Updating my previous comments.
[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 displayPageHeader( $color, $mailbox );
64 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
65 $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
66 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
67 plain_error_message( $errormessage );
68 echo '</body></html>';
69 exit;
70 }
71 if (count($flags)) {
72 foreach ($flags as $flag) {
73 //FIXME: please document why it is we have to check the first char of the flag but we then go ahead and do a full string comparison anyway. Is this a speed enhancement? If not, let's keep it simple and just compare the full string and forget the switch block.
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 = $regs[1];
150 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
151 $ret = $wholemessage;
152 } else {
153 global $where, $what, $mailbox, $passed_id, $startMessage;
154 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
155 if (isset($where) && isset($what)) {
156 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
157 } else {
158 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
159 }
160 $par .= '&amp;response=' . urlencode($response) .
161 '&amp;message=' . urlencode($message) .
162 '&amp;topline=' . urlencode($topline);
163
164 echo '<tt><br />' .
165 '<table width="80%"><tr>' .
166 '<tr><td colspan="2">' .
167 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
168 '</td></tr>' .
169 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
170 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
171 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
172 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
173 "</table><br /></tt></font><hr />";
174
175 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
176 array_shift($data);
177 $wholemessage = implode('', $data);
178
179 $ret = $wholemessage;
180 }
181 return $ret;
182 }
183
184 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout') {
185
186 /* Don't kill the connection if the browser is over a dialup
187 * and it would take over 30 seconds to download it.
188 * Don't call set_time_limit in safe mode.
189 */
190
191 if (!ini_get('safe_mode')) {
192 set_time_limit(0);
193 }
194 /* in case of base64 encoded attachments, do not buffer them.
195 Instead, echo the decoded attachment directly to screen */
196 if (strtolower($encoding) == 'base64') {
197 if (!$ent_id) {
198 $query = "FETCH $id BODY[]";
199 } else {
200 $query = "FETCH $id BODY[$ent_id]";
201 }
202 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
203 } else {
204 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
205 if (is_resource($rStream)) {
206 fputs($rStream,decodeBody($body,$encoding));
207 } else {
208 echo decodeBody($body, $encoding);
209 }
210 }
211
212 /*
213 TODO, use the same method for quoted printable.
214 However, I assume that quoted printable attachments aren't that large
215 so the performancegain / memory usage drop will be minimal.
216 If we decide to add that then we need to adapt sqimap_fread because
217 we need to split te result on \n and fread doesn't stop at \n. That
218 means we also should provide $results from sqimap_fread (by ref) to
219 te function and set $no_return to false. The $filter function for
220 quoted printable should handle unsetting of $results.
221 */
222 /*
223 TODO 2: find out how we write to the output stream php://stdout. fwrite
224 doesn't work because 'php://stdout isn't a stream.
225 */
226
227 return;
228 }
229
230 /* -[ END MIME DECODING ]----------------------------------------------------------- */
231
232 /* This is here for debugging purposes. It will print out a list
233 * of all the entity IDs that are in the $message object.
234 */
235 function listEntities ($message) {
236 if ($message) {
237 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
238 for ($i = 0; isset($message->entities[$i]); $i++) {
239 echo "$i : ";
240 $msg = listEntities($message->entities[$i]);
241
242 if ($msg) {
243 echo "return: ";
244 return $msg;
245 }
246 }
247 }
248 }
249
250 function getPriorityStr($priority) {
251 $priority_level = substr($priority,0,1);
252
253 switch($priority_level) {
254 /* Check for a higher then normal priority. */
255 case '1':
256 case '2':
257 $priority_string = _("High");
258 break;
259
260 /* Check for a lower then normal priority. */
261 case '4':
262 case '5':
263 $priority_string = _("Low");
264 break;
265
266 /* Check for a normal priority. */
267 case '3':
268 default:
269 $priority_level = '3';
270 $priority_string = _("Normal");
271 break;
272
273 }
274 return $priority_string;
275 }
276
277 /* returns a $message object for a particular entity id */
278 function getEntity ($message, $ent_id) {
279 return $message->getEntity($ent_id);
280 }
281
282 /* translateText
283 * Extracted from strings.php 23/03/2002
284 */
285
286 function translateText(&$body, $wrap_at, $charset) {
287 global $where, $what; /* from searching */
288 global $color; /* color theme */
289
290 // require_once(SM_PATH . 'functions/url_parser.php');
291
292 $body_ary = explode("\n", $body);
293 for ($i=0; $i < count($body_ary); $i++) {
294 $line = $body_ary[$i];
295 if (strlen($line) - 2 >= $wrap_at) {
296 sqWordWrap($line, $wrap_at, $charset);
297 }
298 $line = charset_decode($charset, $line);
299 $line = str_replace("\t", ' ', $line);
300
301 parseUrl ($line);
302
303 $quotes = 0;
304 $pos = 0;
305 $j = strlen($line);
306
307 while ($pos < $j) {
308 if ($line[$pos] == ' ') {
309 $pos++;
310 } else if (strpos($line, '&gt;', $pos) === $pos) {
311 $pos += 4;
312 $quotes++;
313 } else {
314 break;
315 }
316 }
317
318 if ($quotes % 2) {
319 $line = '<span class="quote1">' . $line . '</span>';
320 } elseif ($quotes) {
321 $line = '<span class="quote2">' . $line . '</span>';
322 }
323
324 $body_ary[$i] = $line;
325 }
326 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
327 }
328
329 /**
330 * This returns a parsed string called $body. That string can then
331 * be displayed as the actual message in the HTML. It contains
332 * everything needed, including HTML Tags, Attachments at the
333 * bottom, etc.
334 *
335 * Since 1.2.0 function uses message_body hook.
336 * Till 1.3.0 function included output of formatAttachments().
337 *
338 * @param resource $imap_stream imap connection resource
339 * @param object $message squirrelmail message object
340 * @param array $color squirrelmail color theme array
341 * @param integer $wrap_at number of characters per line
342 * @param string $ent_num (since 1.3.0) message part id
343 * @param integer $id (since 1.3.0) message id
344 * @param string $mailbox (since 1.3.0) imap folder name
345 * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
346 * @return string html formated message text
347 */
348 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
349 /* This if statement checks for the entity to show as the
350 * primary message. To add more of them, just put them in the
351 * order that is their priority.
352 */
353 global $startMessage, $languages, $squirrelmail_language,
354 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
355 $use_iframe, $iframe_height, $download_and_unsafe_link,
356 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
357 $oTemplate, $nbsp;
358
359 // workaround for not updated config.php
360 if (! isset($use_iframe)) $use_iframe = false;
361
362 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
363 // images off by default.
364 // FIXME: Update this code to use the default value FALSE.
365 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
366 $view_unsafe_images = false;
367 }
368
369 $body = '';
370 $urlmailbox = urlencode($mailbox);
371 $body_message = getEntity($message, $ent_num);
372 if (($body_message->header->type0 == 'text') ||
373 ($body_message->header->type0 == 'rfc822')) {
374 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
375 $body = decodeBody($body, $body_message->header->encoding);
376
377 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
378 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
379 if (mb_detect_encoding($body) != 'ASCII') {
380 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
381 }
382 }
383
384 /* As of 1.5.2, $body is passed (and modified) by reference */
385 do_hook('message_body', $body);
386
387 /* If there are other types that shouldn't be formatted, add
388 * them here.
389 */
390
391 if ($body_message->header->type1 == 'html') {
392 if ($show_html_default <> 1) {
393 $entity_conv = array('&nbsp;' => ' ',
394 '<p>' => "\n",
395 '<P>' => "\n",
396 '<br>' => "\n",
397 '<BR>' => "\n",
398 '<br />' => "\n",
399 '<BR />' => "\n",
400 '&gt;' => '>',
401 '&lt;' => '<');
402 $body = strtr($body, $entity_conv);
403 $body = strip_tags($body);
404 $body = trim($body);
405 translateText($body, $wrap_at,
406 $body_message->header->getParameter('charset'));
407 } elseif ($use_iframe && ! $clean) {
408 // $clean is used to remove iframe in printable view.
409
410 /**
411 * If we don't add html message between iframe tags,
412 * we must detect unsafe images and modify $has_unsafe_images.
413 */
414 $html_body = magicHTML($body, $id, $message, $mailbox);
415 // Convert character set in order to display html mails in different character set
416 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
417
418 // creating iframe url
419 $iframeurl=sqm_baseuri().'src/view_html.php?'
420 . 'mailbox=' . $urlmailbox
421 . '&amp;passed_id=' . $id
422 . '&amp;ent_id=' . $ent_num
423 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
424
425 global $oTemplate;
426 $oTemplate->assign('iframe_url', $iframeurl);
427 $oTemplate->assign('iframe_height', $iframe_height);
428 $oTemplate->assign('html_body', $html_body);
429
430 $body = $oTemplate->fetch('read_html_iframe.tpl');
431 } else {
432 // old way of html rendering
433 /**
434 * convert character set. charset_decode does not remove html special chars
435 * applied by magicHTML functions and does not sanitize them second time if
436 * fourth argument is true.
437 */
438 $charset = $body_message->header->getParameter('charset');
439 if (!empty($charset)) {
440 $body = charset_decode($charset,$body,false,true);
441 }
442 $body = magicHTML($body, $id, $message, $mailbox);
443 }
444 } else {
445 translateText($body, $wrap_at,
446 $body_message->header->getParameter('charset'));
447 }
448
449 // if this is the clean display (i.e. printer friendly), stop here.
450 if ( $clean ) {
451 return $body;
452 }
453
454 /*
455 * Previously the links for downloading and unsafe images were printed
456 * under the mail. By putting the links in a global variable we can
457 * print it in the toolbar where it belongs. Since the original code was
458 * in this place it's left here. It might be possible to move it to some
459 * other place if that makes sense. The possibility to do so has not
460 * been evaluated yet.
461 */
462
463 // Initialize the global variable to an empty string.
464 // FIXME: To have $download_and_unsafe_link as a global variable might not be needed since the use of separate variables ($download_href, $unsafe_image_toggle_href, and $unsafe_image_toggle_text) for the templates was introduced.
465 $download_and_unsafe_link = '';
466
467 // Prepare and build a link for downloading the mail.
468 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
469 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
470 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
471 if (isset($passed_ent_id)) {
472 $link .= '&amp;passed_ent_id='.$passed_ent_id;
473 }
474 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
475
476 // Always add the link for downloading the mail as a file to the global
477 // variable.
478 $download_and_unsafe_link .= "$nbsp|$nbsp"
479 . create_hyperlink($download_href, _("Download this as a file"));
480
481 // Find out the right text to use in the link depending on the
482 // circumstances. If the unsafe images are displayed the link should
483 // hide them, if they aren't displayed the link should only appear if
484 // the mail really contains unsafe images.
485 if ($view_unsafe_images) {
486 $text = _("Hide Unsafe Images");
487 } else {
488 if (isset($has_unsafe_images) && $has_unsafe_images) {
489 $link .= '&amp;view_unsafe_images=1';
490 $text = _("View Unsafe Images");
491 } else {
492 $text = '';
493 }
494 }
495
496 // Only create a link for unsafe images if there's need for one. If so:
497 // add it to the global variable.
498 if($text != '') {
499 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
500 $unsafe_image_toggle_text = $text;
501 $download_and_unsafe_link .= "$nbsp|$nbsp"
502 . create_hyperlink($unsafe_image_toggle_href, $text);
503 }
504 }
505 return $body;
506 }
507
508 /**
509 * Generate attachments array for passing to templates. Separated from
510 * formatAttachments() below so that the same array can be given to the
511 * print-friendly version.
512 *
513 * @since 1.5.2
514 * @param object $message SquirrelMail message object
515 * @param array $exclude_id message parts that are not attachments.
516 * @param string $mailbox mailbox name
517 * @param integer $id message id
518 */
519 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
520 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
521
522 $att_ar = $message->getAttachments($exclude_id);
523 $urlMailbox = urlencode($mailbox);
524
525 $attachments = array();
526 foreach ($att_ar as $att) {
527 $ent = $att->entity_id;
528 $header = $att->header;
529 $type0 = strtolower($header->type0);
530 $type1 = strtolower($header->type1);
531 $name = '';
532 $links = array();
533 $links['download link']['text'] = _("Download");
534 $links['download link']['href'] = $base_uri .
535 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
536
537 if ($type0 =='message' && $type1 == 'rfc822') {
538 $default_page = $base_uri . 'src/read_body.php';
539 $rfc822_header = $att->rfc822_header;
540 $filename = $rfc822_header->subject;
541 if (trim( $filename ) == '') {
542 $filename = 'untitled-[' . $ent . ']' ;
543 }
544 $from_o = $rfc822_header->from;
545 if (is_object($from_o)) {
546 $from_name = decodeHeader($from_o->getAddress(false));
547 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
548 // something weird happens when a digest message is opened and you return to the digest
549 // now the from object is part of an array. Probably the parseHeader call overwrites the info
550 // retrieved from the bodystructure in a different way. We need to fix this later.
551 // possible starting point, do not fetch header we already have and inspect how
552 // the rfc822_header object behaves.
553 $from_name = decodeHeader($from_o[0]->getAddress(false));
554 } else {
555 $from_name = _("Unknown sender");
556 }
557 $description = _("From").': '.$from_name;
558 } else {
559 $default_page = $base_uri . 'src/download.php';
560 $filename = $att->getFilename();
561 if ($header->description) {
562 $description = decodeHeader($header->description);
563 } else {
564 $description = '';
565 }
566 }
567
568 $display_filename = $filename;
569 if (isset($passed_ent_id)) {
570 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
571 } else {
572 $passed_ent_id_link = '';
573 }
574 $defaultlink = $default_page . "?startMessage=$startMessage"
575 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
576 . '&amp;ent_id='.$ent.$passed_ent_id_link;
577 if ($where && $what) {
578 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
579 }
580 // IE does make use of mime content sniffing. Forcing a download
581 // prohibit execution of XSS inside an application/octet-stream attachment
582 if ($type0 == 'application' && $type1 == 'octet-stream') {
583 $defaultlink .= '&amp;absolute_dl=true';
584 }
585
586 /* This executes the attachment hook with a specific MIME-type.
587 * If that doesn't have results, it tries if there's a rule
588 * for a more generic type. Finally, a hook for ALL attachment
589 * types is run as well.
590 */
591 // First remember the default link.
592 $defaultlink_orig = $defaultlink;
593
594 /* The API for this hook has changed as of 1.5.2 so that all plugin
595 arguments are passed in an array instead of each their own plugin
596 argument, and arguments are passed by reference, so instead of
597 returning any changes, changes should simply be made to the original
598 arguments themselves. */
599 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
600 &$defaultlink, &$display_filename, &$where, &$what);
601 do_hook("attachment $type0/$type1", $temp);
602 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
603 /* The API for this hook has changed as of 1.5.2 so that all plugin
604 arguments are passed in an array instead of each their own plugin
605 argument, and arguments are passed by reference, so instead of
606 returning any changes, changes should simply be made to the original
607 arguments themselves. */
608 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
609 &$defaultlink, &$display_filename, &$where, &$what);
610 do_hook("attachment $type0/*", $temp);
611 }
612 /* The API for this hook has changed as of 1.5.2 so that all plugin
613 arguments are passed in an array instead of each their own plugin
614 argument, and arguments are passed by reference, so instead of
615 returning any changes, changes should simply be made to the original
616 arguments themselves. */
617 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
618 &$defaultlink, &$display_filename, &$where, &$what);
619 // Do not let a generic plugin change the default link if a more
620 // specialized one already did it...
621 if ($defaultlink != $defaultlink_orig) {
622 $dummy = '';
623 $temp[5] = &$dummy;
624 }
625 do_hook("attachment */*", $temp);
626
627 $this_attachment = array();
628 $this_attachment['Name'] = decodeHeader($display_filename);
629 $this_attachment['Description'] = $description;
630 $this_attachment['DefaultHREF'] = $defaultlink;
631 $this_attachment['DownloadHREF'] = $links['download link']['href'];
632 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
633 $this_attachment['Size'] = $header->size;
634 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
635 $this_attachment['OtherLinks'] = array();
636 foreach ($links as $val) {
637 if ($val['text']==_("Download") || $val['text'] == _("View"))
638 continue;
639 if (empty($val['text']) && empty($val['extra']))
640 continue;
641
642 $temp = array();
643 $temp['HREF'] = $val['href'];
644 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
645 $this_attachment['OtherLinks'][] = $temp;
646 }
647 $attachments[] = $this_attachment;
648
649 unset($links);
650 }
651
652 return $attachments;
653 }
654
655 /**
656 * Displays attachment links and information
657 *
658 * Since 1.3.0 function is not included in formatBody() call.
659 *
660 * Since 1.0.2 uses attachment $type0/$type1 hook.
661 * Since 1.2.5 uses attachment $type0/* hook.
662 * Since 1.5.0 uses attachments_bottom hook.
663 * Since 1.5.2 uses templates and does *not* return a value.
664 *
665 * @param object $message SquirrelMail message object
666 * @param array $exclude_id message parts that are not attachments.
667 * @param string $mailbox mailbox name
668 * @param integer $id message id
669 */
670 function formatAttachments($message, $exclude_id, $mailbox, $id) {
671 global $oTemplate;
672
673 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
674
675 $oTemplate->assign('attachments', $attach);
676 $oTemplate->display('read_attachments.tpl');
677 }
678
679 function sqimap_base64_decode(&$string) {
680
681 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
682 // fly decoding (to reduce memory usage) you have to check if the
683 // data has incomplete pairs
684
685 // Remove the noise in order to check if the 4 bytes pairs are complete
686 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
687
688 $sStringRem = '';
689 $iMod = strlen($string) % 4;
690 if ($iMod) {
691 $sStringRem = substr($string,-$iMod);
692 // Check if $sStringRem contains padding characters
693 if (substr($sStringRem,-1) != '=') {
694 $string = substr($string,0,-$iMod);
695 } else {
696 $sStringRem = '';
697 }
698 }
699 $string = base64_decode($string);
700 return $sStringRem;
701 }
702
703 /**
704 * Decodes encoded message body
705 *
706 * This function decodes the body depending on the encoding type.
707 * Currently quoted-printable and base64 encodings are supported.
708 * decode_body hook was added to this function in 1.4.2/1.5.0
709 * @param string $body encoded message body
710 * @param string $encoding used encoding
711 * @return string decoded string
712 * @since 1.0
713 */
714 function decodeBody($body, $encoding) {
715
716 $body = str_replace("\r\n", "\n", $body);
717 $encoding = strtolower($encoding);
718
719 $encoding_handler = do_hook('decode_body', $encoding);
720
721
722 // plugins get first shot at decoding the body
723 //
724 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
725 $body = $encoding_handler('decode', $body);
726
727 } elseif ($encoding == 'quoted-printable' ||
728 $encoding == 'quoted_printable') {
729 /**
730 * quoted_printable_decode() function is broken in older
731 * php versions. Text with \r\n decoding was fixed only
732 * in php 4.3.0. Minimal code requirement 4.0.4 +
733 * str_replace("\r\n", "\n", $body); call.
734 */
735 $body = quoted_printable_decode($body);
736 } elseif ($encoding == 'base64') {
737 $body = base64_decode($body);
738 }
739
740 // All other encodings are returned raw.
741 return $body;
742 }
743
744 /**
745 * Decodes headers
746 *
747 * This function decodes strings that are encoded according to
748 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
749 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
750 *
751 * @param string $string header string that has to be made readable
752 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
753 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
754 * @param boolean $decide decide if string can be utfencoded. defaults to false
755 * @return string decoded header string
756 */
757 function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
758 global $languages, $squirrelmail_language,$default_charset;
759 if (is_array($string)) {
760 $string = implode("\n", $string);
761 }
762
763 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
764 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
765 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
766 // Do we need to return at this point?
767 // return $string;
768 }
769 $i = 0;
770 $iLastMatch = -2;
771 $encoded = true;
772
773 $aString = explode(' ',$string);
774 $ret = '';
775 foreach ($aString as $chunk) {
776 if ($encoded && $chunk === '') {
777 continue;
778 } elseif ($chunk === '') {
779 $ret .= ' ';
780 continue;
781 }
782 $encoded = false;
783 /* if encoded words are not separated by a linear-space-white we still catch them */
784 $j = $i-1;
785
786 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
787 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
788 if ($iLastMatch !== $j) {
789 if ($htmlsafe) {
790 $ret .= '&#32;';
791 } else {
792 $ret .= ' ';
793 }
794 }
795 $iLastMatch = $i;
796 $j = $i;
797 if ($htmlsafe) {
798 $ret .= htmlspecialchars($res[1]);
799 } else {
800 $ret .= $res[1];
801 }
802 $encoding = ucfirst($res[3]);
803
804 /* decide about valid decoding */
805 if ($decide && is_conversion_safe($res[2])) {
806 $utfencode=true;
807 $can_be_encoded=true;
808 } else {
809 $can_be_encoded=false;
810 }
811 switch ($encoding)
812 {
813 case 'B':
814 $replace = base64_decode($res[4]);
815 if ($utfencode) {
816 if ($can_be_encoded) {
817 /* convert string to different charset,
818 * if functions asks for it (usually in compose)
819 */
820 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
821 } else {
822 // convert string to html codes in order to display it
823 $ret .= charset_decode($res[2],$replace);
824 }
825 } else {
826 if ($htmlsafe) {
827 $replace = htmlspecialchars($replace);
828 }
829 $ret.= $replace;
830 }
831 break;
832 case 'Q':
833 $replace = str_replace('_', ' ', $res[4]);
834 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
835 $replace);
836 if ($utfencode) {
837 if ($can_be_encoded) {
838 /* convert string to different charset,
839 * if functions asks for it (usually in compose)
840 */
841 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
842 } else {
843 // convert string to html codes in order to display it
844 $replace = charset_decode($res[2], $replace);
845 }
846 } else {
847 if ($htmlsafe) {
848 $replace = htmlspecialchars($replace);
849 }
850 }
851 $ret .= $replace;
852 break;
853 default:
854 break;
855 }
856 $chunk = $res[5];
857 $encoded = true;
858 }
859 if (!$encoded) {
860 if ($htmlsafe) {
861 $ret .= '&#32;';
862 } else {
863 $ret .= ' ';
864 }
865 }
866
867 if (!$encoded && $htmlsafe) {
868 $ret .= htmlspecialchars($chunk);
869 } else {
870 $ret .= $chunk;
871 }
872 ++$i;
873 }
874 /* remove the first added space */
875 if ($ret) {
876 if ($htmlsafe) {
877 $ret = substr($ret,5);
878 } else {
879 $ret = substr($ret,1);
880 }
881 }
882
883 return $ret;
884 }
885
886 /**
887 * Encodes header
888 *
889 * Function uses XTRA_CODE _encodeheader function, if such function exists.
890 *
891 * Function uses Q encoding by default and encodes a string according to RFC
892 * 1522 for use in headers if it contains 8-bit characters or anything that
893 * looks like it should be encoded.
894 *
895 * Function switches to B encoding and encodeHeaderBase64() function, if
896 * string is 8bit and multibyte character set supported by mbstring extension
897 * is used. It can cause E_USER_NOTICE errors, if interface is used with
898 * multibyte character set unsupported by mbstring extension.
899 *
900 * @param string $string header string, that has to be encoded
901 * @return string quoted-printable encoded string
902 * @todo make $mb_charsets system wide constant
903 */
904 function encodeHeader ($string) {
905 global $default_charset, $languages, $squirrelmail_language;
906
907 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
908 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
909 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
910 }
911
912 // Use B encoding for multibyte charsets
913 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
914 if (in_array($default_charset,$mb_charsets) &&
915 in_array($default_charset,sq_mb_list_encodings()) &&
916 sq_is8bit($string)) {
917 return encodeHeaderBase64($string,$default_charset);
918 } elseif (in_array($default_charset,$mb_charsets) &&
919 sq_is8bit($string) &&
920 ! in_array($default_charset,sq_mb_list_encodings())) {
921 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
922 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
923 }
924
925 // Encode only if the string contains 8-bit characters or =?
926 $j = strlen($string);
927 $max_l = 75 - strlen($default_charset) - 7;
928 $aRet = array();
929 $ret = '';
930 $iEncStart = $enc_init = false;
931 $cur_l = $iOffset = 0;
932 for($i = 0; $i < $j; ++$i) {
933 switch($string{$i})
934 {
935 case '=':
936 case '<':
937 case '>':
938 case ',':
939 case '?':
940 case '_':
941 if ($iEncStart === false) {
942 $iEncStart = $i;
943 }
944 $cur_l+=3;
945 if ($cur_l > ($max_l-2)) {
946 /* if there is an stringpart that doesn't need encoding, add it */
947 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
948 $aRet[] = "=?$default_charset?Q?$ret?=";
949 $iOffset = $i;
950 $cur_l = 0;
951 $ret = '';
952 $iEncStart = false;
953 } else {
954 $ret .= sprintf("=%02X",ord($string{$i}));
955 }
956 break;
957 case '(':
958 case ')':
959 if ($iEncStart !== false) {
960 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
961 $aRet[] = "=?$default_charset?Q?$ret?=";
962 $iOffset = $i;
963 $cur_l = 0;
964 $ret = '';
965 $iEncStart = false;
966 }
967 break;
968 case ' ':
969 if ($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 $iOffset = $i;
975 $cur_l = 0;
976 $ret = '';
977 $iEncStart = false;
978 } else {
979 $ret .= '_';
980 }
981 }
982 break;
983 default:
984 $k = ord($string{$i});
985 if ($k > 126) {
986 if ($iEncStart === false) {
987 // do not start encoding in the middle of a string, also take the rest of the word.
988 $sLeadString = substr($string,0,$i);
989 $aLeadString = explode(' ',$sLeadString);
990 $sToBeEncoded = array_pop($aLeadString);
991 $iEncStart = $i - strlen($sToBeEncoded);
992 $ret .= $sToBeEncoded;
993 $cur_l += strlen($sToBeEncoded);
994 }
995 $cur_l += 3;
996 /* first we add the encoded string that reached it's max size */
997 if ($cur_l > ($max_l-2)) {
998 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
999 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1000 $cur_l = 3;
1001 $ret = '';
1002 $iOffset = $i;
1003 $iEncStart = $i;
1004 }
1005 $enc_init = true;
1006 $ret .= sprintf("=%02X", $k);
1007 } else {
1008 if ($iEncStart !== false) {
1009 $cur_l++;
1010 if ($cur_l > $max_l) {
1011 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1012 $aRet[] = "=?$default_charset?Q?$ret?=";
1013 $iEncStart = false;
1014 $iOffset = $i;
1015 $cur_l = 0;
1016 $ret = '';
1017 } else {
1018 $ret .= $string{$i};
1019 }
1020 }
1021 }
1022 break;
1023 }
1024 }
1025
1026 if ($enc_init) {
1027 if ($iEncStart !== false) {
1028 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1029 $aRet[] = "=?$default_charset?Q?$ret?=";
1030 } else {
1031 $aRet[] = substr($string,$iOffset);
1032 }
1033 $string = implode('',$aRet);
1034 }
1035 return $string;
1036 }
1037
1038 /**
1039 * Encodes string according to rfc2047 B encoding header formating rules
1040 *
1041 * It is recommended way to encode headers with character sets that store
1042 * symbols in more than one byte.
1043 *
1044 * Function requires mbstring support. If required mbstring functions are missing,
1045 * function returns false and sets E_USER_WARNING level error message.
1046 *
1047 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1048 * that mbstring functions will generate E_WARNING errors, if unsupported
1049 * character set is used. mb_encode_mimeheader function provided by php
1050 * mbstring extension is not used in order to get better control of header
1051 * encoding.
1052 *
1053 * Used php code functions - function_exists(), trigger_error(), strlen()
1054 * (is used with charset names and base64 strings). Used php mbstring
1055 * functions - mb_strlen and mb_substr.
1056 *
1057 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1058 * encoding), rfc 2822 (header folding)
1059 *
1060 * @param string $string header string that must be encoded
1061 * @param string $charset character set. Must be supported by mbstring extension.
1062 * Use sq_mb_list_encodings() to detect supported charsets.
1063 * @return string string encoded according to rfc2047 B encoding formating rules
1064 * @since 1.5.1
1065 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1066 * @todo Do we want to control max length of header?
1067 * @todo Do we want to control EOL (end-of-line) marker?
1068 * @todo Do we want to translate error message?
1069 */
1070 function encodeHeaderBase64($string,$charset) {
1071 /**
1072 * Check mbstring function requirements.
1073 */
1074 if (! function_exists('mb_strlen') ||
1075 ! function_exists('mb_substr')) {
1076 // set E_USER_WARNING
1077 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1078 // return false
1079 return false;
1080 }
1081
1082 // initial return array
1083 $aRet = array();
1084
1085 /**
1086 * header length = 75 symbols max (same as in encodeHeader)
1087 * remove $charset length
1088 * remove =? ? ?= (5 chars)
1089 * remove 2 more chars (\r\n ?)
1090 */
1091 $iMaxLength = 75 - strlen($charset) - 7;
1092
1093 // set first character position
1094 $iStartCharNum = 0;
1095
1096 // loop through all characters. count characters and not bytes.
1097 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1098 // encode string from starting character to current character.
1099 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1100
1101 // Check encoded string length
1102 if(strlen($encoded_string)>$iMaxLength) {
1103 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1104 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1105
1106 // set new starting character
1107 $iStartCharNum = $iCharNum-1;
1108
1109 // encode last char (in case it is last character in string)
1110 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1111 } // if string is shorter than max length - add next character
1112 }
1113
1114 // add last encoded string to array
1115 $aRet[] = $encoded_string;
1116
1117 // set initial return string
1118 $sRet = '';
1119
1120 // loop through encoded strings
1121 foreach($aRet as $string) {
1122 // TODO: Do we want to control EOL (end-of-line) marker
1123 if ($sRet!='') $sRet.= " ";
1124
1125 // add header tags and encoded string to return string
1126 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1127 }
1128
1129 return $sRet;
1130 }
1131
1132 /* This function trys to locate the entity_id of a specific mime element */
1133 function find_ent_id($id, $message) {
1134 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1135 if ($message->entities[$i]->header->type0 == 'multipart') {
1136 $ret = find_ent_id($id, $message->entities[$i]);
1137 } else {
1138 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1139 // if (sq_check_save_extension($message->entities[$i])) {
1140 return $message->entities[$i]->entity_id;
1141 // }
1142 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1143 /**
1144 * This is part of a fix for Outlook Express 6.x generating
1145 * cid URLs without creating content-id headers
1146 * @@JA - 20050207
1147 */
1148 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1149 return $message->entities[$i]->entity_id;
1150 }
1151 }
1152 }
1153 }
1154 return $ret;
1155 }
1156
1157 function sq_check_save_extension($message) {
1158 $filename = $message->getFilename();
1159 $ext = substr($filename, strrpos($filename,'.')+1);
1160 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1161 return in_array($ext, $save_extensions);
1162 }
1163
1164
1165 /**
1166 ** HTMLFILTER ROUTINES
1167 */
1168
1169 /**
1170 * This function checks attribute values for entity-encoded values
1171 * and returns them translated into 8-bit strings so we can run
1172 * checks on them.
1173 *
1174 * @param $attvalue A string to run entity check against.
1175 * @return Nothing, modifies a reference value.
1176 */
1177 function sq_defang(&$attvalue){
1178 $me = 'sq_defang';
1179 /**
1180 * Skip this if there aren't ampersands or backslashes.
1181 */
1182 if (strpos($attvalue, '&') === false
1183 && strpos($attvalue, '\\') === false){
1184 return;
1185 }
1186 $m = false;
1187 // before deent, translate the dangerous unicode characters and ... to safe values
1188 // otherwise the regular expressions do not match.
1189
1190
1191
1192 do {
1193 $m = false;
1194 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1195 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1196 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1197 } while ($m == true);
1198 $attvalue = stripslashes($attvalue);
1199 }
1200
1201 /**
1202 * Kill any tabs, newlines, or carriage returns. Our friends the
1203 * makers of the browser with 95% market value decided that it'd
1204 * be funny to make "java[tab]script" be just as good as "javascript".
1205 *
1206 * @param attvalue The attribute value before extraneous spaces removed.
1207 * @return attvalue Nothing, modifies a reference value.
1208 */
1209 function sq_unspace(&$attvalue){
1210 $me = 'sq_unspace';
1211 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1212 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1213 Array('', '', '', '', ''), $attvalue);
1214 }
1215 }
1216
1217 /**
1218 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1219 * IE as regular characters.
1220 *
1221 * @param attvalue The attribute value before dangerous characters are translated.
1222 * @return attvalue Nothing, modifies a reference value.
1223 * @author Marc Groot Koerkamp.
1224 */
1225 function sq_fixIE_idiocy(&$attvalue) {
1226 // remove NUL
1227 $attvalue = str_replace("\0", "", $attvalue);
1228 // remove comments
1229 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1230
1231 // IE has the evil habit of accepting every possible value for the attribute expression.
1232 // The table below contains characters which are parsed by IE if they are used in the "expression"
1233 // attribute value.
1234 $aDangerousCharsReplacementTable = array(
1235 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1236 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1237 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
1238 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1239 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1240 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1241 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1242 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1243 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1244 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1245 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1246 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1247 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1248 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1249 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1250 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1251 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1252 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1253 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1254 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1255 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1256 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1257 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1258 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1259 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1260 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1261 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1262 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1263 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1264 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1265 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1266 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1267 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1268 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1269 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1270 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1271 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1272 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1273 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1274 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1275 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1276 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1277 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1278 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1279 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1280 "\xCA\x9F", /* L UNICODE IPA Extension */
1281 "\xCA\x80", /* R UNICODE IPA Extension */
1282 "\xC9\xB4"), /* N UNICODE IPA Extension */
1283 array('l', 'l', 'r','r','n','n',
1284 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1285 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1286 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1287 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1288
1289 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1290 // used for XSS.
1291 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1292 }
1293
1294 /**
1295 * This function returns the final tag out of the tag name, an array
1296 * of attributes, and the type of the tag. This function is called by
1297 * sq_sanitize internally.
1298 *
1299 * @param $tagname the name of the tag.
1300 * @param $attary the array of attributes and their values
1301 * @param $tagtype The type of the tag (see in comments).
1302 * @return a string with the final tag representation.
1303 */
1304 function sq_tagprint($tagname, $attary, $tagtype){
1305 $me = 'sq_tagprint';
1306
1307 if ($tagtype == 2){
1308 $fulltag = '</' . $tagname . '>';
1309 } else {
1310 $fulltag = '<' . $tagname;
1311 if (is_array($attary) && sizeof($attary)){
1312 $atts = Array();
1313 while (list($attname, $attvalue) = each($attary)){
1314 array_push($atts, "$attname=$attvalue");
1315 }
1316 $fulltag .= ' ' . join(" ", $atts);
1317 }
1318 if ($tagtype == 3){
1319 $fulltag .= ' /';
1320 }
1321 $fulltag .= '>';
1322 }
1323 return $fulltag;
1324 }
1325
1326 /**
1327 * A small helper function to use with array_walk. Modifies a by-ref
1328 * value and makes it lowercase.
1329 *
1330 * @param $val a value passed by-ref.
1331 * @return void since it modifies a by-ref value.
1332 */
1333 function sq_casenormalize(&$val){
1334 $val = strtolower($val);
1335 }
1336
1337 /**
1338 * This function skips any whitespace from the current position within
1339 * a string and to the next non-whitespace value.
1340 *
1341 * @param $body the string
1342 * @param $offset the offset within the string where we should start
1343 * looking for the next non-whitespace character.
1344 * @return the location within the $body where the next
1345 * non-whitespace char is located.
1346 */
1347 function sq_skipspace($body, $offset){
1348 $me = 'sq_skipspace';
1349 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1350 if (sizeof($matches{1})){
1351 $count = strlen($matches{1});
1352 $offset += $count;
1353 }
1354 return $offset;
1355 }
1356
1357 /**
1358 * This function looks for the next character within a string. It's
1359 * really just a glorified "strpos", except it catches if failures
1360 * nicely.
1361 *
1362 * @param $body The string to look for needle in.
1363 * @param $offset Start looking from this position.
1364 * @param $needle The character/string to look for.
1365 * @return location of the next occurance of the needle, or
1366 * strlen($body) if needle wasn't found.
1367 */
1368 function sq_findnxstr($body, $offset, $needle){
1369 $me = 'sq_findnxstr';
1370 $pos = strpos($body, $needle, $offset);
1371 if ($pos === FALSE){
1372 $pos = strlen($body);
1373 }
1374 return $pos;
1375 }
1376
1377 /**
1378 * This function takes a PCRE-style regexp and tries to match it
1379 * within the string.
1380 *
1381 * @param $body The string to look for needle in.
1382 * @param $offset Start looking from here.
1383 * @param $reg A PCRE-style regex to match.
1384 * @return Returns a false if no matches found, or an array
1385 * with the following members:
1386 * - integer with the location of the match within $body
1387 * - string with whatever content between offset and the match
1388 * - string with whatever it is we matched
1389 */
1390 function sq_findnxreg($body, $offset, $reg){
1391 $me = 'sq_findnxreg';
1392 $matches = Array();
1393 $retarr = Array();
1394 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1395 if (!isset($matches{0}) || !$matches{0}){
1396 $retarr = false;
1397 } else {
1398 $retarr{0} = $offset + strlen($matches{1});
1399 $retarr{1} = $matches{1};
1400 $retarr{2} = $matches{2};
1401 }
1402 return $retarr;
1403 }
1404
1405 /**
1406 * This function looks for the next tag.
1407 *
1408 * @param $body String where to look for the next tag.
1409 * @param $offset Start looking from here.
1410 * @return false if no more tags exist in the body, or
1411 * an array with the following members:
1412 * - string with the name of the tag
1413 * - array with attributes and their values
1414 * - integer with tag type (1, 2, or 3)
1415 * - integer where the tag starts (starting "<")
1416 * - integer where the tag ends (ending ">")
1417 * first three members will be false, if the tag is invalid.
1418 */
1419 function sq_getnxtag($body, $offset){
1420 $me = 'sq_getnxtag';
1421 if ($offset > strlen($body)){
1422 return false;
1423 }
1424 $lt = sq_findnxstr($body, $offset, "<");
1425 if ($lt == strlen($body)){
1426 return false;
1427 }
1428 /**
1429 * We are here:
1430 * blah blah <tag attribute="value">
1431 * \---------^
1432 */
1433 $pos = sq_skipspace($body, $lt+1);
1434 if ($pos >= strlen($body)){
1435 return Array(false, false, false, $lt, strlen($body));
1436 }
1437 /**
1438 * There are 3 kinds of tags:
1439 * 1. Opening tag, e.g.:
1440 * <a href="blah">
1441 * 2. Closing tag, e.g.:
1442 * </a>
1443 * 3. XHTML-style content-less tag, e.g.:
1444 * <img src="blah" />
1445 */
1446 $tagtype = false;
1447 switch (substr($body, $pos, 1)){
1448 case '/':
1449 $tagtype = 2;
1450 $pos++;
1451 break;
1452 case '!':
1453 /**
1454 * A comment or an SGML declaration.
1455 */
1456 if (substr($body, $pos+1, 2) == "--"){
1457 $gt = strpos($body, "-->", $pos);
1458 if ($gt === false){
1459 $gt = strlen($body);
1460 } else {
1461 $gt += 2;
1462 }
1463 return Array(false, false, false, $lt, $gt);
1464 } else {
1465 $gt = sq_findnxstr($body, $pos, ">");
1466 return Array(false, false, false, $lt, $gt);
1467 }
1468 break;
1469 default:
1470 /**
1471 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1472 * later.
1473 */
1474 $tagtype = 1;
1475 break;
1476 }
1477
1478 $tag_start = $pos;
1479 $tagname = '';
1480 /**
1481 * Look for next [\W-_], which will indicate the end of the tag name.
1482 */
1483 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1484 if ($regary == false){
1485 return Array(false, false, false, $lt, strlen($body));
1486 }
1487 list($pos, $tagname, $match) = $regary;
1488 $tagname = strtolower($tagname);
1489
1490 /**
1491 * $match can be either of these:
1492 * '>' indicating the end of the tag entirely.
1493 * '\s' indicating the end of the tag name.
1494 * '/' indicating that this is type-3 xhtml tag.
1495 *
1496 * Whatever else we find there indicates an invalid tag.
1497 */
1498 switch ($match){
1499 case '/':
1500 /**
1501 * This is an xhtml-style tag with a closing / at the
1502 * end, like so: <img src="blah" />. Check if it's followed
1503 * by the closing bracket. If not, then this tag is invalid
1504 */
1505 if (substr($body, $pos, 2) == "/>"){
1506 $pos++;
1507 $tagtype = 3;
1508 } else {
1509 $gt = sq_findnxstr($body, $pos, ">");
1510 $retary = Array(false, false, false, $lt, $gt);
1511 return $retary;
1512 }
1513 case '>':
1514 return Array($tagname, false, $tagtype, $lt, $pos);
1515 break;
1516 default:
1517 /**
1518 * Check if it's whitespace
1519 */
1520 if (!preg_match('/\s/', $match)){
1521 /**
1522 * This is an invalid tag! Look for the next closing ">".
1523 */
1524 $gt = sq_findnxstr($body, $lt, ">");
1525 return Array(false, false, false, $lt, $gt);
1526 }
1527 break;
1528 }
1529
1530 /**
1531 * At this point we're here:
1532 * <tagname attribute='blah'>
1533 * \-------^
1534 *
1535 * At this point we loop in order to find all attributes.
1536 */
1537 $attname = '';
1538 $atttype = false;
1539 $attary = Array();
1540
1541 while ($pos <= strlen($body)){
1542 $pos = sq_skipspace($body, $pos);
1543 if ($pos == strlen($body)){
1544 /**
1545 * Non-closed tag.
1546 */
1547 return Array(false, false, false, $lt, $pos);
1548 }
1549 /**
1550 * See if we arrived at a ">" or "/>", which means that we reached
1551 * the end of the tag.
1552 */
1553 $matches = Array();
1554 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1555 /**
1556 * Yep. So we did.
1557 */
1558 $pos += strlen($matches{1});
1559 if ($matches{2} == "/>"){
1560 $tagtype = 3;
1561 $pos++;
1562 }
1563 return Array($tagname, $attary, $tagtype, $lt, $pos);
1564 }
1565
1566 /**
1567 * There are several types of attributes, with optional
1568 * [:space:] between members.
1569 * Type 1:
1570 * attrname[:space:]=[:space:]'CDATA'
1571 * Type 2:
1572 * attrname[:space:]=[:space:]"CDATA"
1573 * Type 3:
1574 * attr[:space:]=[:space:]CDATA
1575 * Type 4:
1576 * attrname
1577 *
1578 * We leave types 1 and 2 the same, type 3 we check for
1579 * '"' and convert to "&quot" if needed, then wrap in
1580 * double quotes. Type 4 we convert into:
1581 * attrname="yes".
1582 */
1583 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1584 if ($regary == false){
1585 /**
1586 * Looks like body ended before the end of tag.
1587 */
1588 return Array(false, false, false, $lt, strlen($body));
1589 }
1590 list($pos, $attname, $match) = $regary;
1591 $attname = strtolower($attname);
1592 /**
1593 * We arrived at the end of attribute name. Several things possible
1594 * here:
1595 * '>' means the end of the tag and this is attribute type 4
1596 * '/' if followed by '>' means the same thing as above
1597 * '\s' means a lot of things -- look what it's followed by.
1598 * anything else means the attribute is invalid.
1599 */
1600 switch($match){
1601 case '/':
1602 /**
1603 * This is an xhtml-style tag with a closing / at the
1604 * end, like so: <img src="blah" />. Check if it's followed
1605 * by the closing bracket. If not, then this tag is invalid
1606 */
1607 if (substr($body, $pos, 2) == "/>"){
1608 $pos++;
1609 $tagtype = 3;
1610 } else {
1611 $gt = sq_findnxstr($body, $pos, ">");
1612 $retary = Array(false, false, false, $lt, $gt);
1613 return $retary;
1614 }
1615 case '>':
1616 $attary{$attname} = '"yes"';
1617 return Array($tagname, $attary, $tagtype, $lt, $pos);
1618 break;
1619 default:
1620 /**
1621 * Skip whitespace and see what we arrive at.
1622 */
1623 $pos = sq_skipspace($body, $pos);
1624 $char = substr($body, $pos, 1);
1625 /**
1626 * Two things are valid here:
1627 * '=' means this is attribute type 1 2 or 3.
1628 * \w means this was attribute type 4.
1629 * anything else we ignore and re-loop. End of tag and
1630 * invalid stuff will be caught by our checks at the beginning
1631 * of the loop.
1632 */
1633 if ($char == "="){
1634 $pos++;
1635 $pos = sq_skipspace($body, $pos);
1636 /**
1637 * Here are 3 possibilities:
1638 * "'" attribute type 1
1639 * '"' attribute type 2
1640 * everything else is the content of tag type 3
1641 */
1642 $quot = substr($body, $pos, 1);
1643 if ($quot == "'"){
1644 $regary = sq_findnxreg($body, $pos+1, "\'");
1645 if ($regary == false){
1646 return Array(false, false, false, $lt, strlen($body));
1647 }
1648 list($pos, $attval, $match) = $regary;
1649 $pos++;
1650 $attary{$attname} = "'" . $attval . "'";
1651 } else if ($quot == '"'){
1652 $regary = sq_findnxreg($body, $pos+1, '\"');
1653 if ($regary == false){
1654 return Array(false, false, false, $lt, strlen($body));
1655 }
1656 list($pos, $attval, $match) = $regary;
1657 $pos++;
1658 $attary{$attname} = '"' . $attval . '"';
1659 } else {
1660 /**
1661 * These are hateful. Look for \s, or >.
1662 */
1663 $regary = sq_findnxreg($body, $pos, "[\s>]");
1664 if ($regary == false){
1665 return Array(false, false, false, $lt, strlen($body));
1666 }
1667 list($pos, $attval, $match) = $regary;
1668 /**
1669 * If it's ">" it will be caught at the top.
1670 */
1671 $attval = preg_replace("/\"/s", "&quot;", $attval);
1672 $attary{$attname} = '"' . $attval . '"';
1673 }
1674 } else if (preg_match("|[\w/>]|", $char)) {
1675 /**
1676 * That was attribute type 4.
1677 */
1678 $attary{$attname} = '"yes"';
1679 } else {
1680 /**
1681 * An illegal character. Find next '>' and return.
1682 */
1683 $gt = sq_findnxstr($body, $pos, ">");
1684 return Array(false, false, false, $lt, $gt);
1685 }
1686 break;
1687 }
1688 }
1689 /**
1690 * The fact that we got here indicates that the tag end was never
1691 * found. Return invalid tag indication so it gets stripped.
1692 */
1693 return Array(false, false, false, $lt, strlen($body));
1694 }
1695
1696 /**
1697 * Translates entities into literal values so they can be checked.
1698 *
1699 * @param $attvalue the by-ref value to check.
1700 * @param $regex the regular expression to check against.
1701 * @param $hex whether the entites are hexadecimal.
1702 * @return True or False depending on whether there were matches.
1703 */
1704 function sq_deent(&$attvalue, $regex, $hex=false){
1705 $me = 'sq_deent';
1706 $ret_match = false;
1707 // remove comments
1708 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1709 preg_match_all($regex, $attvalue, $matches);
1710 if (is_array($matches) && sizeof($matches[0]) > 0){
1711 $repl = Array();
1712 for ($i = 0; $i < sizeof($matches[0]); $i++){
1713 $numval = $matches[1][$i];
1714 if ($hex){
1715 $numval = hexdec($numval);
1716 }
1717 $repl{$matches[0][$i]} = chr($numval);
1718 }
1719 $attvalue = strtr($attvalue, $repl);
1720 return true;
1721 } else {
1722 return false;
1723 }
1724 }
1725
1726 /**
1727 * This function runs various checks against the attributes.
1728 *
1729 * @param $tagname String with the name of the tag.
1730 * @param $attary Array with all tag attributes.
1731 * @param $rm_attnames See description for sq_sanitize
1732 * @param $bad_attvals See description for sq_sanitize
1733 * @param $add_attr_to_tag See description for sq_sanitize
1734 * @param $message message object
1735 * @param $id message id
1736 * @return Array with modified attributes.
1737 */
1738 function sq_fixatts($tagname,
1739 $attary,
1740 $rm_attnames,
1741 $bad_attvals,
1742 $add_attr_to_tag,
1743 $message,
1744 $id,
1745 $mailbox
1746 ){
1747 $me = 'sq_fixatts';
1748 while (list($attname, $attvalue) = each($attary)){
1749 /**
1750 * See if this attribute should be removed.
1751 */
1752 foreach ($rm_attnames as $matchtag=>$matchattrs){
1753 if (preg_match($matchtag, $tagname)){
1754 foreach ($matchattrs as $matchattr){
1755 if (preg_match($matchattr, $attname)){
1756 unset($attary{$attname});
1757 continue;
1758 }
1759 }
1760 }
1761 }
1762 /**
1763 * Workaround for IE quirks
1764 */
1765 sq_fixIE_idiocy($attvalue);
1766
1767 /**
1768 * Remove any backslashes, entities, and extraneous whitespace.
1769 */
1770
1771 $oldattvalue = $attvalue;
1772 sq_defang($attvalue);
1773 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1774 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1775 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1776 $attvalue = "idiocy";
1777 $attary{$attname} = $attvalue;
1778 }
1779 sq_unspace($attvalue);
1780
1781 /**
1782 * Now let's run checks on the attvalues.
1783 * I don't expect anyone to comprehend this. If you do,
1784 * get in touch with me so I can drive to where you live and
1785 * shake your hand personally. :)
1786 */
1787 foreach ($bad_attvals as $matchtag=>$matchattrs){
1788 if (preg_match($matchtag, $tagname)){
1789 foreach ($matchattrs as $matchattr=>$valary){
1790 if (preg_match($matchattr, $attname)){
1791 /**
1792 * There are two arrays in valary.
1793 * First is matches.
1794 * Second one is replacements
1795 */
1796 list($valmatch, $valrepl) = $valary;
1797 $newvalue =
1798 preg_replace($valmatch, $valrepl, $attvalue);
1799 if ($newvalue != $attvalue){
1800 $attary{$attname} = $newvalue;
1801 $attvalue = $newvalue;
1802 }
1803 }
1804 }
1805 }
1806 }
1807 if ($attname == 'style') {
1808 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1809 // 8bit and control characters in style attribute values can be used for XSS, remove them
1810 $attary{$attname} = '"disallowed character"';
1811 }
1812 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1813 if (count($aMatch)) {
1814 foreach($aMatch[1] as $sMatch) {
1815 // url value
1816 $urlvalue = $sMatch;
1817 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1818 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1819 }
1820 }
1821 }
1822 /**
1823 * Use white list based filtering on attributes which can contain url's
1824 */
1825 else if ($attname == 'href' || $attname == 'src' || $attname == 'background') {
1826 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1827 $attary{$attname} = $attvalue;
1828 }
1829 }
1830 /**
1831 * See if we need to append any attributes to this tag.
1832 */
1833 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1834 if (preg_match($matchtag, $tagname)){
1835 $attary = array_merge($attary, $addattary);
1836 }
1837 }
1838 return $attary;
1839 }
1840
1841 /**
1842 * This function filters url's
1843 *
1844 * @param $attvalue String with attribute value to filter
1845 * @param $message message object
1846 * @param $id message id
1847 * @param $mailbox mailbox
1848 * @param $sQuote quoting characters around url's
1849 */
1850 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1851 $attvalue = trim($attvalue);
1852 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1853 // remove the double quotes
1854 $sQuote = $attvalue[0];
1855 $attvalue = trim(substr($attvalue,1,-1));
1856 }
1857
1858 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1859 // images off by default.
1860 // FIXME: Update this code to use the default value FALSE.
1861 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1862 $view_unsafe_images = false;
1863 }
1864
1865 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1866
1867 /**
1868 * Replace empty src tags with the blank image. src is only used
1869 * for frames, images, and image inputs. Doing a replace should
1870 * not affect them working as should be, however it will stop
1871 * IE from being kicked off when src for img tags are not set
1872 */
1873 if ($attvalue == '') {
1874 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1875 } else {
1876 // first, disallow 8 bit characters and control characters
1877 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1878 switch ($attname) {
1879 case 'href':
1880 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1881 break;
1882 default:
1883 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1884 break;
1885 }
1886 } else {
1887 $aUrl = parse_url($attvalue);
1888 if (isset($aUrl['scheme'])) {
1889 switch(strtolower($aUrl['scheme'])) {
1890 case 'mailto':
1891 case 'http':
1892 case 'https':
1893 case 'ftp':
1894 if ($attname != 'href') {
1895 if ($view_unsafe_images == false) {
1896 $attvalue = $sQuote . $secremoveimg . $sQuote;
1897 } else {
1898 if (isset($aUrl['path'])) {
1899 // validate image extension.
1900 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1901 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1902 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1903 }
1904 } else {
1905 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1906 }
1907 }
1908 }
1909 break;
1910 case 'outbind':
1911 /**
1912 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1913 * One day MS might actually make it match something useful, for now, falling
1914 * back to using cid2http, so we can grab the blank.png.
1915 */
1916 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1917 break;
1918 case 'cid':
1919 /**
1920 * Turn cid: urls into http-friendly ones.
1921 */
1922 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1923 break;
1924 default:
1925 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
1926 break;
1927 }
1928 } else {
1929 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
1930 // parse_url did not lead to satisfying result
1931 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
1932 }
1933 }
1934 }
1935 }
1936 }
1937
1938 /**
1939 * This function edits the style definition to make them friendly and
1940 * usable in SquirrelMail.
1941 *
1942 * @param $message the message object
1943 * @param $id the message id
1944 * @param $content a string with whatever is between <style> and </style>
1945 * @param $mailbox the message mailbox
1946 * @return a string with edited content.
1947 */
1948 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1949 $me = 'sq_fixstyle';
1950 // workaround for </style> in between comments
1951 $iCurrentPos = $pos;
1952 $content = '';
1953 $sToken = '';
1954 $bSucces = false;
1955 $bEndTag = false;
1956 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
1957 $char = $body{$i};
1958 switch ($char) {
1959 case '<':
1960 $sToken = $char;
1961 break;
1962 case '/':
1963 if ($sToken == '<') {
1964 $sToken .= $char;
1965 $bEndTag = true;
1966 } else {
1967 $content .= $char;
1968 }
1969 break;
1970 case '>':
1971 if ($bEndTag) {
1972 $sToken .= $char;
1973 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
1974 $newpos = $i + 1;
1975 $bSucces = true;
1976 break 2;
1977 } else {
1978 $content .= $sToken;
1979 }
1980 $bEndTag = false;
1981 } else {
1982 $content .= $char;
1983 }
1984 break;
1985 case '!':
1986 if ($sToken == '<') {
1987 // possible comment
1988 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
1989 $i = strpos($body,'-->',$i+3);
1990 if ($i === false) { // no end comment
1991 $i = strlen($body);
1992 }
1993 $sToken = '';
1994 }
1995 } else {
1996 $content .= $char;
1997 }
1998 break;
1999 default:
2000 if ($bEndTag) {
2001 $sToken .= $char;
2002 } else {
2003 $content .= $char;
2004 }
2005 break;
2006 }
2007 }
2008 if ($bSucces == FALSE){
2009 return array(FALSE, strlen($body));
2010 }
2011
2012
2013
2014 /**
2015 * First look for general BODY style declaration, which would be
2016 * like so:
2017 * body {background: blah-blah}
2018 * and change it to .bodyclass so we can just assign it to a <div>
2019 */
2020 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
2021 $secremoveimg = '../images/' . _("sec_remove_eng.png");
2022 /**
2023 * Fix url('blah') declarations.
2024 */
2025 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2026 // "url(\\1$secremoveimg\\2)", $content);
2027
2028 // first check for 8bit sequences and disallowed control characters
2029 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2030 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2031 return array($content, $newpos);
2032 }
2033
2034 // IE Sucks hard. We have a special function for it.
2035 sq_fixIE_idiocy($content);
2036
2037 // remove @import line
2038 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2039
2040 // translate ur\l and variations (IE parses that)
2041 // TODO check if the sq_fixIE_idiocy function already handles this.
2042 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2043 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2044 if (count($aMatch)) {
2045 $aValue = $aReplace = array();
2046 foreach($aMatch[1] as $sMatch) {
2047 // url value
2048 $urlvalue = $sMatch;
2049 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2050 $aValue[] = $sMatch;
2051 $aReplace[] = $urlvalue;
2052 }
2053 $content = str_replace($aValue,$aReplace,$content);
2054 }
2055
2056 /**
2057 * Remove any backslashes, entities, and extraneous whitespace.
2058 */
2059 $contentTemp = $content;
2060 sq_defang($contentTemp);
2061 sq_unspace($contentTemp);
2062
2063 /**
2064 * Fix stupid css declarations which lead to vulnerabilities
2065 * in IE.
2066 */
2067 $match = Array('/\/\*.*\*\//',
2068 '/expression/i',
2069 '/behaviou*r/i',
2070 '/binding/i',
2071 '/include-source/i',
2072 '/javascript/i',
2073 '/script/i');
2074 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
2075 $contentNew = preg_replace($match, $replace, $contentTemp);
2076 if ($contentNew !== $contentTemp) {
2077 // insecure css declarations are used. From now on we don't care
2078 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2079 $content = $contentNew;
2080 }
2081 return array($content, $newpos);
2082 }
2083
2084
2085 /**
2086 * This function converts cid: url's into the ones that can be viewed in
2087 * the browser.
2088 *
2089 * @param $message the message object
2090 * @param $id the message id
2091 * @param $cidurl the cid: url.
2092 * @param $mailbox the message mailbox
2093 * @return a string with a http-friendly url
2094 */
2095 function sq_cid2http($message, $id, $cidurl, $mailbox){
2096 /**
2097 * Get rid of quotes.
2098 */
2099 $quotchar = substr($cidurl, 0, 1);
2100 if ($quotchar == '"' || $quotchar == "'"){
2101 $cidurl = str_replace($quotchar, "", $cidurl);
2102 } else {
2103 $quotchar = '';
2104 }
2105 $cidurl = substr(trim($cidurl), 4);
2106
2107 $match_str = '/\{.*?\}\//';
2108 $str_rep = '';
2109 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2110
2111 $linkurl = find_ent_id($cidurl, $message);
2112 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2113 unsafe link image */
2114 $httpurl = '';
2115
2116 /**
2117 * This is part of a fix for Outlook Express 6.x generating
2118 * cid URLs without creating content-id headers. These images are
2119 * not part of the multipart/related html mail. The html contains
2120 * <img src="cid:{some_id}/image_filename.ext"> references to
2121 * attached images with as goal to render them inline although
2122 * the attachment disposition property is not inline.
2123 */
2124
2125 if (empty($linkurl)) {
2126 if (preg_match('/{.*}\//', $cidurl)) {
2127 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2128 if (!empty($cidurl)) {
2129 $linkurl = find_ent_id($cidurl, $message);
2130 }
2131 }
2132 }
2133
2134 if (!empty($linkurl)) {
2135 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
2136 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2137 '&amp;ent_id=' . $linkurl . $quotchar;
2138 } else {
2139 /**
2140 * If we couldn't generate a proper img url, drop in a blank image
2141 * instead of sending back empty, otherwise it causes unusual behaviour
2142 */
2143 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
2144 }
2145
2146 return $httpurl;
2147 }
2148
2149 /**
2150 * This function changes the <body> tag into a <div> tag since we
2151 * can't really have a body-within-body.
2152 *
2153 * @param $attary an array of attributes and values of <body>
2154 * @param $mailbox mailbox we're currently reading (for cid2http)
2155 * @param $message current message (for cid2http)
2156 * @param $id current message id (for cid2http)
2157 * @return a modified array of attributes to be set for <div>
2158 */
2159 function sq_body2div($attary, $mailbox, $message, $id){
2160 $me = 'sq_body2div';
2161 $divattary = Array('class' => "'bodyclass'");
2162 $text = '#000000';
2163 $has_bgc_stl = $has_txt_stl = false;
2164 $styledef = '';
2165 if (is_array($attary) && sizeof($attary) > 0){
2166 foreach ($attary as $attname=>$attvalue){
2167 $quotchar = substr($attvalue, 0, 1);
2168 $attvalue = str_replace($quotchar, "", $attvalue);
2169 switch ($attname){
2170 case 'background':
2171 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2172 $styledef .= "background-image: url('$attvalue'); ";
2173 break;
2174 case 'bgcolor':
2175 $has_bgc_stl = true;
2176 $styledef .= "background-color: $attvalue; ";
2177 break;
2178 case 'text':
2179 $has_txt_stl = true;
2180 $styledef .= "color: $attvalue; ";
2181 break;
2182 }
2183 }
2184 // Outlook defines a white bgcolor and no text color. This can lead to
2185 // white text on a white bg with certain themes.
2186 if ($has_bgc_stl && !$has_txt_stl) {
2187 $styledef .= "color: $text; ";
2188 }
2189 if (strlen($styledef) > 0){
2190 $divattary{"style"} = "\"$styledef\"";
2191 }
2192 }
2193 return $divattary;
2194 }
2195
2196 /**
2197 * This is the main function and the one you should actually be calling.
2198 * There are several variables you should be aware of an which need
2199 * special description.
2200 *
2201 * Since the description is quite lengthy, see it here:
2202 * http://linux.duke.edu/projects/mini/htmlfilter/
2203 *
2204 * @param $body the string with HTML you wish to filter
2205 * @param $tag_list see description above
2206 * @param $rm_tags_with_content see description above
2207 * @param $self_closing_tags see description above
2208 * @param $force_tag_closing see description above
2209 * @param $rm_attnames see description above
2210 * @param $bad_attvals see description above
2211 * @param $add_attr_to_tag see description above
2212 * @param $message message object
2213 * @param $id message id
2214 * @return sanitized html safe to show on your pages.
2215 */
2216 function sq_sanitize($body,
2217 $tag_list,
2218 $rm_tags_with_content,
2219 $self_closing_tags,
2220 $force_tag_closing,
2221 $rm_attnames,
2222 $bad_attvals,
2223 $add_attr_to_tag,
2224 $message,
2225 $id,
2226 $mailbox
2227 ){
2228 $me = 'sq_sanitize';
2229 $rm_tags = array_shift($tag_list);
2230 /**
2231 * Normalize rm_tags and rm_tags_with_content.
2232 */
2233 @array_walk($tag_list, 'sq_casenormalize');
2234 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2235 @array_walk($self_closing_tags, 'sq_casenormalize');
2236 /**
2237 * See if tag_list is of tags to remove or tags to allow.
2238 * false means remove these tags
2239 * true means allow these tags
2240 */
2241 $curpos = 0;
2242 $open_tags = Array();
2243 $trusted = "\n<!-- begin sanitized html -->\n";
2244 $skip_content = false;
2245 /**
2246 * Take care of netscape's stupid javascript entities like
2247 * &{alert('boo')};
2248 */
2249 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
2250
2251 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2252 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2253 $free_content = substr($body, $curpos, $lt-$curpos);
2254 /**
2255 * Take care of <style>
2256 */
2257 if ($tagname == "style" && $tagtype == 1){
2258 list($free_content, $curpos) =
2259 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
2260 if ($free_content != FALSE){
2261 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2262 $trusted .= $free_content;
2263 $trusted .= sq_tagprint($tagname, false, 2);
2264 }
2265 continue;
2266 }
2267 if ($skip_content == false){
2268 $trusted .= $free_content;
2269 }
2270 if ($tagname != FALSE){
2271 if ($tagtype == 2){
2272 if ($skip_content == $tagname){
2273 /**
2274 * Got to the end of tag we needed to remove.
2275 */
2276 $tagname = false;
2277 $skip_content = false;
2278 } else {
2279 if ($skip_content == false){
2280 if ($tagname == "body"){
2281 $tagname = "div";
2282 }
2283 if (isset($open_tags{$tagname}) &&
2284 $open_tags{$tagname} > 0){
2285 $open_tags{$tagname}--;
2286 } else {
2287 $tagname = false;
2288 }
2289 }
2290 }
2291 } else {
2292 /**
2293 * $rm_tags_with_content
2294 */
2295 if ($skip_content == false){
2296 /**
2297 * See if this is a self-closing type and change
2298 * tagtype appropriately.
2299 */
2300 if ($tagtype == 1
2301 && in_array($tagname, $self_closing_tags)){
2302 $tagtype = 3;
2303 }
2304 /**
2305 * See if we should skip this tag and any content
2306 * inside it.
2307 */
2308 if ($tagtype == 1 &&
2309 in_array($tagname, $rm_tags_with_content)){
2310 $skip_content = $tagname;
2311 } else {
2312 if (($rm_tags == false
2313 && in_array($tagname, $tag_list)) ||
2314 ($rm_tags == true &&
2315 !in_array($tagname, $tag_list))){
2316 $tagname = false;
2317 } else {
2318 /**
2319 * Convert body into div.
2320 */
2321 if ($tagname == "body"){
2322 $tagname = "div";
2323 $attary = sq_body2div($attary, $mailbox,
2324 $message, $id);
2325 }
2326 if ($tagtype == 1){
2327 if (isset($open_tags{$tagname})){
2328 $open_tags{$tagname}++;
2329 } else {
2330 $open_tags{$tagname}=1;
2331 }
2332 }
2333 /**
2334 * This is where we run other checks.
2335 */
2336 if (is_array($attary) && sizeof($attary) > 0){
2337 $attary = sq_fixatts($tagname,
2338 $attary,
2339 $rm_attnames,
2340 $bad_attvals,
2341 $add_attr_to_tag,
2342 $message,
2343 $id,
2344 $mailbox
2345 );
2346 }
2347 }
2348 }
2349 }
2350 }
2351 if ($tagname != false && $skip_content == false){
2352 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2353 }
2354 }
2355 $curpos = $gt+1;
2356 }
2357 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2358 if ($force_tag_closing == true){
2359 foreach ($open_tags as $tagname=>$opentimes){
2360 while ($opentimes > 0){
2361 $trusted .= '</' . $tagname . '>';
2362 $opentimes--;
2363 }
2364 }
2365 $trusted .= "\n";
2366 }
2367 $trusted .= "<!-- end sanitized html -->\n";
2368 return $trusted;
2369 }
2370
2371 /**
2372 * This is a wrapper function to call html sanitizing routines.
2373 *
2374 * @param $body the body of the message
2375 * @param $id the id of the message
2376
2377 * @param $message
2378 * @param $mailbox
2379 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2380 * into internal SM compose links
2381 * (optional; default = TRUE)
2382 * @return a string with html safe to display in the browser.
2383 */
2384 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2385
2386 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2387
2388 global $attachment_common_show_images, $view_unsafe_images,
2389 $has_unsafe_images;
2390 /**
2391 * Don't display attached images in HTML mode.
2392 *
2393 * SB: why?
2394 */
2395 $attachment_common_show_images = false;
2396 $tag_list = Array(
2397 false,
2398 "object",
2399 "meta",
2400 "html",
2401 "head",
2402 "base",
2403 "link",
2404 "frame",
2405 "iframe",
2406 "plaintext",
2407 "marquee"
2408 );
2409
2410 $rm_tags_with_content = Array(
2411 "script",
2412 "applet",
2413 "embed",
2414 "title",
2415 "frameset",
2416 "xmp",
2417 "xml"
2418 );
2419
2420 $self_closing_tags = Array(
2421 "img",
2422 "br",
2423 "hr",
2424 "input",
2425 "outbind"
2426 );
2427
2428 $force_tag_closing = true;
2429
2430 $rm_attnames = Array(
2431 "/.*/" =>
2432 Array(
2433 "/target/i",
2434 "/^on.*/i",
2435 "/^dynsrc/i",
2436 "/^data.*/i",
2437 "/^lowsrc.*/i"
2438 )
2439 );
2440
2441 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2442 $bad_attvals = Array(
2443 "/.*/" =>
2444 Array(
2445 "/^src|background/i" =>
2446 Array(
2447 Array(
2448 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2449 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2450 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2451 ),
2452 Array(
2453 "\\1$secremoveimg\\2",
2454 "\\1$secremoveimg\\2",
2455 "\\1$secremoveimg\\2",
2456 )
2457 ),
2458 "/^href|action/i" =>
2459 Array(
2460 Array(
2461 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2462 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2463 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2464 ),
2465 Array(
2466 "\\1#\\1",
2467 "\\1#\\1",
2468 "\\1#\\1"
2469 )
2470 ),
2471 "/^style/i" =>
2472 Array(
2473 Array(
2474 "/\/\*.*\*\//",
2475 "/expression/i",
2476 "/binding/i",
2477 "/behaviou*r/i",
2478 "/include-source/i",
2479 "/position\s*:\s*absolute/i",
2480 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2481 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2482 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2483 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2484 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2485 ),
2486 Array(
2487 "",
2488 "idiocy",
2489 "idiocy",
2490 "idiocy",
2491 "idiocy",
2492 "idiocy",
2493 "url",
2494 "url(\\1#\\1)",
2495 "url(\\1#\\1)",
2496 "url(\\1#\\1)",
2497 "\\1:url(\\2#\\3)"
2498 )
2499 )
2500 )
2501 );
2502
2503 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2504 // images off by default.
2505 // FIXME: Update this code to use the default value FALSE.
2506 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2507 $view_unsafe_images = false;
2508 }
2509
2510 if (!$view_unsafe_images){
2511 /**
2512 * Remove any references to http/https if view_unsafe_images set
2513 * to false.
2514 */
2515 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2516 '/^([\'\"])\s*https*:.*([\'\"])/si');
2517 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2518 "\\1$secremoveimg\\1");
2519 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2520 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2521 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2522 "url(\\1$secremoveimg\\1)");
2523 }
2524
2525 $add_attr_to_tag = Array(
2526 "/^a$/i" =>
2527 Array('target'=>'"_blank"',
2528 'title'=>'"'._("This external link will open in a new window").'"'
2529 )
2530 );
2531 $trusted = sq_sanitize($body,
2532 $tag_list,
2533 $rm_tags_with_content,
2534 $self_closing_tags,
2535 $force_tag_closing,
2536 $rm_attnames,
2537 $bad_attvals,
2538 $add_attr_to_tag,
2539 $message,
2540 $id,
2541 $mailbox
2542 );
2543 if (strpos($trusted,$secremoveimg)){
2544 $has_unsafe_images = true;
2545 }
2546
2547 // we want to parse mailto's in HTML output, change to SM compose links
2548 // this is a modified version of code from url_parser.php... but Marc is
2549 // right: we need a better filtering implementation; adding this randomly
2550 // here is not a great solution
2551 //
2552 if ($take_mailto_links) {
2553 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2554 global $MailTo_PReg_Match;
2555 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2556 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2557 foreach ($regs[0] as $i => $mailto_before) {
2558 $mailto_params = $regs[10][$i];
2559 // get rid of any tailing quote since we have to add send_to to the end
2560 //
2561 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2562 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2563 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2564 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2565
2566 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2567 $to = 'to=' . $regs[1][$i];
2568 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2569 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2570 else {
2571 if ($mailto_params) //already some params, append to them
2572 $mailto_params .= '&amp;' . $to;
2573 else
2574 $mailto_params .= '?' . $to;
2575 }
2576 }
2577
2578 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2579
2580 // we'll already have target=_blank, no need to allow comp_in_new
2581 // here (which would be a lot more work anyway)
2582 //
2583 global $compose_new_win;
2584 $temp_comp_in_new = $compose_new_win;
2585 $compose_new_win = 0;
2586 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2587 $compose_new_win = $temp_comp_in_new;
2588
2589 // remove <a href=" and anything after the next quote (we only
2590 // need the uri, not the link HTML) in compose uri
2591 //
2592 $comp_uri = substr($comp_uri, 9);
2593 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2594 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2595 }
2596 }
2597 }
2598
2599 return $trusted;
2600 }
2601
2602 /**
2603 * function SendDownloadHeaders - send file to the browser
2604 *
2605 * Original Source: SM core src/download.php
2606 * moved here to make it available to other code, and separate
2607 * front end from back end functionality.
2608 *
2609 * @param string $type0 first half of mime type
2610 * @param string $type1 second half of mime type
2611 * @param string $filename filename to tell the browser for downloaded file
2612 * @param boolean $force whether to force the download dialog to pop
2613 * @param optional integer $filesize send the Content-Header and length to the browser
2614 * @return void
2615 */
2616 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2617 global $languages, $squirrelmail_language;
2618 $isIE = $isIE6plus = false;
2619
2620 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2621
2622 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2623 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2624 $isIE = true;
2625 }
2626
2627 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2628 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2629 $isIE6plus = true;
2630 }
2631
2632 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2633 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2634 $filename =
2635 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2636 } else {
2637 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2638 }
2639
2640 // A Pox on Microsoft and it's Internet Explorer!
2641 //
2642 // IE has lots of bugs with file downloads.
2643 // It also has problems with SSL. Both of these cause problems
2644 // for us in this function.
2645 //
2646 // See this article on Cache Control headers and SSL
2647 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2648 //
2649 // The best thing you can do for IE is to upgrade to the latest
2650 // version
2651 //set all the Cache Control Headers for IE
2652 if ($isIE) {
2653 $filename=rawurlencode($filename);
2654 header ("Pragma: public");
2655 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2656 header ("Cache-Control: post-check=0, pre-check=0", false);
2657 header ("Cache-Control: private");
2658
2659 //set the inline header for IE, we'll add the attachment header later if we need it
2660 header ("Content-Disposition: inline; filename=$filename");
2661 }
2662
2663 if (!$force) {
2664 // Try to show in browser window
2665 header ("Content-Disposition: inline; filename=\"$filename\"");
2666 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2667 } else {
2668 // Try to pop up the "save as" box
2669
2670 // IE makes this hard. It pops up 2 save boxes, or none.
2671 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2672 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2673 // But, according to Microsoft, it is "RFC compliant but doesn't
2674 // take into account some deviations that allowed within the
2675 // specification." Doesn't that mean RFC non-compliant?
2676 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2677
2678 // all browsers need the application/octet-stream header for this
2679 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2680
2681 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2682 // Do not have quotes around filename, but that applied to
2683 // "attachment"... does it apply to inline too?
2684 header ("Content-Disposition: attachment; filename=\"$filename\"");
2685
2686 if ($isIE && !$isIE6plus) {
2687 // This combination seems to work mostly. IE 5.5 SP 1 has
2688 // known issues (see the Microsoft Knowledge Base)
2689
2690 // This works for most types, but doesn't work with Word files
2691 header ("Content-Type: application/download; name=\"$filename\"");
2692 header ("Content-Type: application/force-download; name=\"$filename\"");
2693 // These are spares, just in case. :-)
2694 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2695 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2696 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2697 } else if ($isIE) {
2698 // This is to prevent IE for MIME sniffing and auto open a file in IE
2699 header ("Content-Type: application/force-download; name=\"$filename\"");
2700 } else {
2701 // another application/octet-stream forces download for Netscape
2702 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2703 }
2704 }
2705
2706 //send the content-length header if the calling function provides it
2707 if ($filesize > 0) {
2708 header("Content-Length: $filesize");
2709 }
2710
2711 } // end fn SendDownloadHeaders