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