projects / squirrelmail.git / blob
? search:


071e90cbfa92564751a2dd1cd14648502b1f77e3
[squirrelmail.git] / functions / mime.php
1 <?php
2
3 /**
4 * mime.php
5 *
6 * This contains the functions necessary to detect and decode MIME
7 * messages.
8 *
9 * @copyright &copy; 1999-2006 The SquirrelMail Project Team
10 * @license http://opensource.org/licenses/gpl-license.php GNU Public License
11 * @version $Id$
12 * @package squirrelmail
13 */
14
15 /**
16 * dependency information
17 functions dependency
18 mime_structure
19 class/mime/Message.class.php
20 Message::parseStructure
21 functions/page_header.php
22 displayPageHeader
23 functions/display_messages.php
24 plain_error_message
25 mime_fetch_body
26 functions/imap_general.php
27 sqimap_run_command
28 mime_print_body_lines
29
30
31
32 functions/imap.php
33 functions/attachment_common.php
34 functions/display_messages.php
35
36 magicHtml => url_parser
37 translateText => url_parser
38
39 */
40
41
42 /* -------------------------------------------------------------------------- */
43 /* MIME DECODING */
44 /* -------------------------------------------------------------------------- */
45
46 /**
47 * Get the MIME structure
48 *
49 * This function gets the structure of a message and stores it in the "message" class.
50 * It will return this object for use with all relevant header information and
51 * fully parsed into the standard "message" object format.
52 */
53 function mime_structure ($bodystructure, $flags=array()) {
54
55 /* Isolate the body structure and remove beginning and end parenthesis. */
56 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
57 $read = trim(substr ($read, 0, -1));
58 $i = 0;
59 $msg = Message::parseStructure($read,$i);
60
61 if (!is_object($msg)) {
62 global $color, $mailbox;
63 /* removed urldecode because $_GET is auto urldecoded ??? */
64 displayPageHeader( $color, $mailbox );
65 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
66 $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
67 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
68 plain_error_message( $errormessage, $color );
69 echo '</body></html>';
70 exit;
71 }
72 if (count($flags)) {
73 foreach ($flags as $flag) {
74 $char = strtoupper($flag{1});
75 switch ($char) {
76 case 'S':
77 if (strtolower($flag) == '\\seen') {
78 $msg->is_seen = true;
79 }
80 break;
81 case 'A':
82 if (strtolower($flag) == '\\answered') {
83 $msg->is_answered = true;
84 }
85 break;
86 case 'D':
87 if (strtolower($flag) == '\\deleted') {
88 $msg->is_deleted = true;
89 }
90 break;
91 case 'F':
92 if (strtolower($flag) == '\\flagged') {
93 $msg->is_flagged = true;
94 }
95 break;
96 case 'M':
97 if (strtolower($flag) == '$mdnsent') {
98 $msg->is_mdnsent = true;
99 }
100 break;
101 default:
102 break;
103 }
104 }
105 }
106 // listEntities($msg);
107 return $msg;
108 }
109
110
111
112 /* This starts the parsing of a particular structure. It is called recursively,
113 * so it can be passed different structures. It returns an object of type
114 * $message.
115 * First, it checks to see if it is a multipart message. If it is, then it
116 * handles that as it sees is necessary. If it is just a regular entity,
117 * then it parses it and adds the necessary header information (by calling out
118 * to mime_get_elements()
119 */
120
121 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
122 /* Do a bit of error correction. If we couldn't find the entity id, just guess
123 * that it is the first one. That is usually the case anyway.
124 */
125
126 if (!$ent_id) {
127 $cmd = "FETCH $id BODY[]";
128 } else {
129 $cmd = "FETCH $id BODY[$ent_id]";
130 }
131
132 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
133
134 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
135 do {
136 $topline = trim(array_shift($data));
137 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
138
139 $wholemessage = implode('', $data);
140 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
141 $ret = substr($wholemessage, 0, $regs[1]);
142 /* There is some information in the content info header that could be important
143 * in order to parse html messages. Let's get them here.
144 */
145 // if ($ret{0} == '<') {
146 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
147 // }
148 } else if (ereg('"([^"]*)"', $topline, $regs)) {
149 $ret = $regs[1];
150 } else {
151 global $where, $what, $mailbox, $passed_id, $startMessage;
152 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
153 if (isset($where) && isset($what)) {
154 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
155 } else {
156 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
157 }
158 $par .= '&amp;response=' . urlencode($response) .
159 '&amp;message=' . urlencode($message) .
160 '&amp;topline=' . urlencode($topline);
161
162 echo '<tt><br />' .
163 '<table width="80%"><tr>' .
164 '<tr><td colspan="2">' .
165 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
166 '</td></tr>' .
167 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
168 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
169 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
170 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
171 "</table><br /></tt></font><hr />";
172
173 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
174 array_shift($data);
175 $wholemessage = implode('', $data);
176
177 $ret = $wholemessage;
178 }
179 return $ret;
180 }
181
182 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout') {
183
184 /* Don't kill the connection if the browser is over a dialup
185 * and it would take over 30 seconds to download it.
186 * Don't call set_time_limit in safe mode.
187 */
188
189 if (!ini_get('safe_mode')) {
190 set_time_limit(0);
191 }
192 /* in case of base64 encoded attachments, do not buffer them.
193 Instead, echo the decoded attachment directly to screen */
194 if (strtolower($encoding) == 'base64') {
195 if (!$ent_id) {
196 $query = "FETCH $id BODY[]";
197 } else {
198 $query = "FETCH $id BODY[$ent_id]";
199 }
200 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
201 } else {
202 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
203 if (is_resource($rStream)) {
204 fputs($rStream,decodeBody($body,$encoding));
205 } else {
206 echo decodeBody($body, $encoding);
207 }
208 }
209
210 /*
211 TODO, use the same method for quoted printable.
212 However, I assume that quoted printable attachments aren't that large
213 so the performancegain / memory usage drop will be minimal.
214 If we decide to add that then we need to adapt sqimap_fread because
215 we need to split te result on \n and fread doesn't stop at \n. That
216 means we also should provide $results from sqimap_fread (by ref) to
217 te function and set $no_return to false. The $filter function for
218 quoted printable should handle unsetting of $results.
219 */
220 /*
221 TODO 2: find out how we write to the output stream php://stdout. fwrite
222 doesn't work because 'php://stdout isn't a stream.
223 */
224
225 return;
226 }
227
228 /* -[ END MIME DECODING ]----------------------------------------------------------- */
229
230 /* This is here for debugging purposes. It will print out a list
231 * of all the entity IDs that are in the $message object.
232 */
233 function listEntities ($message) {
234 if ($message) {
235 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
236 for ($i = 0; isset($message->entities[$i]); $i++) {
237 echo "$i : ";
238 $msg = listEntities($message->entities[$i]);
239
240 if ($msg) {
241 echo "return: ";
242 return $msg;
243 }
244 }
245 }
246 }
247
248 function getPriorityStr($priority) {
249 $priority_level = substr($priority,0,1);
250
251 switch($priority_level) {
252 /* Check for a higher then normal priority. */
253 case '1':
254 case '2':
255 $priority_string = _("High");
256 break;
257
258 /* Check for a lower then normal priority. */
259 case '4':
260 case '5':
261 $priority_string = _("Low");
262 break;
263
264 /* Check for a normal priority. */
265 case '3':
266 default:
267 $priority_level = '3';
268 $priority_string = _("Normal");
269 break;
270
271 }
272 return $priority_string;
273 }
274
275 /* returns a $message object for a particular entity id */
276 function getEntity ($message, $ent_id) {
277 return $message->getEntity($ent_id);
278 }
279
280 /* translateText
281 * Extracted from strings.php 23/03/2002
282 */
283
284 function translateText(&$body, $wrap_at, $charset) {
285 global $where, $what; /* from searching */
286 global $color; /* color theme */
287
288 // require_once(SM_PATH . 'functions/url_parser.php');
289
290 $body_ary = explode("\n", $body);
291 for ($i=0; $i < count($body_ary); $i++) {
292 $line = $body_ary[$i];
293 if (strlen($line) - 2 >= $wrap_at) {
294 sqWordWrap($line, $wrap_at, $charset);
295 }
296 $line = charset_decode($charset, $line);
297 $line = str_replace("\t", ' ', $line);
298
299 parseUrl ($line);
300
301 $quotes = 0;
302 $pos = 0;
303 $j = strlen($line);
304
305 while ($pos < $j) {
306 if ($line[$pos] == ' ') {
307 $pos++;
308 } else if (strpos($line, '&gt;', $pos) === $pos) {
309 $pos += 4;
310 $quotes++;
311 } else {
312 break;
313 }
314 }
315
316 if ($quotes % 2) {
317 $line = '<span class="quote1">' . $line . '</span>';
318 } elseif ($quotes) {
319 $line = '<span class="quote2">' . $line . '</span>';
320 }
321
322 $body_ary[$i] = $line;
323 }
324 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
325 }
326
327 /**
328 * This returns a parsed string called $body. That string can then
329 * be displayed as the actual message in the HTML. It contains
330 * everything needed, including HTML Tags, Attachments at the
331 * bottom, etc.
332 *
333 * Since 1.2.0 function uses message_body hook.
334 * Till 1.3.0 function included output of formatAttachments().
335 *
336 * @param resource $imap_stream imap connection resource
337 * @param object $message squirrelmail message object
338 * @param array $color squirrelmail color theme array
339 * @param integer $wrap_at number of characters per line
340 * @param string $ent_num (since 1.3.0) message part id
341 * @param integer $id (since 1.3.0) message id
342 * @param string $mailbox (since 1.3.0) imap folder name
343 * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
344 * @return string html formated message text
345 */
346 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
347 /* This if statement checks for the entity to show as the
348 * primary message. To add more of them, just put them in the
349 * order that is their priority.
350 */
351 global $startMessage, $languages, $squirrelmail_language,
352 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
353 $use_iframe, $iframe_height, $download_and_unsafe_link,
354 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text;
355
356 // workaround for not updated config.php
357 if (! isset($use_iframe)) $use_iframe = false;
358
359 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
360 $view_unsafe_images = false;
361 }
362
363 $body = '';
364 $urlmailbox = urlencode($mailbox);
365 $body_message = getEntity($message, $ent_num);
366 if (($body_message->header->type0 == 'text') ||
367 ($body_message->header->type0 == 'rfc822')) {
368 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
369 $body = decodeBody($body, $body_message->header->encoding);
370
371 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
372 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
373 if (mb_detect_encoding($body) != 'ASCII') {
374 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
375 }
376 }
377 $hookResults = do_hook("message_body", $body);
378 $body = $hookResults[1];
379
380 /* If there are other types that shouldn't be formatted, add
381 * them here.
382 */
383
384 if ($body_message->header->type1 == 'html') {
385 if ($show_html_default <> 1) {
386 $entity_conv = array('&nbsp;' => ' ',
387 '<p>' => "\n",
388 '<P>' => "\n",
389 '<br>' => "\n",
390 '<BR>' => "\n",
391 '<br />' => "\n",
392 '<BR />' => "\n",
393 '&gt;' => '>',
394 '&lt;' => '<');
395 $body = strtr($body, $entity_conv);
396 $body = strip_tags($body);
397 $body = trim($body);
398 translateText($body, $wrap_at,
399 $body_message->header->getParameter('charset'));
400 } elseif ($use_iframe && ! $clean) {
401 // $clean is used to remove iframe in printable view.
402
403 /**
404 * If we don't add html message between iframe tags,
405 * we must detect unsafe images and modify $has_unsafe_images.
406 */
407 $html_body = magicHTML($body, $id, $message, $mailbox);
408 // Convert character set in order to display html mails in different character set
409 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
410
411 // creating iframe url
412 $iframeurl=sqm_baseuri().'src/view_html.php?'
413 . 'mailbox=' . $urlmailbox
414 . '&amp;passed_id=' . $id
415 . '&amp;ent_id=' . $ent_num
416 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
417
418 global $oTemplate;
419 $oTemplate->assign('iframe_url', $iframeurl);
420 $oTemplate->assign('html_body', $html_body);
421
422 $body = $oTemplate->fetch('read_html_iframe.tpl');
423 } else {
424 // old way of html rendering
425 $body = magicHTML($body, $id, $message, $mailbox);
426 /**
427 * convert character set. charset_decode does not remove html special chars
428 * applied by magicHTML functions and does not sanitize them second time if
429 * fourth argument is true.
430 */
431 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
432 }
433 } else {
434 translateText($body, $wrap_at,
435 $body_message->header->getParameter('charset'));
436 }
437
438 // if this is the clean display (i.e. printer friendly), stop here.
439 if ( $clean ) {
440 return $body;
441 }
442
443 $download_and_unsafe_link = '';
444
445 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
446 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
447 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
448 if (isset($passed_ent_id)) {
449 $link .= '&amp;passed_ent_id='.$passed_ent_id;
450 }
451 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
452 $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $download_href .'">' . _("Download this as a file") . '</a>';
453 if ($view_unsafe_images) {
454 $text = _("Hide Unsafe Images");
455 } else {
456 if (isset($has_unsafe_images) && $has_unsafe_images) {
457 $link .= '&amp;view_unsafe_images=1';
458 $text = _("View Unsafe Images");
459 } else {
460 $text = '';
461 }
462 }
463 if($text != '') {
464 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
465 $unsafe_image_toggle_text = $text;
466 $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $unsafe_image_toggle_href .'">' . $text . '</a>';
467 }
468 }
469 return $body;
470 }
471
472 /**
473 * Generate attachments array for passing to templates. Separated from
474 * formatAttachments() below so that the same array can be given to the
475 * print-friendly version.
476 *
477 * @since 1.5.2
478 * @param object $message SquirrelMail message object
479 * @param array $exclude_id message parts that are not attachments.
480 * @param string $mailbox mailbox name
481 * @param integer $id message id
482 */
483 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
484 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
485
486 $att_ar = $message->getAttachments($exclude_id);
487 $urlMailbox = urlencode($mailbox);
488
489 $attachments = array();
490 foreach ($att_ar as $att) {
491 $ent = $att->entity_id;
492 $header = $att->header;
493 $type0 = strtolower($header->type0);
494 $type1 = strtolower($header->type1);
495 $name = '';
496 $links = array();
497 $links['download link']['text'] = _("Download");
498 $links['download link']['href'] = $base_uri .
499 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
500
501 if ($type0 =='message' && $type1 == 'rfc822') {
502 $default_page = $base_uri . 'src/read_body.php';
503 $rfc822_header = $att->rfc822_header;
504 $filename = $rfc822_header->subject;
505 if (trim( $filename ) == '') {
506 $filename = 'untitled-[' . $ent . ']' ;
507 }
508 $from_o = $rfc822_header->from;
509 if (is_object($from_o)) {
510 $from_name = decodeHeader($from_o->getAddress(false));
511 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
512 // something weird happens when a digest message is opened and you return to the digest
513 // now the from object is part of an array. Probably the parseHeader call overwrites the info
514 // retrieved from the bodystructure in a different way. We need to fix this later.
515 // possible starting point, do not fetch header we already have and inspect how
516 // the rfc822_header object behaves.
517 $from_name = decodeHeader($from_o[0]->getAddress(false));
518 } else {
519 $from_name = _("Unknown sender");
520 }
521 $description = _("From").': '.$from_name;
522 } else {
523 $default_page = $base_uri . 'src/download.php';
524 $filename = $att->getFilename();
525 if ($header->description) {
526 $description = decodeHeader($header->description);
527 } else {
528 $description = '';
529 }
530 }
531
532 $display_filename = $filename;
533 if (isset($passed_ent_id)) {
534 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
535 } else {
536 $passed_ent_id_link = '';
537 }
538 $defaultlink = $default_page . "?startMessage=$startMessage"
539 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
540 . '&amp;ent_id='.$ent.$passed_ent_id_link;
541 if ($where && $what) {
542 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
543 }
544 // IE does make use of mime content sniffing. Forcing a download
545 // prohibit execution of XSS inside an application/octet-stream attachment
546 if ($type0 == 'application' && $type1 == 'octet-stream') {
547 $defaultlink .= '&amp;absolute_dl=true';
548 }
549
550 /* This executes the attachment hook with a specific MIME-type.
551 * If that doesn't have results, it tries if there's a rule
552 * for a more generic type. Finally, a hook for ALL attachment
553 * types is run as well.
554 */
555 $hookresults = do_hook("attachment $type0/$type1", $links,
556 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
557 $display_filename, $where, $what);
558 if(count($hookresults[1]) <= 1) {
559 $hookresults = do_hook("attachment $type0/*", $links,
560 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
561 $display_filename, $where, $what);
562 }
563 $hookresults = do_hook("attachment */*", $hookresults[1],
564 $startMessage, $id, $urlMailbox, $ent, $hookresults[6],
565 $display_filename, $where, $what);
566
567 $links = $hookresults[1];
568 $defaultlink = $hookresults[6];
569
570 $this_attachment = array();
571 $this_attachment['Name'] = decodeHeader($display_filename);
572 $this_attachment['Description'] = $description;
573 $this_attachment['DefaultHREF'] = $defaultlink;
574 $this_attachment['DownloadHREF'] = $links['download link']['href'];
575 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
576 $this_attachment['Size'] = $header->size;
577 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
578 $this_attachment['OtherLinks'] = array();
579 foreach ($links as $val) {
580 if ($val['text']==_("Download") || $val['text'] == _("View"))
581 continue;
582 if (empty($val['text']) && empty($val['extra']))
583 continue;
584
585 $temp = array();
586 $temp['HREF'] = $val['href'];
587 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
588 $this_attachment['OtherLinks'][] = $temp;
589 }
590 $attachments[] = $this_attachment;
591
592 unset($links);
593 }
594
595 return $attachments;
596 }
597
598 /**
599 * Displays attachment links and information
600 *
601 * Since 1.3.0 function is not included in formatBody() call.
602 *
603 * Since 1.0.2 uses attachment $type0/$type1 hook.
604 * Since 1.2.5 uses attachment $type0/* hook.
605 * Since 1.5.0 uses attachments_bottom hook.
606 * Since 1.5.2 uses templates and does *not* return a value.
607 *
608 * @param object $message SquirrelMail message object
609 * @param array $exclude_id message parts that are not attachments.
610 * @param string $mailbox mailbox name
611 * @param integer $id message id
612 */
613 function formatAttachments($message, $exclude_id, $mailbox, $id) {
614 global $oTemplate;
615
616 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
617
618 $oTemplate->assign('attachments', $attach);
619 $oTemplate->display('read_attachments.tpl');
620 }
621
622 function sqimap_base64_decode(&$string) {
623
624 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
625 // fly decoding (to reduce memory usage) you have to check if the
626 // data has incomplete pairs
627
628 // Remove the noise in order to check if the 4 bytes pairs are complete
629 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
630
631 $sStringRem = '';
632 $iMod = strlen($string) % 4;
633 if ($iMod) {
634 $sStringRem = substr($string,-$iMod);
635 // Check if $sStringRem contains padding characters
636 if (substr($sStringRem,-1) != '=') {
637 $string = substr($string,0,-$iMod);
638 } else {
639 $sStringRem = '';
640 }
641 }
642 $string = base64_decode($string);
643 return $sStringRem;
644 }
645
646 /**
647 * Decodes encoded message body
648 *
649 * This function decodes the body depending on the encoding type.
650 * Currently quoted-printable and base64 encodings are supported.
651 * decode_body hook was added to this function in 1.4.2/1.5.0
652 * @param string $body encoded message body
653 * @param string $encoding used encoding
654 * @return string decoded string
655 * @since 1.0
656 */
657 function decodeBody($body, $encoding) {
658
659 $body = str_replace("\r\n", "\n", $body);
660 $encoding = strtolower($encoding);
661
662 $encoding_handler = do_hook_function('decode_body', $encoding);
663
664
665 // plugins get first shot at decoding the body
666 //
667 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
668 $body = $encoding_handler('decode', $body);
669
670 } elseif ($encoding == 'quoted-printable' ||
671 $encoding == 'quoted_printable') {
672 /**
673 * quoted_printable_decode() function is broken in older
674 * php versions. Text with \r\n decoding was fixed only
675 * in php 4.3.0. Minimal code requirement 4.0.4 +
676 * str_replace("\r\n", "\n", $body); call.
677 */
678 $body = quoted_printable_decode($body);
679 } elseif ($encoding == 'base64') {
680 $body = base64_decode($body);
681 }
682
683 // All other encodings are returned raw.
684 return $body;
685 }
686
687 /**
688 * Decodes headers
689 *
690 * This functions decode strings that is encoded according to
691 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
692 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
693 *
694 * @param string $string header string that has to be made readable
695 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
696 * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
697 * @param boolean $decide decide if string can be utfencoded. defaults to false
698 * @return string decoded header string
699 */
700 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
701 global $languages, $squirrelmail_language,$default_charset;
702 if (is_array($string)) {
703 $string = implode("\n", $string);
704 }
705
706 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
707 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
708 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
709 // Do we need to return at this point?
710 // return $string;
711 }
712 $i = 0;
713 $iLastMatch = -2;
714 $encoded = true;
715
716 $aString = explode(' ',$string);
717 $ret = '';
718 foreach ($aString as $chunk) {
719 if ($encoded && $chunk === '') {
720 continue;
721 } elseif ($chunk === '') {
722 $ret .= ' ';
723 continue;
724 }
725 $encoded = false;
726 /* if encoded words are not separated by a linear-space-white we still catch them */
727 $j = $i-1;
728
729 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
730 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
731 if ($iLastMatch !== $j) {
732 if ($htmlsave) {
733 $ret .= '&#32;';
734 } else {
735 $ret .= ' ';
736 }
737 }
738 $iLastMatch = $i;
739 $j = $i;
740 if ($htmlsave) {
741 $ret .= htmlspecialchars($res[1]);
742 } else {
743 $ret .= $res[1];
744 }
745 $encoding = ucfirst($res[3]);
746
747 /* decide about valid decoding */
748 if ($decide && is_conversion_safe($res[2])) {
749 $utfencode=true;
750 $can_be_encoded=true;
751 } else {
752 $can_be_encoded=false;
753 }
754 switch ($encoding)
755 {
756 case 'B':
757 $replace = base64_decode($res[4]);
758 if ($utfencode) {
759 if ($can_be_encoded) {
760 /* convert string to different charset,
761 * if functions asks for it (usually in compose)
762 */
763 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
764 } else {
765 // convert string to html codes in order to display it
766 $ret .= charset_decode($res[2],$replace);
767 }
768 } else {
769 if ($htmlsave) {
770 $replace = htmlspecialchars($replace);
771 }
772 $ret.= $replace;
773 }
774 break;
775 case 'Q':
776 $replace = str_replace('_', ' ', $res[4]);
777 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
778 $replace);
779 if ($utfencode) {
780 if ($can_be_encoded) {
781 /* convert string to different charset,
782 * if functions asks for it (usually in compose)
783 */
784 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
785 } else {
786 // convert string to html codes in order to display it
787 $replace = charset_decode($res[2], $replace);
788 }
789 } else {
790 if ($htmlsave) {
791 $replace = htmlspecialchars($replace);
792 }
793 }
794 $ret .= $replace;
795 break;
796 default:
797 break;
798 }
799 $chunk = $res[5];
800 $encoded = true;
801 }
802 if (!$encoded) {
803 if ($htmlsave) {
804 $ret .= '&#32;';
805 } else {
806 $ret .= ' ';
807 }
808 }
809
810 if (!$encoded && $htmlsave) {
811 $ret .= htmlspecialchars($chunk);
812 } else {
813 $ret .= $chunk;
814 }
815 ++$i;
816 }
817 /* remove the first added space */
818 if ($ret) {
819 if ($htmlsave) {
820 $ret = substr($ret,5);
821 } else {
822 $ret = substr($ret,1);
823 }
824 }
825
826 return $ret;
827 }
828
829 /**
830 * Encodes header
831 *
832 * Function uses XTRA_CODE _encodeheader function, if such function exists.
833 *
834 * Function uses Q encoding by default and encodes a string according to RFC
835 * 1522 for use in headers if it contains 8-bit characters or anything that
836 * looks like it should be encoded.
837 *
838 * Function switches to B encoding and encodeHeaderBase64() function, if
839 * string is 8bit and multibyte character set supported by mbstring extension
840 * is used. It can cause E_USER_NOTICE errors, if interface is used with
841 * multibyte character set unsupported by mbstring extension.
842 *
843 * @param string $string header string, that has to be encoded
844 * @return string quoted-printable encoded string
845 * @todo make $mb_charsets system wide constant
846 */
847 function encodeHeader ($string) {
848 global $default_charset, $languages, $squirrelmail_language;
849
850 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
851 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
852 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
853 }
854
855 // Use B encoding for multibyte charsets
856 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
857 if (in_array($default_charset,$mb_charsets) &&
858 in_array($default_charset,sq_mb_list_encodings()) &&
859 sq_is8bit($string)) {
860 return encodeHeaderBase64($string,$default_charset);
861 } elseif (in_array($default_charset,$mb_charsets) &&
862 sq_is8bit($string) &&
863 ! in_array($default_charset,sq_mb_list_encodings())) {
864 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
865 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
866 }
867
868 // Encode only if the string contains 8-bit characters or =?
869 $j = strlen($string);
870 $max_l = 75 - strlen($default_charset) - 7;
871 $aRet = array();
872 $ret = '';
873 $iEncStart = $enc_init = false;
874 $cur_l = $iOffset = 0;
875 for($i = 0; $i < $j; ++$i) {
876 switch($string{$i})
877 {
878 case '=':
879 case '<':
880 case '>':
881 case ',':
882 case '?':
883 case '_':
884 if ($iEncStart === false) {
885 $iEncStart = $i;
886 }
887 $cur_l+=3;
888 if ($cur_l > ($max_l-2)) {
889 /* if there is an stringpart that doesn't need encoding, add it */
890 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
891 $aRet[] = "=?$default_charset?Q?$ret?=";
892 $iOffset = $i;
893 $cur_l = 0;
894 $ret = '';
895 $iEncStart = false;
896 } else {
897 $ret .= sprintf("=%02X",ord($string{$i}));
898 }
899 break;
900 case '(':
901 case ')':
902 if ($iEncStart !== false) {
903 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
904 $aRet[] = "=?$default_charset?Q?$ret?=";
905 $iOffset = $i;
906 $cur_l = 0;
907 $ret = '';
908 $iEncStart = false;
909 }
910 break;
911 case ' ':
912 if ($iEncStart !== false) {
913 $cur_l++;
914 if ($cur_l > $max_l) {
915 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
916 $aRet[] = "=?$default_charset?Q?$ret?=";
917 $iOffset = $i;
918 $cur_l = 0;
919 $ret = '';
920 $iEncStart = false;
921 } else {
922 $ret .= '_';
923 }
924 }
925 break;
926 default:
927 $k = ord($string{$i});
928 if ($k > 126) {
929 if ($iEncStart === false) {
930 // do not start encoding in the middle of a string, also take the rest of the word.
931 $sLeadString = substr($string,0,$i);
932 $aLeadString = explode(' ',$sLeadString);
933 $sToBeEncoded = array_pop($aLeadString);
934 $iEncStart = $i - strlen($sToBeEncoded);
935 $ret .= $sToBeEncoded;
936 $cur_l += strlen($sToBeEncoded);
937 }
938 $cur_l += 3;
939 /* first we add the encoded string that reached it's max size */
940 if ($cur_l > ($max_l-2)) {
941 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
942 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
943 $cur_l = 3;
944 $ret = '';
945 $iOffset = $i;
946 $iEncStart = $i;
947 }
948 $enc_init = true;
949 $ret .= sprintf("=%02X", $k);
950 } else {
951 if ($iEncStart !== false) {
952 $cur_l++;
953 if ($cur_l > $max_l) {
954 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
955 $aRet[] = "=?$default_charset?Q?$ret?=";
956 $iEncStart = false;
957 $iOffset = $i;
958 $cur_l = 0;
959 $ret = '';
960 } else {
961 $ret .= $string{$i};
962 }
963 }
964 }
965 break;
966 }
967 }
968
969 if ($enc_init) {
970 if ($iEncStart !== false) {
971 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
972 $aRet[] = "=?$default_charset?Q?$ret?=";
973 } else {
974 $aRet[] = substr($string,$iOffset);
975 }
976 $string = implode('',$aRet);
977 }
978 return $string;
979 }
980
981 /**
982 * Encodes string according to rfc2047 B encoding header formating rules
983 *
984 * It is recommended way to encode headers with character sets that store
985 * symbols in more than one byte.
986 *
987 * Function requires mbstring support. If required mbstring functions are missing,
988 * function returns false and sets E_USER_WARNING level error message.
989 *
990 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
991 * that mbstring functions will generate E_WARNING errors, if unsupported
992 * character set is used. mb_encode_mimeheader function provided by php
993 * mbstring extension is not used in order to get better control of header
994 * encoding.
995 *
996 * Used php code functions - function_exists(), trigger_error(), strlen()
997 * (is used with charset names and base64 strings). Used php mbstring
998 * functions - mb_strlen and mb_substr.
999 *
1000 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1001 * encoding), rfc 2822 (header folding)
1002 *
1003 * @param string $string header string that must be encoded
1004 * @param string $charset character set. Must be supported by mbstring extension.
1005 * Use sq_mb_list_encodings() to detect supported charsets.
1006 * @return string string encoded according to rfc2047 B encoding formating rules
1007 * @since 1.5.1
1008 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1009 * @todo Do we want to control max length of header?
1010 * @todo Do we want to control EOL (end-of-line) marker?
1011 * @todo Do we want to translate error message?
1012 */
1013 function encodeHeaderBase64($string,$charset) {
1014 /**
1015 * Check mbstring function requirements.
1016 */
1017 if (! function_exists('mb_strlen') ||
1018 ! function_exists('mb_substr')) {
1019 // set E_USER_WARNING
1020 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1021 // return false
1022 return false;
1023 }
1024
1025 // initial return array
1026 $aRet = array();
1027
1028 /**
1029 * header length = 75 symbols max (same as in encodeHeader)
1030 * remove $charset length
1031 * remove =? ? ?= (5 chars)
1032 * remove 2 more chars (\r\n ?)
1033 */
1034 $iMaxLength = 75 - strlen($charset) - 7;
1035
1036 // set first character position
1037 $iStartCharNum = 0;
1038
1039 // loop through all characters. count characters and not bytes.
1040 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1041 // encode string from starting character to current character.
1042 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1043
1044 // Check encoded string length
1045 if(strlen($encoded_string)>$iMaxLength) {
1046 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1047 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1048
1049 // set new starting character
1050 $iStartCharNum = $iCharNum-1;
1051
1052 // encode last char (in case it is last character in string)
1053 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1054 } // if string is shorter than max length - add next character
1055 }
1056
1057 // add last encoded string to array
1058 $aRet[] = $encoded_string;
1059
1060 // set initial return string
1061 $sRet = '';
1062
1063 // loop through encoded strings
1064 foreach($aRet as $string) {
1065 // TODO: Do we want to control EOL (end-of-line) marker
1066 if ($sRet!='') $sRet.= " ";
1067
1068 // add header tags and encoded string to return string
1069 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1070 }
1071
1072 return $sRet;
1073 }
1074
1075 /* This function trys to locate the entity_id of a specific mime element */
1076 function find_ent_id($id, $message) {
1077 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1078 if ($message->entities[$i]->header->type0 == 'multipart') {
1079 $ret = find_ent_id($id, $message->entities[$i]);
1080 } else {
1081 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1082 // if (sq_check_save_extension($message->entities[$i])) {
1083 return $message->entities[$i]->entity_id;
1084 // }
1085 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1086 /**
1087 * This is part of a fix for Outlook Express 6.x generating
1088 * cid URLs without creating content-id headers
1089 * @@JA - 20050207
1090 */
1091 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1092 return $message->entities[$i]->entity_id;
1093 }
1094 }
1095 }
1096 }
1097 return $ret;
1098 }
1099
1100 function sq_check_save_extension($message) {
1101 $filename = $message->getFilename();
1102 $ext = substr($filename, strrpos($filename,'.')+1);
1103 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1104 return in_array($ext, $save_extensions);
1105 }
1106
1107
1108 /**
1109 ** HTMLFILTER ROUTINES
1110 */
1111
1112 /**
1113 * This function checks attribute values for entity-encoded values
1114 * and returns them translated into 8-bit strings so we can run
1115 * checks on them.
1116 *
1117 * @param $attvalue A string to run entity check against.
1118 * @return Nothing, modifies a reference value.
1119 */
1120 function sq_defang(&$attvalue){
1121 $me = 'sq_defang';
1122 /**
1123 * Skip this if there aren't ampersands or backslashes.
1124 */
1125 if (strpos($attvalue, '&') === false
1126 && strpos($attvalue, '\\') === false){
1127 return;
1128 }
1129 $m = false;
1130 // before deent, translate the dangerous unicode characters and ... to safe values
1131 // otherwise the regular expressions do not match.
1132
1133
1134
1135 do {
1136 $m = false;
1137 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1138 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1139 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1140 } while ($m == true);
1141 $attvalue = stripslashes($attvalue);
1142 }
1143
1144 /**
1145 * Kill any tabs, newlines, or carriage returns. Our friends the
1146 * makers of the browser with 95% market value decided that it'd
1147 * be funny to make "java[tab]script" be just as good as "javascript".
1148 *
1149 * @param attvalue The attribute value before extraneous spaces removed.
1150 * @return attvalue Nothing, modifies a reference value.
1151 */
1152 function sq_unspace(&$attvalue){
1153 $me = 'sq_unspace';
1154 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1155 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1156 Array('', '', '', '', ''), $attvalue);
1157 }
1158 }
1159
1160 /**
1161 * Translate all dangerous Unicode or Shift_JIS characters which are acepted by
1162 * IE as regular characters.
1163 *
1164 * @param attvalue The attribute value before dangerous characters are translated.
1165 * @return attvalue Nothing, modifies a reference value.
1166 * @author Marc Groot Koerkamp.
1167 */
1168 function sq_fixIE_idiocy(&$attvalue) {
1169 // remove NUL
1170 $attvalue = str_replace("\0", "", $attvalue);
1171 // remove comments
1172 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1173
1174 // IE has the evil habit of excepting every possible value for the attribute expression
1175 // The table below contain characters which are valid in IE if they are used in the "expression"
1176 // attribute value.
1177 $aDangerousCharsReplacementTable = array(
1178 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1179 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1180 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
1181 '&#xFF25;', '&#65317' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1182 '&#xFF45;', '&#65349' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1183 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1184 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1185 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1186 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1187 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1188 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1189 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1190 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1191 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1192 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1193 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1194 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1195 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1196 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1197 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1198 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1199 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1200 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1201 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1202 '&#x8264;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some chinese char range
1203 '&#x8285;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1204 '&#x8277;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1205 '&#x8298;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1206 '&#x826F;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1207 '&#x8290;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1208 '&#x8271;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1209 '&#x8292;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1210 '&#x8272;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1211 '&#x8293;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1212 '&#x8268;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1213 '&#x8289;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1214 '&#x826E;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1215 '&#x828F;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1216 '&#x826D;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1217 '&#x828E;'), /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1218 array('l', 'l', 'r','r','n','n',
1219 'E','E','e','e','X','X','x','x','P','P','p','p','S','S','s','s','I','I',
1220 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n',
1221 'E','e','X','x','P','p','S','s','I','i','O','o','N','n'));
1222 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1223
1224 // Escapes are usefull for special characters like "{}[]()'&. In other cases they are
1225 // used for XSS
1226 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1227 }
1228
1229 /**
1230 * This function returns the final tag out of the tag name, an array
1231 * of attributes, and the type of the tag. This function is called by
1232 * sq_sanitize internally.
1233 *
1234 * @param $tagname the name of the tag.
1235 * @param $attary the array of attributes and their values
1236 * @param $tagtype The type of the tag (see in comments).
1237 * @return a string with the final tag representation.
1238 */
1239 function sq_tagprint($tagname, $attary, $tagtype){
1240 $me = 'sq_tagprint';
1241
1242 if ($tagtype == 2){
1243 $fulltag = '</' . $tagname . '>';
1244 } else {
1245 $fulltag = '<' . $tagname;
1246 if (is_array($attary) && sizeof($attary)){
1247 $atts = Array();
1248 while (list($attname, $attvalue) = each($attary)){
1249 array_push($atts, "$attname=$attvalue");
1250 }
1251 $fulltag .= ' ' . join(" ", $atts);
1252 }
1253 if ($tagtype == 3){
1254 $fulltag .= ' /';
1255 }
1256 $fulltag .= '>';
1257 }
1258 return $fulltag;
1259 }
1260
1261 /**
1262 * A small helper function to use with array_walk. Modifies a by-ref
1263 * value and makes it lowercase.
1264 *
1265 * @param $val a value passed by-ref.
1266 * @return void since it modifies a by-ref value.
1267 */
1268 function sq_casenormalize(&$val){
1269 $val = strtolower($val);
1270 }
1271
1272 /**
1273 * This function skips any whitespace from the current position within
1274 * a string and to the next non-whitespace value.
1275 *
1276 * @param $body the string
1277 * @param $offset the offset within the string where we should start
1278 * looking for the next non-whitespace character.
1279 * @return the location within the $body where the next
1280 * non-whitespace char is located.
1281 */
1282 function sq_skipspace($body, $offset){
1283 $me = 'sq_skipspace';
1284 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1285 if (sizeof($matches{1})){
1286 $count = strlen($matches{1});
1287 $offset += $count;
1288 }
1289 return $offset;
1290 }
1291
1292 /**
1293 * This function looks for the next character within a string. It's
1294 * really just a glorified "strpos", except it catches if failures
1295 * nicely.
1296 *
1297 * @param $body The string to look for needle in.
1298 * @param $offset Start looking from this position.
1299 * @param $needle The character/string to look for.
1300 * @return location of the next occurance of the needle, or
1301 * strlen($body) if needle wasn't found.
1302 */
1303 function sq_findnxstr($body, $offset, $needle){
1304 $me = 'sq_findnxstr';
1305 $pos = strpos($body, $needle, $offset);
1306 if ($pos === FALSE){
1307 $pos = strlen($body);
1308 }
1309 return $pos;
1310 }
1311
1312 /**
1313 * This function takes a PCRE-style regexp and tries to match it
1314 * within the string.
1315 *
1316 * @param $body The string to look for needle in.
1317 * @param $offset Start looking from here.
1318 * @param $reg A PCRE-style regex to match.
1319 * @return Returns a false if no matches found, or an array
1320 * with the following members:
1321 * - integer with the location of the match within $body
1322 * - string with whatever content between offset and the match
1323 * - string with whatever it is we matched
1324 */
1325 function sq_findnxreg($body, $offset, $reg){
1326 $me = 'sq_findnxreg';
1327 $matches = Array();
1328 $retarr = Array();
1329 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1330 if (!isset($matches{0}) || !$matches{0}){
1331 $retarr = false;
1332 } else {
1333 $retarr{0} = $offset + strlen($matches{1});
1334 $retarr{1} = $matches{1};
1335 $retarr{2} = $matches{2};
1336 }
1337 return $retarr;
1338 }
1339
1340 /**
1341 * This function looks for the next tag.
1342 *
1343 * @param $body String where to look for the next tag.
1344 * @param $offset Start looking from here.
1345 * @return false if no more tags exist in the body, or
1346 * an array with the following members:
1347 * - string with the name of the tag
1348 * - array with attributes and their values
1349 * - integer with tag type (1, 2, or 3)
1350 * - integer where the tag starts (starting "<")
1351 * - integer where the tag ends (ending ">")
1352 * first three members will be false, if the tag is invalid.
1353 */
1354 function sq_getnxtag($body, $offset){
1355 $me = 'sq_getnxtag';
1356 if ($offset > strlen($body)){
1357 return false;
1358 }
1359 $lt = sq_findnxstr($body, $offset, "<");
1360 if ($lt == strlen($body)){
1361 return false;
1362 }
1363 /**
1364 * We are here:
1365 * blah blah <tag attribute="value">
1366 * \---------^
1367 */
1368 $pos = sq_skipspace($body, $lt+1);
1369 if ($pos >= strlen($body)){
1370 return Array(false, false, false, $lt, strlen($body));
1371 }
1372 /**
1373 * There are 3 kinds of tags:
1374 * 1. Opening tag, e.g.:
1375 * <a href="blah">
1376 * 2. Closing tag, e.g.:
1377 * </a>
1378 * 3. XHTML-style content-less tag, e.g.:
1379 * <img src="blah" />
1380 */
1381 $tagtype = false;
1382 switch (substr($body, $pos, 1)){
1383 case '/':
1384 $tagtype = 2;
1385 $pos++;
1386 break;
1387 case '!':
1388 /**
1389 * A comment or an SGML declaration.
1390 */
1391 if (substr($body, $pos+1, 2) == "--"){
1392 $gt = strpos($body, "-->", $pos);
1393 if ($gt === false){
1394 $gt = strlen($body);
1395 } else {
1396 $gt += 2;
1397 }
1398 return Array(false, false, false, $lt, $gt);
1399 } else {
1400 $gt = sq_findnxstr($body, $pos, ">");
1401 return Array(false, false, false, $lt, $gt);
1402 }
1403 break;
1404 default:
1405 /**
1406 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1407 * later.
1408 */
1409 $tagtype = 1;
1410 break;
1411 }
1412
1413 $tag_start = $pos;
1414 $tagname = '';
1415 /**
1416 * Look for next [\W-_], which will indicate the end of the tag name.
1417 */
1418 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1419 if ($regary == false){
1420 return Array(false, false, false, $lt, strlen($body));
1421 }
1422 list($pos, $tagname, $match) = $regary;
1423 $tagname = strtolower($tagname);
1424
1425 /**
1426 * $match can be either of these:
1427 * '>' indicating the end of the tag entirely.
1428 * '\s' indicating the end of the tag name.
1429 * '/' indicating that this is type-3 xhtml tag.
1430 *
1431 * Whatever else we find there indicates an invalid tag.
1432 */
1433 switch ($match){
1434 case '/':
1435 /**
1436 * This is an xhtml-style tag with a closing / at the
1437 * end, like so: <img src="blah" />. Check if it's followed
1438 * by the closing bracket. If not, then this tag is invalid
1439 */
1440 if (substr($body, $pos, 2) == "/>"){
1441 $pos++;
1442 $tagtype = 3;
1443 } else {
1444 $gt = sq_findnxstr($body, $pos, ">");
1445 $retary = Array(false, false, false, $lt, $gt);
1446 return $retary;
1447 }
1448 case '>':
1449 return Array($tagname, false, $tagtype, $lt, $pos);
1450 break;
1451 default:
1452 /**
1453 * Check if it's whitespace
1454 */
1455 if (!preg_match('/\s/', $match)){
1456 /**
1457 * This is an invalid tag! Look for the next closing ">".
1458 */
1459 $gt = sq_findnxstr($body, $lt, ">");
1460 return Array(false, false, false, $lt, $gt);
1461 }
1462 break;
1463 }
1464
1465 /**
1466 * At this point we're here:
1467 * <tagname attribute='blah'>
1468 * \-------^
1469 *
1470 * At this point we loop in order to find all attributes.
1471 */
1472 $attname = '';
1473 $atttype = false;
1474 $attary = Array();
1475
1476 while ($pos <= strlen($body)){
1477 $pos = sq_skipspace($body, $pos);
1478 if ($pos == strlen($body)){
1479 /**
1480 * Non-closed tag.
1481 */
1482 return Array(false, false, false, $lt, $pos);
1483 }
1484 /**
1485 * See if we arrived at a ">" or "/>", which means that we reached
1486 * the end of the tag.
1487 */
1488 $matches = Array();
1489 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1490 /**
1491 * Yep. So we did.
1492 */
1493 $pos += strlen($matches{1});
1494 if ($matches{2} == "/>"){
1495 $tagtype = 3;
1496 $pos++;
1497 }
1498 return Array($tagname, $attary, $tagtype, $lt, $pos);
1499 }
1500
1501 /**
1502 * There are several types of attributes, with optional
1503 * [:space:] between members.
1504 * Type 1:
1505 * attrname[:space:]=[:space:]'CDATA'
1506 * Type 2:
1507 * attrname[:space:]=[:space:]"CDATA"
1508 * Type 3:
1509 * attr[:space:]=[:space:]CDATA
1510 * Type 4:
1511 * attrname
1512 *
1513 * We leave types 1 and 2 the same, type 3 we check for
1514 * '"' and convert to "&quot" if needed, then wrap in
1515 * double quotes. Type 4 we convert into:
1516 * attrname="yes".
1517 */
1518 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1519 if ($regary == false){
1520 /**
1521 * Looks like body ended before the end of tag.
1522 */
1523 return Array(false, false, false, $lt, strlen($body));
1524 }
1525 list($pos, $attname, $match) = $regary;
1526 $attname = strtolower($attname);
1527 /**
1528 * We arrived at the end of attribute name. Several things possible
1529 * here:
1530 * '>' means the end of the tag and this is attribute type 4
1531 * '/' if followed by '>' means the same thing as above
1532 * '\s' means a lot of things -- look what it's followed by.
1533 * anything else means the attribute is invalid.
1534 */
1535 switch($match){
1536 case '/':
1537 /**
1538 * This is an xhtml-style tag with a closing / at the
1539 * end, like so: <img src="blah" />. Check if it's followed
1540 * by the closing bracket. If not, then this tag is invalid
1541 */
1542 if (substr($body, $pos, 2) == "/>"){
1543 $pos++;
1544 $tagtype = 3;
1545 } else {
1546 $gt = sq_findnxstr($body, $pos, ">");
1547 $retary = Array(false, false, false, $lt, $gt);
1548 return $retary;
1549 }
1550 case '>':
1551 $attary{$attname} = '"yes"';
1552 return Array($tagname, $attary, $tagtype, $lt, $pos);
1553 break;
1554 default:
1555 /**
1556 * Skip whitespace and see what we arrive at.
1557 */
1558 $pos = sq_skipspace($body, $pos);
1559 $char = substr($body, $pos, 1);
1560 /**
1561 * Two things are valid here:
1562 * '=' means this is attribute type 1 2 or 3.
1563 * \w means this was attribute type 4.
1564 * anything else we ignore and re-loop. End of tag and
1565 * invalid stuff will be caught by our checks at the beginning
1566 * of the loop.
1567 */
1568 if ($char == "="){
1569 $pos++;
1570 $pos = sq_skipspace($body, $pos);
1571 /**
1572 * Here are 3 possibilities:
1573 * "'" attribute type 1
1574 * '"' attribute type 2
1575 * everything else is the content of tag type 3
1576 */
1577 $quot = substr($body, $pos, 1);
1578 if ($quot == "'"){
1579 $regary = sq_findnxreg($body, $pos+1, "\'");
1580 if ($regary == false){
1581 return Array(false, false, false, $lt, strlen($body));
1582 }
1583 list($pos, $attval, $match) = $regary;
1584 $pos++;
1585 $attary{$attname} = "'" . $attval . "'";
1586 } else if ($quot == '"'){
1587 $regary = sq_findnxreg($body, $pos+1, '\"');
1588 if ($regary == false){
1589 return Array(false, false, false, $lt, strlen($body));
1590 }
1591 list($pos, $attval, $match) = $regary;
1592 $pos++;
1593 $attary{$attname} = '"' . $attval . '"';
1594 } else {
1595 /**
1596 * These are hateful. Look for \s, or >.
1597 */
1598 $regary = sq_findnxreg($body, $pos, "[\s>]");
1599 if ($regary == false){
1600 return Array(false, false, false, $lt, strlen($body));
1601 }
1602 list($pos, $attval, $match) = $regary;
1603 /**
1604 * If it's ">" it will be caught at the top.
1605 */
1606 $attval = preg_replace("/\"/s", "&quot;", $attval);
1607 $attary{$attname} = '"' . $attval . '"';
1608 }
1609 } else if (preg_match("|[\w/>]|", $char)) {
1610 /**
1611 * That was attribute type 4.
1612 */
1613 $attary{$attname} = '"yes"';
1614 } else {
1615 /**
1616 * An illegal character. Find next '>' and return.
1617 */
1618 $gt = sq_findnxstr($body, $pos, ">");
1619 return Array(false, false, false, $lt, $gt);
1620 }
1621 break;
1622 }
1623 }
1624 /**
1625 * The fact that we got here indicates that the tag end was never
1626 * found. Return invalid tag indication so it gets stripped.
1627 */
1628 return Array(false, false, false, $lt, strlen($body));
1629 }
1630
1631 /**
1632 * Translates entities into literal values so they can be checked.
1633 *
1634 * @param $attvalue the by-ref value to check.
1635 * @param $regex the regular expression to check against.
1636 * @param $hex whether the entites are hexadecimal.
1637 * @return True or False depending on whether there were matches.
1638 */
1639 function sq_deent(&$attvalue, $regex, $hex=false){
1640 $me = 'sq_deent';
1641 $ret_match = false;
1642 // remove comments
1643 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1644 preg_match_all($regex, $attvalue, $matches);
1645 if (is_array($matches) && sizeof($matches[0]) > 0){
1646 $repl = Array();
1647 for ($i = 0; $i < sizeof($matches[0]); $i++){
1648 $numval = $matches[1][$i];
1649 if ($hex){
1650 $numval = hexdec($numval);
1651 }
1652 $repl{$matches[0][$i]} = chr($numval);
1653 }
1654 $attvalue = strtr($attvalue, $repl);
1655 return true;
1656 } else {
1657 return false;
1658 }
1659 }
1660
1661 /**
1662 * This function runs various checks against the attributes.
1663 *
1664 * @param $tagname String with the name of the tag.
1665 * @param $attary Array with all tag attributes.
1666 * @param $rm_attnames See description for sq_sanitize
1667 * @param $bad_attvals See description for sq_sanitize
1668 * @param $add_attr_to_tag See description for sq_sanitize
1669 * @param $message message object
1670 * @param $id message id
1671 * @return Array with modified attributes.
1672 */
1673 function sq_fixatts($tagname,
1674 $attary,
1675 $rm_attnames,
1676 $bad_attvals,
1677 $add_attr_to_tag,
1678 $message,
1679 $id,
1680 $mailbox
1681 ){
1682 $me = 'sq_fixatts';
1683 while (list($attname, $attvalue) = each($attary)){
1684 /**
1685 * See if this attribute should be removed.
1686 */
1687 foreach ($rm_attnames as $matchtag=>$matchattrs){
1688 if (preg_match($matchtag, $tagname)){
1689 foreach ($matchattrs as $matchattr){
1690 if (preg_match($matchattr, $attname)){
1691 unset($attary{$attname});
1692 continue;
1693 }
1694 }
1695 }
1696 }
1697 /**
1698 * Workaround for IE quirks
1699 */
1700 sq_fixIE_idiocy($attvalue);
1701
1702 /**
1703 * Remove any backslashes, entities, and extraneous whitespace.
1704 */
1705
1706 $oldattvalue = $attvalue;
1707 sq_defang($attvalue);
1708 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1709 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1710 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1711 $attvalue = "idiocy";
1712 $attary{$attname} = $attvalue;
1713 }
1714 sq_unspace($attvalue);
1715
1716 /**
1717 * Now let's run checks on the attvalues.
1718 * I don't expect anyone to comprehend this. If you do,
1719 * get in touch with me so I can drive to where you live and
1720 * shake your hand personally. :)
1721 */
1722 foreach ($bad_attvals as $matchtag=>$matchattrs){
1723 if (preg_match($matchtag, $tagname)){
1724 foreach ($matchattrs as $matchattr=>$valary){
1725 if (preg_match($matchattr, $attname)){
1726 /**
1727 * There are two arrays in valary.
1728 * First is matches.
1729 * Second one is replacements
1730 */
1731 list($valmatch, $valrepl) = $valary;
1732 $newvalue =
1733 preg_replace($valmatch, $valrepl, $attvalue);
1734 if ($newvalue != $attvalue){
1735 $attary{$attname} = $newvalue;
1736 }
1737 }
1738 }
1739 }
1740 }
1741
1742 /**
1743 * Replace empty src tags with the blank image. src is only used
1744 * for frames, images, and image inputs. Doing a replace should
1745 * not affect them working as should be, however it will stop
1746 * IE from being kicked off when src for img tags are not set
1747 */
1748 if (($attname == 'src') && ($attvalue == '""')) {
1749 $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1750 }
1751
1752 /**
1753 * Turn cid: urls into http-friendly ones.
1754 */
1755 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1756 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1757 }
1758
1759 /**
1760 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1761 * One day MS might actually make it match something useful, for now, falling
1762 * back to using cid2http, so we can grab the blank.png.
1763 */
1764 if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1765 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1766 }
1767
1768 }
1769 /**
1770 * See if we need to append any attributes to this tag.
1771 */
1772 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1773 if (preg_match($matchtag, $tagname)){
1774 $attary = array_merge($attary, $addattary);
1775 }
1776 }
1777 return $attary;
1778 }
1779
1780 /**
1781 * This function edits the style definition to make them friendly and
1782 * usable in SquirrelMail.
1783 *
1784 * @param $message the message object
1785 * @param $id the message id
1786 * @param $content a string with whatever is between <style> and </style>
1787 * @param $mailbox the message mailbox
1788 * @return a string with edited content.
1789 */
1790 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1791 global $view_unsafe_images;
1792 $me = 'sq_fixstyle';
1793 // workaround for </style> in between comments
1794 $iCurrentPos = $pos;
1795 $content = '';
1796 $sToken = '';
1797 $bSucces = false;
1798 $bEndTag = false;
1799 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
1800 $char = $body{$i};
1801 switch ($char) {
1802 case '<':
1803 $sToken .= $char;
1804 break;
1805 case '/':
1806 if ($sToken == '<') {
1807 $sToken .= $char;
1808 $bEndTag = true;
1809 } else {
1810 $content .= $char;
1811 }
1812 break;
1813 case '>':
1814 if ($bEndTag) {
1815 $sToken .= $char;
1816 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
1817 $newpos = $i + 1;
1818 $bSucces = true;
1819 break 2;
1820 } else {
1821 $content .= $sToken;
1822 }
1823 $bEndTag = false;
1824 } else {
1825 $content .= $char;
1826 }
1827 break;
1828 case '!':
1829 if ($sToken == '<') {
1830 // possible comment
1831 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
1832 $i = strpos($body,'-->',$i+3);
1833 if ($i === false) { // no end comment
1834 $i = strlen($body);
1835 }
1836 $sToken = '';
1837 }
1838 } else {
1839 $content .= $char;
1840 }
1841 break;
1842 default:
1843 if ($bEndTag) {
1844 $sToken .= $char;
1845 } else {
1846 $content .= $char;
1847 }
1848 break;
1849 }
1850 }
1851 if ($bSucces == FALSE){
1852 return array(FALSE, strlen($body));
1853 }
1854
1855
1856
1857 /**
1858 * First look for general BODY style declaration, which would be
1859 * like so:
1860 * body {background: blah-blah}
1861 * and change it to .bodyclass so we can just assign it to a <div>
1862 */
1863 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1864 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1865 /**
1866 * Fix url('blah') declarations.
1867 */
1868 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1869 // "url(\\1$secremoveimg\\2)", $content);
1870
1871 // IE Sucks hard. We have a special function for it.
1872 sq_fixIE_idiocy($content);
1873
1874 // remove @import line
1875 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
1876
1877 // translate ur\l and variations (IE parses that)
1878 // TODO check if the sq_fixIE_idiocy function already handles this.
1879 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
1880 // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1881 while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1882 $sProto = strtolower($matches[1]);
1883 switch ($sProto) {
1884 /**
1885 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1886 * is false.
1887 */
1888 case 'https':
1889 case 'http':
1890 if (!$view_unsafe_images){
1891
1892 $sExpr = "/url\s*\(\s*[\'\"]?\s*$sProto*:.*[\'\"]?\s*\)/si";
1893 $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1894
1895 } else {
1896 $content = preg_replace('/url/i',"u\0r\0l",$content);
1897 }
1898 break;
1899 /**
1900 * Fix urls that refer to cid:
1901 */
1902 case 'cid':
1903 $cidurl = 'cid:'. $matches[2];
1904 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1905 // escape parentheses that can modify the regular expression
1906 $cidurl = str_replace(array('(',')'),array('\\(','\\)'),$cidurl);
1907 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1908 "u\0r\0l($httpurl)", $content);
1909 break;
1910 default:
1911 /**
1912 * replace url with protocol other then the white list
1913 * http,https and cid by an empty string.
1914 */
1915 $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1916 "", $content);
1917 break;
1918 }
1919 }
1920 // remove NUL
1921 $content = str_replace("\0", "", $content);
1922 /**
1923 * Remove any backslashes, entities, and extraneous whitespace.
1924 */
1925 $contentTemp = $content;
1926 sq_defang($contentTemp);
1927 sq_unspace($contentTemp);
1928
1929 /**
1930 * Fix stupid css declarations which lead to vulnerabilities
1931 * in IE.
1932 */
1933 $match = Array('/\/\*.*\*\//',
1934 '/expression/i',
1935 '/behaviou*r/i',
1936 '/binding/i',
1937 '/include-source/i',
1938 '/javascript/i',
1939 '/script/i');
1940 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
1941 $contentNew = preg_replace($match, $replace, $contentTemp);
1942 if ($contentNew !== $contentTemp) {
1943 // insecure css declarations are used. From now on we don't care
1944 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1945 $content = $contentNew;
1946 }
1947 return array($content, $newpos);
1948 }
1949
1950
1951 /**
1952 * This function converts cid: url's into the ones that can be viewed in
1953 * the browser.
1954 *
1955 * @param $message the message object
1956 * @param $id the message id
1957 * @param $cidurl the cid: url.
1958 * @param $mailbox the message mailbox
1959 * @return a string with a http-friendly url
1960 */
1961 function sq_cid2http($message, $id, $cidurl, $mailbox){
1962 /**
1963 * Get rid of quotes.
1964 */
1965 $quotchar = substr($cidurl, 0, 1);
1966 if ($quotchar == '"' || $quotchar == "'"){
1967 $cidurl = str_replace($quotchar, "", $cidurl);
1968 } else {
1969 $quotchar = '';
1970 }
1971 $cidurl = substr(trim($cidurl), 4);
1972
1973 $match_str = '/\{.*?\}\//';
1974 $str_rep = '';
1975 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1976
1977 $linkurl = find_ent_id($cidurl, $message);
1978 /* in case of non-save cid links $httpurl should be replaced by a sort of
1979 unsave link image */
1980 $httpurl = '';
1981
1982 /**
1983 * This is part of a fix for Outlook Express 6.x generating
1984 * cid URLs without creating content-id headers. These images are
1985 * not part of the multipart/related html mail. The html contains
1986 * <img src="cid:{some_id}/image_filename.ext"> references to
1987 * attached images with as goal to render them inline although
1988 * the attachment disposition property is not inline.
1989 */
1990
1991 if (empty($linkurl)) {
1992 if (preg_match('/{.*}\//', $cidurl)) {
1993 $cidurl = preg_replace('/{.*}\//','', $cidurl);
1994 if (!empty($cidurl)) {
1995 $linkurl = find_ent_id($cidurl, $message);
1996 }
1997 }
1998 }
1999
2000 if (!empty($linkurl)) {
2001 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
2002 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2003 '&amp;ent_id=' . $linkurl . $quotchar;
2004 } else {
2005 /**
2006 * If we couldn't generate a proper img url, drop in a blank image
2007 * instead of sending back empty, otherwise it causes unusual behaviour
2008 */
2009 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
2010 }
2011
2012 return $httpurl;
2013 }
2014
2015 /**
2016 * This function changes the <body> tag into a <div> tag since we
2017 * can't really have a body-within-body.
2018 *
2019 * @param $attary an array of attributes and values of <body>
2020 * @param $mailbox mailbox we're currently reading (for cid2http)
2021 * @param $message current message (for cid2http)
2022 * @param $id current message id (for cid2http)
2023 * @return a modified array of attributes to be set for <div>
2024 */
2025 function sq_body2div($attary, $mailbox, $message, $id){
2026 $me = 'sq_body2div';
2027 $divattary = Array('class' => "'bodyclass'");
2028 $text = '#000000';
2029 $has_bgc_stl = $has_txt_stl = false;
2030 $styledef = '';
2031 if (is_array($attary) && sizeof($attary) > 0){
2032 foreach ($attary as $attname=>$attvalue){
2033 $quotchar = substr($attvalue, 0, 1);
2034 $attvalue = str_replace($quotchar, "", $attvalue);
2035 switch ($attname){
2036 case 'background':
2037 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2038 $styledef .= "background-image: url('$attvalue'); ";
2039 break;
2040 case 'bgcolor':
2041 $has_bgc_stl = true;
2042 $styledef .= "background-color: $attvalue; ";
2043 break;
2044 case 'text':
2045 $has_txt_stl = true;
2046 $styledef .= "color: $attvalue; ";
2047 break;
2048 }
2049 }
2050 // Outlook defines a white bgcolor and no text color. This can lead to
2051 // white text on a white bg with certain themes.
2052 if ($has_bgc_stl && !$has_txt_stl) {
2053 $styledef .= "color: $text; ";
2054 }
2055 if (strlen($styledef) > 0){
2056 $divattary{"style"} = "\"$styledef\"";
2057 }
2058 }
2059 return $divattary;
2060 }
2061
2062 /**
2063 * This is the main function and the one you should actually be calling.
2064 * There are several variables you should be aware of an which need
2065 * special description.
2066 *
2067 * Since the description is quite lengthy, see it here:
2068 * http://linux.duke.edu/projects/mini/htmlfilter/
2069 *
2070 * @param $body the string with HTML you wish to filter
2071 * @param $tag_list see description above
2072 * @param $rm_tags_with_content see description above
2073 * @param $self_closing_tags see description above
2074 * @param $force_tag_closing see description above
2075 * @param $rm_attnames see description above
2076 * @param $bad_attvals see description above
2077 * @param $add_attr_to_tag see description above
2078 * @param $message message object
2079 * @param $id message id
2080 * @return sanitized html safe to show on your pages.
2081 */
2082 function sq_sanitize($body,
2083 $tag_list,
2084 $rm_tags_with_content,
2085 $self_closing_tags,
2086 $force_tag_closing,
2087 $rm_attnames,
2088 $bad_attvals,
2089 $add_attr_to_tag,
2090 $message,
2091 $id,
2092 $mailbox
2093 ){
2094 $me = 'sq_sanitize';
2095 $rm_tags = array_shift($tag_list);
2096 /**
2097 * Normalize rm_tags and rm_tags_with_content.
2098 */
2099 @array_walk($tag_list, 'sq_casenormalize');
2100 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2101 @array_walk($self_closing_tags, 'sq_casenormalize');
2102 /**
2103 * See if tag_list is of tags to remove or tags to allow.
2104 * false means remove these tags
2105 * true means allow these tags
2106 */
2107 $curpos = 0;
2108 $open_tags = Array();
2109 $trusted = "\n<!-- begin sanitized html -->\n";
2110 $skip_content = false;
2111 /**
2112 * Take care of netscape's stupid javascript entities like
2113 * &{alert('boo')};
2114 */
2115 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
2116
2117 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2118 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2119 $free_content = substr($body, $curpos, $lt-$curpos);
2120 /**
2121 * Take care of <style>
2122 */
2123 if ($tagname == "style" && $tagtype == 1){
2124 list($free_content, $curpos) =
2125 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
2126 if ($free_content != FALSE){
2127 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2128 $trusted .= $free_content;
2129 $trusted .= sq_tagprint($tagname, false, 2);
2130 }
2131 continue;
2132 }
2133 if ($skip_content == false){
2134 $trusted .= $free_content;
2135 }
2136 if ($tagname != FALSE){
2137 if ($tagtype == 2){
2138 if ($skip_content == $tagname){
2139 /**
2140 * Got to the end of tag we needed to remove.
2141 */
2142 $tagname = false;
2143 $skip_content = false;
2144 } else {
2145 if ($skip_content == false){
2146 if ($tagname == "body"){
2147 $tagname = "div";
2148 }
2149 if (isset($open_tags{$tagname}) &&
2150 $open_tags{$tagname} > 0){
2151 $open_tags{$tagname}--;
2152 } else {
2153 $tagname = false;
2154 }
2155 }
2156 }
2157 } else {
2158 /**
2159 * $rm_tags_with_content
2160 */
2161 if ($skip_content == false){
2162 /**
2163 * See if this is a self-closing type and change
2164 * tagtype appropriately.
2165 */
2166 if ($tagtype == 1
2167 && in_array($tagname, $self_closing_tags)){
2168 $tagtype = 3;
2169 }
2170 /**
2171 * See if we should skip this tag and any content
2172 * inside it.
2173 */
2174 if ($tagtype == 1 &&
2175 in_array($tagname, $rm_tags_with_content)){
2176 $skip_content = $tagname;
2177 } else {
2178 if (($rm_tags == false
2179 && in_array($tagname, $tag_list)) ||
2180 ($rm_tags == true &&
2181 !in_array($tagname, $tag_list))){
2182 $tagname = false;
2183 } else {
2184 /**
2185 * Convert body into div.
2186 */
2187 if ($tagname == "body"){
2188 $tagname = "div";
2189 $attary = sq_body2div($attary, $mailbox,
2190 $message, $id);
2191 }
2192 if ($tagtype == 1){
2193 if (isset($open_tags{$tagname})){
2194 $open_tags{$tagname}++;
2195 } else {
2196 $open_tags{$tagname}=1;
2197 }
2198 }
2199 /**
2200 * This is where we run other checks.
2201 */
2202 if (is_array($attary) && sizeof($attary) > 0){
2203 $attary = sq_fixatts($tagname,
2204 $attary,
2205 $rm_attnames,
2206 $bad_attvals,
2207 $add_attr_to_tag,
2208 $message,
2209 $id,
2210 $mailbox
2211 );
2212 }
2213 }
2214 }
2215 }
2216 }
2217 if ($tagname != false && $skip_content == false){
2218 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2219 }
2220 }
2221 $curpos = $gt+1;
2222 }
2223 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2224 if ($force_tag_closing == true){
2225 foreach ($open_tags as $tagname=>$opentimes){
2226 while ($opentimes > 0){
2227 $trusted .= '</' . $tagname . '>';
2228 $opentimes--;
2229 }
2230 }
2231 $trusted .= "\n";
2232 }
2233 $trusted .= "<!-- end sanitized html -->\n";
2234 return $trusted;
2235 }
2236
2237 /**
2238 * This is a wrapper function to call html sanitizing routines.
2239 *
2240 * @param $body the body of the message
2241 * @param $id the id of the message
2242
2243 * @param $message
2244 * @param $mailbox
2245 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2246 * into internal SM compose links
2247 * (optional; default = TRUE)
2248 * @return a string with html safe to display in the browser.
2249 */
2250 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2251
2252 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2253
2254 global $attachment_common_show_images, $view_unsafe_images,
2255 $has_unsafe_images;
2256 /**
2257 * Don't display attached images in HTML mode.
2258 *
2259 * SB: why?
2260 */
2261 $attachment_common_show_images = false;
2262 $tag_list = Array(
2263 false,
2264 "object",
2265 "meta",
2266 "html",
2267 "head",
2268 "base",
2269 "link",
2270 "frame",
2271 "iframe",
2272 "plaintext",
2273 "marquee"
2274 );
2275
2276 $rm_tags_with_content = Array(
2277 "script",
2278 "applet",
2279 "embed",
2280 "title",
2281 "frameset",
2282 "xmp",
2283 "xml"
2284 );
2285
2286 $self_closing_tags = Array(
2287 "img",
2288 "br",
2289 "hr",
2290 "input",
2291 "outbind"
2292 );
2293
2294 $force_tag_closing = true;
2295
2296 $rm_attnames = Array(
2297 "/.*/" =>
2298 Array(
2299 "/target/i",
2300 "/^on.*/i",
2301 "/^dynsrc/i",
2302 "/^data.*/i",
2303 "/^lowsrc.*/i"
2304 )
2305 );
2306
2307 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2308 $bad_attvals = Array(
2309 "/.*/" =>
2310 Array(
2311 "/^src|background/i" =>
2312 Array(
2313 Array(
2314 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2315 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2316 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2317 ),
2318 Array(
2319 "\\1$secremoveimg\\2",
2320 "\\1$secremoveimg\\2",
2321 "\\1$secremoveimg\\2",
2322 )
2323 ),
2324 "/^href|action/i" =>
2325 Array(
2326 Array(
2327 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2328 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2329 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2330 ),
2331 Array(
2332 "\\1#\\1",
2333 "\\1#\\1",
2334 "\\1#\\1"
2335 )
2336 ),
2337 "/^style/i" =>
2338 Array(
2339 Array(
2340 "/\/\*.*\*\//",
2341 "/expression/i",
2342 "/binding/i",
2343 "/behaviou*r/i",
2344 "/include-source/i",
2345 "/position\s*:\s*absolute/i",
2346 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2347 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2348 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2349 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2350 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2351 ),
2352 Array(
2353 "",
2354 "idiocy",
2355 "idiocy",
2356 "idiocy",
2357 "idiocy",
2358 "",
2359 "url",
2360 "url(\\1#\\1)",
2361 "url(\\1#\\1)",
2362 "url(\\1#\\1)",
2363 "\\1:url(\\2#\\3)"
2364 )
2365 )
2366 )
2367 );
2368 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2369 $view_unsafe_images = false;
2370 }
2371 if (!$view_unsafe_images){
2372 /**
2373 * Remove any references to http/https if view_unsafe_images set
2374 * to false.
2375 */
2376 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2377 '/^([\'\"])\s*https*:.*([\'\"])/si');
2378 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2379 "\\1$secremoveimg\\1");
2380 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2381 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2382 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2383 "url(\\1$secremoveimg\\1)");
2384 }
2385
2386 $add_attr_to_tag = Array(
2387 "/^a$/i" =>
2388 Array('target'=>'"_blank"',
2389 'title'=>'"'._("This external link will open in a new window").'"'
2390 )
2391 );
2392 $trusted = sq_sanitize($body,
2393 $tag_list,
2394 $rm_tags_with_content,
2395 $self_closing_tags,
2396 $force_tag_closing,
2397 $rm_attnames,
2398 $bad_attvals,
2399 $add_attr_to_tag,
2400 $message,
2401 $id,
2402 $mailbox
2403 );
2404 if (preg_match("|$secremoveimg|i", $trusted)){
2405 $has_unsafe_images = true;
2406 }
2407
2408 // we want to parse mailto's in HTML output, change to SM compose links
2409 // this is a modified version of code from url_parser.php... but Marc is
2410 // right: we need a better filtering implementation; adding this randomly
2411 // here is not a great solution
2412 //
2413 if ($take_mailto_links) {
2414 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2415 global $MailTo_PReg_Match;
2416 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2417 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2418 foreach ($regs[0] as $i => $mailto_before) {
2419 $mailto_params = $regs[10][$i];
2420 // get rid of any tailing quote since we have to add send_to to the end
2421 //
2422 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2423 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2424 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2425 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2426
2427 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2428 $to = 'to=' . $regs[1][$i];
2429 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2430 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2431 else {
2432 if ($mailto_params) //already some params, append to them
2433 $mailto_params .= '&amp;' . $to;
2434 else
2435 $mailto_params .= '?' . $to;
2436 }
2437 }
2438
2439 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2440
2441 // we'll already have target=_blank, no need to allow comp_in_new
2442 // here (which would be a lot more work anyway)
2443 //
2444 global $compose_new_win;
2445 $temp_comp_in_new = $compose_new_win;
2446 $compose_new_win = 0;
2447 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2448 $compose_new_win = $temp_comp_in_new;
2449
2450 // remove <a href=" and anything after the next quote (we only
2451 // need the uri, not the link HTML) in compose uri
2452 //
2453 $comp_uri = substr($comp_uri, 9);
2454 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2455 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2456 }
2457 }
2458 }
2459
2460 return $trusted;
2461 }
2462
2463 /**
2464 * function SendDownloadHeaders - send file to the browser
2465 *
2466 * Original Source: SM core src/download.php
2467 * moved here to make it available to other code, and separate
2468 * front end from back end functionality.
2469 *
2470 * @param string $type0 first half of mime type
2471 * @param string $type1 second half of mime type
2472 * @param string $filename filename to tell the browser for downloaded file
2473 * @param boolean $force whether to force the download dialog to pop
2474 * @param optional integer $filesize send the Content-Header and length to the browser
2475 * @return void
2476 */
2477 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2478 global $languages, $squirrelmail_language;
2479 $isIE = $isIE6plus = false;
2480
2481 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2482
2483 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2484 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2485 $isIE = true;
2486 }
2487
2488 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2489 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2490 $isIE6plus = true;
2491 }
2492
2493 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2494 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2495 $filename =
2496 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2497 } else {
2498 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2499 }
2500
2501 // A Pox on Microsoft and it's Internet Explorer!
2502 //
2503 // IE has lots of bugs with file downloads.
2504 // It also has problems with SSL. Both of these cause problems
2505 // for us in this function.
2506 //
2507 // See this article on Cache Control headers and SSL
2508 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2509 //
2510 // The best thing you can do for IE is to upgrade to the latest
2511 // version
2512 //set all the Cache Control Headers for IE
2513 if ($isIE) {
2514 $filename=rawurlencode($filename);
2515 header ("Pragma: public");
2516 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2517 header ("Cache-Control: post-check=0, pre-check=0", false);
2518 header ("Cache-Control: private");
2519
2520 //set the inline header for IE, we'll add the attachment header later if we need it
2521 header ("Content-Disposition: inline; filename=$filename");
2522 }
2523
2524 if (!$force) {
2525 // Try to show in browser window
2526 header ("Content-Disposition: inline; filename=\"$filename\"");
2527 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2528 } else {
2529 // Try to pop up the "save as" box
2530
2531 // IE makes this hard. It pops up 2 save boxes, or none.
2532 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2533 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2534 // But, according to Microsoft, it is "RFC compliant but doesn't
2535 // take into account some deviations that allowed within the
2536 // specification." Doesn't that mean RFC non-compliant?
2537 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2538
2539 // all browsers need the application/octet-stream header for this
2540 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2541
2542 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2543 // Do not have quotes around filename, but that applied to
2544 // "attachment"... does it apply to inline too?
2545 header ("Content-Disposition: attachment; filename=\"$filename\"");
2546
2547 if ($isIE && !$isIE6plus) {
2548 // This combination seems to work mostly. IE 5.5 SP 1 has
2549 // known issues (see the Microsoft Knowledge Base)
2550
2551 // This works for most types, but doesn't work with Word files
2552 header ("Content-Type: application/download; name=\"$filename\"");
2553 header ("Content-Type: application/force-download; name=\"$filename\"");
2554 // These are spares, just in case. :-)
2555 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2556 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2557 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2558 } else if ($isIE) {
2559 // This is to prevent IE for MIME sniffing and auto open a file in IE
2560 header ("Content-Type: application/force-download; name=\"$filename\"");
2561 } else {
2562 // another application/octet-stream forces download for Netscape
2563 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2564 }
2565 }
2566
2567 //send the content-length header if the calling function provides it
2568 if ($filesize > 0) {
2569 header("Content-Length: $filesize");
2570 }
2571
2572 } // end fn SendDownloadHeaders
Unnamed repository; edit this file 'description' to name the repository.