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