Swap colors, because new %2 test by Jon is true when the number is NOT
[squirrelmail.git] / functions / mime.php
1 <?php
2
3 /**
4 * mime.php
5 *
6 * Copyright (c) 1999-2003 The SquirrelMail Project Team
7 * Licensed under the GNU GPL. For full terms see the file COPYING.
8 *
9 * This contains the functions necessary to detect and decode MIME
10 * messages.
11 *
12 * $Id$
13 */
14
15 require_once(SM_PATH . 'functions/imap.php');
16 require_once(SM_PATH . 'functions/attachment_common.php');
17
18 /* --------------------------------------------------------------------------------- */
19 /* MIME DECODING */
20 /* --------------------------------------------------------------------------------- */
21
22 /* This function gets the structure of a message and stores it in the "message" class.
23 * It will return this object for use with all relevant header information and
24 * fully parsed into the standard "message" object format.
25 */
26
27 function mime_structure ($bodystructure, $flags=array()) {
28
29 /* Isolate the body structure and remove beginning and end parenthesis. */
30 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
31 $read = trim(substr ($read, 0, -1));
32 $i = 0;
33 $msg = Message::parseStructure($read,$i);
34 if (!is_object($msg)) {
35 include_once(SM_PATH . 'functions/display_messages.php');
36 global $color, $mailbox;
37 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, $color );
44 echo '</body></html>';
45 exit;
46 }
47 if (count($flags)) {
48 foreach ($flags as $flag) {
49 $char = strtoupper($flag{1});
50 switch ($char) {
51 case 'S':
52 if (strtolower($flag) == '\\seen') {
53 $msg->is_seen = true;
54 }
55 break;
56 case 'A':
57 if (strtolower($flag) == '\\answered') {
58 $msg->is_answered = true;
59 }
60 break;
61 case 'D':
62 if (strtolower($flag) == '\\deleted') {
63 $msg->is_deleted = true;
64 }
65 break;
66 case 'F':
67 if (strtolower($flag) == '\\flagged') {
68 $msg->is_flagged = true;
69 }
70 break;
71 case 'M':
72 if (strtolower($flag) == '$mdnsent') {
73 $msg->is_mdnsent = true;
74 }
75 break;
76 default:
77 break;
78 }
79 }
80 }
81 // listEntities($msg);
82 return $msg;
83 }
84
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=1) {
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 $cmd = "FETCH $id BODY[]";
103 } else {
104 $cmd = "FETCH $id BODY[$ent_id]";
105 }
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=1, $encoding) {
156 global $uid_support;
157
158 $sid = sqimap_session_id($uid_support);
159 /* Don't kill the connection if the browser is over a dialup
160 * and it would take over 30 seconds to download it.
161 * DonĀ“t call set_time_limit in safe mode.
162 */
163
164 if (!ini_get('safe_mode')) {
165 set_time_limit(0);
166 }
167 if ($uid_support) {
168 $sid_s = substr($sid,0,strpos($sid, ' '));
169 } else {
170 $sid_s = $sid;
171 }
172
173 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
174 echo decodeBody($body, $encoding);
175 return;
176 /*
177 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
178 $cnt = 0;
179 $continue = true;
180 $read = fgets ($imap_stream,8192);
181
182
183 // This could be bad -- if the section has sqimap_session_id() . ' OK'
184 // or similar, it will kill the download.
185 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
186 if (trim($read) == ')==') {
187 $read1 = $read;
188 $read = fgets ($imap_stream,4096);
189 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
190 return;
191 } else {
192 echo decodeBody($read1, $encoding) .
193 decodeBody($read, $encoding);
194 }
195 } else if ($cnt) {
196 echo decodeBody($read, $encoding);
197 }
198 $read = fgets ($imap_stream,4096);
199 $cnt++;
200 // break;
201 }
202 */
203 }
204
205 /* -[ END MIME DECODING ]----------------------------------------------------------- */
206
207 /* This is here for debugging purposes. It will print out a list
208 * of all the entity IDs that are in the $message object.
209 */
210 function listEntities ($message) {
211 if ($message) {
212 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
213 for ($i = 0; isset($message->entities[$i]); $i++) {
214 echo "$i : ";
215 $msg = listEntities($message->entities[$i]);
216
217 if ($msg) {
218 echo "return: ";
219 return $msg;
220 }
221 }
222 }
223 }
224
225 function getPriorityStr($priority) {
226 $priority_level = substr($priority,0,1);
227
228 switch($priority_level) {
229 /* Check for a higher then normal priority. */
230 case '1':
231 case '2':
232 $priority_string = _("High");
233 break;
234
235 /* Check for a lower then normal priority. */
236 case '4':
237 case '5':
238 $priority_string = _("Low");
239 break;
240
241 /* Check for a normal priority. */
242 case '3':
243 default:
244 $priority_level = '3';
245 $priority_string = _("Normal");
246 break;
247
248 }
249 return $priority_string;
250 }
251
252 /* returns a $message object for a particular entity id */
253 function getEntity ($message, $ent_id) {
254 return $message->getEntity($ent_id);
255 }
256
257 /* translateText
258 * Extracted from strings.php 23/03/2002
259 */
260
261 function translateText(&$body, $wrap_at, $charset) {
262 global $where, $what; /* from searching */
263 global $color; /* color theme */
264
265 require_once(SM_PATH . 'functions/url_parser.php');
266
267 $body_ary = explode("\n", $body);
268 for ($i=0; $i < count($body_ary); $i++) {
269 $line = $body_ary[$i];
270 if (strlen($line) - 2 >= $wrap_at) {
271 sqWordWrap($line, $wrap_at);
272 }
273 $line = charset_decode($charset, $line);
274 $line = htmlspecialchars($line);
275 $line = str_replace("\t", ' ', $line);
276
277 parseUrl ($line);
278
279 $quotes = 0;
280 $pos = 0;
281 $j = strlen($line);
282
283 while ($pos < $j) {
284 if ($line[$pos] == ' ') {
285 $pos++;
286 } else if (strpos($line, '&gt;', $pos) === $pos) {
287 $pos += 4;
288 $quotes++;
289 } else {
290 break;
291 }
292 }
293
294 if ($quotes % 2) {
295 if (!isset($color[13])) {
296 $color[13] = '#800000';
297 }
298 $line = '<font color="' . $color[13] . '">' . $line . '</font>';
299 } elseif ($quotes) {
300 if (!isset($color[14])) {
301 $color[14] = '#FF0000';
302 }
303 $line = '<font color="' . $color[14] . '">' . $line . '</font>';
304 }
305
306 $body_ary[$i] = $line;
307 }
308 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
309 }
310
311
312 /* This returns a parsed string called $body. That string can then
313 * be displayed as the actual message in the HTML. It contains
314 * everything needed, including HTML Tags, Attachments at the
315 * bottom, etc.
316 */
317 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
318 /* This if statement checks for the entity to show as the
319 * primary message. To add more of them, just put them in the
320 * order that is their priority.
321 */
322 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
323 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
324
325 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
326 $view_unsafe_images = false;
327 }
328
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 $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
365 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
366 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
367 if (isset($passed_ent_id)) {
368 $link .= '&amp;passed_ent_id='.$passed_ent_id;
369 }
370 if ($view_unsafe_images) {
371 $text = _("Hide Unsafe Images");
372 } else {
373 if (isset($has_unsafe_images) && $has_unsafe_images) {
374 $link .= '&amp;view_unsafe_images=1';
375 $text = _("View Unsafe Images");
376 } else {
377 $text = '';
378 }
379 }
380 $body .= '<center><small><a href="'.$link.'">'.$text.
381 '</a></small></center><br>' . "\n";
382 }
383 return $body;
384 }
385
386
387 function formatAttachments($message, $exclude_id, $mailbox, $id) {
388 global $where, $what, $startMessage, $color;
389 static $ShownHTML = 0;
390
391 $att_ar = $message->getAttachments($exclude_id);
392
393 if (!count($att_ar)) return '';
394
395 $attachments = '';
396
397 $urlMailbox = urlencode($mailbox);
398
399 foreach ($att_ar as $att) {
400 $ent = urldecode($att->entity_id);
401 $header = $att->header;
402 $type0 = strtolower($header->type0);
403 $type1 = strtolower($header->type1);
404 $name = '';
405 $links['download link']['text'] = _("download");
406 $links['download link']['href'] =
407 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
408 $ImageURL = '';
409 if ($type0 =='message' && $type1 == 'rfc822') {
410 $default_page = '../src/read_body.php';
411 $rfc822_header = $att->rfc822_header;
412 $filename = decodeHeader($rfc822_header->subject);
413 if (trim( $filename ) == '') {
414 $filename = 'untitled-[' . $ent . ']' ;
415 }
416 $from_o = $rfc822_header->from;
417 if (is_object($from_o)) {
418 $from_name = $from_o->getAddress(false);
419 } else {
420 $from_name = _("Unknown sender");
421 }
422 $from_name = decodeHeader(htmlspecialchars($from_name));
423 $description = $from_name;
424 } else {
425 $default_page = '../src/download.php';
426 if (is_object($header->disposition)) {
427 $filename = decodeHeader($header->disposition->getProperty('filename'));
428 if (trim($filename) == '') {
429 $name = decodeHeader($header->disposition->getProperty('name'));
430 if (trim($name) == '') {
431 $name = decodeHeader($header->getParameter('name'));
432 if(trim($name) == '') {
433 if (trim( $header->id ) == '') {
434 $filename = 'untitled-[' . $ent . ']' ;
435 } else {
436 $filename = 'cid: ' . $header->id;
437 }
438 } else {
439 $filename = $name;
440 }
441 } else {
442 $filename = $name;
443 }
444 }
445 } else {
446 $filename = decodeHeader($header->getParameter('name'));
447 if (!trim($filename)) {
448 if (trim( $header->id ) == '') {
449 $filename = 'untitled-[' . $ent . ']' ;
450 } else {
451 $filename = 'cid: ' . $header->id;
452 }
453 }
454 }
455 if ($header->description) {
456 $description = htmlspecialchars($header->description);
457 } else {
458 $description = '';
459 }
460 }
461
462 $display_filename = $filename;
463 if (isset($passed_ent_id)) {
464 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
465 } else {
466 $passed_ent_id_link = '';
467 }
468 $defaultlink = $default_page . "?startMessage=$startMessage"
469 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
470 . '&amp;ent_id='.$ent.$passed_ent_id_link.'&amp;absolute_dl=true';
471 if ($where && $what) {
472 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
473 }
474 /* This executes the attachment hook with a specific MIME-type.
475 * If that doesn't have results, it tries if there's a rule
476 * for a more generic type.
477 */
478 $hookresults = do_hook("attachment $type0/$type1", $links,
479 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
480 $display_filename, $where, $what);
481 if(count($hookresults[1]) <= 1) {
482 $hookresults = do_hook("attachment $type0/*", $links,
483 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
484 $display_filename, $where, $what);
485 }
486
487 $links = $hookresults[1];
488 $defaultlink = $hookresults[6];
489
490 $attachments .= '<TR><TD>' .
491 '<A HREF="'.$defaultlink.'">'.htmlspecialchars($display_filename).'</A>&nbsp;</TD>' .
492 '<TD><SMALL><b>' . show_readable_size($header->size) .
493 '</b>&nbsp;&nbsp;</small></TD>' .
494 "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
495 '<TD><SMALL>';
496 $attachments .= '<b>' . $description . '</b>';
497 $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
498
499 $skipspaces = 1;
500 foreach ($links as $val) {
501 if ($skipspaces) {
502 $skipspaces = 0;
503 } else {
504 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
505 }
506 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
507 }
508 unset($links);
509 $attachments .= "</TD></TR>\n";
510 }
511 return $attachments;
512 }
513
514 /* This function decodes the body depending on the encoding type. */
515 function decodeBody($body, $encoding) {
516 global $languages, $squirrelmail_language;
517 global $show_html_default;
518
519 $body = str_replace("\r\n", "\n", $body);
520 $encoding = strtolower($encoding);
521
522 if ($encoding == 'quoted-printable' ||
523 $encoding == 'quoted_printable') {
524 $body = quoted_printable_decode($body);
525
526 while (ereg("=\n", $body)) {
527 $body = ereg_replace ("=\n", '', $body);
528 }
529
530 } else if ($encoding == 'base64') {
531 $body = base64_decode($body);
532 }
533
534 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
535 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
536 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
537 }
538
539 // All other encodings are returned raw.
540 return $body;
541 }
542
543 /*
544 * This functions decode strings that is encoded according to
545 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
546 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
547 */
548 function decodeHeader ($string, $utfencode=true) {
549 global $languages, $squirrelmail_language;
550 if (is_array($string)) {
551 $string = implode("\n", $string);
552 }
553
554 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
555 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
556 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
557 }
558
559 $i = 0;
560 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
561 $string, $res)) {
562 $prefix = $res[1];
563 /* Ignore white-space between consecutive encoded-words. */
564 if (strspn($res[2], " \t") != strlen($res[2])) {
565 $prefix .= $res[2];
566 }
567
568 if (ucfirst($res[4]) == 'B') {
569 $replace = base64_decode($res[5]);
570 $replace = charset_decode($res[3],$replace);
571
572 } else {
573 $replace = str_replace('_', ' ', $res[5]);
574 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
575 $replace);
576 /* Only encode into entities by default. Some places
577 * don't need the encoding, like the compose form.
578 */
579 if ($utfencode) {
580 $replace = charset_decode($res[3], $replace);
581 }
582 }
583 $string = $prefix . $replace . substr($string, strlen($res[0]));
584 $i = strlen($prefix) + strlen($replace);
585 }
586 return $string;
587 }
588
589 /*
590 * Encode a string according to RFC 1522 for use in headers if it
591 * contains 8-bit characters or anything that looks like it should
592 * be encoded.
593 */
594 function encodeHeader ($string) {
595 global $default_charset, $languages, $squirrelmail_language;
596
597 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
598 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
599 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
600 }
601
602 // Encode only if the string contains 8-bit characters or =?
603 $j = strlen($string);
604 $l = strstr($string, '=?'); // Must be encoded ?
605 $ret = '';
606 for($i = 0; $i < $j; ++$i) {
607 switch($string{$i}) {
608 case '=':
609 $ret .= '=3D';
610 break;
611 case '?':
612 $ret .= '=3F';
613 break;
614 case '_':
615 $ret .= '=5F';
616 break;
617 case ' ':
618 $ret .= '_';
619 break;
620 default:
621 $k = ord($string{$i});
622 if ($k > 126) {
623 $ret .= sprintf("=%02X", $k);
624 $l = TRUE;
625 } else {
626 $ret .= $string{$i};
627 }
628 break;
629 }
630 }
631
632 if ($l) {
633 $string = "=?$default_charset?Q?$ret?=";
634 }
635
636 return $string;
637 }
638
639 /* This function trys to locate the entity_id of a specific mime element */
640 function find_ent_id($id, $message) {
641 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
642 if ($message->entities[$i]->header->type0 == 'multipart') {
643 $ret = find_ent_id($id, $message->entities[$i]);
644 } else {
645 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
646 // if (sq_check_save_extension($message->entities[$i])) {
647 return $message->entities[$i]->entity_id;
648 // }
649 }
650 }
651 }
652 return $ret;
653 }
654
655 function sq_check_save_extension($message) {
656 $filename = $message->getFilename();
657 $ext = substr($filename, strrpos($filename,'.')+1);
658 $save_extensions = array('jpg','jpeg','gif','png','bmp');
659 return in_array($ext, $save_extensions);
660 }
661
662
663 /**
664 ** HTMLFILTER ROUTINES
665 */
666
667 /**
668 * This function returns the final tag out of the tag name, an array
669 * of attributes, and the type of the tag. This function is called by
670 * sq_sanitize internally.
671 *
672 * @param $tagname the name of the tag.
673 * @param $attary the array of attributes and their values
674 * @param $tagtype The type of the tag (see in comments).
675 * @return a string with the final tag representation.
676 */
677 function sq_tagprint($tagname, $attary, $tagtype){
678 $me = 'sq_tagprint';
679
680 if ($tagtype == 2){
681 $fulltag = '</' . $tagname . '>';
682 } else {
683 $fulltag = '<' . $tagname;
684 if (is_array($attary) && sizeof($attary)){
685 $atts = Array();
686 while (list($attname, $attvalue) = each($attary)){
687 array_push($atts, "$attname=$attvalue");
688 }
689 $fulltag .= ' ' . join(" ", $atts);
690 }
691 if ($tagtype == 3){
692 $fulltag .= ' /';
693 }
694 $fulltag .= '>';
695 }
696 return $fulltag;
697 }
698
699 /**
700 * A small helper function to use with array_walk. Modifies a by-ref
701 * value and makes it lowercase.
702 *
703 * @param $val a value passed by-ref.
704 * @return void since it modifies a by-ref value.
705 */
706 function sq_casenormalize(&$val){
707 $val = strtolower($val);
708 }
709
710 /**
711 * This function skips any whitespace from the current position within
712 * a string and to the next non-whitespace value.
713 *
714 * @param $body the string
715 * @param $offset the offset within the string where we should start
716 * looking for the next non-whitespace character.
717 * @return the location within the $body where the next
718 * non-whitespace char is located.
719 */
720 function sq_skipspace($body, $offset){
721 $me = 'sq_skipspace';
722 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
723 if (sizeof($matches{1})){
724 $count = strlen($matches{1});
725 $offset += $count;
726 }
727 return $offset;
728 }
729
730 /**
731 * This function looks for the next character within a string. It's
732 * really just a glorified "strpos", except it catches if failures
733 * nicely.
734 *
735 * @param $body The string to look for needle in.
736 * @param $offset Start looking from this position.
737 * @param $needle The character/string to look for.
738 * @return location of the next occurance of the needle, or
739 * strlen($body) if needle wasn't found.
740 */
741 function sq_findnxstr($body, $offset, $needle){
742 $me = 'sq_findnxstr';
743 $pos = strpos($body, $needle, $offset);
744 if ($pos === FALSE){
745 $pos = strlen($body);
746 }
747 return $pos;
748 }
749
750 /**
751 * This function takes a PCRE-style regexp and tries to match it
752 * within the string.
753 *
754 * @param $body The string to look for needle in.
755 * @param $offset Start looking from here.
756 * @param $reg A PCRE-style regex to match.
757 * @return Returns a false if no matches found, or an array
758 * with the following members:
759 * - integer with the location of the match within $body
760 * - string with whatever content between offset and the match
761 * - string with whatever it is we matched
762 */
763 function sq_findnxreg($body, $offset, $reg){
764 $me = 'sq_findnxreg';
765 $matches = Array();
766 $retarr = Array();
767 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
768 if (!$matches{0}){
769 $retarr = false;
770 } else {
771 $retarr{0} = $offset + strlen($matches{1});
772 $retarr{1} = $matches{1};
773 $retarr{2} = $matches{2};
774 }
775 return $retarr;
776 }
777
778 /**
779 * This function looks for the next tag.
780 *
781 * @param $body String where to look for the next tag.
782 * @param $offset Start looking from here.
783 * @return false if no more tags exist in the body, or
784 * an array with the following members:
785 * - string with the name of the tag
786 * - array with attributes and their values
787 * - integer with tag type (1, 2, or 3)
788 * - integer where the tag starts (starting "<")
789 * - integer where the tag ends (ending ">")
790 * first three members will be false, if the tag is invalid.
791 */
792 function sq_getnxtag($body, $offset){
793 $me = 'sq_getnxtag';
794 if ($offset > strlen($body)){
795 return false;
796 }
797 $lt = sq_findnxstr($body, $offset, "<");
798 if ($lt == strlen($body)){
799 return false;
800 }
801 /**
802 * We are here:
803 * blah blah <tag attribute="value">
804 * \---------^
805 */
806 $pos = sq_skipspace($body, $lt+1);
807 if ($pos >= strlen($body)){
808 return Array(false, false, false, $lt, strlen($body));
809 }
810 /**
811 * There are 3 kinds of tags:
812 * 1. Opening tag, e.g.:
813 * <a href="blah">
814 * 2. Closing tag, e.g.:
815 * </a>
816 * 3. XHTML-style content-less tag, e.g.:
817 * <img src="blah"/>
818 */
819 $tagtype = false;
820 switch (substr($body, $pos, 1)){
821 case '/':
822 $tagtype = 2;
823 $pos++;
824 break;
825 case '!':
826 /**
827 * A comment or an SGML declaration.
828 */
829 if (substr($body, $pos+1, 2) == "--"){
830 $gt = strpos($body, "-->", $pos);
831 if ($gt === false){
832 $gt = strlen($body);
833 } else {
834 $gt += 2;
835 }
836 return Array(false, false, false, $lt, $gt);
837 } else {
838 $gt = sq_findnxstr($body, $pos, ">");
839 return Array(false, false, false, $lt, $gt);
840 }
841 break;
842 default:
843 /**
844 * Assume tagtype 1 for now. If it's type 3, we'll switch values
845 * later.
846 */
847 $tagtype = 1;
848 break;
849 }
850
851 $tag_start = $pos;
852 $tagname = '';
853 /**
854 * Look for next [\W-_], which will indicate the end of the tag name.
855 */
856 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
857 if ($regary == false){
858 return Array(false, false, false, $lt, strlen($body));
859 }
860 list($pos, $tagname, $match) = $regary;
861 $tagname = strtolower($tagname);
862
863 /**
864 * $match can be either of these:
865 * '>' indicating the end of the tag entirely.
866 * '\s' indicating the end of the tag name.
867 * '/' indicating that this is type-3 xhtml tag.
868 *
869 * Whatever else we find there indicates an invalid tag.
870 */
871 switch ($match){
872 case '/':
873 /**
874 * This is an xhtml-style tag with a closing / at the
875 * end, like so: <img src="blah"/>. Check if it's followed
876 * by the closing bracket. If not, then this tag is invalid
877 */
878 if (substr($body, $pos, 2) == "/>"){
879 $pos++;
880 $tagtype = 3;
881 } else {
882 $gt = sq_findnxstr($body, $pos, ">");
883 $retary = Array(false, false, false, $lt, $gt);
884 return $retary;
885 }
886 case '>':
887 return Array($tagname, false, $tagtype, $lt, $pos);
888 break;
889 default:
890 /**
891 * Check if it's whitespace
892 */
893 if (!preg_match('/\s/', $match)){
894 /**
895 * This is an invalid tag! Look for the next closing ">".
896 */
897 $gt = sq_findnxstr($body, $offset, ">");
898 return Array(false, false, false, $lt, $gt);
899 }
900 break;
901 }
902
903 /**
904 * At this point we're here:
905 * <tagname attribute='blah'>
906 * \-------^
907 *
908 * At this point we loop in order to find all attributes.
909 */
910 $attname = '';
911 $atttype = false;
912 $attary = Array();
913
914 while ($pos <= strlen($body)){
915 $pos = sq_skipspace($body, $pos);
916 if ($pos == strlen($body)){
917 /**
918 * Non-closed tag.
919 */
920 return Array(false, false, false, $lt, $pos);
921 }
922 /**
923 * See if we arrived at a ">" or "/>", which means that we reached
924 * the end of the tag.
925 */
926 $matches = Array();
927 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
928 /**
929 * Yep. So we did.
930 */
931 $pos += strlen($matches{1});
932 if ($matches{2} == "/>"){
933 $tagtype = 3;
934 $pos++;
935 }
936 return Array($tagname, $attary, $tagtype, $lt, $pos);
937 }
938
939 /**
940 * There are several types of attributes, with optional
941 * [:space:] between members.
942 * Type 1:
943 * attrname[:space:]=[:space:]'CDATA'
944 * Type 2:
945 * attrname[:space:]=[:space:]"CDATA"
946 * Type 3:
947 * attr[:space:]=[:space:]CDATA
948 * Type 4:
949 * attrname
950 *
951 * We leave types 1 and 2 the same, type 3 we check for
952 * '"' and convert to "&quot" if needed, then wrap in
953 * double quotes. Type 4 we convert into:
954 * attrname="yes".
955 */
956 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
957 if ($regary == false){
958 /**
959 * Looks like body ended before the end of tag.
960 */
961 return Array(false, false, false, $lt, strlen($body));
962 }
963 list($pos, $attname, $match) = $regary;
964 $attname = strtolower($attname);
965 /**
966 * We arrived at the end of attribute name. Several things possible
967 * here:
968 * '>' means the end of the tag and this is attribute type 4
969 * '/' if followed by '>' means the same thing as above
970 * '\s' means a lot of things -- look what it's followed by.
971 * anything else means the attribute is invalid.
972 */
973 switch($match){
974 case '/':
975 /**
976 * This is an xhtml-style tag with a closing / at the
977 * end, like so: <img src="blah"/>. Check if it's followed
978 * by the closing bracket. If not, then this tag is invalid
979 */
980 if (substr($body, $pos, 2) == "/>"){
981 $pos++;
982 $tagtype = 3;
983 } else {
984 $gt = sq_findnxstr($body, $pos, ">");
985 $retary = Array(false, false, false, $lt, $gt);
986 return $retary;
987 }
988 case '>':
989 $attary{$attname} = '"yes"';
990 return Array($tagname, $attary, $tagtype, $lt, $pos);
991 break;
992 default:
993 /**
994 * Skip whitespace and see what we arrive at.
995 */
996 $pos = sq_skipspace($body, $pos);
997 $char = substr($body, $pos, 1);
998 /**
999 * Two things are valid here:
1000 * '=' means this is attribute type 1 2 or 3.
1001 * \w means this was attribute type 4.
1002 * anything else we ignore and re-loop. End of tag and
1003 * invalid stuff will be caught by our checks at the beginning
1004 * of the loop.
1005 */
1006 if ($char == "="){
1007 $pos++;
1008 $pos = sq_skipspace($body, $pos);
1009 /**
1010 * Here are 3 possibilities:
1011 * "'" attribute type 1
1012 * '"' attribute type 2
1013 * everything else is the content of tag type 3
1014 */
1015 $quot = substr($body, $pos, 1);
1016 if ($quot == "'"){
1017 $regary = sq_findnxreg($body, $pos+1, "\'");
1018 if ($regary == false){
1019 return Array(false, false, false, $lt, strlen($body));
1020 }
1021 list($pos, $attval, $match) = $regary;
1022 $pos++;
1023 $attary{$attname} = "'" . $attval . "'";
1024 } else if ($quot == '"'){
1025 $regary = sq_findnxreg($body, $pos+1, '\"');
1026 if ($regary == false){
1027 return Array(false, false, false, $lt, strlen($body));
1028 }
1029 list($pos, $attval, $match) = $regary;
1030 $pos++;
1031 $attary{$attname} = '"' . $attval . '"';
1032 } else {
1033 /**
1034 * These are hateful. Look for \s, or >.
1035 */
1036 $regary = sq_findnxreg($body, $pos, "[\s>]");
1037 if ($regary == false){
1038 return Array(false, false, false, $lt, strlen($body));
1039 }
1040 list($pos, $attval, $match) = $regary;
1041 /**
1042 * If it's ">" it will be caught at the top.
1043 */
1044 $attval = preg_replace("/\"/s", "&quot;", $attval);
1045 $attary{$attname} = '"' . $attval . '"';
1046 }
1047 } else if (preg_match("|[\w/>]|", $char)) {
1048 /**
1049 * That was attribute type 4.
1050 */
1051 $attary{$attname} = '"yes"';
1052 } else {
1053 /**
1054 * An illegal character. Find next '>' and return.
1055 */
1056 $gt = sq_findnxstr($body, $pos, ">");
1057 return Array(false, false, false, $lt, $gt);
1058 }
1059 break;
1060 }
1061 }
1062 /**
1063 * The fact that we got here indicates that the tag end was never
1064 * found. Return invalid tag indication so it gets stripped.
1065 */
1066 return Array(false, false, false, $lt, strlen($body));
1067 }
1068
1069 /**
1070 * This function checks attribute values for entity-encoded values
1071 * and returns them translated into 8-bit strings so we can run
1072 * checks on them.
1073 *
1074 * @param $attvalue A string to run entity check against.
1075 * @return Translated value.
1076 */
1077 function sq_deent($attvalue){
1078 $me = 'sq_deent';
1079 /**
1080 * See if we have to run the checks first. All entities must start
1081 * with "&".
1082 */
1083 if (strpos($attvalue, "&") === false){
1084 return $attvalue;
1085 }
1086 /**
1087 * Check named entities first.
1088 */
1089 $trans = get_html_translation_table(HTML_ENTITIES);
1090 /**
1091 * Leave &quot; in, as it can mess us up.
1092 */
1093 $trans = array_flip($trans);
1094 unset($trans{"&quot;"});
1095 while (list($ent, $val) = each($trans)){
1096 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1097 }
1098 /**
1099 * Now translate numbered entities from 1 to 255 if needed.
1100 */
1101 if (strpos($attvalue, "#") !== false){
1102 $omit = Array(34, 39);
1103 for ($asc=1; $asc<256; $asc++){
1104 if (!in_array($asc, $omit)){
1105 $chr = chr($asc);
1106 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1107 $attvalue);
1108 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1109 "$chr\\1", $attvalue);
1110 }
1111 }
1112 }
1113 return $attvalue;
1114 }
1115
1116 /**
1117 * This function runs various checks against the attributes.
1118 *
1119 * @param $tagname String with the name of the tag.
1120 * @param $attary Array with all tag attributes.
1121 * @param $rm_attnames See description for sq_sanitize
1122 * @param $bad_attvals See description for sq_sanitize
1123 * @param $add_attr_to_tag See description for sq_sanitize
1124 * @param $message message object
1125 * @param $id message id
1126 * @return Array with modified attributes.
1127 */
1128 function sq_fixatts($tagname,
1129 $attary,
1130 $rm_attnames,
1131 $bad_attvals,
1132 $add_attr_to_tag,
1133 $message,
1134 $id,
1135 $mailbox
1136 ){
1137 $me = 'sq_fixatts';
1138 while (list($attname, $attvalue) = each($attary)){
1139 /**
1140 * See if this attribute should be removed.
1141 */
1142 foreach ($rm_attnames as $matchtag=>$matchattrs){
1143 if (preg_match($matchtag, $tagname)){
1144 foreach ($matchattrs as $matchattr){
1145 if (preg_match($matchattr, $attname)){
1146 unset($attary{$attname});
1147 continue;
1148 }
1149 }
1150 }
1151 }
1152 /**
1153 * Remove any entities.
1154 */
1155 $attvalue = sq_deent($attvalue);
1156
1157 /**
1158 * Now let's run checks on the attvalues.
1159 * I don't expect anyone to comprehend this. If you do,
1160 * get in touch with me so I can drive to where you live and
1161 * shake your hand personally. :)
1162 */
1163 foreach ($bad_attvals as $matchtag=>$matchattrs){
1164 if (preg_match($matchtag, $tagname)){
1165 foreach ($matchattrs as $matchattr=>$valary){
1166 if (preg_match($matchattr, $attname)){
1167 /**
1168 * There are two arrays in valary.
1169 * First is matches.
1170 * Second one is replacements
1171 */
1172 list($valmatch, $valrepl) = $valary;
1173 $newvalue =
1174 preg_replace($valmatch, $valrepl, $attvalue);
1175 if ($newvalue != $attvalue){
1176 $attary{$attname} = $newvalue;
1177 }
1178 }
1179 }
1180 }
1181 }
1182 /**
1183 * Turn cid: urls into http-friendly ones.
1184 */
1185 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1186 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1187 }
1188 }
1189 /**
1190 * See if we need to append any attributes to this tag.
1191 */
1192 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1193 if (preg_match($matchtag, $tagname)){
1194 $attary = array_merge($attary, $addattary);
1195 }
1196 }
1197 return $attary;
1198 }
1199
1200 /**
1201 * This function edits the style definition to make them friendly and
1202 * usable in squirrelmail.
1203 *
1204 * @param $message the message object
1205 * @param $id the message id
1206 * @param $content a string with whatever is between <style> and </style>
1207 * @return a string with edited content.
1208 */
1209 function sq_fixstyle($message, $id, $content){
1210 global $view_unsafe_images;
1211 $me = 'sq_fixstyle';
1212 /**
1213 * First look for general BODY style declaration, which would be
1214 * like so:
1215 * body {background: blah-blah}
1216 * and change it to .bodyclass so we can just assign it to a <div>
1217 */
1218 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1219 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1220 /**
1221 * Fix url('blah') declarations.
1222 */
1223 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1224 "url(\\1$secremoveimg\\2)", $content);
1225 /**
1226 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1227 * is false.
1228 */
1229 if (!$view_unsafe_images){
1230 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1231 "url(\\1$secremoveimg\\2)", $content);
1232 }
1233
1234 /**
1235 * Fix urls that refer to cid:
1236 */
1237 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1238 $matches)){
1239 $cidurl = $matches{1};
1240 $httpurl = sq_cid2http($message, $id, $cidurl);
1241 $content = preg_replace("|url\($cidurl\)|si",
1242 "url($httpurl)", $content);
1243 }
1244
1245 /**
1246 * Fix stupid css declarations which lead to vulnerabilities
1247 * in IE.
1248 */
1249 $match = Array('/expression/si',
1250 '/behaviou*r/si',
1251 '/binding/si');
1252 $replace = Array('idiocy', 'idiocy', 'idiocy');
1253 $content = preg_replace($match, $replace, $content);
1254 return $content;
1255 }
1256
1257 /**
1258 * This function converts cid: url's into the ones that can be viewed in
1259 * the browser.
1260 *
1261 * @param $message the message object
1262 * @param $id the message id
1263 * @param $cidurl the cid: url.
1264 * @return a string with a http-friendly url
1265 */
1266 function sq_cid2http($message, $id, $cidurl, $mailbox){
1267 /**
1268 * Get rid of quotes.
1269 */
1270 $quotchar = substr($cidurl, 0, 1);
1271 $cidurl = str_replace($quotchar, "", $cidurl);
1272 $cidurl = substr(trim($cidurl), 4);
1273 $linkurl = find_ent_id($cidurl, $message);
1274 /* in case of non-save cid links $httpurl should be replaced by a sort of
1275 unsave link image */
1276 $httpurl = '';
1277 if ($linkurl) {
1278 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&amp;' .
1279 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1280 '&amp;ent_id=' . $linkurl . $quotchar;
1281 }
1282 return $httpurl;
1283 }
1284
1285 /**
1286 * This function changes the <body> tag into a <div> tag since we
1287 * can't really have a body-within-body.
1288 *
1289 * @param $attary an array of attributes and values of <body>
1290 * @return a modified array of attributes to be set for <div>
1291 */
1292 function sq_body2div($attary){
1293 $me = 'sq_body2div';
1294 $divattary = Array('class' => "'bodyclass'");
1295 $bgcolor = '#ffffff';
1296 $text = '#000000';
1297 $styledef = '';
1298 if (is_array($attary) && sizeof($attary) > 0){
1299 foreach ($attary as $attname=>$attvalue){
1300 $quotchar = substr($attvalue, 0, 1);
1301 $attvalue = str_replace($quotchar, "", $attvalue);
1302 switch ($attname){
1303 case 'background':
1304 $styledef .= "background-image: url('$attvalue'); ";
1305 break;
1306 case 'bgcolor':
1307 $styledef .= "background-color: $attvalue; ";
1308 break;
1309 case 'text':
1310 $styledef .= "color: $attvalue; ";
1311 break;
1312 }
1313 }
1314 if (strlen($styledef) > 0){
1315 $divattary{"style"} = "\"$styledef\"";
1316 }
1317 }
1318 return $divattary;
1319 }
1320
1321 /**
1322 * This is the main function and the one you should actually be calling.
1323 * There are several variables you should be aware of an which need
1324 * special description.
1325 *
1326 * Since the description is quite lengthy, see it here:
1327 * http://www.mricon.com/html/phpfilter.html
1328 *
1329 * @param $body the string with HTML you wish to filter
1330 * @param $tag_list see description above
1331 * @param $rm_tags_with_content see description above
1332 * @param $self_closing_tags see description above
1333 * @param $force_tag_closing see description above
1334 * @param $rm_attnames see description above
1335 * @param $bad_attvals see description above
1336 * @param $add_attr_to_tag see description above
1337 * @param $message message object
1338 * @param $id message id
1339 * @return sanitized html safe to show on your pages.
1340 */
1341 function sq_sanitize($body,
1342 $tag_list,
1343 $rm_tags_with_content,
1344 $self_closing_tags,
1345 $force_tag_closing,
1346 $rm_attnames,
1347 $bad_attvals,
1348 $add_attr_to_tag,
1349 $message,
1350 $id,
1351 $mailbox
1352 ){
1353 $me = 'sq_sanitize';
1354 /**
1355 * Normalize rm_tags and rm_tags_with_content.
1356 */
1357 @array_walk($rm_tags, 'sq_casenormalize');
1358 @array_walk($rm_tags_with_content, 'sq_casenormalize');
1359 @array_walk($self_closing_tags, 'sq_casenormalize');
1360 /**
1361 * See if tag_list is of tags to remove or tags to allow.
1362 * false means remove these tags
1363 * true means allow these tags
1364 */
1365 $rm_tags = array_shift($tag_list);
1366 $curpos = 0;
1367 $open_tags = Array();
1368 $trusted = "<!-- begin sanitized html -->\n";
1369 $skip_content = false;
1370 /**
1371 * Take care of netscape's stupid javascript entities like
1372 * &{alert('boo')};
1373 */
1374 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1375
1376 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1377 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1378 $free_content = substr($body, $curpos, $lt-$curpos);
1379 /**
1380 * Take care of <style>
1381 */
1382 if ($tagname == "style" && $tagtype == 2){
1383 /**
1384 * This is a closing </style>. Edit the
1385 * content before we apply it.
1386 */
1387 $free_content = sq_fixstyle($message, $id, $free_content);
1388 }
1389 if ($skip_content == false){
1390 $trusted .= $free_content;
1391 }
1392 if ($tagname != FALSE){
1393 if ($tagtype == 2){
1394 if ($skip_content == $tagname){
1395 /**
1396 * Got to the end of tag we needed to remove.
1397 */
1398 $tagname = false;
1399 $skip_content = false;
1400 } else {
1401 if ($skip_content == false){
1402 if ($tagname == "body"){
1403 $tagname = "div";
1404 } else {
1405 if (isset($open_tags{$tagname}) &&
1406 $open_tags{$tagname} > 0){
1407 $open_tags{$tagname}--;
1408 } else {
1409 $tagname = false;
1410 }
1411 }
1412 }
1413 }
1414 } else {
1415 /**
1416 * $rm_tags_with_content
1417 */
1418 if ($skip_content == false){
1419 /**
1420 * See if this is a self-closing type and change
1421 * tagtype appropriately.
1422 */
1423 if ($tagtype == 1
1424 && in_array($tagname, $self_closing_tags)){
1425 $tagtype=3;
1426 }
1427 /**
1428 * See if we should skip this tag and any content
1429 * inside it.
1430 */
1431 if ($tagtype == 1 &&
1432 in_array($tagname, $rm_tags_with_content)){
1433 $skip_content = $tagname;
1434 } else {
1435 if (($rm_tags == false
1436 && in_array($tagname, $tag_list)) ||
1437 ($rm_tags == true &&
1438 !in_array($tagname, $tag_list))){
1439 $tagname = false;
1440 } else {
1441 if ($tagtype == 1){
1442 if (isset($open_tags{$tagname})){
1443 $open_tags{$tagname}++;
1444 } else {
1445 $open_tags{$tagname}=1;
1446 }
1447 }
1448 /**
1449 * This is where we run other checks.
1450 */
1451 if (is_array($attary) && sizeof($attary) > 0){
1452 $attary = sq_fixatts($tagname,
1453 $attary,
1454 $rm_attnames,
1455 $bad_attvals,
1456 $add_attr_to_tag,
1457 $message,
1458 $id,
1459 $mailbox
1460 );
1461 }
1462 /**
1463 * Convert body into div.
1464 */
1465 if ($tagname == "body"){
1466 $tagname = "div";
1467 $attary = sq_body2div($attary, $message, $id);
1468 }
1469 }
1470 }
1471 }
1472 }
1473 if ($tagname != false && $skip_content == false){
1474 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1475 }
1476 }
1477 $curpos = $gt+1;
1478 }
1479 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1480 if ($force_tag_closing == true){
1481 foreach ($open_tags as $tagname=>$opentimes){
1482 while ($opentimes > 0){
1483 $trusted .= '</' . $tagname . '>';
1484 $opentimes--;
1485 }
1486 }
1487 $trusted .= "\n";
1488 }
1489 $trusted .= "<!-- end sanitized html -->\n";
1490 return $trusted;
1491 }
1492
1493 /**
1494 * This is a wrapper function to call html sanitizing routines.
1495 *
1496 * @param $body the body of the message
1497 * @param $id the id of the message
1498 * @return a string with html safe to display in the browser.
1499 */
1500 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1501 global $attachment_common_show_images, $view_unsafe_images,
1502 $has_unsafe_images;
1503 /**
1504 * Don't display attached images in HTML mode.
1505 */
1506 $attachment_common_show_images = false;
1507 $tag_list = Array(
1508 false,
1509 "object",
1510 "meta",
1511 "html",
1512 "head",
1513 "base",
1514 "link",
1515 "frame",
1516 "iframe"
1517 );
1518
1519 $rm_tags_with_content = Array(
1520 "script",
1521 "applet",
1522 "embed",
1523 "title"
1524 );
1525
1526 $self_closing_tags = Array(
1527 "img",
1528 "br",
1529 "hr",
1530 "input"
1531 );
1532
1533 $force_tag_closing = false;
1534
1535 $rm_attnames = Array(
1536 "/.*/" =>
1537 Array(
1538 "/target/si",
1539 "/^on.*/si",
1540 "/^dynsrc/si",
1541 "/^data.*/si"
1542 )
1543 );
1544
1545 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1546 $bad_attvals = Array(
1547 "/.*/" =>
1548 Array(
1549 "/^src|background/i" =>
1550 Array(
1551 Array(
1552 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1553 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1554 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1555 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1556 ),
1557 Array(
1558 "\\1$secremoveimg\\2",
1559 "\\1$secremoveimg\\2",
1560 "\\1$secremoveimg\\2",
1561 "\\1$secremoveimg\\2"
1562 )
1563 ),
1564 "/^href|action/i" =>
1565 Array(
1566 Array(
1567 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1568 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1569 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1570 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1571 ),
1572 Array(
1573 "\\1#\\2",
1574 "\\1#\\2",
1575 "\\1#\\2",
1576 "\\1#\\2"
1577 )
1578 ),
1579 "/^style/si" =>
1580 Array(
1581 Array(
1582 "/expression/si",
1583 "/binding/si",
1584 "/behaviou*r/si",
1585 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1586 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1587 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1588 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1589 ),
1590 Array(
1591 "idiocy",
1592 "idiocy",
1593 "idiocy",
1594 "url(\\1#\\2)",
1595 "url(\\1#\\2)",
1596 "url(\\1#\\2)",
1597 "url(\\1#\\2)"
1598 )
1599 )
1600 )
1601 );
1602 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1603 $view_unsafe_images = false;
1604 }
1605 if (!$view_unsafe_images){
1606 /**
1607 * Remove any references to http/https if view_unsafe_images set
1608 * to false.
1609 */
1610 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1611 '/^([\'\"])\s*https*:.*([\'\"])/si');
1612 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1613 "\\1$secremoveimg\\2");
1614 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1615 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1616 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1617 "url(\\1$secremoveimg\\2)");
1618 }
1619
1620 $add_attr_to_tag = Array(
1621 "/^a$/si" => Array('target'=>'"_new"')
1622 );
1623 $trusted = sq_sanitize($body,
1624 $tag_list,
1625 $rm_tags_with_content,
1626 $self_closing_tags,
1627 $force_tag_closing,
1628 $rm_attnames,
1629 $bad_attvals,
1630 $add_attr_to_tag,
1631 $message,
1632 $id,
1633 $mailbox
1634 );
1635 if (preg_match("|$secremoveimg|si", $trusted)){
1636 $has_unsafe_images = true;
1637 }
1638 return $trusted;
1639 }
1640
1641 ?>