{1});
52 switch ($char) {
53 case 'S':
54 if (strtolower($flag) == '\\seen') {
55 $msg->is_seen = true;
56 }
57 break;
58 case 'A':
59 if (strtolower($flag) == '\\answered') {
60 $msg->is_answered = true;
61 }
62 break;
63 case 'D':
64 if (strtolower($flag) == '\\deleted') {
65 $msg->is_deleted = true;
66 }
67 break;
68 case 'F':
69 if (strtolower($flag) == '\\flagged') {
70 $msg->is_flagged = true;
71 }
72 break;
73 case 'M':
74 if (strtolower($flag) == '$mdnsent') {
75 $msg->is_mdnsent = true;
76 }
77 break;
78 default:
79 break;
80 }
81 }
82 }
83 // listEntities($msg);
84 return $msg;
85 }
86
87 /* This starts the parsing of a particular structure. It is called recursively,
88 * so it can be passed different structures. It returns an object of type
89 * $message.
90 * First, it checks to see if it is a multipart message. If it is, then it
91 * handles that as it sees is necessary. If it is just a regular entity,
92 * then it parses it and adds the necessary header information (by calling out
93 * to mime_get_elements()
94 */
95
96 function mime_fetch_body($imap_stream, $id, $ent_id) {
97 global $uid_support;
98 /* Do a bit of error correction. If we couldn't find the entity id, just guess
99 * that it is the first one. That is usually the case anyway.
100 */
101 if (!$ent_id) {
102 $ent_id = 1;
103 }
104 $cmd = "FETCH $id BODY[$ent_id]";
105
106 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
107 do {
108 $topline = trim(array_shift($data));
109 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
110
111 $wholemessage = implode('', $data);
112 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
113 $ret = substr($wholemessage, 0, $regs[1]);
114 /* There is some information in the content info header that could be important
115 * in order to parse html messages. Let's get them here.
116 */
117 if ($ret{0} == '<') {
118 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
119 }
120 } else if (ereg('"([^"]*)"', $topline, $regs)) {
121 $ret = $regs[1];
122 } else {
123 global $where, $what, $mailbox, $passed_id, $startMessage;
124 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
125 if (isset($where) && isset($what)) {
126 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
127 } else {
128 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
129 }
130 $par .= '&amp;response=' . urlencode($response) .
131 '&amp;message=' . urlencode($message) .
132 '&amp;topline=' . urlencode($topline);
133
134 echo '<tt><br>' .
135 '<table width="80%"><tr>' .
136 '<tr><td colspan=2>' .
137 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
138 '</td></tr>' .
139 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
140 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
141 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
142 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
143 "</table><BR></tt></font><hr>";
144
145 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
146 array_shift($data);
147 $wholemessage = implode('', $data);
148
149 $ret = $wholemessage;
150 }
151 return $ret;
152 }
153
154 function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) {
155 global $uid_support;
156 /* Do a bit of error correction. If we couldn't find the entity id, just guess
157 * that it is the first one. That is usually the case anyway.
158 */
159 if (!$ent_id) {
160 $ent_id = 1;
161 }
162 $sid = sqimap_session_id($uid_support);
163 /* Don't kill the connection if the browser is over a dialup
164 * and it would take over 30 seconds to download it.
165 * Don´t call set_time_limit in safe mode.
166 */
167
168 if (!ini_get('safe_mode')) {
169 set_time_limit(0);
170 }
171 if ($uid_support) {
172 $sid_s = substr($sid,0,strpos($sid, ' '));
173 } else {
174 $sid_s = $sid;
175 }
176
177 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
178 echo decodeBody($body, $encoding);
179 return;
180 /*
181 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
182 $cnt = 0;
183 $continue = true;
184 $read = fgets ($imap_stream,8192);
185
186
187 // This could be bad -- if the section has sqimap_session_id() . ' OK'
188 // or similar, it will kill the download.
189 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
190 if (trim($read) == ')==') {
191 $read1 = $read;
192 $read = fgets ($imap_stream,4096);
193 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
194 return;
195 } else {
196 echo decodeBody($read1, $encoding) .
197 decodeBody($read, $encoding);
198 }
199 } else if ($cnt) {
200 echo decodeBody($read, $encoding);
201 }
202 $read = fgets ($imap_stream,4096);
203 $cnt++;
204 // break;
205 }
206 */
207 }
208
209 /* -[ END MIME DECODING ]----------------------------------------------------------- */
210
211 /* This is here for debugging purposes. It will print out a list
212 * of all the entity IDs that are in the $message object.
213 */
214 function listEntities ($message) {
215 if ($message) {
216 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
217 for ($i = 0; isset($message->entities[$i]); $i++) {
218 echo "$i : ";
219 $msg = listEntities($message->entities[$i]);
220
221 if ($msg) {
222 echo "return: ";
223 return $msg;
224 }
225 }
226 }
227 }
228
229 function getPriorityStr($priority) {
230 $priority_level = substr($priority,0,1);
231
232 switch($priority_level) {
233 /* Check for a higher then normal priority. */
234 case '1':
235 case '2':
236 $priority_string = _("High");
237 break;
238
239 /* Check for a lower then normal priority. */
240 case '4':
241 case '5':
242 $priority_string = _("Low");
243 break;
244
245 /* Check for a normal priority. */
246 case '3':
247 default:
248 $priority_level = '3';
249 $priority_string = _("Normal");
250 break;
251
252 }
253 return $priority_string;
254 }
255
256 /* returns a $message object for a particular entity id */
257 function getEntity ($message, $ent_id) {
258 return $message->getEntity($ent_id);
259 }
260
261 /* translateText
262 * Extracted from strings.php 23/03/2002
263 */
264
265 function translateText(&$body, $wrap_at, $charset) {
266 global $where, $what; /* from searching */
267 global $color; /* color theme */
268
269 require_once(SM_PATH . 'functions/url_parser.php');
270
271 $body_ary = explode("\n", $body);
272 for ($i=0; $i < count($body_ary); $i++) {
273 $line = $body_ary[$i];
274 if (strlen($line) - 2 >= $wrap_at) {
275 sqWordWrap($line, $wrap_at);
276 }
277 $line = charset_decode($charset, $line);
278 $line = str_replace("\t", ' ', $line);
279
280 parseUrl ($line);
281
282 $quotes = 0;
283 $pos = 0;
284 $j = strlen($line);
285
286 while ($pos < $j) {
287 if ($line[$pos] == ' ') {
288 $pos++;
289 } else if (strpos($line, '&gt;', $pos) === $pos) {
290 $pos += 4;
291 $quotes++;
292 } else {
293 break;
294 }
295 }
296
297 if ($quotes > 1) {
298 if (!isset($color[14])) {
299 $color[14] = '#FF0000';
300 }
301 $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
302 } elseif ($quotes) {
303 if (!isset($color[13])) {
304 $color[13] = '#800000';
305 }
306 $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
307 }
308
309 $body_ary[$i] = $line;
310 }
311 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
312 }
313
314
315 /* This returns a parsed string called $body. That string can then
316 * be displayed as the actual message in the HTML. It contains
317 * everything needed, including HTML Tags, Attachments at the
318 * bottom, etc.
319 */
320 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
321 /* This if statement checks for the entity to show as the
322 * primary message. To add more of them, just put them in the
323 * order that is their priority.
324 */
325 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
326 $show_html_default, $has_unsafe_images, $view_unsafe_images, $sort;
327
328 $has_unsafe_images= 0;
329 $body = '';
330 $urlmailbox = urlencode($mailbox);
331 $body_message = getEntity($message, $ent_num);
332 if (($body_message->header->type0 == 'text') ||
333 ($body_message->header->type0 == 'rfc822')) {
334 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
335 $body = decodeBody($body, $body_message->header->encoding);
336 $hookResults = do_hook("message_body", $body);
337 $body = $hookResults[1];
338
339 /* If there are other types that shouldn't be formatted, add
340 * them here.
341 */
342
343 if ($body_message->header->type1 == 'html') {
344 if ($show_html_default <> 1) {
345 $entity_conv = array('&nbsp;' => ' ',
346 '<p>' => "\n",
347 '<br>' => "\n",
348 '<P>' => "\n",
349 '<BR>' => "\n",
350 '&gt;' => '>',
351 '&lt;' => '<');
352 $body = strtr($body, $entity_conv);
353 $body = strip_tags($body);
354 $body = trim($body);
355 translateText($body, $wrap_at,
356 $body_message->header->getParameter('charset'));
357 } else {
358 $body = magicHTML($body, $id, $message, $mailbox);
359 }
360 } else {
361 translateText($body, $wrap_at,
362 $body_message->header->getParameter('charset'));
363 }
364
365 if ($has_unsafe_images) {
366 if ($view_unsafe_images) {
367 $untext = '">' . _("Hide Unsafe Images");
368 } else {
369 $untext = '&amp;view_unsafe_images=1">' . _("View Unsafe Images");
370 }
371 $body .= '<center><small><a href="read_body.php?passed_id=' . $id .
372 '&amp;passed_ent_id=' . $message->entity_id . '&amp;mailbox=' . $urlmailbox .
373 '&amp;sort=' . $sort . '&amp;startMessage=' . $startMessage . '&amp;show_more=0' .
374 $untext . '</a></small></center><br>' . "\n";
375 }
376 }
377 return $body;
378 }
379
380
381 function formatAttachments($message, $exclude_id, $mailbox, $id) {
382 global $where, $what, $startMessage, $color;
383 static $ShownHTML = 0;
384
385 $att_ar = $message->getAttachments($exclude_id);
386
387 if (!count($att_ar)) return '';
388
389 $attachments = '';
390
391 $urlMailbox = urlencode($mailbox);
392
393 foreach ($att_ar as $att) {
394 $ent = urldecode($att->entity_id);
395 $header = $att->header;
396 $type0 = strtolower($header->type0);
397 $type1 = strtolower($header->type1);
398 $name = '';
399 $links['download link']['text'] = _("download");
400 $links['download link']['href'] =
401 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
402 $ImageURL = '';
403 if ($type0 =='message' && $type1 == 'rfc822') {
404 $default_page = '../src/read_body.php';
405 $rfc822_header = $att->rfc822_header;
406 $filename = decodeHeader($rfc822_header->subject);
407
408 $from_o = $rfc822_header->from;
409 if (is_object($from_o)) {
410 $from_name = $from_o->getAddress(false);
411 } else {
412 $from_name = _("Unknown sender");
413 }
414 $from_name = decodeHeader(htmlspecialchars($from_name));
415 $description = $from_name;
416 } else {
417 $default_page = '../src/download.php';
418 if (is_object($header->disposition)) {
419 $filename = decodeHeader($header->disposition->getProperty('filename'));
420 if (trim($filename) == '') {
421 $name = decodeHeader($header->disposition->getProperty('name'));
422 if (trim($name) == '') {
423 if (trim( $header->id ) == '') {
424 $filename = 'untitled-[' . $ent . ']' ;
425 } else {
426 $filename = 'cid: ' . $header->id;
427 }
428 } else {
429 $filename = $name;
430 }
431 }
432 } else {
433 if (trim( $header->id ) == '') {
434 $filename = 'untitled-[' . $ent . ']' ;
435 } else {
436 $filename = 'cid: ' . $header->id;
437 }
438 }
439
440 if ($header->description) {
441 $description = htmlspecialchars($header->description);
442 } else {
443 $description = '';
444 }
445 }
446
447 $display_filename = $filename;
448 if (isset($passed_ent_id)) {
449 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
450 } else {
451 $passed_ent_id_link = '';
452 }
453 $defaultlink = $default_page . "?startMessage=$startMessage"
454 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
455 . '&amp;ent_id='.$ent.$passed_ent_id_link;
456 if ($where && $what) {
457 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
458 }
459 /* This executes the attachment hook with a specific MIME-type.
460 * If that doesn't have results, it tries if there's a rule
461 * for a more generic type.
462 */
463 $hookresults = do_hook("attachment $type0/$type1", $links,
464 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
465 $display_filename, $where, $what);
466 if(count($hookresults[1]) <= 1) {
467 $hookresults = do_hook("attachment $type0/*", $links,
468 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
469 $display_filename, $where, $what);
470 }
471
472 $links = $hookresults[1];
473 $defaultlink = $hookresults[6];
474
475 $attachments .= '<TR><TD>' .
476 "<A HREF=\"$defaultlink\">$display_filename</A>&nbsp;</TD>" .
477 '<TD><SMALL><b>' . show_readable_size($header->size) .
478 '</b>&nbsp;&nbsp;</small></TD>' .
479 "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
480 '<TD><SMALL>';
481 $attachments .= '<b>' . $description . '</b>';
482 $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
483
484 $skipspaces = 1;
485 foreach ($links as $val) {
486 if ($skipspaces) {
487 $skipspaces = 0;
488 } else {
489 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
490 }
491 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
492 }
493 unset($links);
494 $attachments .= "</TD></TR>\n";
495 }
496 return $attachments;
497 }
498
499 /* This function decodes the body depending on the encoding type. */
500 function decodeBody($body, $encoding) {
501 global $languages, $squirrelmail_language;
502 global $show_html_default;
503
504 $body = str_replace("\r\n", "\n", $body);
505 $encoding = strtolower($encoding);
506
507 if ($encoding == 'quoted-printable' ||
508 $encoding == 'quoted_printable') {
509 $body = quoted_printable_decode($body);
510
511 while (ereg("=\n", $body)) {
512 $body = ereg_replace ("=\n", '', $body);
513 }
514
515 } else if ($encoding == 'base64') {
516 $body = base64_decode($body);
517 }
518
519 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
520 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
521 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
522 }
523
524 // All other encodings are returned raw.
525 return $body;
526 }
527
528 /*
529 * This functions decode strings that is encoded according to
530 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
531 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
532 */
533 function decodeHeader ($string, $utfencode=true) {
534 global $languages, $squirrelmail_language;
535 if (is_array($string)) {
536 $string = implode("\n", $string);
537 }
538
539 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
540 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
541 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
542 }
543
544 $i = 0;
545 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
546 $string, $res)) {
547 $prefix = $res[1];
548 /* Ignore white-space between consecutive encoded-words. */
549 if (strspn($res[2], " \t") != strlen($res[2])) {
550 $prefix .= $res[2];
551 }
552
553 if (ucfirst($res[4]) == 'B') {
554 $replace = base64_decode($res[5]);
555 } else {
556 $replace = str_replace('_', ' ', $res[5]);
557 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
558 $replace);
559 /* Only encode into entities by default. Some places
560 * don't need the encoding, like the compose form.
561 */
562 if ($utfencode) {
563 $replace = charset_decode($res[3], $replace);
564 }
565 }
566 $string = $prefix . $replace . substr($string, strlen($res[0]));
567 $i = strlen($prefix) + strlen($replace);
568 }
569 return $string;
570 }
571
572 /*
573 * Encode a string according to RFC 1522 for use in headers if it
574 * contains 8-bit characters or anything that looks like it should
575 * be encoded.
576 */
577 function encodeHeader ($string) {
578 global $default_charset, $languages, $squirrelmail_language;
579
580 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
581 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
582 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
583 }
584
585 // Encode only if the string contains 8-bit characters or =?
586 $j = strlen($string);
587 $l = strstr($string, '=?'); // Must be encoded ?
588 $ret = '';
589 for($i = 0; $i < $j; ++$i) {
590 switch($string{$i}) {
591 case '=':
592 $ret .= '=3D';
593 break;
594 case '?':
595 $ret .= '=3F';
596 break;
597 case '_':
598 $ret .= '=5F';
599 break;
600 case ' ':
601 $ret .= '_';
602 break;
603 default:
604 $k = ord($string{$i});
605 if ($k > 126) {
606 $ret .= sprintf("=%02X", $k);
607 $l = TRUE;
608 } else {
609 $ret .= $string{$i};
610 }
611 break;
612 }
613 }
614
615 if ($l) {
616 $string = "=?$default_charset?Q?$ret?=";
617 }
618
619 return $string;
620 }
621
622 /* This function trys to locate the entity_id of a specific mime element */
623 function find_ent_id($id, $message) {
624 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
625 if ($message->entities[$i]->header->type0 == 'multipart') {
626 $ret = find_ent_id($id, $message->entities[$i]);
627 } else {
628 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
629 if (sq_check_save_extension($message->entities[$i])) {
630 return $message->entities[$i]->entity_id;
631 }
632 }
633 }
634 }
635 return $ret;
636 }
637
638 function sq_check_save_extension($message) {
639 $filename = $message->getFilename();
640 $ext = substr($filename, strrpos($filename,'.')+1);
641 $save_extensions = array('jpg','jpeg','gif','png','bmp');
642 return in_array($ext, $save_extensions);
643 }
644
645
646 /**
647 ** HTMLFILTER ROUTINES
648 */
649
650 /**
651 * This function returns the final tag out of the tag name, an array
652 * of attributes, and the type of the tag. This function is called by
653 * sq_sanitize internally.
654 *
655 * @param $tagname the name of the tag.
656 * @param $attary the array of attributes and their values
657 * @param $tagtype The type of the tag (see in comments).
658 * @return a string with the final tag representation.
659 */
660 function sq_tagprint($tagname, $attary, $tagtype){
661 $me = 'sq_tagprint';
662
663 if ($tagtype == 2){
664 $fulltag = '</' . $tagname . '>';
665 } else {
666 $fulltag = '<' . $tagname;
667 if (is_array($attary) && sizeof($attary)){
668 $atts = Array();
669 while (list($attname, $attvalue) = each($attary)){
670 array_push($atts, "$attname=$attvalue");
671 }
672 $fulltag .= ' ' . join(" ", $atts);
673 }
674 if ($tagtype == 3){
675 $fulltag .= ' /';
676 }
677 $fulltag .= '>';
678 }
679 return $fulltag;
680 }
681
682 /**
683 * A small helper function to use with array_walk. Modifies a by-ref
684 * value and makes it lowercase.
685 *
686 * @param $val a value passed by-ref.
687 * @return void since it modifies a by-ref value.
688 */
689 function sq_casenormalize(&$val){
690 $val = strtolower($val);
691 }
692
693 /**
694 * This function skips any whitespace from the current position within
695 * a string and to the next non-whitespace value.
696 *
697 * @param $body the string
698 * @param $offset the offset within the string where we should start
699 * looking for the next non-whitespace character.
700 * @return the location within the $body where the next
701 * non-whitespace char is located.
702 */
703 function sq_skipspace($body, $offset){
704 $me = 'sq_skipspace';
705 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
706 if (sizeof($matches{1})){
707 $count = strlen($matches{1});
708 $offset += $count;
709 }
710 return $offset;
711 }
712
713 /**
714 * This function looks for the next character within a string. It's
715 * really just a glorified "strpos", except it catches if failures
716 * nicely.
717 *
718 * @param $body The string to look for needle in.
719 * @param $offset Start looking from this position.
720 * @param $needle The character/string to look for.
721 * @return location of the next occurance of the needle, or
722 * strlen($body) if needle wasn't found.
723 */
724 function sq_findnxstr($body, $offset, $needle){
725 $me = 'sq_findnxstr';
726 $pos = strpos($body, $needle, $offset);
727 if ($pos === FALSE){
728 $pos = strlen($body);
729 }
730 return $pos;
731 }
732
733 /**
734 * This function takes a PCRE-style regexp and tries to match it
735 * within the string.
736 *
737 * @param $body The string to look for needle in.
738 * @param $offset Start looking from here.
739 * @param $reg A PCRE-style regex to match.
740 * @return Returns a false if no matches found, or an array
741 * with the following members:
742 * - integer with the location of the match within $body
743 * - string with whatever content between offset and the match
744 * - string with whatever it is we matched
745 */
746 function sq_findnxreg($body, $offset, $reg){
747 $me = 'sq_findnxreg';
748 $matches = Array();
749 $retarr = Array();
750 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
751 if (!$matches{0}){
752 $retarr = false;
753 } else {
754 $retarr{0} = $offset + strlen($matches{1});
755 $retarr{1} = $matches{1};
756 $retarr{2} = $matches{2};
757 }
758 return $retarr;
759 }
760
761 /**
762 * This function looks for the next tag.
763 *
764 * @param $body String where to look for the next tag.
765 * @param $offset Start looking from here.
766 * @return false if no more tags exist in the body, or
767 * an array with the following members:
768 * - string with the name of the tag
769 * - array with attributes and their values
770 * - integer with tag type (1, 2, or 3)
771 * - integer where the tag starts (starting "<")
772 * - integer where the tag ends (ending ">")
773 * first three members will be false, if the tag is invalid.
774 */
775 function sq_getnxtag($body, $offset){
776 $me = 'sq_getnxtag';
777 if ($offset > strlen($body)){
778 return false;
779 }
780 $lt = sq_findnxstr($body, $offset, "<");
781 if ($lt == strlen($body)){
782 return false;
783 }
784 /**
785 * We are here:
786 * blah blah <tag attribute="value">
787 * \---------^
788 */
789 $pos = sq_skipspace($body, $lt+1);
790 if ($pos >= strlen($body)){
791 return Array(false, false, false, $lt, strlen($body));
792 }
793 /**
794 * There are 3 kinds of tags:
795 * 1. Opening tag, e.g.:
796 * <a href="blah">
797 * 2. Closing tag, e.g.:
798 * </a>
799 * 3. XHTML-style content-less tag, e.g.:
800 * <img src="blah"/>
801 */
802 $tagtype = false;
803 switch (substr($body, $pos, 1)){
804 case '/':
805 $tagtype = 2;
806 $pos++;
807 break;
808 case '!':
809 /**
810 * A comment or an SGML declaration.
811 */
812 if (substr($body, $pos+1, 2) == "--"){
813 $gt = strpos($body, "-->", $pos);
814 if ($gt === false){
815 $gt = strlen($body);
816 } else {
817 $gt += 2;
818 }
819 return Array(false, false, false, $lt, $gt);
820 } else {
821 $gt = sq_findnxstr($body, $pos, ">");
822 return Array(false, false, false, $lt, $gt);
823 }
824 break;
825 default:
826 /**
827 * Assume tagtype 1 for now. If it's type 3, we'll switch values
828 * later.
829 */
830 $tagtype = 1;
831 break;
832 }
833
834 $tag_start = $pos;
835 $tagname = '';
836 /**
837 * Look for next [\W-_], which will indicate the end of the tag name.
838 */
839 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
840 if ($regary == false){
841 return Array(false, false, false, $lt, strlen($body));
842 }
843 list($pos, $tagname, $match) = $regary;
844 $tagname = strtolower($tagname);
845
846 /**
847 * $match can be either of these:
848 * '>' indicating the end of the tag entirely.
849 * '\s' indicating the end of the tag name.
850 * '/' indicating that this is type-3 xhtml tag.
851 *
852 * Whatever else we find there indicates an invalid tag.
853 */
854 switch ($match){
855 case '/':
856 /**
857 * This is an xhtml-style tag with a closing / at the
858 * end, like so: <img src="blah"/>. Check if it's followed
859 * by the closing bracket. If not, then this tag is invalid
860 */
861 if (substr($body, $pos, 2) == "/>"){
862 $pos++;
863 $tagtype = 3;
864 } else {
865 $gt = sq_findnxstr($body, $pos, ">");
866 $retary = Array(false, false, false, $lt, $gt);
867 return $retary;
868 }
869 case '>':
870 return Array($tagname, false, $tagtype, $lt, $pos);
871 break;
872 default:
873 /**
874 * Check if it's whitespace
875 */
876 if (!preg_match('/\s/', $match)){
877 /**
878 * This is an invalid tag! Look for the next closing ">".
879 */
880 $gt = sq_findnxstr($body, $offset, ">");
881 return Array(false, false, false, $lt, $gt);
882 }
883 break;
884 }
885
886 /**
887 * At this point we're here:
888 * <tagname attribute='blah'>
889 * \-------^
890 *
891 * At this point we loop in order to find all attributes.
892 */
893 $attname = '';
894 $atttype = false;
895 $attary = Array();
896
897 while ($pos <= strlen($body)){
898 $pos = sq_skipspace($body, $pos);
899 if ($pos == strlen($body)){
900 /**
901 * Non-closed tag.
902 */
903 return Array(false, false, false, $lt, $pos);
904 }
905 /**
906 * See if we arrived at a ">" or "/>", which means that we reached
907 * the end of the tag.
908 */
909 $matches = Array();
910 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
911 /**
912 * Yep. So we did.
913 */
914 $pos += strlen($matches{1});
915 if ($matches{2} == "/>"){
916 $tagtype = 3;
917 $pos++;
918 }
919 return Array($tagname, $attary, $tagtype, $lt, $pos);
920 }
921
922 /**
923 * There are several types of attributes, with optional
924 * [:space:] between members.
925 * Type 1:
926 * attrname[:space:]=[:space:]'CDATA'
927 * Type 2:
928 * attrname[:space:]=[:space:]"CDATA"
929 * Type 3:
930 * attr[:space:]=[:space:]CDATA
931 * Type 4:
932 * attrname
933 *
934 * We leave types 1 and 2 the same, type 3 we check for
935 * '"' and convert to "&quot" if needed, then wrap in
936 * double quotes. Type 4 we convert into:
937 * attrname="yes".
938 */
939 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
940 if ($regary == false){
941 /**
942 * Looks like body ended before the end of tag.
943 */
944 return Array(false, false, false, $lt, strlen($body));
945 }
946 list($pos, $attname, $match) = $regary;
947 $attname = strtolower($attname);
948 /**
949 * We arrived at the end of attribute name. Several things possible
950 * here:
951 * '>' means the end of the tag and this is attribute type 4
952 * '/' if followed by '>' means the same thing as above
953 * '\s' means a lot of things -- look what it's followed by.
954 * anything else means the attribute is invalid.
955 */
956 switch($match){
957 case '/':
958 /**
959 * This is an xhtml-style tag with a closing / at the
960 * end, like so: <img src="blah"/>. Check if it's followed
961 * by the closing bracket. If not, then this tag is invalid
962 */
963 if (substr($body, $pos, 2) == "/>"){
964 $pos++;
965 $tagtype = 3;
966 } else {
967 $gt = sq_findnxstr($body, $pos, ">");
968 $retary = Array(false, false, false, $lt, $gt);
969 return $retary;
970 }
971 case '>':
972 $attary{$attname} = '"yes"';
973 return Array($tagname, $attary, $tagtype, $lt, $pos);
974 break;
975 default:
976 /**
977 * Skip whitespace and see what we arrive at.
978 */
979 $pos = sq_skipspace($body, $pos);
980 $char = substr($body, $pos, 1);
981 /**
982 * Two things are valid here:
983 * '=' means this is attribute type 1 2 or 3.
984 * \w means this was attribute type 4.
985 * anything else we ignore and re-loop. End of tag and
986 * invalid stuff will be caught by our checks at the beginning
987 * of the loop.
988 */
989 if ($char == "="){
990 $pos++;
991 $pos = sq_skipspace($body, $pos);
992 /**
993 * Here are 3 possibilities:
994 * "'" attribute type 1
995 * '"' attribute type 2
996 * everything else is the content of tag type 3
997 */
998 $quot = substr($body, $pos, 1);
999 if ($quot == "'"){
1000 $regary = sq_findnxreg($body, $pos+1, "\'");
1001 if ($regary == false){
1002 return Array(false, false, false, $lt, strlen($body));
1003 }
1004 list($pos, $attval, $match) = $regary;
1005 $pos++;
1006 $attary{$attname} = "'" . $attval . "'";
1007 } else if ($quot == '"'){
1008 $regary = sq_findnxreg($body, $pos+1, '\"');
1009 if ($regary == false){
1010 return Array(false, false, false, $lt, strlen($body));
1011 }
1012 list($pos, $attval, $match) = $regary;
1013 $pos++;
1014 $attary{$attname} = '"' . $attval . '"';
1015 } else {
1016 /**
1017 * These are hateful. Look for \s, or >.
1018 */
1019 $regary = sq_findnxreg($body, $pos, "[\s>]");
1020 if ($regary == false){
1021 return Array(false, false, false, $lt, strlen($body));
1022 }
1023 list($pos, $attval, $match) = $regary;
1024 /**
1025 * If it's ">" it will be caught at the top.
1026 */
1027 $attval = preg_replace("/\"/s", "&quot;", $attval);
1028 $attary{$attname} = '"' . $attval . '"';
1029 }
1030 } else if (preg_match("|[\w/>]|", $char)) {
1031 /**
1032 * That was attribute type 4.
1033 */
1034 $attary{$attname} = '"yes"';
1035 } else {
1036 /**
1037 * An illegal character. Find next '>' and return.
1038 */
1039 $gt = sq_findnxstr($body, $pos, ">");
1040 return Array(false, false, false, $lt, $gt);
1041 }
1042 break;
1043 }
1044 }
1045 /**
1046 * The fact that we got here indicates that the tag end was never
1047 * found. Return invalid tag indication so it gets stripped.
1048 */
1049 return Array(false, false, false, $lt, strlen($body));
1050 }
1051
1052 /**
1053 * This function checks attribute values for entity-encoded values
1054 * and returns them translated into 8-bit strings so we can run
1055 * checks on them.
1056 *
1057 * @param $attvalue A string to run entity check against.
1058 * @return Translated value.
1059 */
1060 function sq_deent($attvalue){
1061 $me = 'sq_deent';
1062 /**
1063 * See if we have to run the checks first. All entities must start
1064 * with "&".
1065 */
1066 if (strpos($attvalue, "&") === false){
1067 return $attvalue;
1068 }
1069 /**
1070 * Check named entities first.
1071 */
1072 $trans = get_html_translation_table(HTML_ENTITIES);
1073 /**
1074 * Leave &quot; in, as it can mess us up.
1075 */
1076 $trans = array_flip($trans);
1077 unset($trans{"&quot;"});
1078 while (list($ent, $val) = each($trans)){
1079 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1080 }
1081 /**
1082 * Now translate numbered entities from 1 to 255 if needed.
1083 */
1084 if (strpos($attvalue, "#") !== false){
1085 $omit = Array(34, 39);
1086 for ($asc=1; $asc<256; $asc++){
1087 if (!in_array($asc, $omit)){
1088 $chr = chr($asc);
1089 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1090 $attvalue);
1091 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1092 "$chr\\1", $attvalue);
1093 }
1094 }
1095 }
1096 return $attvalue;
1097 }
1098
1099 /**
1100 * This function runs various checks against the attributes.
1101 *
1102 * @param $tagname String with the name of the tag.
1103 * @param $attary Array with all tag attributes.
1104 * @param $rm_attnames See description for sq_sanitize
1105 * @param $bad_attvals See description for sq_sanitize
1106 * @param $add_attr_to_tag See description for sq_sanitize
1107 * @param $message message object
1108 * @param $id message id
1109 * @return Array with modified attributes.
1110 */
1111 function sq_fixatts($tagname,
1112 $attary,
1113 $rm_attnames,
1114 $bad_attvals,
1115 $add_attr_to_tag,
1116 $message,
1117 $id,
1118 $mailbox
1119 ){
1120 $me = 'sq_fixatts';
1121 while (list($attname, $attvalue) = each($attary)){
1122 /**
1123 * See if this attribute should be removed.
1124 */
1125 foreach ($rm_attnames as $matchtag=>$matchattrs){
1126 if (preg_match($matchtag, $tagname)){
1127 foreach ($matchattrs as $matchattr){
1128 if (preg_match($matchattr, $attname)){
1129 unset($attary{$attname});
1130 continue;
1131 }
1132 }
1133 }
1134 }
1135 /**
1136 * Remove any entities.
1137 */
1138 $attvalue = sq_deent($attvalue);
1139
1140 /**
1141 * Now let's run checks on the attvalues.
1142 * I don't expect anyone to comprehend this. If you do,
1143 * get in touch with me so I can drive to where you live and
1144 * shake your hand personally. :)
1145 */
1146 foreach ($bad_attvals as $matchtag=>$matchattrs){
1147 if (preg_match($matchtag, $tagname)){
1148 foreach ($matchattrs as $matchattr=>$valary){
1149 if (preg_match($matchattr, $attname)){
1150 /**
1151 * There are two arrays in valary.
1152 * First is matches.
1153 * Second one is replacements
1154 */
1155 list($valmatch, $valrepl) = $valary;
1156 $newvalue =
1157 preg_replace($valmatch, $valrepl, $attvalue);
1158 if ($newvalue != $attvalue){
1159 $attary{$attname} = $newvalue;
1160 }
1161 }
1162 }
1163 }
1164 }
1165 /**
1166 * Turn cid: urls into http-friendly ones.
1167 */
1168 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1169 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1170 }
1171 }
1172 /**
1173 * See if we need to append any attributes to this tag.
1174 */
1175 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1176 if (preg_match($matchtag, $tagname)){
1177 $attary = array_merge($attary, $addattary);
1178 }
1179 }
1180 return $attary;
1181 }
1182
1183 /**
1184 * This function edits the style definition to make them friendly and
1185 * usable in squirrelmail.
1186 *
1187 * @param $message the message object
1188 * @param $id the message id
1189 * @param $content a string with whatever is between <style> and </style>
1190 * @return a string with edited content.
1191 */
1192 function sq_fixstyle($message, $id, $content){
1193 global $view_unsafe_images;
1194 $me = 'sq_fixstyle';
1195 /**
1196 * First look for general BODY style declaration, which would be
1197 * like so:
1198 * body {background: blah-blah}
1199 * and change it to .bodyclass so we can just assign it to a <div>
1200 */
1201 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1202 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1203 /**
1204 * Fix url('blah') declarations.
1205 */
1206 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1207 "url(\\1$secremoveimg\\2)", $content);
1208 /**
1209 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1210 * is false.
1211 */
1212 if (!$view_unsafe_images){
1213 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1214 "url(\\1$secremoveimg\\2)", $content);
1215 }
1216
1217 /**
1218 * Fix urls that refer to cid:
1219 */
1220 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1221 $matches)){
1222 $cidurl = $matches{1};
1223 $httpurl = sq_cid2http($message, $id, $cidurl);
1224 $content = preg_replace("|url\($cidurl\)|si",
1225 "url($httpurl)", $content);
1226 }
1227
1228 /**
1229 * Fix stupid css declarations which lead to vulnerabilities
1230 * in IE.
1231 */
1232 $match = Array('/expression/si',
1233 '/behaviou*r/si',
1234 '/binding/si');
1235 $replace = Array('idiocy', 'idiocy', 'idiocy');
1236 $content = preg_replace($match, $replace, $content);
1237 return $content;
1238 }
1239
1240 /**
1241 * This function converts cid: url's into the ones that can be viewed in
1242 * the browser.
1243 *
1244 * @param $message the message object
1245 * @param $id the message id
1246 * @param $cidurl the cid: url.
1247 * @return a string with a http-friendly url
1248 */
1249 function sq_cid2http($message, $id, $cidurl, $mailbox){
1250 /**
1251 * Get rid of quotes.
1252 */
1253 $quotchar = substr($cidurl, 0, 1);
1254 $cidurl = str_replace($quotchar, "", $cidurl);
1255 $cidurl = substr(trim($cidurl), 4);
1256 $linkurl = find_ent_id($cidurl, $message);
1257 /* in case of non-save cid links $httpurl should be replaced by a sort of
1258 unsave link image */
1259 $httpurl = '';
1260 if ($linkurl) {
1261 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&amp;' .
1262 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1263 '&amp;ent_id=' . $linkurl . $quotchar;
1264 }
1265 return $httpurl;
1266 }
1267
1268 /**
1269 * This function changes the <body> tag into a <div> tag since we
1270 * can't really have a body-within-body.
1271 *
1272 * @param $attary an array of attributes and values of <body>
1273 * @return a modified array of attributes to be set for <div>
1274 */
1275 function sq_body2div($attary){
1276 $me = 'sq_body2div';
1277 $divattary = Array('class' => "'bodyclass'");
1278 $bgcolor = '#ffffff';
1279 $text = '#000000';
1280 $styledef = '';
1281 if (is_array($attary) && sizeof($attary) > 0){
1282 foreach ($attary as $attname=>$attvalue){
1283 $quotchar = substr($attvalue, 0, 1);
1284 $attvalue = str_replace($quotchar, "", $attvalue);
1285 switch ($attname){
1286 case 'background':
1287 $styledef .= "background-image: url('$attvalue'); ";
1288 break;
1289 case 'bgcolor':
1290 $styledef .= "background-color: $attvalue; ";
1291 break;
1292 case 'text':
1293 $styledef .= "color: $attvalue; ";
1294 break;
1295 }
1296 }
1297 if (strlen($styledef) > 0){
1298 $divattary{"style"} = "\"$styledef\"";
1299 }
1300 }
1301 return $divattary;
1302 }
1303
1304 /**
1305 * This is the main function and the one you should actually be calling.
1306 * There are several variables you should be aware of an which need
1307 * special description.
1308 *
1309 * Since the description is quite lengthy, see it here:
1310 * http://www.mricon.com/html/phpfilter.html
1311 *
1312 * @param $body the string with HTML you wish to filter
1313 * @param $tag_list see description above
1314 * @param $rm_tags_with_content see description above
1315 * @param $self_closing_tags see description above
1316 * @param $force_tag_closing see description above
1317 * @param $rm_attnames see description above
1318 * @param $bad_attvals see description above
1319 * @param $add_attr_to_tag see description above
1320 * @param $message message object
1321 * @param $id message id
1322 * @return sanitized html safe to show on your pages.
1323 */
1324 function sq_sanitize($body,
1325 $tag_list,
1326 $rm_tags_with_content,
1327 $self_closing_tags,
1328 $force_tag_closing,
1329 $rm_attnames,
1330 $bad_attvals,
1331 $add_attr_to_tag,
1332 $message,
1333 $id,
1334 $mailbox
1335 ){
1336 $me = 'sq_sanitize';
1337 /**
1338 * Normalize rm_tags and rm_tags_with_content.
1339 */
1340 @array_walk($rm_tags, 'sq_casenormalize');
1341 @array_walk($rm_tags_with_content, 'sq_casenormalize');
1342 @array_walk($self_closing_tags, 'sq_casenormalize');
1343 /**
1344 * See if tag_list is of tags to remove or tags to allow.
1345 * false means remove these tags
1346 * true means allow these tags
1347 */
1348 $rm_tags = array_shift($tag_list);
1349 $curpos = 0;
1350 $open_tags = Array();
1351 $trusted = "<!-- begin sanitized html -->\n";
1352 $skip_content = false;
1353 /**
1354 * Take care of netscape's stupid javascript entities like
1355 * &{alert('boo')};
1356 */
1357 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1358
1359 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1360 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1361 $free_content = substr($body, $curpos, $lt-$curpos);
1362 /**
1363 * Take care of <style>
1364 */
1365 if ($tagname == "style" && $tagtype == 2){
1366 /**
1367 * This is a closing </style>. Edit the
1368 * content before we apply it.
1369 */
1370 $free_content = sq_fixstyle($message, $id, $free_content);
1371 }
1372 if ($skip_content == false){
1373 $trusted .= $free_content;
1374 }
1375 if ($tagname != FALSE){
1376 if ($tagtype == 2){
1377 if ($skip_content == $tagname){
1378 /**
1379 * Got to the end of tag we needed to remove.
1380 */
1381 $tagname = false;
1382 $skip_content = false;
1383 } else {
1384 if ($skip_content == false){
1385 if ($tagname == "body"){
1386 $tagname = "div";
1387 } else {
1388 if (isset($open_tags{$tagname}) &&
1389 $open_tags{$tagname} > 0){
1390 $open_tags{$tagname}--;
1391 } else {
1392 $tagname = false;
1393 }
1394 }
1395 }
1396 }
1397 } else {
1398 /**
1399 * $rm_tags_with_content
1400 */
1401 if ($skip_content == false){
1402 /**
1403 * See if this is a self-closing type and change
1404 * tagtype appropriately.
1405 */
1406 if ($tagtype == 1
1407 && in_array($tagname, $self_closing_tags)){
1408 $tagtype=3;
1409 }
1410 /**
1411 * See if we should skip this tag and any content
1412 * inside it.
1413 */
1414 if ($tagtype == 1 &&
1415 in_array($tagname, $rm_tags_with_content)){
1416 $skip_content = $tagname;
1417 } else {
1418 if (($rm_tags == false
1419 && in_array($tagname, $tag_list)) ||
1420 ($rm_tags == true &&
1421 !in_array($tagname, $tag_list))){
1422 $tagname = false;
1423 } else {
1424 if ($tagtype == 1){
1425 if (isset($open_tags{$tagname})){
1426 $open_tags{$tagname}++;
1427 } else {
1428 $open_tags{$tagname}=1;
1429 }
1430 }
1431 /**
1432 * This is where we run other checks.
1433 */
1434 if (is_array($attary) && sizeof($attary) > 0){
1435 $attary = sq_fixatts($tagname,
1436 $attary,
1437 $rm_attnames,
1438 $bad_attvals,
1439 $add_attr_to_tag,
1440 $message,
1441 $id,
1442 $mailbox
1443 );
1444 }
1445 /**
1446 * Convert body into div.
1447 */
1448 if ($tagname == "body"){
1449 $tagname = "div";
1450 $attary = sq_body2div($attary, $message, $id);
1451 }
1452 }
1453 }
1454 }
1455 }
1456 if ($tagname != false && $skip_content == false){
1457 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1458 }
1459 }
1460 $curpos = $gt+1;
1461 }
1462 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1463 if ($force_tag_closing == true){
1464 foreach ($open_tags as $tagname=>$opentimes){
1465 while ($opentimes > 0){
1466 $trusted .= '</' . $tagname . '>';
1467 $opentimes--;
1468 }
1469 }
1470 $trusted .= "\n";
1471 }
1472 $trusted .= "<!-- end sanitized html -->\n";
1473 return $trusted;
1474 }
1475
1476 /**
1477 * This is a wrapper function to call html sanitizing routines.
1478 *
1479 * @param $body the body of the message
1480 * @param $id the id of the message
1481 * @return a string with html safe to display in the browser.
1482 */
1483 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1484 global $attachment_common_show_images, $view_unsafe_images,
1485 $has_unsafe_images;
1486 /**
1487 * Don't display attached images in HTML mode.
1488 */
1489 $attachment_common_show_images = false;
1490 $tag_list = Array(
1491 false,
1492 "object",
1493 "meta",
1494 "html",
1495 "head",
1496 "base",
1497 "link",
1498 "frame",
1499 "iframe"
1500 );
1501
1502 $rm_tags_with_content = Array(
1503 "script",
1504 "applet",
1505 "embed",
1506 "title"
1507 );
1508
1509 $self_closing_tags = Array(
1510 "img",
1511 "br",
1512 "hr",
1513 "input"
1514 );
1515
1516 $force_tag_closing = false;
1517
1518 $rm_attnames = Array(
1519 "/.*/" =>
1520 Array(
1521 "/target/si",
1522 "/^on.*/si",
1523 "/^dynsrc/si",
1524 "/^data.*/si"
1525 )
1526 );
1527
1528 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1529 $bad_attvals = Array(
1530 "/.*/" =>
1531 Array(
1532 "/^src|background/i" =>
1533 Array(
1534 Array(
1535 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1536 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1537 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1538 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1539 ),
1540 Array(
1541 "\\1$secremoveimg\\2",
1542 "\\1$secremoveimg\\2",
1543 "\\1$secremoveimg\\2",
1544 "\\1$secremoveimg\\2"
1545 )
1546 ),
1547 "/^href|action/i" =>
1548 Array(
1549 Array(
1550 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1551 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1552 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1553 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1554 ),
1555 Array(
1556 "\\1#\\2",
1557 "\\1#\\2",
1558 "\\1#\\2",
1559 "\\1#\\2"
1560 )
1561 ),
1562 "/^style/si" =>
1563 Array(
1564 Array(
1565 "/expression/si",
1566 "/binding/si",
1567 "/behaviou*r/si",
1568 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1569 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1570 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1571 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1572 ),
1573 Array(
1574 "idiocy",
1575 "idiocy",
1576 "idiocy",
1577 "url(\\1#\\2)",
1578 "url(\\1#\\2)",
1579 "url(\\1#\\2)",
1580 "url(\\1#\\2)"
1581 )
1582 )
1583 )
1584 );
1585 if (!$view_unsafe_images){
1586 /**
1587 * Remove any references to http/https if view_unsafe_images set
1588 * to false.
1589 */
1590 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1591 '/^([\'\"])\s*https*:.*([\'\"])/si');
1592 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1593 "\\1$secremoveimg\\2");
1594 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1595 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1596 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1597 "url(\\1$secremoveimg\\2)");
1598 }
1599
1600 $add_attr_to_tag = Array(
1601 "/^a$/si" => Array('target'=>'"_new"')
1602 );
1603 $trusted = sq_sanitize($body,
1604 $tag_list,
1605 $rm_tags_with_content,
1606 $self_closing_tags,
1607 $force_tag_closing,
1608 $rm_attnames,
1609 $bad_attvals,
1610 $add_attr_to_tag,
1611 $message,
1612 $id,
1613 $mailbox
1614 );
1615 if (preg_match("|$secremoveimg|si", $trusted)){
1616 $has_unsafe_images = true;
1617 }
1618 return $trusted;
1619 }
1620
1621 ?>