projects / squirrelmail.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
Generate links using templates
[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; 1999-2007 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 $oTemplate;
356
357 $nbsp = $oTemplate->fetch('non_breaking_space.tpl');
358
359 // workaround for not updated config.php
360 if (! isset($use_iframe)) $use_iframe = false;
361
362 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
363 $view_unsafe_images = false;
364 }
365
366 $body = '';
367 $urlmailbox = urlencode($mailbox);
368 $body_message = getEntity($message, $ent_num);
369 if (($body_message->header->type0 == 'text') ||
370 ($body_message->header->type0 == 'rfc822')) {
371 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
372 $body = decodeBody($body, $body_message->header->encoding);
373
374 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
375 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
376 if (mb_detect_encoding($body) != 'ASCII') {
377 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
378 }
379 }
380
381 /* As of 1.5.2, $body is passed (and modified) by reference */
382 do_hook('message_body', $body);
383
384 /* If there are other types that shouldn't be formatted, add
385 * them here.
386 */
387
388 if ($body_message->header->type1 == 'html') {
389 if ($show_html_default <> 1) {
390 $entity_conv = array('&nbsp;' => ' ',
391 '<p>' => "\n",
392 '<P>' => "\n",
393 '<br>' => "\n",
394 '<BR>' => "\n",
395 '<br />' => "\n",
396 '<BR />' => "\n",
397 '&gt;' => '>',
398 '&lt;' => '<');
399 $body = strtr($body, $entity_conv);
400 $body = strip_tags($body);
401 $body = trim($body);
402 translateText($body, $wrap_at,
403 $body_message->header->getParameter('charset'));
404 } elseif ($use_iframe && ! $clean) {
405 // $clean is used to remove iframe in printable view.
406
407 /**
408 * If we don't add html message between iframe tags,
409 * we must detect unsafe images and modify $has_unsafe_images.
410 */
411 $html_body = magicHTML($body, $id, $message, $mailbox);
412 // Convert character set in order to display html mails in different character set
413 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
414
415 // creating iframe url
416 $iframeurl=sqm_baseuri().'src/view_html.php?'
417 . 'mailbox=' . $urlmailbox
418 . '&amp;passed_id=' . $id
419 . '&amp;ent_id=' . $ent_num
420 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
421
422 global $oTemplate;
423 $oTemplate->assign('iframe_url', $iframeurl);
424 $oTemplate->assign('html_body', $html_body);
425
426 $body = $oTemplate->fetch('read_html_iframe.tpl');
427 } else {
428 // old way of html rendering
429 $body = magicHTML($body, $id, $message, $mailbox);
430 /**
431 * convert character set. charset_decode does not remove html special chars
432 * applied by magicHTML functions and does not sanitize them second time if
433 * fourth argument is true.
434 */
435 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
436 }
437 } else {
438 translateText($body, $wrap_at,
439 $body_message->header->getParameter('charset'));
440 }
441
442 // if this is the clean display (i.e. printer friendly), stop here.
443 if ( $clean ) {
444 return $body;
445 }
446
447 $download_and_unsafe_link = '';
448
449 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
450 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
451 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
452 if (isset($passed_ent_id)) {
453 $link .= '&amp;passed_ent_id='.$passed_ent_id;
454 }
455 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
456 $download_and_unsafe_link .= "$nbsp|$nbsp"
457 . create_hyperlink($download_href, _("Download this as a file"));
458 if ($view_unsafe_images) {
459 $text = _("Hide Unsafe Images");
460 } else {
461 if (isset($has_unsafe_images) && $has_unsafe_images) {
462 $link .= '&amp;view_unsafe_images=1';
463 $text = _("View Unsafe Images");
464 } else {
465 $text = '';
466 }
467 }
468 if($text != '') {
469 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
470 $unsafe_image_toggle_text = $text;
471 $download_and_unsafe_link .= "$nbsp|$nbsp"
472 . create_hyperlink($unsafe_image_toggle_href, $text);
473 }
474 }
475 return $body;
476 }
477
478 /**
479 * Generate attachments array for passing to templates. Separated from
480 * formatAttachments() below so that the same array can be given to the
481 * print-friendly version.
482 *
483 * @since 1.5.2
484 * @param object $message SquirrelMail message object
485 * @param array $exclude_id message parts that are not attachments.
486 * @param string $mailbox mailbox name
487 * @param integer $id message id
488 */
489 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
490 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
491
492 $att_ar = $message->getAttachments($exclude_id);
493 $urlMailbox = urlencode($mailbox);
494
495 $attachments = array();
496 foreach ($att_ar as $att) {
497 $ent = $att->entity_id;
498 $header = $att->header;
499 $type0 = strtolower($header->type0);
500 $type1 = strtolower($header->type1);
501 $name = '';
502 $links = array();
503 $links['download link']['text'] = _("Download");
504 $links['download link']['href'] = $base_uri .
505 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
506
507 if ($type0 =='message' && $type1 == 'rfc822') {
508 $default_page = $base_uri . 'src/read_body.php';
509 $rfc822_header = $att->rfc822_header;
510 $filename = $rfc822_header->subject;
511 if (trim( $filename ) == '') {
512 $filename = 'untitled-[' . $ent . ']' ;
513 }
514 $from_o = $rfc822_header->from;
515 if (is_object($from_o)) {
516 $from_name = decodeHeader($from_o->getAddress(false));
517 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
518 // something weird happens when a digest message is opened and you return to the digest
519 // now the from object is part of an array. Probably the parseHeader call overwrites the info
520 // retrieved from the bodystructure in a different way. We need to fix this later.
521 // possible starting point, do not fetch header we already have and inspect how
522 // the rfc822_header object behaves.
523 $from_name = decodeHeader($from_o[0]->getAddress(false));
524 } else {
525 $from_name = _("Unknown sender");
526 }
527 $description = _("From").': '.$from_name;
528 } else {
529 $default_page = $base_uri . 'src/download.php';
530 $filename = $att->getFilename();
531 if ($header->description) {
532 $description = decodeHeader($header->description);
533 } else {
534 $description = '';
535 }
536 }
537
538 $display_filename = $filename;
539 if (isset($passed_ent_id)) {
540 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
541 } else {
542 $passed_ent_id_link = '';
543 }
544 $defaultlink = $default_page . "?startMessage=$startMessage"
545 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
546 . '&amp;ent_id='.$ent.$passed_ent_id_link;
547 if ($where && $what) {
548 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
549 }
550 // IE does make use of mime content sniffing. Forcing a download
551 // prohibit execution of XSS inside an application/octet-stream attachment
552 if ($type0 == 'application' && $type1 == 'octet-stream') {
553 $defaultlink .= '&amp;absolute_dl=true';
554 }
555
556 /* This executes the attachment hook with a specific MIME-type.
557 * If that doesn't have results, it tries if there's a rule
558 * for a more generic type. Finally, a hook for ALL attachment
559 * types is run as well.
560 */
561 /* The API for this hook has changed as of 1.5.2 so that all plugin
562 arguments are passed in an array instead of each their own plugin
563 argument, and arguments are passed by reference, so instead of
564 returning any changes, changes should simply be made to the original
565 arguments themselves. */
566 do_hook("attachment $type0/$type1", $temp=array(&$links,
567 &$startMessage, &$id, &$urlMailbox, &$ent, &$defaultlink,
568 &$display_filename, &$where, &$what));
569 if(count($links) <= 1) {
570 /* The API for this hook has changed as of 1.5.2 so that all plugin
571 arguments are passed in an array instead of each their own plugin
572 argument, and arguments are passed by reference, so instead of
573 returning any changes, changes should simply be made to the original
574 arguments themselves. */
575 do_hook("attachment $type0/*", $temp=array(&$links,
576 &$startMessage, &$id, &$urlMailbox, &$ent, &$defaultlink,
577 &$display_filename, &$where, &$what));
578 }
579 /* The API for this hook has changed as of 1.5.2 so that all plugin
580 arguments are passed in an array instead of each their own plugin
581 argument, and arguments are passed by reference, so instead of
582 returning any changes, changes should simply be made to the original
583 arguments themselves. */
584 do_hook("attachment */*", $temp=array(&$links,
585 &$startMessage, &$id, &$urlMailbox, &$ent, &$defaultlink,
586 &$display_filename, &$where, &$what));
587
588 $this_attachment = array();
589 $this_attachment['Name'] = decodeHeader($display_filename);
590 $this_attachment['Description'] = $description;
591 $this_attachment['DefaultHREF'] = $defaultlink;
592 $this_attachment['DownloadHREF'] = $links['download link']['href'];
593 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
594 $this_attachment['Size'] = $header->size;
595 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
596 $this_attachment['OtherLinks'] = array();
597 foreach ($links as $val) {
598 if ($val['text']==_("Download") || $val['text'] == _("View"))
599 continue;
600 if (empty($val['text']) && empty($val['extra']))
601 continue;
602
603 $temp = array();
604 $temp['HREF'] = $val['href'];
605 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
606 $this_attachment['OtherLinks'][] = $temp;
607 }
608 $attachments[] = $this_attachment;
609
610 unset($links);
611 }
612
613 return $attachments;
614 }
615
616 /**
617 * Displays attachment links and information
618 *
619 * Since 1.3.0 function is not included in formatBody() call.
620 *
621 * Since 1.0.2 uses attachment $type0/$type1 hook.
622 * Since 1.2.5 uses attachment $type0/* hook.
623 * Since 1.5.0 uses attachments_bottom hook.
624 * Since 1.5.2 uses templates and does *not* return a value.
625 *
626 * @param object $message SquirrelMail message object
627 * @param array $exclude_id message parts that are not attachments.
628 * @param string $mailbox mailbox name
629 * @param integer $id message id
630 */
631 function formatAttachments($message, $exclude_id, $mailbox, $id) {
632 global $oTemplate;
633
634 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
635
636 $oTemplate->assign('attachments', $attach);
637 $oTemplate->display('read_attachments.tpl');
638 }
639
640 function sqimap_base64_decode(&$string) {
641
642 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
643 // fly decoding (to reduce memory usage) you have to check if the
644 // data has incomplete pairs
645
646 // Remove the noise in order to check if the 4 bytes pairs are complete
647 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
648
649 $sStringRem = '';
650 $iMod = strlen($string) % 4;
651 if ($iMod) {
652 $sStringRem = substr($string,-$iMod);
653 // Check if $sStringRem contains padding characters
654 if (substr($sStringRem,-1) != '=') {
655 $string = substr($string,0,-$iMod);
656 } else {
657 $sStringRem = '';
658 }
659 }
660 $string = base64_decode($string);
661 return $sStringRem;
662 }
663
664 /**
665 * Decodes encoded message body
666 *
667 * This function decodes the body depending on the encoding type.
668 * Currently quoted-printable and base64 encodings are supported.
669 * decode_body hook was added to this function in 1.4.2/1.5.0
670 * @param string $body encoded message body
671 * @param string $encoding used encoding
672 * @return string decoded string
673 * @since 1.0
674 */
675 function decodeBody($body, $encoding) {
676
677 $body = str_replace("\r\n", "\n", $body);
678 $encoding = strtolower($encoding);
679
680 $encoding_handler = do_hook('decode_body', $encoding);
681
682
683 // plugins get first shot at decoding the body
684 //
685 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
686 $body = $encoding_handler('decode', $body);
687
688 } elseif ($encoding == 'quoted-printable' ||
689 $encoding == 'quoted_printable') {
690 /**
691 * quoted_printable_decode() function is broken in older
692 * php versions. Text with \r\n decoding was fixed only
693 * in php 4.3.0. Minimal code requirement 4.0.4 +
694 * str_replace("\r\n", "\n", $body); call.
695 */
696 $body = quoted_printable_decode($body);
697 } elseif ($encoding == 'base64') {
698 $body = base64_decode($body);
699 }
700
701 // All other encodings are returned raw.
702 return $body;
703 }
704
705 /**
706 * Decodes headers
707 *
708 * This functions decode strings that is encoded according to
709 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
710 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
711 *
712 * @param string $string header string that has to be made readable
713 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
714 * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
715 * @param boolean $decide decide if string can be utfencoded. defaults to false
716 * @return string decoded header string
717 */
718 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
719 global $languages, $squirrelmail_language,$default_charset;
720 if (is_array($string)) {
721 $string = implode("\n", $string);
722 }
723
724 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
725 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
726 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
727 // Do we need to return at this point?
728 // return $string;
729 }
730 $i = 0;
731 $iLastMatch = -2;
732 $encoded = true;
733
734 $aString = explode(' ',$string);
735 $ret = '';
736 foreach ($aString as $chunk) {
737 if ($encoded && $chunk === '') {
738 continue;
739 } elseif ($chunk === '') {
740 $ret .= ' ';
741 continue;
742 }
743 $encoded = false;
744 /* if encoded words are not separated by a linear-space-white we still catch them */
745 $j = $i-1;
746
747 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
748 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
749 if ($iLastMatch !== $j) {
750 if ($htmlsave) {
751 $ret .= '&#32;';
752 } else {
753 $ret .= ' ';
754 }
755 }
756 $iLastMatch = $i;
757 $j = $i;
758 if ($htmlsave) {
759 $ret .= htmlspecialchars($res[1]);
760 } else {
761 $ret .= $res[1];
762 }
763 $encoding = ucfirst($res[3]);
764
765 /* decide about valid decoding */
766 if ($decide && is_conversion_safe($res[2])) {
767 $utfencode=true;
768 $can_be_encoded=true;
769 } else {
770 $can_be_encoded=false;
771 }
772 switch ($encoding)
773 {
774 case 'B':
775 $replace = base64_decode($res[4]);
776 if ($utfencode) {
777 if ($can_be_encoded) {
778 /* convert string to different charset,
779 * if functions asks for it (usually in compose)
780 */
781 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
782 } else {
783 // convert string to html codes in order to display it
784 $ret .= charset_decode($res[2],$replace);
785 }
786 } else {
787 if ($htmlsave) {
788 $replace = htmlspecialchars($replace);
789 }
790 $ret.= $replace;
791 }
792 break;
793 case 'Q':
794 $replace = str_replace('_', ' ', $res[4]);
795 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
796 $replace);
797 if ($utfencode) {
798 if ($can_be_encoded) {
799 /* convert string to different charset,
800 * if functions asks for it (usually in compose)
801 */
802 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
803 } else {
804 // convert string to html codes in order to display it
805 $replace = charset_decode($res[2], $replace);
806 }
807 } else {
808 if ($htmlsave) {
809 $replace = htmlspecialchars($replace);
810 }
811 }
812 $ret .= $replace;
813 break;
814 default:
815 break;
816 }
817 $chunk = $res[5];
818 $encoded = true;
819 }
820 if (!$encoded) {
821 if ($htmlsave) {
822 $ret .= '&#32;';
823 } else {
824 $ret .= ' ';
825 }
826 }
827
828 if (!$encoded && $htmlsave) {
829 $ret .= htmlspecialchars($chunk);
830 } else {
831 $ret .= $chunk;
832 }
833 ++$i;
834 }
835 /* remove the first added space */
836 if ($ret) {
837 if ($htmlsave) {
838 $ret = substr($ret,5);
839 } else {
840 $ret = substr($ret,1);
841 }
842 }
843
844 return $ret;
845 }
846
847 /**
848 * Encodes header
849 *
850 * Function uses XTRA_CODE _encodeheader function, if such function exists.
851 *
852 * Function uses Q encoding by default and encodes a string according to RFC
853 * 1522 for use in headers if it contains 8-bit characters or anything that
854 * looks like it should be encoded.
855 *
856 * Function switches to B encoding and encodeHeaderBase64() function, if
857 * string is 8bit and multibyte character set supported by mbstring extension
858 * is used. It can cause E_USER_NOTICE errors, if interface is used with
859 * multibyte character set unsupported by mbstring extension.
860 *
861 * @param string $string header string, that has to be encoded
862 * @return string quoted-printable encoded string
863 * @todo make $mb_charsets system wide constant
864 */
865 function encodeHeader ($string) {
866 global $default_charset, $languages, $squirrelmail_language;
867
868 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
869 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
870 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
871 }
872
873 // Use B encoding for multibyte charsets
874 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
875 if (in_array($default_charset,$mb_charsets) &&
876 in_array($default_charset,sq_mb_list_encodings()) &&
877 sq_is8bit($string)) {
878 return encodeHeaderBase64($string,$default_charset);
879 } elseif (in_array($default_charset,$mb_charsets) &&
880 sq_is8bit($string) &&
881 ! in_array($default_charset,sq_mb_list_encodings())) {
882 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
883 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
884 }
885
886 // Encode only if the string contains 8-bit characters or =?
887 $j = strlen($string);
888 $max_l = 75 - strlen($default_charset) - 7;
889 $aRet = array();
890 $ret = '';
891 $iEncStart = $enc_init = false;
892 $cur_l = $iOffset = 0;
893 for($i = 0; $i < $j; ++$i) {
894 switch($string{$i})
895 {
896 case '=':
897 case '<':
898 case '>':
899 case ',':
900 case '?':
901 case '_':
902 if ($iEncStart === false) {
903 $iEncStart = $i;
904 }
905 $cur_l+=3;
906 if ($cur_l > ($max_l-2)) {
907 /* if there is an stringpart that doesn't need encoding, add it */
908 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
909 $aRet[] = "=?$default_charset?Q?$ret?=";
910 $iOffset = $i;
911 $cur_l = 0;
912 $ret = '';
913 $iEncStart = false;
914 } else {
915 $ret .= sprintf("=%02X",ord($string{$i}));
916 }
917 break;
918 case '(':
919 case ')':
920 if ($iEncStart !== false) {
921 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
922 $aRet[] = "=?$default_charset?Q?$ret?=";
923 $iOffset = $i;
924 $cur_l = 0;
925 $ret = '';
926 $iEncStart = false;
927 }
928 break;
929 case ' ':
930 if ($iEncStart !== false) {
931 $cur_l++;
932 if ($cur_l > $max_l) {
933 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
934 $aRet[] = "=?$default_charset?Q?$ret?=";
935 $iOffset = $i;
936 $cur_l = 0;
937 $ret = '';
938 $iEncStart = false;
939 } else {
940 $ret .= '_';
941 }
942 }
943 break;
944 default:
945 $k = ord($string{$i});
946 if ($k > 126) {
947 if ($iEncStart === false) {
948 // do not start encoding in the middle of a string, also take the rest of the word.
949 $sLeadString = substr($string,0,$i);
950 $aLeadString = explode(' ',$sLeadString);
951 $sToBeEncoded = array_pop($aLeadString);
952 $iEncStart = $i - strlen($sToBeEncoded);
953 $ret .= $sToBeEncoded;
954 $cur_l += strlen($sToBeEncoded);
955 }
956 $cur_l += 3;
957 /* first we add the encoded string that reached it's max size */
958 if ($cur_l > ($max_l-2)) {
959 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
960 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
961 $cur_l = 3;
962 $ret = '';
963 $iOffset = $i;
964 $iEncStart = $i;
965 }
966 $enc_init = true;
967 $ret .= sprintf("=%02X", $k);
968 } else {
969 if ($iEncStart !== false) {
970 $cur_l++;
971 if ($cur_l > $max_l) {
972 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
973 $aRet[] = "=?$default_charset?Q?$ret?=";
974 $iEncStart = false;
975 $iOffset = $i;
976 $cur_l = 0;
977 $ret = '';
978 } else {
979 $ret .= $string{$i};
980 }
981 }
982 }
983 break;
984 }
985 }
986
987 if ($enc_init) {
988 if ($iEncStart !== false) {
989 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
990 $aRet[] = "=?$default_charset?Q?$ret?=";
991 } else {
992 $aRet[] = substr($string,$iOffset);
993 }
994 $string = implode('',$aRet);
995 }
996 return $string;
997 }
998
999 /**
1000 * Encodes string according to rfc2047 B encoding header formating rules
1001 *
1002 * It is recommended way to encode headers with character sets that store
1003 * symbols in more than one byte.
1004 *
1005 * Function requires mbstring support. If required mbstring functions are missing,
1006 * function returns false and sets E_USER_WARNING level error message.
1007 *
1008 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1009 * that mbstring functions will generate E_WARNING errors, if unsupported
1010 * character set is used. mb_encode_mimeheader function provided by php
1011 * mbstring extension is not used in order to get better control of header
1012 * encoding.
1013 *
1014 * Used php code functions - function_exists(), trigger_error(), strlen()
1015 * (is used with charset names and base64 strings). Used php mbstring
1016 * functions - mb_strlen and mb_substr.
1017 *
1018 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1019 * encoding), rfc 2822 (header folding)
1020 *
1021 * @param string $string header string that must be encoded
1022 * @param string $charset character set. Must be supported by mbstring extension.
1023 * Use sq_mb_list_encodings() to detect supported charsets.
1024 * @return string string encoded according to rfc2047 B encoding formating rules
1025 * @since 1.5.1
1026 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1027 * @todo Do we want to control max length of header?
1028 * @todo Do we want to control EOL (end-of-line) marker?
1029 * @todo Do we want to translate error message?
1030 */
1031 function encodeHeaderBase64($string,$charset) {
1032 /**
1033 * Check mbstring function requirements.
1034 */
1035 if (! function_exists('mb_strlen') ||
1036 ! function_exists('mb_substr')) {
1037 // set E_USER_WARNING
1038 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1039 // return false
1040 return false;
1041 }
1042
1043 // initial return array
1044 $aRet = array();
1045
1046 /**
1047 * header length = 75 symbols max (same as in encodeHeader)
1048 * remove $charset length
1049 * remove =? ? ?= (5 chars)
1050 * remove 2 more chars (\r\n ?)
1051 */
1052 $iMaxLength = 75 - strlen($charset) - 7;
1053
1054 // set first character position
1055 $iStartCharNum = 0;
1056
1057 // loop through all characters. count characters and not bytes.
1058 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1059 // encode string from starting character to current character.
1060 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1061
1062 // Check encoded string length
1063 if(strlen($encoded_string)>$iMaxLength) {
1064 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1065 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1066
1067 // set new starting character
1068 $iStartCharNum = $iCharNum-1;
1069
1070 // encode last char (in case it is last character in string)
1071 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1072 } // if string is shorter than max length - add next character
1073 }
1074
1075 // add last encoded string to array
1076 $aRet[] = $encoded_string;
1077
1078 // set initial return string
1079 $sRet = '';
1080
1081 // loop through encoded strings
1082 foreach($aRet as $string) {
1083 // TODO: Do we want to control EOL (end-of-line) marker
1084 if ($sRet!='') $sRet.= " ";
1085
1086 // add header tags and encoded string to return string
1087 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1088 }
1089
1090 return $sRet;
1091 }
1092
1093 /* This function trys to locate the entity_id of a specific mime element */
1094 function find_ent_id($id, $message) {
1095 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1096 if ($message->entities[$i]->header->type0 == 'multipart') {
1097 $ret = find_ent_id($id, $message->entities[$i]);
1098 } else {
1099 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1100 // if (sq_check_save_extension($message->entities[$i])) {
1101 return $message->entities[$i]->entity_id;
1102 // }
1103 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1104 /**
1105 * This is part of a fix for Outlook Express 6.x generating
1106 * cid URLs without creating content-id headers
1107 * @@JA - 20050207
1108 */
1109 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1110 return $message->entities[$i]->entity_id;
1111 }
1112 }
1113 }
1114 }
1115 return $ret;
1116 }
1117
1118 function sq_check_save_extension($message) {
1119 $filename = $message->getFilename();
1120 $ext = substr($filename, strrpos($filename,'.')+1);
1121 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1122 return in_array($ext, $save_extensions);
1123 }
1124
1125
1126 /**
1127 ** HTMLFILTER ROUTINES
1128 */
1129
1130 /**
1131 * This function checks attribute values for entity-encoded values
1132 * and returns them translated into 8-bit strings so we can run
1133 * checks on them.
1134 *
1135 * @param $attvalue A string to run entity check against.
1136 * @return Nothing, modifies a reference value.
1137 */
1138 function sq_defang(&$attvalue){
1139 $me = 'sq_defang';
1140 /**
1141 * Skip this if there aren't ampersands or backslashes.
1142 */
1143 if (strpos($attvalue, '&') === false
1144 && strpos($attvalue, '\\') === false){
1145 return;
1146 }
1147 $m = false;
1148 // before deent, translate the dangerous unicode characters and ... to safe values
1149 // otherwise the regular expressions do not match.
1150
1151
1152
1153 do {
1154 $m = false;
1155 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1156 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1157 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1158 } while ($m == true);
1159 $attvalue = stripslashes($attvalue);
1160 }
1161
1162 /**
1163 * Kill any tabs, newlines, or carriage returns. Our friends the
1164 * makers of the browser with 95% market value decided that it'd
1165 * be funny to make "java[tab]script" be just as good as "javascript".
1166 *
1167 * @param attvalue The attribute value before extraneous spaces removed.
1168 * @return attvalue Nothing, modifies a reference value.
1169 */
1170 function sq_unspace(&$attvalue){
1171 $me = 'sq_unspace';
1172 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1173 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1174 Array('', '', '', '', ''), $attvalue);
1175 }
1176 }
1177
1178 /**
1179 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1180 * IE as regular characters.
1181 *
1182 * @param attvalue The attribute value before dangerous characters are translated.
1183 * @return attvalue Nothing, modifies a reference value.
1184 * @author Marc Groot Koerkamp.
1185 */
1186 function sq_fixIE_idiocy(&$attvalue) {
1187 // remove NUL
1188 $attvalue = str_replace("\0", "", $attvalue);
1189 // remove comments
1190 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1191
1192 // IE has the evil habit of accepting every possible value for the attribute expression.
1193 // The table below contains characters which are parsed by IE if they are used in the "expression"
1194 // attribute value.
1195 $aDangerousCharsReplacementTable = array(
1196 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1197 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1198 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
1199 '&#xFF25;', '&#65317' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1200 '&#xFF45;', '&#65349' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1201 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1202 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1203 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1204 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1205 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1206 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1207 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1208 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1209 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1210 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1211 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1212 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1213 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1214 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1215 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1216 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1217 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1218 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1219 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1220 '&#x8264;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1221 '&#x8285;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1222 '&#x8277;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1223 '&#x8298;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1224 '&#x826F;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1225 '&#x8290;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1226 '&#x8271;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1227 '&#x8292;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1228 '&#x8272;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1229 '&#x8293;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1230 '&#x8268;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1231 '&#x8289;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1232 '&#x826E;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1233 '&#x828F;', /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1234 '&#x826D;', /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1235 '&#x828E;'), /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1236 array('l', 'l', 'r','r','n','n',
1237 'E','E','e','e','X','X','x','x','P','P','p','p','S','S','s','s','I','I',
1238 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n',
1239 'E','e','X','x','P','p','S','s','I','i','O','o','N','n'));
1240 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1241
1242 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1243 // used for XSS.
1244 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1245 }
1246
1247 /**
1248 * This function returns the final tag out of the tag name, an array
1249 * of attributes, and the type of the tag. This function is called by
1250 * sq_sanitize internally.
1251 *
1252 * @param $tagname the name of the tag.
1253 * @param $attary the array of attributes and their values
1254 * @param $tagtype The type of the tag (see in comments).
1255 * @return a string with the final tag representation.
1256 */
1257 function sq_tagprint($tagname, $attary, $tagtype){
1258 $me = 'sq_tagprint';
1259
1260 if ($tagtype == 2){
1261 $fulltag = '</' . $tagname . '>';
1262 } else {
1263 $fulltag = '<' . $tagname;
1264 if (is_array($attary) && sizeof($attary)){
1265 $atts = Array();
1266 while (list($attname, $attvalue) = each($attary)){
1267 array_push($atts, "$attname=$attvalue");
1268 }
1269 $fulltag .= ' ' . join(" ", $atts);
1270 }
1271 if ($tagtype == 3){
1272 $fulltag .= ' /';
1273 }
1274 $fulltag .= '>';
1275 }
1276 return $fulltag;
1277 }
1278
1279 /**
1280 * A small helper function to use with array_walk. Modifies a by-ref
1281 * value and makes it lowercase.
1282 *
1283 * @param $val a value passed by-ref.
1284 * @return void since it modifies a by-ref value.
1285 */
1286 function sq_casenormalize(&$val){
1287 $val = strtolower($val);
1288 }
1289
1290 /**
1291 * This function skips any whitespace from the current position within
1292 * a string and to the next non-whitespace value.
1293 *
1294 * @param $body the string
1295 * @param $offset the offset within the string where we should start
1296 * looking for the next non-whitespace character.
1297 * @return the location within the $body where the next
1298 * non-whitespace char is located.
1299 */
1300 function sq_skipspace($body, $offset){
1301 $me = 'sq_skipspace';
1302 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1303 if (sizeof($matches{1})){
1304 $count = strlen($matches{1});
1305 $offset += $count;
1306 }
1307 return $offset;
1308 }
1309
1310 /**
1311 * This function looks for the next character within a string. It's
1312 * really just a glorified "strpos", except it catches if failures
1313 * nicely.
1314 *
1315 * @param $body The string to look for needle in.
1316 * @param $offset Start looking from this position.
1317 * @param $needle The character/string to look for.
1318 * @return location of the next occurance of the needle, or
1319 * strlen($body) if needle wasn't found.
1320 */
1321 function sq_findnxstr($body, $offset, $needle){
1322 $me = 'sq_findnxstr';
1323 $pos = strpos($body, $needle, $offset);
1324 if ($pos === FALSE){
1325 $pos = strlen($body);
1326 }
1327 return $pos;
1328 }
1329
1330 /**
1331 * This function takes a PCRE-style regexp and tries to match it
1332 * within the string.
1333 *
1334 * @param $body The string to look for needle in.
1335 * @param $offset Start looking from here.
1336 * @param $reg A PCRE-style regex to match.
1337 * @return Returns a false if no matches found, or an array
1338 * with the following members:
1339 * - integer with the location of the match within $body
1340 * - string with whatever content between offset and the match
1341 * - string with whatever it is we matched
1342 */
1343 function sq_findnxreg($body, $offset, $reg){
1344 $me = 'sq_findnxreg';
1345 $matches = Array();
1346 $retarr = Array();
1347 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1348 if (!isset($matches{0}) || !$matches{0}){
1349 $retarr = false;
1350 } else {
1351 $retarr{0} = $offset + strlen($matches{1});
1352 $retarr{1} = $matches{1};
1353 $retarr{2} = $matches{2};
1354 }
1355 return $retarr;
1356 }
1357
1358 /**
1359 * This function looks for the next tag.
1360 *
1361 * @param $body String where to look for the next tag.
1362 * @param $offset Start looking from here.
1363 * @return false if no more tags exist in the body, or
1364 * an array with the following members:
1365 * - string with the name of the tag
1366 * - array with attributes and their values
1367 * - integer with tag type (1, 2, or 3)
1368 * - integer where the tag starts (starting "<")
1369 * - integer where the tag ends (ending ">")
1370 * first three members will be false, if the tag is invalid.
1371 */
1372 function sq_getnxtag($body, $offset){
1373 $me = 'sq_getnxtag';
1374 if ($offset > strlen($body)){
1375 return false;
1376 }
1377 $lt = sq_findnxstr($body, $offset, "<");
1378 if ($lt == strlen($body)){
1379 return false;
1380 }
1381 /**
1382 * We are here:
1383 * blah blah <tag attribute="value">
1384 * \---------^
1385 */
1386 $pos = sq_skipspace($body, $lt+1);
1387 if ($pos >= strlen($body)){
1388 return Array(false, false, false, $lt, strlen($body));
1389 }
1390 /**
1391 * There are 3 kinds of tags:
1392 * 1. Opening tag, e.g.:
1393 * <a href="blah">
1394 * 2. Closing tag, e.g.:
1395 * </a>
1396 * 3. XHTML-style content-less tag, e.g.:
1397 * <img src="blah" />
1398 */
1399 $tagtype = false;
1400 switch (substr($body, $pos, 1)){
1401 case '/':
1402 $tagtype = 2;
1403 $pos++;
1404 break;
1405 case '!':
1406 /**
1407 * A comment or an SGML declaration.
1408 */
1409 if (substr($body, $pos+1, 2) == "--"){
1410 $gt = strpos($body, "-->", $pos);
1411 if ($gt === false){
1412 $gt = strlen($body);
1413 } else {
1414 $gt += 2;
1415 }
1416 return Array(false, false, false, $lt, $gt);
1417 } else {
1418 $gt = sq_findnxstr($body, $pos, ">");
1419 return Array(false, false, false, $lt, $gt);
1420 }
1421 break;
1422 default:
1423 /**
1424 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1425 * later.
1426 */
1427 $tagtype = 1;
1428 break;
1429 }
1430
1431 $tag_start = $pos;
1432 $tagname = '';
1433 /**
1434 * Look for next [\W-_], which will indicate the end of the tag name.
1435 */
1436 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1437 if ($regary == false){
1438 return Array(false, false, false, $lt, strlen($body));
1439 }
1440 list($pos, $tagname, $match) = $regary;
1441 $tagname = strtolower($tagname);
1442
1443 /**
1444 * $match can be either of these:
1445 * '>' indicating the end of the tag entirely.
1446 * '\s' indicating the end of the tag name.
1447 * '/' indicating that this is type-3 xhtml tag.
1448 *
1449 * Whatever else we find there indicates an invalid tag.
1450 */
1451 switch ($match){
1452 case '/':
1453 /**
1454 * This is an xhtml-style tag with a closing / at the
1455 * end, like so: <img src="blah" />. Check if it's followed
1456 * by the closing bracket. If not, then this tag is invalid
1457 */
1458 if (substr($body, $pos, 2) == "/>"){
1459 $pos++;
1460 $tagtype = 3;
1461 } else {
1462 $gt = sq_findnxstr($body, $pos, ">");
1463 $retary = Array(false, false, false, $lt, $gt);
1464 return $retary;
1465 }
1466 case '>':
1467 return Array($tagname, false, $tagtype, $lt, $pos);
1468 break;
1469 default:
1470 /**
1471 * Check if it's whitespace
1472 */
1473 if (!preg_match('/\s/', $match)){
1474 /**
1475 * This is an invalid tag! Look for the next closing ">".
1476 */
1477 $gt = sq_findnxstr($body, $lt, ">");
1478 return Array(false, false, false, $lt, $gt);
1479 }
1480 break;
1481 }
1482
1483 /**
1484 * At this point we're here:
1485 * <tagname attribute='blah'>
1486 * \-------^
1487 *
1488 * At this point we loop in order to find all attributes.
1489 */
1490 $attname = '';
1491 $atttype = false;
1492 $attary = Array();
1493
1494 while ($pos <= strlen($body)){
1495 $pos = sq_skipspace($body, $pos);
1496 if ($pos == strlen($body)){
1497 /**
1498 * Non-closed tag.
1499 */
1500 return Array(false, false, false, $lt, $pos);
1501 }
1502 /**
1503 * See if we arrived at a ">" or "/>", which means that we reached
1504 * the end of the tag.
1505 */
1506 $matches = Array();
1507 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1508 /**
1509 * Yep. So we did.
1510 */
1511 $pos += strlen($matches{1});
1512 if ($matches{2} == "/>"){
1513 $tagtype = 3;
1514 $pos++;
1515 }
1516 return Array($tagname, $attary, $tagtype, $lt, $pos);
1517 }
1518
1519 /**
1520 * There are several types of attributes, with optional
1521 * [:space:] between members.
1522 * Type 1:
1523 * attrname[:space:]=[:space:]'CDATA'
1524 * Type 2:
1525 * attrname[:space:]=[:space:]"CDATA"
1526 * Type 3:
1527 * attr[:space:]=[:space:]CDATA
1528 * Type 4:
1529 * attrname
1530 *
1531 * We leave types 1 and 2 the same, type 3 we check for
1532 * '"' and convert to "&quot" if needed, then wrap in
1533 * double quotes. Type 4 we convert into:
1534 * attrname="yes".
1535 */
1536 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1537 if ($regary == false){
1538 /**
1539 * Looks like body ended before the end of tag.
1540 */
1541 return Array(false, false, false, $lt, strlen($body));
1542 }
1543 list($pos, $attname, $match) = $regary;
1544 $attname = strtolower($attname);
1545 /**
1546 * We arrived at the end of attribute name. Several things possible
1547 * here:
1548 * '>' means the end of the tag and this is attribute type 4
1549 * '/' if followed by '>' means the same thing as above
1550 * '\s' means a lot of things -- look what it's followed by.
1551 * anything else means the attribute is invalid.
1552 */
1553 switch($match){
1554 case '/':
1555 /**
1556 * This is an xhtml-style tag with a closing / at the
1557 * end, like so: <img src="blah" />. Check if it's followed
1558 * by the closing bracket. If not, then this tag is invalid
1559 */
1560 if (substr($body, $pos, 2) == "/>"){
1561 $pos++;
1562 $tagtype = 3;
1563 } else {
1564 $gt = sq_findnxstr($body, $pos, ">");
1565 $retary = Array(false, false, false, $lt, $gt);
1566 return $retary;
1567 }
1568 case '>':
1569 $attary{$attname} = '"yes"';
1570 return Array($tagname, $attary, $tagtype, $lt, $pos);
1571 break;
1572 default:
1573 /**
1574 * Skip whitespace and see what we arrive at.
1575 */
1576 $pos = sq_skipspace($body, $pos);
1577 $char = substr($body, $pos, 1);
1578 /**
1579 * Two things are valid here:
1580 * '=' means this is attribute type 1 2 or 3.
1581 * \w means this was attribute type 4.
1582 * anything else we ignore and re-loop. End of tag and
1583 * invalid stuff will be caught by our checks at the beginning
1584 * of the loop.
1585 */
1586 if ($char == "="){
1587 $pos++;
1588 $pos = sq_skipspace($body, $pos);
1589 /**
1590 * Here are 3 possibilities:
1591 * "'" attribute type 1
1592 * '"' attribute type 2
1593 * everything else is the content of tag type 3
1594 */
1595 $quot = substr($body, $pos, 1);
1596 if ($quot == "'"){
1597 $regary = sq_findnxreg($body, $pos+1, "\'");
1598 if ($regary == false){
1599 return Array(false, false, false, $lt, strlen($body));
1600 }
1601 list($pos, $attval, $match) = $regary;
1602 $pos++;
1603 $attary{$attname} = "'" . $attval . "'";
1604 } else if ($quot == '"'){
1605 $regary = sq_findnxreg($body, $pos+1, '\"');
1606 if ($regary == false){
1607 return Array(false, false, false, $lt, strlen($body));
1608 }
1609 list($pos, $attval, $match) = $regary;
1610 $pos++;
1611 $attary{$attname} = '"' . $attval . '"';
1612 } else {
1613 /**
1614 * These are hateful. Look for \s, or >.
1615 */
1616 $regary = sq_findnxreg($body, $pos, "[\s>]");
1617 if ($regary == false){
1618 return Array(false, false, false, $lt, strlen($body));
1619 }
1620 list($pos, $attval, $match) = $regary;
1621 /**
1622 * If it's ">" it will be caught at the top.
1623 */
1624 $attval = preg_replace("/\"/s", "&quot;", $attval);
1625 $attary{$attname} = '"' . $attval . '"';
1626 }
1627 } else if (preg_match("|[\w/>]|", $char)) {
1628 /**
1629 * That was attribute type 4.
1630 */
1631 $attary{$attname} = '"yes"';
1632 } else {
1633 /**
1634 * An illegal character. Find next '>' and return.
1635 */
1636 $gt = sq_findnxstr($body, $pos, ">");
1637 return Array(false, false, false, $lt, $gt);
1638 }
1639 break;
1640 }
1641 }
1642 /**
1643 * The fact that we got here indicates that the tag end was never
1644 * found. Return invalid tag indication so it gets stripped.
1645 */
1646 return Array(false, false, false, $lt, strlen($body));
1647 }
1648
1649 /**
1650 * Translates entities into literal values so they can be checked.
1651 *
1652 * @param $attvalue the by-ref value to check.
1653 * @param $regex the regular expression to check against.
1654 * @param $hex whether the entites are hexadecimal.
1655 * @return True or False depending on whether there were matches.
1656 */
1657 function sq_deent(&$attvalue, $regex, $hex=false){
1658 $me = 'sq_deent';
1659 $ret_match = false;
1660 // remove comments
1661 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1662 preg_match_all($regex, $attvalue, $matches);
1663 if (is_array($matches) && sizeof($matches[0]) > 0){
1664 $repl = Array();
1665 for ($i = 0; $i < sizeof($matches[0]); $i++){
1666 $numval = $matches[1][$i];
1667 if ($hex){
1668 $numval = hexdec($numval);
1669 }
1670 $repl{$matches[0][$i]} = chr($numval);
1671 }
1672 $attvalue = strtr($attvalue, $repl);
1673 return true;
1674 } else {
1675 return false;
1676 }
1677 }
1678
1679 /**
1680 * This function runs various checks against the attributes.
1681 *
1682 * @param $tagname String with the name of the tag.
1683 * @param $attary Array with all tag attributes.
1684 * @param $rm_attnames See description for sq_sanitize
1685 * @param $bad_attvals See description for sq_sanitize
1686 * @param $add_attr_to_tag See description for sq_sanitize
1687 * @param $message message object
1688 * @param $id message id
1689 * @return Array with modified attributes.
1690 */
1691 function sq_fixatts($tagname,
1692 $attary,
1693 $rm_attnames,
1694 $bad_attvals,
1695 $add_attr_to_tag,
1696 $message,
1697 $id,
1698 $mailbox
1699 ){
1700 $me = 'sq_fixatts';
1701 while (list($attname, $attvalue) = each($attary)){
1702 /**
1703 * See if this attribute should be removed.
1704 */
1705 foreach ($rm_attnames as $matchtag=>$matchattrs){
1706 if (preg_match($matchtag, $tagname)){
1707 foreach ($matchattrs as $matchattr){
1708 if (preg_match($matchattr, $attname)){
1709 unset($attary{$attname});
1710 continue;
1711 }
1712 }
1713 }
1714 }
1715 /**
1716 * Workaround for IE quirks
1717 */
1718 sq_fixIE_idiocy($attvalue);
1719
1720 /**
1721 * Remove any backslashes, entities, and extraneous whitespace.
1722 */
1723
1724 $oldattvalue = $attvalue;
1725 sq_defang($attvalue);
1726 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1727 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1728 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1729 $attvalue = "idiocy";
1730 $attary{$attname} = $attvalue;
1731 }
1732 sq_unspace($attvalue);
1733
1734 /**
1735 * Now let's run checks on the attvalues.
1736 * I don't expect anyone to comprehend this. If you do,
1737 * get in touch with me so I can drive to where you live and
1738 * shake your hand personally. :)
1739 */
1740 foreach ($bad_attvals as $matchtag=>$matchattrs){
1741 if (preg_match($matchtag, $tagname)){
1742 foreach ($matchattrs as $matchattr=>$valary){
1743 if (preg_match($matchattr, $attname)){
1744 /**
1745 * There are two arrays in valary.
1746 * First is matches.
1747 * Second one is replacements
1748 */
1749 list($valmatch, $valrepl) = $valary;
1750 $newvalue =
1751 preg_replace($valmatch, $valrepl, $attvalue);
1752 if ($newvalue != $attvalue){
1753 $attary{$attname} = $newvalue;
1754 }
1755 }
1756 }
1757 }
1758 }
1759
1760 /**
1761 * Replace empty src tags with the blank image. src is only used
1762 * for frames, images, and image inputs. Doing a replace should
1763 * not affect them working as should be, however it will stop
1764 * IE from being kicked off when src for img tags are not set
1765 */
1766 if (($attname == 'src') && ($attvalue == '""')) {
1767 $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1768 }
1769
1770 /**
1771 * Turn cid: urls into http-friendly ones.
1772 */
1773 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1774 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1775 }
1776
1777 /**
1778 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1779 * One day MS might actually make it match something useful, for now, falling
1780 * back to using cid2http, so we can grab the blank.png.
1781 */
1782 if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1783 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1784 }
1785
1786 }
1787 /**
1788 * See if we need to append any attributes to this tag.
1789 */
1790 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1791 if (preg_match($matchtag, $tagname)){
1792 $attary = array_merge($attary, $addattary);
1793 }
1794 }
1795 return $attary;
1796 }
1797
1798 /**
1799 * This function edits the style definition to make them friendly and
1800 * usable in SquirrelMail.
1801 *
1802 * @param $message the message object
1803 * @param $id the message id
1804 * @param $content a string with whatever is between <style> and </style>
1805 * @param $mailbox the message mailbox
1806 * @return a string with edited content.
1807 */
1808 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1809 global $view_unsafe_images;
1810 $me = 'sq_fixstyle';
1811 // workaround for </style> in between comments
1812 $iCurrentPos = $pos;
1813 $content = '';
1814 $sToken = '';
1815 $bSucces = false;
1816 $bEndTag = false;
1817 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
1818 $char = $body{$i};
1819 switch ($char) {
1820 case '<':
1821 $sToken .= $char;
1822 break;
1823 case '/':
1824 if ($sToken == '<') {
1825 $sToken .= $char;
1826 $bEndTag = true;
1827 } else {
1828 $content .= $char;
1829 }
1830 break;
1831 case '>':
1832 if ($bEndTag) {
1833 $sToken .= $char;
1834 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
1835 $newpos = $i + 1;
1836 $bSucces = true;
1837 break 2;
1838 } else {
1839 $content .= $sToken;
1840 }
1841 $bEndTag = false;
1842 } else {
1843 $content .= $char;
1844 }
1845 break;
1846 case '!':
1847 if ($sToken == '<') {
1848 // possible comment
1849 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
1850 $i = strpos($body,'-->',$i+3);
1851 if ($i === false) { // no end comment
1852 $i = strlen($body);
1853 }
1854 $sToken = '';
1855 }
1856 } else {
1857 $content .= $char;
1858 }
1859 break;
1860 default:
1861 if ($bEndTag) {
1862 $sToken .= $char;
1863 } else {
1864 $content .= $char;
1865 }
1866 break;
1867 }
1868 }
1869 if ($bSucces == FALSE){
1870 return array(FALSE, strlen($body));
1871 }
1872
1873
1874
1875 /**
1876 * First look for general BODY style declaration, which would be
1877 * like so:
1878 * body {background: blah-blah}
1879 * and change it to .bodyclass so we can just assign it to a <div>
1880 */
1881 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1882 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1883 /**
1884 * Fix url('blah') declarations.
1885 */
1886 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1887 // "url(\\1$secremoveimg\\2)", $content);
1888
1889 // IE Sucks hard. We have a special function for it.
1890 sq_fixIE_idiocy($content);
1891
1892 // remove @import line
1893 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
1894
1895 // translate ur\l and variations (IE parses that)
1896 // TODO check if the sq_fixIE_idiocy function already handles this.
1897 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
1898 // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1899 while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1900 $sProto = strtolower($matches[1]);
1901 switch ($sProto) {
1902 /**
1903 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1904 * is false.
1905 */
1906 case 'https':
1907 case 'http':
1908 if (!$view_unsafe_images){
1909
1910 $sExpr = "/url\s*\(\s*[\'\"]?\s*$sProto*:.*[\'\"]?\s*\)/si";
1911 $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1912
1913 } else {
1914 $content = preg_replace('/url/i',"u\0r\0l",$content);
1915 }
1916 break;
1917 /**
1918 * Fix urls that refer to cid:
1919 */
1920 case 'cid':
1921 $cidurl = 'cid:'. $matches[2];
1922 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1923 // escape parentheses that can modify the regular expression
1924 $cidurl = str_replace(array('(',')'),array('\\(','\\)'),$cidurl);
1925 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1926 "u\0r\0l($httpurl)", $content);
1927 break;
1928 default:
1929 /**
1930 * replace url with protocol other then the white list
1931 * http,https and cid by an empty string.
1932 */
1933 $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1934 "", $content);
1935 break;
1936 }
1937 }
1938 // remove NUL
1939 $content = str_replace("\0", "", $content);
1940 /**
1941 * Remove any backslashes, entities, and extraneous whitespace.
1942 */
1943 $contentTemp = $content;
1944 sq_defang($contentTemp);
1945 sq_unspace($contentTemp);
1946
1947 /**
1948 * Fix stupid css declarations which lead to vulnerabilities
1949 * in IE.
1950 */
1951 $match = Array('/\/\*.*\*\//',
1952 '/expression/i',
1953 '/behaviou*r/i',
1954 '/binding/i',
1955 '/include-source/i',
1956 '/javascript/i',
1957 '/script/i');
1958 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
1959 $contentNew = preg_replace($match, $replace, $contentTemp);
1960 if ($contentNew !== $contentTemp) {
1961 // insecure css declarations are used. From now on we don't care
1962 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1963 $content = $contentNew;
1964 }
1965 return array($content, $newpos);
1966 }
1967
1968
1969 /**
1970 * This function converts cid: url's into the ones that can be viewed in
1971 * the browser.
1972 *
1973 * @param $message the message object
1974 * @param $id the message id
1975 * @param $cidurl the cid: url.
1976 * @param $mailbox the message mailbox
1977 * @return a string with a http-friendly url
1978 */
1979 function sq_cid2http($message, $id, $cidurl, $mailbox){
1980 /**
1981 * Get rid of quotes.
1982 */
1983 $quotchar = substr($cidurl, 0, 1);
1984 if ($quotchar == '"' || $quotchar == "'"){
1985 $cidurl = str_replace($quotchar, "", $cidurl);
1986 } else {
1987 $quotchar = '';
1988 }
1989 $cidurl = substr(trim($cidurl), 4);
1990
1991 $match_str = '/\{.*?\}\//';
1992 $str_rep = '';
1993 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1994
1995 $linkurl = find_ent_id($cidurl, $message);
1996 /* in case of non-save cid links $httpurl should be replaced by a sort of
1997 unsave link image */
1998 $httpurl = '';
1999
2000 /**
2001 * This is part of a fix for Outlook Express 6.x generating
2002 * cid URLs without creating content-id headers. These images are
2003 * not part of the multipart/related html mail. The html contains
2004 * <img src="cid:{some_id}/image_filename.ext"> references to
2005 * attached images with as goal to render them inline although
2006 * the attachment disposition property is not inline.
2007 */
2008
2009 if (empty($linkurl)) {
2010 if (preg_match('/{.*}\//', $cidurl)) {
2011 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2012 if (!empty($cidurl)) {
2013 $linkurl = find_ent_id($cidurl, $message);
2014 }
2015 }
2016 }
2017
2018 if (!empty($linkurl)) {
2019 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
2020 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2021 '&amp;ent_id=' . $linkurl . $quotchar;
2022 } else {
2023 /**
2024 * If we couldn't generate a proper img url, drop in a blank image
2025 * instead of sending back empty, otherwise it causes unusual behaviour
2026 */
2027 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
2028 }
2029
2030 return $httpurl;
2031 }
2032
2033 /**
2034 * This function changes the <body> tag into a <div> tag since we
2035 * can't really have a body-within-body.
2036 *
2037 * @param $attary an array of attributes and values of <body>
2038 * @param $mailbox mailbox we're currently reading (for cid2http)
2039 * @param $message current message (for cid2http)
2040 * @param $id current message id (for cid2http)
2041 * @return a modified array of attributes to be set for <div>
2042 */
2043 function sq_body2div($attary, $mailbox, $message, $id){
2044 $me = 'sq_body2div';
2045 $divattary = Array('class' => "'bodyclass'");
2046 $text = '#000000';
2047 $has_bgc_stl = $has_txt_stl = false;
2048 $styledef = '';
2049 if (is_array($attary) && sizeof($attary) > 0){
2050 foreach ($attary as $attname=>$attvalue){
2051 $quotchar = substr($attvalue, 0, 1);
2052 $attvalue = str_replace($quotchar, "", $attvalue);
2053 switch ($attname){
2054 case 'background':
2055 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2056 $styledef .= "background-image: url('$attvalue'); ";
2057 break;
2058 case 'bgcolor':
2059 $has_bgc_stl = true;
2060 $styledef .= "background-color: $attvalue; ";
2061 break;
2062 case 'text':
2063 $has_txt_stl = true;
2064 $styledef .= "color: $attvalue; ";
2065 break;
2066 }
2067 }
2068 // Outlook defines a white bgcolor and no text color. This can lead to
2069 // white text on a white bg with certain themes.
2070 if ($has_bgc_stl && !$has_txt_stl) {
2071 $styledef .= "color: $text; ";
2072 }
2073 if (strlen($styledef) > 0){
2074 $divattary{"style"} = "\"$styledef\"";
2075 }
2076 }
2077 return $divattary;
2078 }
2079
2080 /**
2081 * This is the main function and the one you should actually be calling.
2082 * There are several variables you should be aware of an which need
2083 * special description.
2084 *
2085 * Since the description is quite lengthy, see it here:
2086 * http://linux.duke.edu/projects/mini/htmlfilter/
2087 *
2088 * @param $body the string with HTML you wish to filter
2089 * @param $tag_list see description above
2090 * @param $rm_tags_with_content see description above
2091 * @param $self_closing_tags see description above
2092 * @param $force_tag_closing see description above
2093 * @param $rm_attnames see description above
2094 * @param $bad_attvals see description above
2095 * @param $add_attr_to_tag see description above
2096 * @param $message message object
2097 * @param $id message id
2098 * @return sanitized html safe to show on your pages.
2099 */
2100 function sq_sanitize($body,
2101 $tag_list,
2102 $rm_tags_with_content,
2103 $self_closing_tags,
2104 $force_tag_closing,
2105 $rm_attnames,
2106 $bad_attvals,
2107 $add_attr_to_tag,
2108 $message,
2109 $id,
2110 $mailbox
2111 ){
2112 $me = 'sq_sanitize';
2113 $rm_tags = array_shift($tag_list);
2114 /**
2115 * Normalize rm_tags and rm_tags_with_content.
2116 */
2117 @array_walk($tag_list, 'sq_casenormalize');
2118 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2119 @array_walk($self_closing_tags, 'sq_casenormalize');
2120 /**
2121 * See if tag_list is of tags to remove or tags to allow.
2122 * false means remove these tags
2123 * true means allow these tags
2124 */
2125 $curpos = 0;
2126 $open_tags = Array();
2127 $trusted = "\n<!-- begin sanitized html -->\n";
2128 $skip_content = false;
2129 /**
2130 * Take care of netscape's stupid javascript entities like
2131 * &{alert('boo')};
2132 */
2133 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
2134
2135 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2136 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2137 $free_content = substr($body, $curpos, $lt-$curpos);
2138 /**
2139 * Take care of <style>
2140 */
2141 if ($tagname == "style" && $tagtype == 1){
2142 list($free_content, $curpos) =
2143 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
2144 if ($free_content != FALSE){
2145 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2146 $trusted .= $free_content;
2147 $trusted .= sq_tagprint($tagname, false, 2);
2148 }
2149 continue;
2150 }
2151 if ($skip_content == false){
2152 $trusted .= $free_content;
2153 }
2154 if ($tagname != FALSE){
2155 if ($tagtype == 2){
2156 if ($skip_content == $tagname){
2157 /**
2158 * Got to the end of tag we needed to remove.
2159 */
2160 $tagname = false;
2161 $skip_content = false;
2162 } else {
2163 if ($skip_content == false){
2164 if ($tagname == "body"){
2165 $tagname = "div";
2166 }
2167 if (isset($open_tags{$tagname}) &&
2168 $open_tags{$tagname} > 0){
2169 $open_tags{$tagname}--;
2170 } else {
2171 $tagname = false;
2172 }
2173 }
2174 }
2175 } else {
2176 /**
2177 * $rm_tags_with_content
2178 */
2179 if ($skip_content == false){
2180 /**
2181 * See if this is a self-closing type and change
2182 * tagtype appropriately.
2183 */
2184 if ($tagtype == 1
2185 && in_array($tagname, $self_closing_tags)){
2186 $tagtype = 3;
2187 }
2188 /**
2189 * See if we should skip this tag and any content
2190 * inside it.
2191 */
2192 if ($tagtype == 1 &&
2193 in_array($tagname, $rm_tags_with_content)){
2194 $skip_content = $tagname;
2195 } else {
2196 if (($rm_tags == false
2197 && in_array($tagname, $tag_list)) ||
2198 ($rm_tags == true &&
2199 !in_array($tagname, $tag_list))){
2200 $tagname = false;
2201 } else {
2202 /**
2203 * Convert body into div.
2204 */
2205 if ($tagname == "body"){
2206 $tagname = "div";
2207 $attary = sq_body2div($attary, $mailbox,
2208 $message, $id);
2209 }
2210 if ($tagtype == 1){
2211 if (isset($open_tags{$tagname})){
2212 $open_tags{$tagname}++;
2213 } else {
2214 $open_tags{$tagname}=1;
2215 }
2216 }
2217 /**
2218 * This is where we run other checks.
2219 */
2220 if (is_array($attary) && sizeof($attary) > 0){
2221 $attary = sq_fixatts($tagname,
2222 $attary,
2223 $rm_attnames,
2224 $bad_attvals,
2225 $add_attr_to_tag,
2226 $message,
2227 $id,
2228 $mailbox
2229 );
2230 }
2231 }
2232 }
2233 }
2234 }
2235 if ($tagname != false && $skip_content == false){
2236 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2237 }
2238 }
2239 $curpos = $gt+1;
2240 }
2241 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2242 if ($force_tag_closing == true){
2243 foreach ($open_tags as $tagname=>$opentimes){
2244 while ($opentimes > 0){
2245 $trusted .= '</' . $tagname . '>';
2246 $opentimes--;
2247 }
2248 }
2249 $trusted .= "\n";
2250 }
2251 $trusted .= "<!-- end sanitized html -->\n";
2252 return $trusted;
2253 }
2254
2255 /**
2256 * This is a wrapper function to call html sanitizing routines.
2257 *
2258 * @param $body the body of the message
2259 * @param $id the id of the message
2260
2261 * @param $message
2262 * @param $mailbox
2263 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2264 * into internal SM compose links
2265 * (optional; default = TRUE)
2266 * @return a string with html safe to display in the browser.
2267 */
2268 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2269
2270 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2271
2272 global $attachment_common_show_images, $view_unsafe_images,
2273 $has_unsafe_images;
2274 /**
2275 * Don't display attached images in HTML mode.
2276 *
2277 * SB: why?
2278 */
2279 $attachment_common_show_images = false;
2280 $tag_list = Array(
2281 false,
2282 "object",
2283 "meta",
2284 "html",
2285 "head",
2286 "base",
2287 "link",
2288 "frame",
2289 "iframe",
2290 "plaintext",
2291 "marquee"
2292 );
2293
2294 $rm_tags_with_content = Array(
2295 "script",
2296 "applet",
2297 "embed",
2298 "title",
2299 "frameset",
2300 "xmp",
2301 "xml"
2302 );
2303
2304 $self_closing_tags = Array(
2305 "img",
2306 "br",
2307 "hr",
2308 "input",
2309 "outbind"
2310 );
2311
2312 $force_tag_closing = true;
2313
2314 $rm_attnames = Array(
2315 "/.*/" =>
2316 Array(
2317 "/target/i",
2318 "/^on.*/i",
2319 "/^dynsrc/i",
2320 "/^data.*/i",
2321 "/^lowsrc.*/i"
2322 )
2323 );
2324
2325 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2326 $bad_attvals = Array(
2327 "/.*/" =>
2328 Array(
2329 "/^src|background/i" =>
2330 Array(
2331 Array(
2332 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2333 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2334 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2335 ),
2336 Array(
2337 "\\1$secremoveimg\\2",
2338 "\\1$secremoveimg\\2",
2339 "\\1$secremoveimg\\2",
2340 )
2341 ),
2342 "/^href|action/i" =>
2343 Array(
2344 Array(
2345 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2346 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2347 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2348 ),
2349 Array(
2350 "\\1#\\1",
2351 "\\1#\\1",
2352 "\\1#\\1"
2353 )
2354 ),
2355 "/^style/i" =>
2356 Array(
2357 Array(
2358 "/\/\*.*\*\//",
2359 "/expression/i",
2360 "/binding/i",
2361 "/behaviou*r/i",
2362 "/include-source/i",
2363 "/position\s*:\s*absolute/i",
2364 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2365 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2366 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2367 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2368 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2369 ),
2370 Array(
2371 "",
2372 "idiocy",
2373 "idiocy",
2374 "idiocy",
2375 "idiocy",
2376 "",
2377 "url",
2378 "url(\\1#\\1)",
2379 "url(\\1#\\1)",
2380 "url(\\1#\\1)",
2381 "\\1:url(\\2#\\3)"
2382 )
2383 )
2384 )
2385 );
2386 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2387 $view_unsafe_images = false;
2388 }
2389 if (!$view_unsafe_images){
2390 /**
2391 * Remove any references to http/https if view_unsafe_images set
2392 * to false.
2393 */
2394 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2395 '/^([\'\"])\s*https*:.*([\'\"])/si');
2396 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2397 "\\1$secremoveimg\\1");
2398 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2399 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2400 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2401 "url(\\1$secremoveimg\\1)");
2402 }
2403
2404 $add_attr_to_tag = Array(
2405 "/^a$/i" =>
2406 Array('target'=>'"_blank"',
2407 'title'=>'"'._("This external link will open in a new window").'"'
2408 )
2409 );
2410 $trusted = sq_sanitize($body,
2411 $tag_list,
2412 $rm_tags_with_content,
2413 $self_closing_tags,
2414 $force_tag_closing,
2415 $rm_attnames,
2416 $bad_attvals,
2417 $add_attr_to_tag,
2418 $message,
2419 $id,
2420 $mailbox
2421 );
2422 if (preg_match("|$secremoveimg|i", $trusted)){
2423 $has_unsafe_images = true;
2424 }
2425
2426 // we want to parse mailto's in HTML output, change to SM compose links
2427 // this is a modified version of code from url_parser.php... but Marc is
2428 // right: we need a better filtering implementation; adding this randomly
2429 // here is not a great solution
2430 //
2431 if ($take_mailto_links) {
2432 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2433 global $MailTo_PReg_Match;
2434 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2435 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2436 foreach ($regs[0] as $i => $mailto_before) {
2437 $mailto_params = $regs[10][$i];
2438 // get rid of any tailing quote since we have to add send_to to the end
2439 //
2440 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2441 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2442 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2443 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2444
2445 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2446 $to = 'to=' . $regs[1][$i];
2447 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2448 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2449 else {
2450 if ($mailto_params) //already some params, append to them
2451 $mailto_params .= '&amp;' . $to;
2452 else
2453 $mailto_params .= '?' . $to;
2454 }
2455 }
2456
2457 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2458
2459 // we'll already have target=_blank, no need to allow comp_in_new
2460 // here (which would be a lot more work anyway)
2461 //
2462 global $compose_new_win;
2463 $temp_comp_in_new = $compose_new_win;
2464 $compose_new_win = 0;
2465 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2466 $compose_new_win = $temp_comp_in_new;
2467
2468 // remove <a href=" and anything after the next quote (we only
2469 // need the uri, not the link HTML) in compose uri
2470 //
2471 $comp_uri = substr($comp_uri, 9);
2472 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2473 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2474 }
2475 }
2476 }
2477
2478 return $trusted;
2479 }
2480
2481 /**
2482 * function SendDownloadHeaders - send file to the browser
2483 *
2484 * Original Source: SM core src/download.php
2485 * moved here to make it available to other code, and separate
2486 * front end from back end functionality.
2487 *
2488 * @param string $type0 first half of mime type
2489 * @param string $type1 second half of mime type
2490 * @param string $filename filename to tell the browser for downloaded file
2491 * @param boolean $force whether to force the download dialog to pop
2492 * @param optional integer $filesize send the Content-Header and length to the browser
2493 * @return void
2494 */
2495 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2496 global $languages, $squirrelmail_language;
2497 $isIE = $isIE6plus = false;
2498
2499 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2500
2501 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2502 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2503 $isIE = true;
2504 }
2505
2506 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2507 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2508 $isIE6plus = true;
2509 }
2510
2511 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2512 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2513 $filename =
2514 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2515 } else {
2516 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2517 }
2518
2519 // A Pox on Microsoft and it's Internet Explorer!
2520 //
2521 // IE has lots of bugs with file downloads.
2522 // It also has problems with SSL. Both of these cause problems
2523 // for us in this function.
2524 //
2525 // See this article on Cache Control headers and SSL
2526 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2527 //
2528 // The best thing you can do for IE is to upgrade to the latest
2529 // version
2530 //set all the Cache Control Headers for IE
2531 if ($isIE) {
2532 $filename=rawurlencode($filename);
2533 header ("Pragma: public");
2534 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2535 header ("Cache-Control: post-check=0, pre-check=0", false);
2536 header ("Cache-Control: private");
2537
2538 //set the inline header for IE, we'll add the attachment header later if we need it
2539 header ("Content-Disposition: inline; filename=$filename");
2540 }
2541
2542 if (!$force) {
2543 // Try to show in browser window
2544 header ("Content-Disposition: inline; filename=\"$filename\"");
2545 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2546 } else {
2547 // Try to pop up the "save as" box
2548
2549 // IE makes this hard. It pops up 2 save boxes, or none.
2550 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2551 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2552 // But, according to Microsoft, it is "RFC compliant but doesn't
2553 // take into account some deviations that allowed within the
2554 // specification." Doesn't that mean RFC non-compliant?
2555 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2556
2557 // all browsers need the application/octet-stream header for this
2558 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2559
2560 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2561 // Do not have quotes around filename, but that applied to
2562 // "attachment"... does it apply to inline too?
2563 header ("Content-Disposition: attachment; filename=\"$filename\"");
2564
2565 if ($isIE && !$isIE6plus) {
2566 // This combination seems to work mostly. IE 5.5 SP 1 has
2567 // known issues (see the Microsoft Knowledge Base)
2568
2569 // This works for most types, but doesn't work with Word files
2570 header ("Content-Type: application/download; name=\"$filename\"");
2571 header ("Content-Type: application/force-download; name=\"$filename\"");
2572 // These are spares, just in case. :-)
2573 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2574 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2575 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2576 } else if ($isIE) {
2577 // This is to prevent IE for MIME sniffing and auto open a file in IE
2578 header ("Content-Type: application/force-download; name=\"$filename\"");
2579 } else {
2580 // another application/octet-stream forces download for Netscape
2581 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2582 }
2583 }
2584
2585 //send the content-length header if the calling function provides it
2586 if ($filesize > 0) {
2587 header("Content-Length: $filesize");
2588 }
2589
2590 } // end fn SendDownloadHeaders
Unnamed repository; edit this file 'description' to name the repository.