Fixed typo.
[squirrelmail.git] / functions / mime.php
1 <?php
2
3 /**
4 * mime.php
5 *
6 * Copyright (c) 1999-2005 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 * @version $Id$
13 * @package squirrelmail
14 */
15
16 /** The typical includes... */
17 require_once(SM_PATH . 'functions/imap.php');
18 require_once(SM_PATH . 'functions/attachment_common.php');
19
20 /* -------------------------------------------------------------------------- */
21 /* MIME DECODING */
22 /* -------------------------------------------------------------------------- */
23
24 /**
25 * Get the MIME structure
26 *
27 * This function gets the structure of a message and stores it in the "message" class.
28 * It will return this object for use with all relevant header information and
29 * fully parsed into the standard "message" object format.
30 */
31 function mime_structure ($bodystructure, $flags=array()) {
32
33 /* Isolate the body structure and remove beginning and end parenthesis. */
34 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
35 $read = trim(substr ($read, 0, -1));
36 $i = 0;
37 $msg = Message::parseStructure($read,$i);
38 if (!is_object($msg)) {
39 include_once(SM_PATH . 'functions/display_messages.php');
40 global $color, $mailbox;
41 /* removed urldecode because $_GET is auto urldecoded ??? */
42 displayPageHeader( $color, $mailbox );
43 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
44 $errormessage .= '<br />'._("the provided bodystructure by your imap-server").':<br /><br />';
45 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
46 plain_error_message( $errormessage, $color );
47 echo '</body></html>';
48 exit;
49 }
50 if (count($flags)) {
51 foreach ($flags as $flag) {
52 $char = strtoupper($flag{1});
53 switch ($char) {
54 case 'S':
55 if (strtolower($flag) == '\\seen') {
56 $msg->is_seen = true;
57 }
58 break;
59 case 'A':
60 if (strtolower($flag) == '\\answered') {
61 $msg->is_answered = true;
62 }
63 break;
64 case 'D':
65 if (strtolower($flag) == '\\deleted') {
66 $msg->is_deleted = true;
67 }
68 break;
69 case 'F':
70 if (strtolower($flag) == '\\flagged') {
71 $msg->is_flagged = true;
72 }
73 break;
74 case 'M':
75 if (strtolower($flag) == '$mdnsent') {
76 $msg->is_mdnsent = true;
77 }
78 break;
79 default:
80 break;
81 }
82 }
83 }
84 // listEntities($msg);
85 return $msg;
86 }
87
88
89
90 /* This starts the parsing of a particular structure. It is called recursively,
91 * so it can be passed different structures. It returns an object of type
92 * $message.
93 * First, it checks to see if it is a multipart message. If it is, then it
94 * handles that as it sees is necessary. If it is just a regular entity,
95 * then it parses it and adds the necessary header information (by calling out
96 * to mime_get_elements()
97 */
98
99 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
100 /* Do a bit of error correction. If we couldn't find the entity id, just guess
101 * that it is the first one. That is usually the case anyway.
102 */
103
104 if (!$ent_id) {
105 $cmd = "FETCH $id BODY[]";
106 } else {
107 $cmd = "FETCH $id BODY[$ent_id]";
108 }
109
110 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
111
112 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
113 do {
114 $topline = trim(array_shift($data));
115 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
116
117 $wholemessage = implode('', $data);
118 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
119 $ret = substr($wholemessage, 0, $regs[1]);
120 /* There is some information in the content info header that could be important
121 * in order to parse html messages. Let's get them here.
122 */
123 // if ($ret{0} == '<') {
124 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
125 // }
126 } else if (ereg('"([^"]*)"', $topline, $regs)) {
127 $ret = $regs[1];
128 } else {
129 global $where, $what, $mailbox, $passed_id, $startMessage;
130 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
131 if (isset($where) && isset($what)) {
132 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
133 } else {
134 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
135 }
136 $par .= '&amp;response=' . urlencode($response) .
137 '&amp;message=' . urlencode($message) .
138 '&amp;topline=' . urlencode($topline);
139
140 echo '<tt><br />' .
141 '<table width="80%"><tr>' .
142 '<tr><td colspan="2">' .
143 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
144 '</td></tr>' .
145 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
146 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
147 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
148 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
149 "</table><br /></tt></font><hr />";
150
151 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
152 array_shift($data);
153 $wholemessage = implode('', $data);
154
155 $ret = $wholemessage;
156 }
157 return $ret;
158 }
159
160 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
161
162 /* Don't kill the connection if the browser is over a dialup
163 * and it would take over 30 seconds to download it.
164 * Don't call set_time_limit in safe mode.
165 */
166
167 if (!ini_get('safe_mode')) {
168 set_time_limit(0);
169 }
170 /* in case of base64 encoded attachments, do not buffer them.
171 Instead, echo the decoded attachment directly to screen */
172 if (strtolower($encoding) == 'base64') {
173 if (!$ent_id) {
174 $query = "FETCH $id BODY[]";
175 } else {
176 $query = "FETCH $id BODY[$ent_id]";
177 }
178 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode','php://stdout',true);
179 } else {
180 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
181 echo decodeBody($body, $encoding);
182 }
183
184 /*
185 TODO, use the same method for quoted printable.
186 However, I assume that quoted printable attachments aren't that large
187 so the performancegain / memory usage drop will be minimal.
188 If we decide to add that then we need to adapt sqimap_fread because
189 we need to split te result on \n and fread doesn't stop at \n. That
190 means we also should provide $results from sqimap_fread (by ref) to
191 te function and set $no_return to false. The $filter function for
192 quoted printable should handle unsetting of $results.
193 */
194 /*
195 TODO 2: find out how we write to the output stream php://stdout. fwrite
196 doesn't work because 'php://stdout isn't a stream.
197 */
198
199 return;
200 /*
201 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
202 $cnt = 0;
203 $continue = true;
204 $read = fgets ($imap_stream,8192);
205
206
207 // This could be bad -- if the section has sqimap_session_id() . ' OK'
208 // or similar, it will kill the download.
209 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
210 if (trim($read) == ')==') {
211 $read1 = $read;
212 $read = fgets ($imap_stream,4096);
213 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
214 return;
215 } else {
216 echo decodeBody($read1, $encoding) .
217 decodeBody($read, $encoding);
218 }
219 } else if ($cnt) {
220 echo decodeBody($read, $encoding);
221 }
222 $read = fgets ($imap_stream,4096);
223 $cnt++;
224 // break;
225 }
226 */
227 }
228
229 /* -[ END MIME DECODING ]----------------------------------------------------------- */
230
231 /* This is here for debugging purposes. It will print out a list
232 * of all the entity IDs that are in the $message object.
233 */
234 function listEntities ($message) {
235 if ($message) {
236 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
237 for ($i = 0; isset($message->entities[$i]); $i++) {
238 echo "$i : ";
239 $msg = listEntities($message->entities[$i]);
240
241 if ($msg) {
242 echo "return: ";
243 return $msg;
244 }
245 }
246 }
247 }
248
249 function getPriorityStr($priority) {
250 $priority_level = substr($priority,0,1);
251
252 switch($priority_level) {
253 /* Check for a higher then normal priority. */
254 case '1':
255 case '2':
256 $priority_string = _("High");
257 break;
258
259 /* Check for a lower then normal priority. */
260 case '4':
261 case '5':
262 $priority_string = _("Low");
263 break;
264
265 /* Check for a normal priority. */
266 case '3':
267 default:
268 $priority_level = '3';
269 $priority_string = _("Normal");
270 break;
271
272 }
273 return $priority_string;
274 }
275
276 /* returns a $message object for a particular entity id */
277 function getEntity ($message, $ent_id) {
278 return $message->getEntity($ent_id);
279 }
280
281 /* translateText
282 * Extracted from strings.php 23/03/2002
283 */
284
285 function translateText(&$body, $wrap_at, $charset) {
286 global $where, $what; /* from searching */
287 global $color; /* color theme */
288
289 require_once(SM_PATH . 'functions/url_parser.php');
290
291 $body_ary = explode("\n", $body);
292 for ($i=0; $i < count($body_ary); $i++) {
293 $line = $body_ary[$i];
294 if (strlen($line) - 2 >= $wrap_at) {
295 sqWordWrap($line, $wrap_at, $charset);
296 }
297 $line = charset_decode($charset, $line);
298 $line = str_replace("\t", ' ', $line);
299
300 parseUrl ($line);
301
302 $quotes = 0;
303 $pos = 0;
304 $j = strlen($line);
305
306 while ($pos < $j) {
307 if ($line[$pos] == ' ') {
308 $pos++;
309 } else if (strpos($line, '&gt;', $pos) === $pos) {
310 $pos += 4;
311 $quotes++;
312 } else {
313 break;
314 }
315 }
316
317 if ($quotes % 2) {
318 if (!isset($color[13])) {
319 $color[13] = '#800000';
320 }
321 $line = '<font color="' . $color[13] . '">' . $line . '</font>';
322 } elseif ($quotes) {
323 if (!isset($color[14])) {
324 $color[14] = '#FF0000';
325 }
326 $line = '<font color="' . $color[14] . '">' . $line . '</font>';
327 }
328
329 $body_ary[$i] = $line;
330 }
331 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
332 }
333
334 /**
335 * This returns a parsed string called $body. That string can then
336 * be displayed as the actual message in the HTML. It contains
337 * everything needed, including HTML Tags, Attachments at the
338 * bottom, etc.
339 * @param clean Do not output stuff that's irrelevant for the printable version.
340 */
341 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
342 /* This if statement checks for the entity to show as the
343 * primary message. To add more of them, just put them in the
344 * order that is their priority.
345 */
346 global $startMessage, $languages, $squirrelmail_language,
347 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
348
349 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
350 $view_unsafe_images = false;
351 }
352
353 $body = '';
354 $urlmailbox = urlencode($mailbox);
355 $body_message = getEntity($message, $ent_num);
356 if (($body_message->header->type0 == 'text') ||
357 ($body_message->header->type0 == 'rfc822')) {
358 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
359 $body = decodeBody($body, $body_message->header->encoding);
360
361 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
362 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
363 if (mb_detect_encoding($body) != 'ASCII') {
364 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
365 }
366 }
367 $hookResults = do_hook("message_body", $body);
368 $body = $hookResults[1];
369
370 /* If there are other types that shouldn't be formatted, add
371 * them here.
372 */
373
374 if ($body_message->header->type1 == 'html') {
375 if ($show_html_default <> 1) {
376 $entity_conv = array('&nbsp;' => ' ',
377 '<p>' => "\n",
378 '<P>' => "\n",
379 '<br>' => "\n",
380 '<BR>' => "\n",
381 '<br />' => "\n",
382 '<BR />' => "\n",
383 '&gt;' => '>',
384 '&lt;' => '<');
385 $body = strtr($body, $entity_conv);
386 $body = strip_tags($body);
387 $body = trim($body);
388 translateText($body, $wrap_at,
389 $body_message->header->getParameter('charset'));
390 } else {
391 $body = magicHTML($body, $id, $message, $mailbox);
392 }
393 } else {
394 translateText($body, $wrap_at,
395 $body_message->header->getParameter('charset'));
396 }
397
398 // if this is the clean display (i.e. printer friendly), stop here.
399 if ( $clean ) {
400 return $body;
401 }
402
403 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
404 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
405 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
406 if (isset($passed_ent_id)) {
407 $link .= '&amp;passed_ent_id='.$passed_ent_id;
408 }
409 $body .= '<center><small><a href="download.php?absolute_dl=true&amp;' .
410 $link . '">' . _("Download this as a file") . '</a>';
411 if ($view_unsafe_images) {
412 $text = _("Hide Unsafe Images");
413 } else {
414 if (isset($has_unsafe_images) && $has_unsafe_images) {
415 $link .= '&amp;view_unsafe_images=1';
416 $text = _("View Unsafe Images");
417 } else {
418 $text = '';
419 }
420 }
421 if($text != '') {
422 $body .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
423 }
424 $body .= '</small></center><br />' . "\n";
425 }
426 return $body;
427 }
428
429
430 function formatAttachments($message, $exclude_id, $mailbox, $id) {
431 global $where, $what, $startMessage, $color, $passed_ent_id;
432
433 $att_ar = $message->getAttachments($exclude_id);
434
435 if (!count($att_ar)) return '';
436
437 $attachments = '';
438
439 $urlMailbox = urlencode($mailbox);
440
441 foreach ($att_ar as $att) {
442 $ent = $att->entity_id;
443 $header = $att->header;
444 $type0 = strtolower($header->type0);
445 $type1 = strtolower($header->type1);
446 $name = '';
447 $links['download link']['text'] = _("Download");
448 $links['download link']['href'] = SM_PATH .
449 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
450 if ($type0 =='message' && $type1 == 'rfc822') {
451 $default_page = SM_PATH . 'src/read_body.php';
452 $rfc822_header = $att->rfc822_header;
453 $filename = $rfc822_header->subject;
454 if (trim( $filename ) == '') {
455 $filename = 'untitled-[' . $ent . ']' ;
456 }
457 $from_o = $rfc822_header->from;
458 if (is_object($from_o)) {
459 $from_name = decodeHeader($from_o->getAddress(false));
460 } else {
461 $from_name = _("Unknown sender");
462 }
463 $description = $from_name;
464 } else {
465 $default_page = SM_PATH . 'src/download.php';
466 $filename = $att->getFilename();
467 if ($header->description) {
468 $description = decodeHeader($header->description);
469 } else {
470 $description = '';
471 }
472 }
473
474 $display_filename = $filename;
475 if (isset($passed_ent_id)) {
476 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
477 } else {
478 $passed_ent_id_link = '';
479 }
480 $defaultlink = $default_page . "?startMessage=$startMessage"
481 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
482 . '&amp;ent_id='.$ent.$passed_ent_id_link;
483 if ($where && $what) {
484 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
485 }
486
487 /* This executes the attachment hook with a specific MIME-type.
488 * If that doesn't have results, it tries if there's a rule
489 * for a more generic type.
490 */
491 $hookresults = do_hook("attachment $type0/$type1", $links,
492 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
493 $display_filename, $where, $what);
494 if(count($hookresults[1]) <= 1) {
495 $hookresults = do_hook("attachment $type0/*", $links,
496 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
497 $display_filename, $where, $what);
498 }
499
500 $links = $hookresults[1];
501 $defaultlink = $hookresults[6];
502
503 $attachments .= '<tr><td>' .
504 '<a href="'.$defaultlink.'">'.decodeHeader($display_filename).'</a>&nbsp;</td>' .
505 '<td><small><b>' . show_readable_size($header->size) .
506 '</b>&nbsp;&nbsp;</small></td>' .
507 '<td><small>[ '.htmlspecialchars($type0).'/'.htmlspecialchars($type1).' ]&nbsp;</small></td>' .
508 '<td><small>';
509 $attachments .= '<b>' . $description . '</b>';
510 $attachments .= '</small></td><td><small>&nbsp;';
511
512 $skipspaces = 1;
513 foreach ($links as $val) {
514 if ($skipspaces) {
515 $skipspaces = 0;
516 } else {
517 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
518 }
519 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
520 }
521 unset($links);
522 $attachments .= "</td></tr>\n";
523 }
524 $attachmentadd = do_hook_function('attachments_bottom',$attachments);
525 if ($attachmentadd != '')
526 $attachments = $attachmentadd;
527 return $attachments;
528 }
529
530 function sqimap_base64_decode(&$string) {
531
532 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
533 // fly decoding (to reduce memory usage) you have to check if the
534 // data has incomplete pairs
535
536 // Remove the noise in order to check if the 4 bytes pairs are complete
537 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
538
539 $sStringRem = '';
540 $iMod = strlen($string) % 4;
541 if ($iMod) {
542 $sStringRem = substr($string,-$iMod);
543 // Check if $sStringRem contains padding characters
544 if (substr($sStringRem,-1) != '=') {
545 $string = substr($string,0,-$iMod);
546 } else {
547 $sStringRem = '';
548 }
549 }
550 $string = base64_decode($string);
551 return $sStringRem;
552 }
553
554
555 /* This function decodes the body depending on the encoding type. */
556 function decodeBody($body, $encoding) {
557 global $show_html_default;
558
559 $body = str_replace("\r\n", "\n", $body);
560 $encoding = strtolower($encoding);
561
562 $encoding_handler = do_hook_function('decode_body', $encoding);
563
564
565 // plugins get first shot at decoding the body
566 //
567 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
568 $body = $encoding_handler('decode', $body);
569
570 } else if ($encoding == 'quoted-printable' ||
571 $encoding == 'quoted_printable') {
572 $body = quoted_printable_decode($body);
573
574 while (ereg("=\n", $body)) {
575 $body = ereg_replace ("=\n", '', $body);
576 }
577
578 } else if ($encoding == 'base64') {
579 $body = base64_decode($body);
580 }
581
582 // All other encodings are returned raw.
583 return $body;
584 }
585
586 /**
587 * Decodes headers
588 *
589 * This functions decode strings that is encoded according to
590 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
591 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
592 *
593 * @param string $string header string that has to be made readable
594 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
595 * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
596 * @param boolean $decide decide if string can be utfencoded. defaults to false
597 * @return string decoded header string
598 */
599 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
600 global $languages, $squirrelmail_language,$default_charset;
601 if (is_array($string)) {
602 $string = implode("\n", $string);
603 }
604
605 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
606 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
607 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
608 // Do we need to return at this point?
609 // return $string;
610 }
611 $i = 0;
612 $iLastMatch = -2;
613 $encoded = true;
614
615 $aString = explode(' ',$string);
616 $ret = '';
617 foreach ($aString as $chunk) {
618 if ($encoded && $chunk === '') {
619 continue;
620 } elseif ($chunk === '') {
621 $ret .= ' ';
622 continue;
623 }
624 $encoded = false;
625 /* if encoded words are not separated by a linear-space-white we still catch them */
626 $j = $i-1;
627
628 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
629 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
630 if ($iLastMatch !== $j) {
631 if ($htmlsave) {
632 $ret .= '&#32;';
633 } else {
634 $ret .= ' ';
635 }
636 }
637 $iLastMatch = $i;
638 $j = $i;
639 if ($htmlsave) {
640 $ret .= htmlspecialchars($res[1]);
641 } else {
642 $ret .= $res[1];
643 }
644 $encoding = ucfirst($res[3]);
645
646 /* decide about valid decoding */
647 if ($decide && is_conversion_safe($res[2])) {
648 $utfencode=true;
649 $can_be_encoded=true;
650 } else {
651 $can_be_encoded=false;
652 }
653 switch ($encoding)
654 {
655 case 'B':
656 $replace = base64_decode($res[4]);
657 if ($utfencode) {
658 if ($can_be_encoded) {
659 /* convert string to different charset,
660 * if functions asks for it (usually in compose)
661 */
662 $ret .= charset_convert($res[2],$replace,$default_charset);
663 } else {
664 // convert string to html codes in order to display it
665 $ret .= charset_decode($res[2],$replace);
666 }
667 } else {
668 if ($htmlsave) {
669 $replace = htmlspecialchars($replace);
670 }
671 $ret.= $replace;
672 }
673 break;
674 case 'Q':
675 $replace = str_replace('_', ' ', $res[4]);
676 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
677 $replace);
678 if ($utfencode) {
679 if ($can_be_encoded) {
680 /* convert string to different charset,
681 * if functions asks for it (usually in compose)
682 */
683 $replace = charset_convert($res[2], $replace,$default_charset);
684 } else {
685 // convert string to html codes in order to display it
686 $replace = charset_decode($res[2], $replace);
687 }
688 } else {
689 if ($htmlsave) {
690 $replace = htmlspecialchars($replace);
691 }
692 }
693 $ret .= $replace;
694 break;
695 default:
696 break;
697 }
698 $chunk = $res[5];
699 $encoded = true;
700 }
701 if (!$encoded) {
702 if ($htmlsave) {
703 $ret .= '&#32;';
704 } else {
705 $ret .= ' ';
706 }
707 }
708
709 if (!$encoded && $htmlsave) {
710 $ret .= htmlspecialchars($chunk);
711 } else {
712 $ret .= $chunk;
713 }
714 ++$i;
715 }
716 /* remove the first added space */
717 if ($ret) {
718 if ($htmlsave) {
719 $ret = substr($ret,5);
720 } else {
721 $ret = substr($ret,1);
722 }
723 }
724
725 return $ret;
726 }
727
728 /**
729 * Encodes header as quoted-printable
730 *
731 * Encode a string according to RFC 1522 for use in headers if it
732 * contains 8-bit characters or anything that looks like it should
733 * be encoded.
734 *
735 * @param string $string header string, that has to be encoded
736 * @return string quoted-printable encoded string
737 */
738 function encodeHeader ($string) {
739 global $default_charset, $languages, $squirrelmail_language;
740
741 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
742 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
743 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
744 }
745
746 // Encode only if the string contains 8-bit characters or =?
747 $j = strlen($string);
748 $max_l = 75 - strlen($default_charset) - 7;
749 $aRet = array();
750 $ret = '';
751 $iEncStart = $enc_init = false;
752 $cur_l = $iOffset = 0;
753 for($i = 0; $i < $j; ++$i) {
754 switch($string{$i})
755 {
756 case '=':
757 case '<':
758 case '>':
759 case ',':
760 case '?':
761 case '_':
762 if ($iEncStart === false) {
763 $iEncStart = $i;
764 }
765 $cur_l+=3;
766 if ($cur_l > ($max_l-2)) {
767 /* if there is an stringpart that doesn't need encoding, add it */
768 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
769 $aRet[] = "=?$default_charset?Q?$ret?=";
770 $iOffset = $i;
771 $cur_l = 0;
772 $ret = '';
773 $iEncStart = false;
774 } else {
775 $ret .= sprintf("=%02X",ord($string{$i}));
776 }
777 break;
778 case '(':
779 case ')':
780 if ($iEncStart !== false) {
781 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
782 $aRet[] = "=?$default_charset?Q?$ret?=";
783 $iOffset = $i;
784 $cur_l = 0;
785 $ret = '';
786 $iEncStart = false;
787 }
788 break;
789 case ' ':
790 if ($iEncStart !== false) {
791 $cur_l++;
792 if ($cur_l > $max_l) {
793 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
794 $aRet[] = "=?$default_charset?Q?$ret?=";
795 $iOffset = $i;
796 $cur_l = 0;
797 $ret = '';
798 $iEncStart = false;
799 } else {
800 $ret .= '_';
801 }
802 }
803 break;
804 default:
805 $k = ord($string{$i});
806 if ($k > 126) {
807 if ($iEncStart === false) {
808 // do not start encoding in the middle of a string, also take the rest of the word.
809 $sLeadString = substr($string,0,$i);
810 $aLeadString = explode(' ',$sLeadString);
811 $sToBeEncoded = array_pop($aLeadString);
812 $iEncStart = $i - strlen($sToBeEncoded);
813 $ret .= $sToBeEncoded;
814 $cur_l += strlen($sToBeEncoded);
815 }
816 $cur_l += 3;
817 /* first we add the encoded string that reached it's max size */
818 if ($cur_l > ($max_l-2)) {
819 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
820 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
821 $cur_l = 3;
822 $ret = '';
823 $iOffset = $i;
824 $iEncStart = $i;
825 }
826 $enc_init = true;
827 $ret .= sprintf("=%02X", $k);
828 } else {
829 if ($iEncStart !== false) {
830 $cur_l++;
831 if ($cur_l > $max_l) {
832 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
833 $aRet[] = "=?$default_charset?Q?$ret?=";
834 $iEncStart = false;
835 $iOffset = $i;
836 $cur_l = 0;
837 $ret = '';
838 } else {
839 $ret .= $string{$i};
840 }
841 }
842 }
843 break;
844 }
845 }
846
847 if ($enc_init) {
848 if ($iEncStart !== false) {
849 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
850 $aRet[] = "=?$default_charset?Q?$ret?=";
851 } else {
852 $aRet[] = substr($string,$iOffset);
853 }
854 $string = implode('',$aRet);
855 }
856 return $string;
857 }
858
859 /* This function trys to locate the entity_id of a specific mime element */
860 function find_ent_id($id, $message) {
861 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
862 if ($message->entities[$i]->header->type0 == 'multipart') {
863 $ret = find_ent_id($id, $message->entities[$i]);
864 } else {
865 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
866 // if (sq_check_save_extension($message->entities[$i])) {
867 return $message->entities[$i]->entity_id;
868 // }
869 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
870 /**
871 * This is part of a fix for Outlook Express 6.x generating
872 * cid URLs without creating content-id headers
873 * @@JA - 20050207
874 */
875 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
876 return $message->entities[$i]->entity_id;
877 }
878 }
879 }
880 }
881 return $ret;
882 }
883
884 function sq_check_save_extension($message) {
885 $filename = $message->getFilename();
886 $ext = substr($filename, strrpos($filename,'.')+1);
887 $save_extensions = array('jpg','jpeg','gif','png','bmp');
888 return in_array($ext, $save_extensions);
889 }
890
891
892 /**
893 ** HTMLFILTER ROUTINES
894 */
895
896 /**
897 * This function is more or less a wrapper around stripslashes. Apparently
898 * Explorer is stupid enough to just remove the backslashes and then
899 * execute the content of the attribute as if nothing happened.
900 * Who does that?
901 *
902 * @param attvalue The value of the attribute
903 * @return attvalue The value of the attribute stripslashed.
904 */
905 function sq_unbackslash($attvalue){
906 /**
907 * Remove any backslashes. See if there are any first.
908 */
909
910 if (strstr($attvalue, '\\') !== false){
911 $attvalue = stripslashes($attvalue);
912 }
913 return $attvalue;
914 }
915
916 /**
917 * Kill any tabs, newlines, or carriage returns. Our friends the
918 * makers of the browser with 95% market value decided that it'd
919 * be funny to make "java[tab]script" be just as good as "javascript".
920 *
921 * @param attvalue The attribute value before extraneous spaces removed.
922 * @return attvalue The attribute value after extraneous spaces removed.
923 */
924 function sq_unspace($attvalue){
925 if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
926 $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
927 $attvalue);
928 }
929 return $attvalue;
930 }
931
932 /**
933 * This function returns the final tag out of the tag name, an array
934 * of attributes, and the type of the tag. This function is called by
935 * sq_sanitize internally.
936 *
937 * @param $tagname the name of the tag.
938 * @param $attary the array of attributes and their values
939 * @param $tagtype The type of the tag (see in comments).
940 * @return a string with the final tag representation.
941 */
942 function sq_tagprint($tagname, $attary, $tagtype){
943 $me = 'sq_tagprint';
944
945 if ($tagtype == 2){
946 $fulltag = '</' . $tagname . '>';
947 } else {
948 $fulltag = '<' . $tagname;
949 if (is_array($attary) && sizeof($attary)){
950 $atts = Array();
951 while (list($attname, $attvalue) = each($attary)){
952 array_push($atts, "$attname=$attvalue");
953 }
954 $fulltag .= ' ' . join(" ", $atts);
955 }
956 if ($tagtype == 3){
957 $fulltag .= ' /';
958 }
959 $fulltag .= '>';
960 }
961 return $fulltag;
962 }
963
964 /**
965 * A small helper function to use with array_walk. Modifies a by-ref
966 * value and makes it lowercase.
967 *
968 * @param $val a value passed by-ref.
969 * @return void since it modifies a by-ref value.
970 */
971 function sq_casenormalize(&$val){
972 $val = strtolower($val);
973 }
974
975 /**
976 * This function skips any whitespace from the current position within
977 * a string and to the next non-whitespace value.
978 *
979 * @param $body the string
980 * @param $offset the offset within the string where we should start
981 * looking for the next non-whitespace character.
982 * @return the location within the $body where the next
983 * non-whitespace char is located.
984 */
985 function sq_skipspace($body, $offset){
986 $me = 'sq_skipspace';
987 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
988 if (sizeof($matches{1})){
989 $count = strlen($matches{1});
990 $offset += $count;
991 }
992 return $offset;
993 }
994
995 /**
996 * This function looks for the next character within a string. It's
997 * really just a glorified "strpos", except it catches if failures
998 * nicely.
999 *
1000 * @param $body The string to look for needle in.
1001 * @param $offset Start looking from this position.
1002 * @param $needle The character/string to look for.
1003 * @return location of the next occurance of the needle, or
1004 * strlen($body) if needle wasn't found.
1005 */
1006 function sq_findnxstr($body, $offset, $needle){
1007 $me = 'sq_findnxstr';
1008 $pos = strpos($body, $needle, $offset);
1009 if ($pos === FALSE){
1010 $pos = strlen($body);
1011 }
1012 return $pos;
1013 }
1014
1015 /**
1016 * This function takes a PCRE-style regexp and tries to match it
1017 * within the string.
1018 *
1019 * @param $body The string to look for needle in.
1020 * @param $offset Start looking from here.
1021 * @param $reg A PCRE-style regex to match.
1022 * @return Returns a false if no matches found, or an array
1023 * with the following members:
1024 * - integer with the location of the match within $body
1025 * - string with whatever content between offset and the match
1026 * - string with whatever it is we matched
1027 */
1028 function sq_findnxreg($body, $offset, $reg){
1029 $me = 'sq_findnxreg';
1030 $matches = Array();
1031 $retarr = Array();
1032 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1033 if (!isset($matches{0}) || !$matches{0}){
1034 $retarr = false;
1035 } else {
1036 $retarr{0} = $offset + strlen($matches{1});
1037 $retarr{1} = $matches{1};
1038 $retarr{2} = $matches{2};
1039 }
1040 return $retarr;
1041 }
1042
1043 /**
1044 * This function looks for the next tag.
1045 *
1046 * @param $body String where to look for the next tag.
1047 * @param $offset Start looking from here.
1048 * @return false if no more tags exist in the body, or
1049 * an array with the following members:
1050 * - string with the name of the tag
1051 * - array with attributes and their values
1052 * - integer with tag type (1, 2, or 3)
1053 * - integer where the tag starts (starting "<")
1054 * - integer where the tag ends (ending ">")
1055 * first three members will be false, if the tag is invalid.
1056 */
1057 function sq_getnxtag($body, $offset){
1058 $me = 'sq_getnxtag';
1059 if ($offset > strlen($body)){
1060 return false;
1061 }
1062 $lt = sq_findnxstr($body, $offset, "<");
1063 if ($lt == strlen($body)){
1064 return false;
1065 }
1066 /**
1067 * We are here:
1068 * blah blah <tag attribute="value">
1069 * \---------^
1070 */
1071 $pos = sq_skipspace($body, $lt+1);
1072 if ($pos >= strlen($body)){
1073 return Array(false, false, false, $lt, strlen($body));
1074 }
1075 /**
1076 * There are 3 kinds of tags:
1077 * 1. Opening tag, e.g.:
1078 * <a href="blah">
1079 * 2. Closing tag, e.g.:
1080 * </a>
1081 * 3. XHTML-style content-less tag, e.g.:
1082 * <img src="blah" />
1083 */
1084 $tagtype = false;
1085 switch (substr($body, $pos, 1)){
1086 case '/':
1087 $tagtype = 2;
1088 $pos++;
1089 break;
1090 case '!':
1091 /**
1092 * A comment or an SGML declaration.
1093 */
1094 if (substr($body, $pos+1, 2) == "--"){
1095 $gt = strpos($body, "-->", $pos);
1096 if ($gt === false){
1097 $gt = strlen($body);
1098 } else {
1099 $gt += 2;
1100 }
1101 return Array(false, false, false, $lt, $gt);
1102 } else {
1103 $gt = sq_findnxstr($body, $pos, ">");
1104 return Array(false, false, false, $lt, $gt);
1105 }
1106 break;
1107 default:
1108 /**
1109 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1110 * later.
1111 */
1112 $tagtype = 1;
1113 break;
1114 }
1115
1116 $tagname = '';
1117 /**
1118 * Look for next [\W-_], which will indicate the end of the tag name.
1119 */
1120 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1121 if ($regary == false){
1122 return Array(false, false, false, $lt, strlen($body));
1123 }
1124 list($pos, $tagname, $match) = $regary;
1125 $tagname = strtolower($tagname);
1126
1127 /**
1128 * $match can be either of these:
1129 * '>' indicating the end of the tag entirely.
1130 * '\s' indicating the end of the tag name.
1131 * '/' indicating that this is type-3 xhtml tag.
1132 *
1133 * Whatever else we find there indicates an invalid tag.
1134 */
1135 switch ($match){
1136 case '/':
1137 /**
1138 * This is an xhtml-style tag with a closing / at the
1139 * end, like so: <img src="blah" />. Check if it's followed
1140 * by the closing bracket. If not, then this tag is invalid
1141 */
1142 if (substr($body, $pos, 2) == "/>"){
1143 $pos++;
1144 $tagtype = 3;
1145 } else {
1146 $gt = sq_findnxstr($body, $pos, ">");
1147 $retary = Array(false, false, false, $lt, $gt);
1148 return $retary;
1149 }
1150 case '>':
1151 return Array($tagname, false, $tagtype, $lt, $pos);
1152 break;
1153 default:
1154 /**
1155 * Check if it's whitespace
1156 */
1157 if (!preg_match('/\s/', $match)){
1158 /**
1159 * This is an invalid tag! Look for the next closing ">".
1160 */
1161 $gt = sq_findnxstr($body, $lt, ">");
1162 return Array(false, false, false, $lt, $gt);
1163 }
1164 break;
1165 }
1166
1167 /**
1168 * At this point we're here:
1169 * <tagname attribute='blah'>
1170 * \-------^
1171 *
1172 * At this point we loop in order to find all attributes.
1173 */
1174 $attname = '';
1175 $attary = Array();
1176
1177 while ($pos <= strlen($body)){
1178 $pos = sq_skipspace($body, $pos);
1179 if ($pos == strlen($body)){
1180 /**
1181 * Non-closed tag.
1182 */
1183 return Array(false, false, false, $lt, $pos);
1184 }
1185 /**
1186 * See if we arrived at a ">" or "/>", which means that we reached
1187 * the end of the tag.
1188 */
1189 $matches = Array();
1190 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1191 /**
1192 * Yep. So we did.
1193 */
1194 $pos += strlen($matches{1});
1195 if ($matches{2} == "/>"){
1196 $tagtype = 3;
1197 $pos++;
1198 }
1199 return Array($tagname, $attary, $tagtype, $lt, $pos);
1200 }
1201
1202 /**
1203 * There are several types of attributes, with optional
1204 * [:space:] between members.
1205 * Type 1:
1206 * attrname[:space:]=[:space:]'CDATA'
1207 * Type 2:
1208 * attrname[:space:]=[:space:]"CDATA"
1209 * Type 3:
1210 * attr[:space:]=[:space:]CDATA
1211 * Type 4:
1212 * attrname
1213 *
1214 * We leave types 1 and 2 the same, type 3 we check for
1215 * '"' and convert to "&quot" if needed, then wrap in
1216 * double quotes. Type 4 we convert into:
1217 * attrname="yes".
1218 */
1219 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1220 if ($regary == false){
1221 /**
1222 * Looks like body ended before the end of tag.
1223 */
1224 return Array(false, false, false, $lt, strlen($body));
1225 }
1226 list($pos, $attname, $match) = $regary;
1227 $attname = strtolower($attname);
1228 /**
1229 * We arrived at the end of attribute name. Several things possible
1230 * here:
1231 * '>' means the end of the tag and this is attribute type 4
1232 * '/' if followed by '>' means the same thing as above
1233 * '\s' means a lot of things -- look what it's followed by.
1234 * anything else means the attribute is invalid.
1235 */
1236 switch($match){
1237 case '/':
1238 /**
1239 * This is an xhtml-style tag with a closing / at the
1240 * end, like so: <img src="blah" />. Check if it's followed
1241 * by the closing bracket. If not, then this tag is invalid
1242 */
1243 if (substr($body, $pos, 2) == "/>"){
1244 $pos++;
1245 $tagtype = 3;
1246 } else {
1247 $gt = sq_findnxstr($body, $pos, ">");
1248 $retary = Array(false, false, false, $lt, $gt);
1249 return $retary;
1250 }
1251 case '>':
1252 $attary{$attname} = '"yes"';
1253 return Array($tagname, $attary, $tagtype, $lt, $pos);
1254 break;
1255 default:
1256 /**
1257 * Skip whitespace and see what we arrive at.
1258 */
1259 $pos = sq_skipspace($body, $pos);
1260 $char = substr($body, $pos, 1);
1261 /**
1262 * Two things are valid here:
1263 * '=' means this is attribute type 1 2 or 3.
1264 * \w means this was attribute type 4.
1265 * anything else we ignore and re-loop. End of tag and
1266 * invalid stuff will be caught by our checks at the beginning
1267 * of the loop.
1268 */
1269 if ($char == "="){
1270 $pos++;
1271 $pos = sq_skipspace($body, $pos);
1272 /**
1273 * Here are 3 possibilities:
1274 * "'" attribute type 1
1275 * '"' attribute type 2
1276 * everything else is the content of tag type 3
1277 */
1278 $quot = substr($body, $pos, 1);
1279 if ($quot == "'"){
1280 $regary = sq_findnxreg($body, $pos+1, "\'");
1281 if ($regary == false){
1282 return Array(false, false, false, $lt, strlen($body));
1283 }
1284 list($pos, $attval, $match) = $regary;
1285 $pos++;
1286 $attary{$attname} = "'" . $attval . "'";
1287 } else if ($quot == '"'){
1288 $regary = sq_findnxreg($body, $pos+1, '\"');
1289 if ($regary == false){
1290 return Array(false, false, false, $lt, strlen($body));
1291 }
1292 list($pos, $attval, $match) = $regary;
1293 $pos++;
1294 $attary{$attname} = '"' . $attval . '"';
1295 } else {
1296 /**
1297 * These are hateful. Look for \s, or >.
1298 */
1299 $regary = sq_findnxreg($body, $pos, "[\s>]");
1300 if ($regary == false){
1301 return Array(false, false, false, $lt, strlen($body));
1302 }
1303 list($pos, $attval, $match) = $regary;
1304 /**
1305 * If it's ">" it will be caught at the top.
1306 */
1307 $attval = preg_replace("/\"/s", "&quot;", $attval);
1308 $attary{$attname} = '"' . $attval . '"';
1309 }
1310 } else if (preg_match("|[\w/>]|", $char)) {
1311 /**
1312 * That was attribute type 4.
1313 */
1314 $attary{$attname} = '"yes"';
1315 } else {
1316 /**
1317 * An illegal character. Find next '>' and return.
1318 */
1319 $gt = sq_findnxstr($body, $pos, ">");
1320 return Array(false, false, false, $lt, $gt);
1321 }
1322 break;
1323 }
1324 }
1325 /**
1326 * The fact that we got here indicates that the tag end was never
1327 * found. Return invalid tag indication so it gets stripped.
1328 */
1329 return Array(false, false, false, $lt, strlen($body));
1330 }
1331
1332 /**
1333 * This function checks attribute values for entity-encoded values
1334 * and returns them translated into 8-bit strings so we can run
1335 * checks on them.
1336 *
1337 * @param $attvalue A string to run entity check against.
1338 * @return Translated value.
1339 */
1340
1341 function sq_deent($attvalue){
1342 $me = 'sq_deent';
1343 /**
1344 * See if we have to run the checks first. All entities must start
1345 * with "&".
1346 */
1347 if (strpos($attvalue, '&') === false){
1348 return $attvalue;
1349 }
1350 /**
1351 * Check named entities first.
1352 */
1353 $trans = get_html_translation_table(HTML_ENTITIES);
1354 /**
1355 * Leave &quot; in, as it can mess us up.
1356 */
1357 $trans = array_flip($trans);
1358 unset($trans{'&quot;'});
1359 while (list($ent, $val) = each($trans)){
1360 $attvalue = preg_replace('/' . $ent . '*/si', $val, $attvalue);
1361 }
1362 /**
1363 * Now translate numbered entities from 1 to 255 if needed.
1364 */
1365 if (strpos($attvalue, '#') !== false){
1366 $omit = Array(34, 39);
1367 for ($asc = 256; $asc >= 0; $asc--){
1368 if (!in_array($asc, $omit)){
1369 $chr = chr($asc);
1370 $octrule = '/\&#0*' . $asc . ';*/si';
1371 $hexrule = '/\&#x0*' . dechex($asc) . ';*/si';
1372 $attvalue = preg_replace($octrule, $chr, $attvalue);
1373 $attvalue = preg_replace($hexrule, $chr, $attvalue);
1374 }
1375 }
1376 }
1377 return $attvalue;
1378 }
1379
1380 /**
1381 * This function runs various checks against the attributes.
1382 *
1383 * @param $tagname String with the name of the tag.
1384 * @param $attary Array with all tag attributes.
1385 * @param $rm_attnames See description for sq_sanitize
1386 * @param $bad_attvals See description for sq_sanitize
1387 * @param $add_attr_to_tag See description for sq_sanitize
1388 * @param $message message object
1389 * @param $id message id
1390 * @return Array with modified attributes.
1391 */
1392 function sq_fixatts($tagname,
1393 $attary,
1394 $rm_attnames,
1395 $bad_attvals,
1396 $add_attr_to_tag,
1397 $message,
1398 $id,
1399 $mailbox
1400 ){
1401 $me = 'sq_fixatts';
1402 while (list($attname, $attvalue) = each($attary)){
1403 /**
1404 * See if this attribute should be removed.
1405 */
1406 foreach ($rm_attnames as $matchtag=>$matchattrs){
1407 if (preg_match($matchtag, $tagname)){
1408 foreach ($matchattrs as $matchattr){
1409 if (preg_match($matchattr, $attname)){
1410 unset($attary{$attname});
1411 continue;
1412 }
1413 }
1414 }
1415 }
1416 /**
1417 * Remove any backslashes, entities, and extraneous whitespace.
1418 */
1419 $attvalue = sq_unbackslash($attvalue);
1420 $attvalue = sq_deent($attvalue);
1421 $attvalue = sq_unspace($attvalue);
1422
1423 /**
1424 * Remove \r \n \t \0 " " "\\"
1425 */
1426 $attvalue = str_replace(Array("\r", "\n", "\t", "\0", " ", "\\"),
1427 Array('', '','','','',''), $attvalue);
1428
1429 /**
1430 * Now let's run checks on the attvalues.
1431 * I don't expect anyone to comprehend this. If you do,
1432 * get in touch with me so I can drive to where you live and
1433 * shake your hand personally. :)
1434 */
1435 foreach ($bad_attvals as $matchtag=>$matchattrs){
1436 if (preg_match($matchtag, $tagname)){
1437 foreach ($matchattrs as $matchattr=>$valary){
1438 if (preg_match($matchattr, $attname)){
1439 /**
1440 * There are two arrays in valary.
1441 * First is matches.
1442 * Second one is replacements
1443 */
1444 list($valmatch, $valrepl) = $valary;
1445 $newvalue =
1446 preg_replace($valmatch, $valrepl, $attvalue);
1447 if ($newvalue != $attvalue){
1448 $attary{$attname} = $newvalue;
1449 }
1450 }
1451 }
1452 }
1453 }
1454
1455
1456 /**
1457 * Replace empty src tags with the blank image. src is only used
1458 * for frames, images, and image inputs. Doing a replace should
1459 * not affect them working as should be, however it will stop
1460 * IE from being kicked off when src for img tags are not set
1461 */
1462 if (($attname == 'src') && ($attvalue == '""')) {
1463 $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1464 }
1465
1466 /**
1467 * Turn cid: urls into http-friendly ones.
1468 */
1469 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1470 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1471 }
1472
1473 /**
1474 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1475 * One day MS might actually make it match something useful, for now, falling
1476 * back to using cid2http, so we can grab the blank.png.
1477 */
1478 if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1479 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1480 }
1481
1482 }
1483 /**
1484 * See if we need to append any attributes to this tag.
1485 */
1486 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1487 if (preg_match($matchtag, $tagname)){
1488 $attary = array_merge($attary, $addattary);
1489 }
1490 }
1491 return $attary;
1492 }
1493
1494 /**
1495 * This function edits the style definition to make them friendly and
1496 * usable in SquirrelMail.
1497 *
1498 * @param $message the message object
1499 * @param $id the message id
1500 * @param $content a string with whatever is between <style> and </style>
1501 * @param $mailbox the message mailbox
1502 * @return a string with edited content.
1503 */
1504 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1505 global $view_unsafe_images;
1506 $me = 'sq_fixstyle';
1507 $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1508 if ($ret == FALSE){
1509 return array(FALSE, strlen($body));
1510 }
1511 $newpos = $ret[0] + strlen($ret[2]);
1512 $content = $ret[1];
1513 /**
1514 * First look for general BODY style declaration, which would be
1515 * like so:
1516 * body {background: blah-blah}
1517 * and change it to .bodyclass so we can just assign it to a <div>
1518 */
1519 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1520 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1521 /**
1522 * Fix url('blah') declarations.
1523 */
1524 $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1525 "url(\\1$secremoveimg\\2)", $content);
1526 /**
1527 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1528 * is false.
1529 */
1530 if (!$view_unsafe_images){
1531 $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1532 "url(\\1$secremoveimg\\2)", $content);
1533 }
1534
1535 /**
1536 * Fix urls that refer to cid:
1537 */
1538 while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1539 $content, $matches)){
1540 $cidurl = $matches{1};
1541 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1542 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1543 "url($httpurl)", $content);
1544 }
1545
1546 /**
1547 * Fix stupid css declarations which lead to vulnerabilities
1548 * in IE.
1549 */
1550 $match = Array('/expression/i',
1551 '/behaviou*r/i',
1552 '/binding/i',
1553 '/include-source/i');
1554 $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1555 $content = preg_replace($match, $replace, $content);
1556 return array($content, $newpos);
1557 }
1558
1559 /**
1560 * This function converts cid: url's into the ones that can be viewed in
1561 * the browser.
1562 *
1563 * @param $message the message object
1564 * @param $id the message id
1565 * @param $cidurl the cid: url.
1566 * @param $mailbox the message mailbox
1567 * @return a string with a http-friendly url
1568 */
1569 function sq_cid2http($message, $id, $cidurl, $mailbox){
1570 /**
1571 * Get rid of quotes.
1572 */
1573 $quotchar = substr($cidurl, 0, 1);
1574 if ($quotchar == '"' || $quotchar == "'"){
1575 $cidurl = str_replace($quotchar, "", $cidurl);
1576 } else {
1577 $quotchar = '';
1578 }
1579 $cidurl = substr(trim($cidurl), 4);
1580 $linkurl = find_ent_id($cidurl, $message);
1581 /* in case of non-save cid links $httpurl should be replaced by a sort of
1582 unsave link image */
1583 $httpurl = '';
1584
1585 /**
1586 * This is part of a fix for Outlook Express 6.x generating
1587 * cid URLs without creating content-id headers. These images are
1588 * not part of the multipart/related html mail. The html contains
1589 * <img src="cid:{some_id}/image_filename.ext"> references to
1590 * attached images with as goal to render them inline although
1591 * the attachment disposition property is not inline.
1592 **/
1593
1594 if (empty($linkurl)) {
1595 if (preg_match('/{.*}\//', $cidurl)) {
1596 $cidurl = preg_replace('/{.*}\//','', $cidurl);
1597 if (!empty($cidurl)) {
1598 $linkurl = find_ent_id($cidurl, $message);
1599 }
1600 }
1601 }
1602
1603 if (!empty($linkurl)) {
1604 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1605 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1606 '&amp;ent_id=' . $linkurl . $quotchar;
1607 } else {
1608 /**
1609 * If we couldn't generate a proper img url, drop in a blank image
1610 * instead of sending back empty, otherwise it causes unusual behaviour
1611 */
1612 $httpurl = $quotechar . SM_PATH . 'images/blank.png';
1613 }
1614
1615 return $httpurl;
1616 }
1617
1618 /**
1619 * This function changes the <body> tag into a <div> tag since we
1620 * can't really have a body-within-body.
1621 *
1622 * @param $attary an array of attributes and values of <body>
1623 * @param $mailbox mailbox we're currently reading (for cid2http)
1624 * @param $message current message (for cid2http)
1625 * @param $id current message id (for cid2http)
1626 * @return a modified array of attributes to be set for <div>
1627 */
1628 function sq_body2div($attary, $mailbox, $message, $id){
1629 $me = 'sq_body2div';
1630 $divattary = Array('class' => "'bodyclass'");
1631 $text = '#000000';
1632 $has_bgc_stl = $has_txt_stl = false;
1633 $styledef = '';
1634 if (is_array($attary) && sizeof($attary) > 0){
1635 foreach ($attary as $attname=>$attvalue){
1636 $quotchar = substr($attvalue, 0, 1);
1637 $attvalue = str_replace($quotchar, "", $attvalue);
1638 switch ($attname){
1639 case 'background':
1640 $attvalue = sq_cid2http($message, $id,
1641 $attvalue, $mailbox);
1642 $styledef .= "background-image: url('$attvalue'); ";
1643 break;
1644 case 'bgcolor':
1645 $has_bgc_stl = true;
1646 $styledef .= "background-color: $attvalue; ";
1647 break;
1648 case 'text':
1649 $has_txt_stl = true;
1650 $styledef .= "color: $attvalue; ";
1651 break;
1652 }
1653 }
1654 // Outlook defines a white bgcolor and no text color. This can lead to
1655 // white text on a white bg with certain themes.
1656 if ($has_bgc_stl && !$has_txt_stl) {
1657 $styledef .= "color: $text; ";
1658 }
1659 if (strlen($styledef) > 0){
1660 $divattary{"style"} = "\"$styledef\"";
1661 }
1662 }
1663 return $divattary;
1664 }
1665
1666 /**
1667 * This is the main function and the one you should actually be calling.
1668 * There are several variables you should be aware of an which need
1669 * special description.
1670 *
1671 * Since the description is quite lengthy, see it here:
1672 * http://linux.duke.edu/projects/mini/htmlfilter/
1673 *
1674 * @param $body the string with HTML you wish to filter
1675 * @param $tag_list see description above
1676 * @param $rm_tags_with_content see description above
1677 * @param $self_closing_tags see description above
1678 * @param $force_tag_closing see description above
1679 * @param $rm_attnames see description above
1680 * @param $bad_attvals see description above
1681 * @param $add_attr_to_tag see description above
1682 * @param $message message object
1683 * @param $id message id
1684 * @return sanitized html safe to show on your pages.
1685 */
1686 function sq_sanitize($body,
1687 $tag_list,
1688 $rm_tags_with_content,
1689 $self_closing_tags,
1690 $force_tag_closing,
1691 $rm_attnames,
1692 $bad_attvals,
1693 $add_attr_to_tag,
1694 $message,
1695 $id,
1696 $mailbox
1697 ){
1698 $me = 'sq_sanitize';
1699 $rm_tags = array_shift($tag_list);
1700 /**
1701 * Normalize rm_tags and rm_tags_with_content.
1702 */
1703 @array_walk($tag_list, 'sq_casenormalize');
1704 @array_walk($rm_tags_with_content, 'sq_casenormalize');
1705 @array_walk($self_closing_tags, 'sq_casenormalize');
1706 /**
1707 * See if tag_list is of tags to remove or tags to allow.
1708 * false means remove these tags
1709 * true means allow these tags
1710 */
1711 $curpos = 0;
1712 $open_tags = Array();
1713 $trusted = "\n<!-- begin sanitized html -->\n";
1714 $skip_content = false;
1715 /**
1716 * Take care of netscape's stupid javascript entities like
1717 * &{alert('boo')};
1718 */
1719 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1720
1721 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1722 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1723 $free_content = substr($body, $curpos, $lt-$curpos);
1724 /**
1725 * Take care of <style>
1726 */
1727 if ($tagname == "style" && $tagtype == 1){
1728 list($free_content, $curpos) =
1729 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1730 if ($free_content != FALSE){
1731 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1732 $trusted .= $free_content;
1733 $trusted .= sq_tagprint($tagname, false, 2);
1734 }
1735 continue;
1736 }
1737 if ($skip_content == false){
1738 $trusted .= $free_content;
1739 }
1740 if ($tagname != FALSE){
1741 if ($tagtype == 2){
1742 if ($skip_content == $tagname){
1743 /**
1744 * Got to the end of tag we needed to remove.
1745 */
1746 $tagname = false;
1747 $skip_content = false;
1748 } else {
1749 if ($skip_content == false){
1750 if ($tagname == "body"){
1751 $tagname = "div";
1752 }
1753 if (isset($open_tags{$tagname}) &&
1754 $open_tags{$tagname} > 0){
1755 $open_tags{$tagname}--;
1756 } else {
1757 $tagname = false;
1758 }
1759 }
1760 }
1761 } else {
1762 /**
1763 * $rm_tags_with_content
1764 */
1765 if ($skip_content == false){
1766 /**
1767 * See if this is a self-closing type and change
1768 * tagtype appropriately.
1769 */
1770 if ($tagtype == 1
1771 && in_array($tagname, $self_closing_tags)){
1772 $tagtype = 3;
1773 }
1774 /**
1775 * See if we should skip this tag and any content
1776 * inside it.
1777 */
1778 if ($tagtype == 1 &&
1779 in_array($tagname, $rm_tags_with_content)){
1780 $skip_content = $tagname;
1781 } else {
1782 if (($rm_tags == false
1783 && in_array($tagname, $tag_list)) ||
1784 ($rm_tags == true &&
1785 !in_array($tagname, $tag_list))){
1786 $tagname = false;
1787 } else {
1788 /**
1789 * Convert body into div.
1790 */
1791 if ($tagname == "body"){
1792 $tagname = "div";
1793 $attary = sq_body2div($attary, $mailbox,
1794 $message, $id);
1795 }
1796 if ($tagtype == 1){
1797 if (isset($open_tags{$tagname})){
1798 $open_tags{$tagname}++;
1799 } else {
1800 $open_tags{$tagname}=1;
1801 }
1802 }
1803 /**
1804 * This is where we run other checks.
1805 */
1806 if (is_array($attary) && sizeof($attary) > 0){
1807 $attary = sq_fixatts($tagname,
1808 $attary,
1809 $rm_attnames,
1810 $bad_attvals,
1811 $add_attr_to_tag,
1812 $message,
1813 $id,
1814 $mailbox
1815 );
1816 }
1817 }
1818 }
1819 }
1820 }
1821 if ($tagname != false && $skip_content == false){
1822 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1823 }
1824 }
1825 $curpos = $gt+1;
1826 }
1827 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1828 if ($force_tag_closing == true){
1829 foreach ($open_tags as $tagname=>$opentimes){
1830 while ($opentimes > 0){
1831 $trusted .= '</' . $tagname . '>';
1832 $opentimes--;
1833 }
1834 }
1835 $trusted .= "\n";
1836 }
1837 $trusted .= "<!-- end sanitized html -->\n";
1838 return $trusted;
1839 }
1840
1841 /**
1842 * This is a wrapper function to call html sanitizing routines.
1843 *
1844 * @param $body the body of the message
1845 * @param $id the id of the message
1846 * @return a string with html safe to display in the browser.
1847 */
1848 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1849 global $attachment_common_show_images, $view_unsafe_images,
1850 $has_unsafe_images;
1851 /**
1852 * Don't display attached images in HTML mode.
1853 */
1854 $attachment_common_show_images = false;
1855 $tag_list = Array(
1856 false,
1857 "object",
1858 "meta",
1859 "html",
1860 "head",
1861 "base",
1862 "link",
1863 "frame",
1864 "iframe",
1865 "plaintext",
1866 "marquee"
1867 );
1868
1869 $rm_tags_with_content = Array(
1870 "script",
1871 "applet",
1872 "embed",
1873 "title",
1874 "frameset",
1875 "xml"
1876 );
1877
1878 $self_closing_tags = Array(
1879 "img",
1880 "br",
1881 "hr",
1882 "input",
1883 "outbind"
1884 );
1885
1886 $force_tag_closing = true;
1887
1888 $rm_attnames = Array(
1889 "/.*/" =>
1890 Array(
1891 "/target/i",
1892 "/^on.*/i",
1893 "/^dynsrc/i",
1894 "/^data.*/i",
1895 "/^lowsrc.*/i"
1896 )
1897 );
1898
1899 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1900 $bad_attvals = Array(
1901 "/.*/" =>
1902 Array(
1903 "/^src|background/i" =>
1904 Array(
1905 Array(
1906 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1907 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1908 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1909 ),
1910 Array(
1911 "\\1$secremoveimg\\2",
1912 "\\1$secremoveimg\\2",
1913 "\\1$secremoveimg\\2",
1914 "\\1$secremoveimg\\2"
1915 )
1916 ),
1917 "/^href|action/i" =>
1918 Array(
1919 Array(
1920 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1921 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1922 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1923 ),
1924 Array(
1925 "\\1#\\1",
1926 "\\1#\\1",
1927 "\\1#\\1",
1928 "\\1#\\1"
1929 )
1930 ),
1931 "/^style/i" =>
1932 Array(
1933 Array(
1934 "/expression/i",
1935 "/binding/i",
1936 "/behaviou*r/i",
1937 "/include-source/i",
1938 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1939 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1940 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
1941 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
1942 ),
1943 Array(
1944 "idiocy",
1945 "idiocy",
1946 "idiocy",
1947 "idiocy",
1948 "url(\\1#\\1)",
1949 "url(\\1#\\1)",
1950 "url(\\1#\\1)",
1951 "url(\\1#\\1)",
1952 "\\1:url(\\2#\\3)"
1953 )
1954 )
1955 )
1956 );
1957 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1958 $view_unsafe_images = false;
1959 }
1960 if (!$view_unsafe_images){
1961 /**
1962 * Remove any references to http/https if view_unsafe_images set
1963 * to false.
1964 */
1965 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1966 '/^([\'\"])\s*https*:.*([\'\"])/si');
1967 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1968 "\\1$secremoveimg\\1");
1969 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1970 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1971 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1972 "url(\\1$secremoveimg\\1)");
1973 }
1974
1975 $add_attr_to_tag = Array(
1976 "/^a$/i" =>
1977 Array('target'=>'"_new"',
1978 'title'=>'"'._("This external link will open in a new window").'"'
1979 )
1980 );
1981 $trusted = sq_sanitize($body,
1982 $tag_list,
1983 $rm_tags_with_content,
1984 $self_closing_tags,
1985 $force_tag_closing,
1986 $rm_attnames,
1987 $bad_attvals,
1988 $add_attr_to_tag,
1989 $message,
1990 $id,
1991 $mailbox
1992 );
1993 if (preg_match("|$secremoveimg|i", $trusted)){
1994 $has_unsafe_images = true;
1995 }
1996 return $trusted;
1997 }
1998
1999 /**
2000 * function SendDownloadHeaders - send file to the browser
2001 *
2002 * Original Source: SM core src/download.php
2003 * moved here to make it available to other code, and separate
2004 * front end from back end functionality.
2005 *
2006 * @param string $type0 first half of mime type
2007 * @param string $type1 second half of mime type
2008 * @param string $filename filename to tell the browser for downloaded file
2009 * @param boolean $force whether to force the download dialog to pop
2010 * @param optional integer $filesize send the Content-Header and length to the browser
2011 * @return void
2012 */
2013 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2014 global $languages, $squirrelmail_language;
2015 $isIE = $isIE6 = 0;
2016
2017 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2018
2019 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2020 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2021 $isIE = 1;
2022 }
2023
2024 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
2025 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2026 $isIE6 = 1;
2027 }
2028
2029 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2030 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2031 $filename =
2032 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2033 } else {
2034 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2035 }
2036
2037 // A Pox on Microsoft and it's Internet Explorer!
2038 //
2039 // IE has lots of bugs with file downloads.
2040 // It also has problems with SSL. Both of these cause problems
2041 // for us in this function.
2042 //
2043 // See this article on Cache Control headers and SSL
2044 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2045 //
2046 // The best thing you can do for IE is to upgrade to the latest
2047 // version
2048 //set all the Cache Control Headers for IE
2049 if ($isIE) {
2050 $filename=rawurlencode($filename);
2051 header ("Pragma: public");
2052 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); # HTTP/1.1
2053 header ("Cache-Control: post-check=0, pre-check=0", false);
2054 header ("Cache-control: private");
2055
2056 //set the inline header for IE, we'll add the attachment header later if we need it
2057 header ("Content-Disposition: inline; filename=$filename");
2058 }
2059
2060 if (!$force) {
2061 // Try to show in browser window
2062 header ("Content-Disposition: inline; filename=\"$filename\"");
2063 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2064 } else {
2065 // Try to pop up the "save as" box
2066
2067 // IE makes this hard. It pops up 2 save boxes, or none.
2068 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2069 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2070 // But, according to Microsoft, it is "RFC compliant but doesn't
2071 // take into account some deviations that allowed within the
2072 // specification." Doesn't that mean RFC non-compliant?
2073 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2074
2075 // all browsers need the application/octet-stream header for this
2076 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2077
2078 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2079 // Do not have quotes around filename, but that applied to
2080 // "attachment"... does it apply to inline too?
2081 header ("Content-Disposition: attachment; filename=\"$filename\"");
2082
2083 if ($isIE && !$isIE6) {
2084 // This combination seems to work mostly. IE 5.5 SP 1 has
2085 // known issues (see the Microsoft Knowledge Base)
2086
2087 // This works for most types, but doesn't work with Word files
2088 header ("Content-Type: application/download; name=\"$filename\"");
2089
2090 // These are spares, just in case. :-)
2091 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2092 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2093 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2094 } else {
2095 // another application/octet-stream forces download for Netscape
2096 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2097 }
2098 }
2099
2100 //send the content-length header if the calling function provides it
2101 if ($filesize > 0) {
2102 header("Content-Length: $filesize");
2103 }
2104
2105 } // end fn SendDownloadHeaders
2106
2107 ?>