6 * Copyright (c) 1999-2003 The SquirrelMail Project Team
7 * Licensed under the GNU GPL. For full terms see the file COPYING.
9 * This contains the functions necessary to detect and decode MIME
15 require_once(SM_PATH
. 'functions/imap.php');
16 require_once(SM_PATH
. 'functions/attachment_common.php');
18 /* -------------------------------------------------------------------------- */
20 /* -------------------------------------------------------------------------- */
22 /* This function gets the structure of a message and stores it in the "message" class.
23 * It will return this object for use with all relevant header information and
24 * fully parsed into the standard "message" object format.
27 function mime_structure ($bodystructure, $flags=array()) {
29 /* Isolate the body structure and remove beginning and end parenthesis. */
30 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') +
13));
31 $read = trim(substr ($read, 0, -1));
33 $msg = Message
::parseStructure($read,$i);
34 if (!is_object($msg)) {
35 include_once(SM_PATH
. 'functions/display_messages.php');
36 global $color, $mailbox;
37 /* removed urldecode because $_GET is auto urldecoded ??? */
38 displayPageHeader( $color, $mailbox );
39 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
40 $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
41 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
42 plain_error_message( $errormessage, $color );
43 echo '</body></html>';
47 foreach ($flags as $flag) {
48 $char = strtoupper($flag{1});
51 if (strtolower($flag) == '\\seen') {
56 if (strtolower($flag) == '\\answered') {
57 $msg->is_answered
= true;
61 if (strtolower($flag) == '\\deleted') {
62 $msg->is_deleted
= true;
66 if (strtolower($flag) == '\\flagged') {
67 $msg->is_flagged
= true;
71 if (strtolower($flag) == '$mdnsent') {
72 $msg->is_mdnsent
= true;
80 // listEntities($msg);
86 /* This starts the parsing of a particular structure. It is called recursively,
87 * so it can be passed different structures. It returns an object of type
89 * First, it checks to see if it is a multipart message. If it is, then it
90 * handles that as it sees is necessary. If it is just a regular entity,
91 * then it parses it and adds the necessary header information (by calling out
92 * to mime_get_elements()
95 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
97 /* Do a bit of error correction. If we couldn't find the entity id, just guess
98 * that it is the first one. That is usually the case anyway.
102 $cmd = "FETCH $id BODY[]";
104 $cmd = "FETCH $id BODY[$ent_id]";
107 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
109 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
111 $topline = trim(array_shift($data));
112 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
114 $wholemessage = implode('', $data);
115 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
116 $ret = substr($wholemessage, 0, $regs[1]);
117 /* There is some information in the content info header that could be important
118 * in order to parse html messages. Let's get them here.
120 // if ($ret{0} == '<') {
121 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
123 } else if (ereg('"([^"]*)"', $topline, $regs)) {
126 global $where, $what, $mailbox, $passed_id, $startMessage;
127 $par = 'mailbox=' . urlencode($mailbox) . '&passed_id=' . $passed_id;
128 if (isset($where) && isset($what)) {
129 $par .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
131 $par .= '&startMessage=' . $startMessage . '&show_more=0';
133 $par .= '&response=' . urlencode($response) .
134 '&message=' . urlencode($message) .
135 '&topline=' . urlencode($topline);
138 '<table width="80%"><tr>' .
139 '<tr><td colspan=2>' .
140 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
142 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
143 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
144 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
145 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
146 "</table><BR></tt></font><hr>";
148 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
150 $wholemessage = implode('', $data);
152 $ret = $wholemessage;
157 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
160 /* Don't kill the connection if the browser is over a dialup
161 * and it would take over 30 seconds to download it.
162 * DonĀ“t call set_time_limit in safe mode.
165 if (!ini_get('safe_mode')) {
168 /* in case of base64 encoded attachments, do not buffer them.
169 Instead, echo the decoded attachment directly to screen */
170 if (strtolower($encoding) == 'base64') {
172 $query = "FETCH $id BODY[]";
174 $query = "FETCH $id BODY[$ent_id]";
176 sqimap_run_command($imap_stream,$query,true,$response,$message,$uid_support,'sqimap_base64_decode','php://stdout',true);
178 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
179 echo decodeBody($body, $encoding);
183 TODO, use the same method for quoted printable.
184 However, I assume that quoted printable attachments aren't that large
185 so the performancegain / memory usage drop will be minimal.
186 If we decide to add that then we need to adapt sqimap_fread because
187 we need to split te result on \n and fread doesn't stop at \n. That
188 means we also should provide $results from sqimap_fread (by ref) to
189 te function and set $no_return to false. The $filter function for
190 quoted printable should handle unsetting of $results.
193 TODO 2: find out how we write to the output stream php://stdout. fwrite
194 doesn't work because 'php://stdout isn't a stream.
199 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
202 $read = fgets ($imap_stream,8192);
205 // This could be bad -- if the section has sqimap_session_id() . ' OK'
206 // or similar, it will kill the download.
207 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
208 if (trim($read) == ')==') {
210 $read = fgets ($imap_stream,4096);
211 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
214 echo decodeBody($read1, $encoding) .
215 decodeBody($read, $encoding);
218 echo decodeBody($read, $encoding);
220 $read = fgets ($imap_stream,4096);
227 /* -[ END MIME DECODING ]----------------------------------------------------------- */
229 /* This is here for debugging purposes. It will print out a list
230 * of all the entity IDs that are in the $message object.
232 function listEntities ($message) {
234 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
235 for ($i = 0; isset($message->entities
[$i]); $i++
) {
237 $msg = listEntities($message->entities
[$i]);
247 function getPriorityStr($priority) {
248 $priority_level = substr($priority,0,1);
250 switch($priority_level) {
251 /* Check for a higher then normal priority. */
254 $priority_string = _("High");
257 /* Check for a lower then normal priority. */
260 $priority_string = _("Low");
263 /* Check for a normal priority. */
266 $priority_level = '3';
267 $priority_string = _("Normal");
271 return $priority_string;
274 /* returns a $message object for a particular entity id */
275 function getEntity ($message, $ent_id) {
276 return $message->getEntity($ent_id);
280 * Extracted from strings.php 23/03/2002
283 function translateText(&$body, $wrap_at, $charset) {
284 global $where, $what; /* from searching */
285 global $color; /* color theme */
287 require_once(SM_PATH
. 'functions/url_parser.php');
289 $body_ary = explode("\n", $body);
290 for ($i=0; $i < count($body_ary); $i++
) {
291 $line = $body_ary[$i];
292 if (strlen($line) - 2 >= $wrap_at) {
293 sqWordWrap($line, $wrap_at);
295 $line = charset_decode($charset, $line);
296 $line = str_replace("\t", ' ', $line);
305 if ($line[$pos] == ' ') {
307 } else if (strpos($line, '>', $pos) === $pos) {
316 if (!isset($color[13])) {
317 $color[13] = '#800000';
319 $line = '<font color="' . $color[13] . '">' . $line . '</font>';
321 if (!isset($color[14])) {
322 $color[14] = '#FF0000';
324 $line = '<font color="' . $color[14] . '">' . $line . '</font>';
327 $body_ary[$i] = $line;
329 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
332 /* This returns a parsed string called $body. That string can then
333 * be displayed as the actual message in the HTML. It contains
334 * everything needed, including HTML Tags, Attachments at the
337 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
338 /* This if statement checks for the entity to show as the
339 * primary message. To add more of them, just put them in the
340 * order that is their priority.
342 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
343 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
344 global $languages, $squirrelmail_language;
346 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
347 $view_unsafe_images = false;
351 $urlmailbox = urlencode($mailbox);
352 $body_message = getEntity($message, $ent_num);
353 if (($body_message->header
->type0
== 'text') ||
354 ($body_message->header
->type0
== 'rfc822')) {
355 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
356 $body = decodeBody($body, $body_message->header
->encoding
);
358 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
359 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
360 if (mb_detect_encoding($body) != 'ASCII') {
361 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
364 $hookResults = do_hook("message_body", $body);
365 $body = $hookResults[1];
367 /* If there are other types that shouldn't be formatted, add
371 if ($body_message->header
->type1
== 'html') {
372 if ($show_html_default <> 1) {
373 $entity_conv = array(' ' => ' ',
380 $body = strtr($body, $entity_conv);
381 $body = strip_tags($body);
383 translateText($body, $wrap_at,
384 $body_message->header
->getParameter('charset'));
386 $body = magicHTML($body, $id, $message, $mailbox);
389 translateText($body, $wrap_at,
390 $body_message->header
->getParameter('charset'));
392 $link = 'read_body.php?passed_id=' . $id . '&ent_id='.$ent_num.
393 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
394 '&startMessage=' . $startMessage . '&show_more=0';
395 if (isset($passed_ent_id)) {
396 $link .= '&passed_ent_id='.$passed_ent_id;
398 if ($view_unsafe_images) {
399 $text = _("Hide Unsafe Images");
401 if (isset($has_unsafe_images) && $has_unsafe_images) {
402 $link .= '&view_unsafe_images=1';
403 $text = _("View Unsafe Images");
408 $body .= '<center><small><a href="'.$link.'">'.$text.
409 '</a></small></center><br>' . "\n";
415 function formatAttachments($message, $exclude_id, $mailbox, $id) {
416 global $where, $what, $startMessage, $color;
417 static $ShownHTML = 0;
419 $att_ar = $message->getAttachments($exclude_id);
421 if (!count($att_ar)) return '';
425 $urlMailbox = urlencode($mailbox);
427 foreach ($att_ar as $att) {
428 $ent = $att->entity_id
;
429 $header = $att->header
;
430 $type0 = strtolower($header->type0
);
431 $type1 = strtolower($header->type1
);
433 $links['download link']['text'] = _("download");
434 $links['download link']['href'] = SM_PATH
.
435 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
437 if ($type0 =='message' && $type1 == 'rfc822') {
438 $default_page = SM_PATH
. 'src/read_body.php';
439 $rfc822_header = $att->rfc822_header
;
440 $filename = $rfc822_header->subject
;
441 if (trim( $filename ) == '') {
442 $filename = 'untitled-[' . $ent . ']' ;
444 $from_o = $rfc822_header->from
;
445 if (is_object($from_o)) {
446 $from_name = decodeHeader($from_o->getAddress(false));
448 $from_name = _("Unknown sender");
450 $description = $from_name;
452 $default_page = SM_PATH
. 'src/download.php';
453 if (is_object($header->disposition
)) {
454 $filename = $header->disposition
->getProperty('filename');
455 if (trim($filename) == '') {
456 $name = decodeHeader($header->disposition
->getProperty('name'));
457 if (trim($name) == '') {
458 $name = $header->getParameter('name');
459 if(trim($name) == '') {
460 if (trim( $header->id
) == '') {
461 $filename = 'untitled-[' . $ent . ']' ;
463 $filename = 'cid: ' . $header->id
;
473 $filename = $header->getParameter('name');
474 if (!trim($filename)) {
475 if (trim( $header->id
) == '') {
476 $filename = 'untitled-[' . $ent . ']' ;
478 $filename = 'cid: ' . $header->id
;
482 if ($header->description
) {
483 $description = decodeHeader($header->description
);
489 $display_filename = $filename;
490 if (isset($passed_ent_id)) {
491 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
493 $passed_ent_id_link = '';
495 $defaultlink = $default_page . "?startMessage=$startMessage"
496 . "&passed_id=$id&mailbox=$urlMailbox"
497 . '&ent_id='.$ent.$passed_ent_id_link;
498 if ($where && $what) {
499 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
501 /* This executes the attachment hook with a specific MIME-type.
502 * If that doesn't have results, it tries if there's a rule
503 * for a more generic type.
505 $hookresults = do_hook("attachment $type0/$type1", $links,
506 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
507 $display_filename, $where, $what);
508 if(count($hookresults[1]) <= 1) {
509 $hookresults = do_hook("attachment $type0/*", $links,
510 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
511 $display_filename, $where, $what);
514 $links = $hookresults[1];
515 $defaultlink = $hookresults[6];
517 $attachments .= '<TR><TD>' .
518 '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A> </TD>' .
519 '<TD><SMALL><b>' . show_readable_size($header->size
) .
520 '</b> </small></TD>' .
521 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
523 $attachments .= '<b>' . $description . '</b>';
524 $attachments .= '</SMALL></TD><TD><SMALL> ';
527 foreach ($links as $val) {
531 $attachments .= ' | ';
533 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
536 $attachments .= "</TD></TR>\n";
538 $attachmentadd = do_hook_function('attachments_bottom',$attachments);
539 if ($attachmentadd != '')
540 $attachments = $attachmentadd;
544 function sqimap_base64_decode(&$string) {
545 $string = str_replace("\r\n", "\n", $string);
546 $string = base64_decode($string);
549 /* This function decodes the body depending on the encoding type. */
550 function decodeBody($body, $encoding) {
551 global $show_html_default;
553 $body = str_replace("\r\n", "\n", $body);
554 $encoding = strtolower($encoding);
556 $encoding_handler = do_hook_function('decode_body', $encoding);
559 // plugins get first shot at decoding the body
561 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
562 $body = $encoding_handler('decode', $body);
564 } else if ($encoding == 'quoted-printable' ||
565 $encoding == 'quoted_printable') {
566 $body = quoted_printable_decode($body);
568 while (ereg("=\n", $body)) {
569 $body = ereg_replace ("=\n", '', $body);
572 } else if ($encoding == 'base64') {
573 $body = base64_decode($body);
576 // All other encodings are returned raw.
581 * This functions decode strings that is encoded according to
582 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
583 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
585 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
586 global $languages, $squirrelmail_language;
587 if (is_array($string)) {
588 $string = implode("\n", $string);
591 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
592 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
593 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
594 // Do we need to return at this point?
601 $aString = explode(' ',$string);
603 foreach ($aString as $chunk) {
604 if ($encoded && $chunk === '') {
606 } elseif ($chunk === '') {
611 /* if encoded words are not separated by a linear-space-white we still catch them */
613 // if ($chunk{0} === '=') { /* performance, saves an unnessecarry preg call */
614 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
615 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
616 if ($iLastMatch !== $j) {
626 $encoding = ucfirst($res[3]);
630 $replace = base64_decode($res[4]);
631 $ret .= charset_decode($res[2],$replace);
634 $replace = str_replace('_', ' ', $res[4]);
635 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
637 /* Only encode into entities by default. Some places
638 * don't need the encoding, like the compose form.
641 $replace = charset_decode($res[2], $replace);
644 $replace = htmlspecialchars($replace);
664 if (!$encoded && $htmlsave) {
665 $ret .= htmlspecialchars($chunk);
671 /* remove the first added space */
674 $ret = substr($ret,6);
676 $ret = substr($ret,1);
684 * Encode a string according to RFC 1522 for use in headers if it
685 * contains 8-bit characters or anything that looks like it should
688 function encodeHeader ($string) {
689 global $default_charset, $languages, $squirrelmail_language;
691 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
692 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
693 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
695 if (strtolower($default_charset) == 'iso-8859-1') {
696 $string = str_replace("\240",' ',$string);
699 // Encode only if the string contains 8-bit characters or =?
700 $j = strlen($string);
701 $max_l = 75 - strlen($default_charset) - 7;
704 $iEncStart = $enc_init = false;
705 $cur_l = $iOffset = 0;
706 for($i = 0; $i < $j; ++
$i) {
715 if ($iEncStart === false) {
719 if ($cur_l > ($max_l-2)) {
720 /* if there is an stringpart that doesn't need encoding, add it */
721 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
722 $aRet[] = "=?$default_charset?Q?$ret?=";
728 $ret .= sprintf("=%02X",ord($string{$i}));
733 if ($iEncStart !== false) {
734 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
735 $aRet[] = "=?$default_charset?Q?$ret?=";
743 if ($iEncStart !== false) {
745 if ($cur_l > $max_l) {
746 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
747 $aRet[] = "=?$default_charset?Q?$ret?=";
758 $k = ord($string{$i});
760 if ($iEncStart === false) {
761 // do not start encoding in the middle of a string, also take the rest of the word.
762 $sLeadString = substr($string,0,$i);
763 $aLeadString = explode(' ',$sLeadString);
764 $sToBeEncoded = array_pop($aLeadString);
765 $iEncStart = $i - strlen($sToBeEncoded);
766 $ret .= $sToBeEncoded;
767 $cur_l +
= strlen($sToBeEncoded);
770 /* first we add the encoded string that reached it's max size */
771 if ($cur_l > ($max_l-2)) {
772 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
773 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
780 $ret .= sprintf("=%02X", $k);
782 if ($iEncStart !== false) {
784 if ($cur_l > $max_l) {
785 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
786 $aRet[] = "=?$default_charset?Q?$ret?=";
801 if ($iEncStart !== false) {
802 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
803 $aRet[] = "=?$default_charset?Q?$ret?=";
805 $aRet[] = substr($string,$iOffset);
807 $string = implode('',$aRet);
812 /* This function trys to locate the entity_id of a specific mime element */
813 function find_ent_id($id, $message) {
814 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
815 if ($message->entities
[$i]->header
->type0
== 'multipart') {
816 $ret = find_ent_id($id, $message->entities
[$i]);
818 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
819 // if (sq_check_save_extension($message->entities[$i])) {
820 return $message->entities
[$i]->entity_id
;
828 function sq_check_save_extension($message) {
829 $filename = $message->getFilename();
830 $ext = substr($filename, strrpos($filename,'.')+
1);
831 $save_extensions = array('jpg','jpeg','gif','png','bmp');
832 return in_array($ext, $save_extensions);
837 ** HTMLFILTER ROUTINES
841 * This function is more or less a wrapper around stripslashes. Apparently
842 * Explorer is stupid enough to just remove the backslashes and then
843 * execute the content of the attribute as if nothing happened.
846 * @param attvalue The value of the attribute
847 * @return attvalue The value of the attribute stripslashed.
849 function sq_unbackslash($attvalue){
851 * Remove any backslashes. See if there are any first.
853 if (strstr($attvalue, '\\') !== false){
854 $attvalue = stripslashes($attvalue);
860 * Kill any tabs, newlines, or carriage returns. Our friends the
861 * makers of the browser with 95% market value decided that it'd
862 * be funny to make "java[tab]script" be just as good as "javascript".
864 * @param attvalue The attribute value before extraneous spaces removed.
865 * @return attvalue The attribute value after extraneous spaces removed.
867 function sq_unspace($attvalue){
868 if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
869 $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
876 * This function returns the final tag out of the tag name, an array
877 * of attributes, and the type of the tag. This function is called by
878 * sq_sanitize internally.
880 * @param $tagname the name of the tag.
881 * @param $attary the array of attributes and their values
882 * @param $tagtype The type of the tag (see in comments).
883 * @return a string with the final tag representation.
885 function sq_tagprint($tagname, $attary, $tagtype){
889 $fulltag = '</' . $tagname . '>';
891 $fulltag = '<' . $tagname;
892 if (is_array($attary) && sizeof($attary)){
894 while (list($attname, $attvalue) = each($attary)){
895 array_push($atts, "$attname=$attvalue");
897 $fulltag .= ' ' . join(" ", $atts);
908 * A small helper function to use with array_walk. Modifies a by-ref
909 * value and makes it lowercase.
911 * @param $val a value passed by-ref.
912 * @return void since it modifies a by-ref value.
914 function sq_casenormalize(&$val){
915 $val = strtolower($val);
919 * This function skips any whitespace from the current position within
920 * a string and to the next non-whitespace value.
922 * @param $body the string
923 * @param $offset the offset within the string where we should start
924 * looking for the next non-whitespace character.
925 * @return the location within the $body where the next
926 * non-whitespace char is located.
928 function sq_skipspace($body, $offset){
929 $me = 'sq_skipspace';
930 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
931 if (sizeof($matches{1})){
932 $count = strlen($matches{1});
939 * This function looks for the next character within a string. It's
940 * really just a glorified "strpos", except it catches if failures
943 * @param $body The string to look for needle in.
944 * @param $offset Start looking from this position.
945 * @param $needle The character/string to look for.
946 * @return location of the next occurance of the needle, or
947 * strlen($body) if needle wasn't found.
949 function sq_findnxstr($body, $offset, $needle){
950 $me = 'sq_findnxstr';
951 $pos = strpos($body, $needle, $offset);
953 $pos = strlen($body);
959 * This function takes a PCRE-style regexp and tries to match it
962 * @param $body The string to look for needle in.
963 * @param $offset Start looking from here.
964 * @param $reg A PCRE-style regex to match.
965 * @return Returns a false if no matches found, or an array
966 * with the following members:
967 * - integer with the location of the match within $body
968 * - string with whatever content between offset and the match
969 * - string with whatever it is we matched
971 function sq_findnxreg($body, $offset, $reg){
972 $me = 'sq_findnxreg';
975 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
976 if (!isset($matches{0}) ||
!$matches{0}){
979 $retarr{0} = $offset +
strlen($matches{1});
980 $retarr{1} = $matches{1};
981 $retarr{2} = $matches{2};
987 * This function looks for the next tag.
989 * @param $body String where to look for the next tag.
990 * @param $offset Start looking from here.
991 * @return false if no more tags exist in the body, or
992 * an array with the following members:
993 * - string with the name of the tag
994 * - array with attributes and their values
995 * - integer with tag type (1, 2, or 3)
996 * - integer where the tag starts (starting "<")
997 * - integer where the tag ends (ending ">")
998 * first three members will be false, if the tag is invalid.
1000 function sq_getnxtag($body, $offset){
1001 $me = 'sq_getnxtag';
1002 if ($offset > strlen($body)){
1005 $lt = sq_findnxstr($body, $offset, "<");
1006 if ($lt == strlen($body)){
1011 * blah blah <tag attribute="value">
1014 $pos = sq_skipspace($body, $lt+
1);
1015 if ($pos >= strlen($body)){
1016 return Array(false, false, false, $lt, strlen($body));
1019 * There are 3 kinds of tags:
1020 * 1. Opening tag, e.g.:
1022 * 2. Closing tag, e.g.:
1024 * 3. XHTML-style content-less tag, e.g.:
1028 switch (substr($body, $pos, 1)){
1035 * A comment or an SGML declaration.
1037 if (substr($body, $pos+
1, 2) == "--"){
1038 $gt = strpos($body, "-->", $pos);
1040 $gt = strlen($body);
1044 return Array(false, false, false, $lt, $gt);
1046 $gt = sq_findnxstr($body, $pos, ">");
1047 return Array(false, false, false, $lt, $gt);
1052 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1062 * Look for next [\W-_], which will indicate the end of the tag name.
1064 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1065 if ($regary == false){
1066 return Array(false, false, false, $lt, strlen($body));
1068 list($pos, $tagname, $match) = $regary;
1069 $tagname = strtolower($tagname);
1072 * $match can be either of these:
1073 * '>' indicating the end of the tag entirely.
1074 * '\s' indicating the end of the tag name.
1075 * '/' indicating that this is type-3 xhtml tag.
1077 * Whatever else we find there indicates an invalid tag.
1082 * This is an xhtml-style tag with a closing / at the
1083 * end, like so: <img src="blah"/>. Check if it's followed
1084 * by the closing bracket. If not, then this tag is invalid
1086 if (substr($body, $pos, 2) == "/>"){
1090 $gt = sq_findnxstr($body, $pos, ">");
1091 $retary = Array(false, false, false, $lt, $gt);
1095 return Array($tagname, false, $tagtype, $lt, $pos);
1099 * Check if it's whitespace
1101 if (!preg_match('/\s/', $match)){
1103 * This is an invalid tag! Look for the next closing ">".
1105 $gt = sq_findnxstr($body, $lt, ">");
1106 return Array(false, false, false, $lt, $gt);
1112 * At this point we're here:
1113 * <tagname attribute='blah'>
1116 * At this point we loop in order to find all attributes.
1122 while ($pos <= strlen($body)){
1123 $pos = sq_skipspace($body, $pos);
1124 if ($pos == strlen($body)){
1128 return Array(false, false, false, $lt, $pos);
1131 * See if we arrived at a ">" or "/>", which means that we reached
1132 * the end of the tag.
1135 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1139 $pos +
= strlen($matches{1});
1140 if ($matches{2} == "/>"){
1144 return Array($tagname, $attary, $tagtype, $lt, $pos);
1148 * There are several types of attributes, with optional
1149 * [:space:] between members.
1151 * attrname[:space:]=[:space:]'CDATA'
1153 * attrname[:space:]=[:space:]"CDATA"
1155 * attr[:space:]=[:space:]CDATA
1159 * We leave types 1 and 2 the same, type 3 we check for
1160 * '"' and convert to """ if needed, then wrap in
1161 * double quotes. Type 4 we convert into:
1164 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1165 if ($regary == false){
1167 * Looks like body ended before the end of tag.
1169 return Array(false, false, false, $lt, strlen($body));
1171 list($pos, $attname, $match) = $regary;
1172 $attname = strtolower($attname);
1174 * We arrived at the end of attribute name. Several things possible
1176 * '>' means the end of the tag and this is attribute type 4
1177 * '/' if followed by '>' means the same thing as above
1178 * '\s' means a lot of things -- look what it's followed by.
1179 * anything else means the attribute is invalid.
1184 * This is an xhtml-style tag with a closing / at the
1185 * end, like so: <img src="blah"/>. Check if it's followed
1186 * by the closing bracket. If not, then this tag is invalid
1188 if (substr($body, $pos, 2) == "/>"){
1192 $gt = sq_findnxstr($body, $pos, ">");
1193 $retary = Array(false, false, false, $lt, $gt);
1197 $attary{$attname} = '"yes"';
1198 return Array($tagname, $attary, $tagtype, $lt, $pos);
1202 * Skip whitespace and see what we arrive at.
1204 $pos = sq_skipspace($body, $pos);
1205 $char = substr($body, $pos, 1);
1207 * Two things are valid here:
1208 * '=' means this is attribute type 1 2 or 3.
1209 * \w means this was attribute type 4.
1210 * anything else we ignore and re-loop. End of tag and
1211 * invalid stuff will be caught by our checks at the beginning
1216 $pos = sq_skipspace($body, $pos);
1218 * Here are 3 possibilities:
1219 * "'" attribute type 1
1220 * '"' attribute type 2
1221 * everything else is the content of tag type 3
1223 $quot = substr($body, $pos, 1);
1225 $regary = sq_findnxreg($body, $pos+
1, "\'");
1226 if ($regary == false){
1227 return Array(false, false, false, $lt, strlen($body));
1229 list($pos, $attval, $match) = $regary;
1231 $attary{$attname} = "'" . $attval . "'";
1232 } else if ($quot == '"'){
1233 $regary = sq_findnxreg($body, $pos+
1, '\"');
1234 if ($regary == false){
1235 return Array(false, false, false, $lt, strlen($body));
1237 list($pos, $attval, $match) = $regary;
1239 $attary{$attname} = '"' . $attval . '"';
1242 * These are hateful. Look for \s, or >.
1244 $regary = sq_findnxreg($body, $pos, "[\s>]");
1245 if ($regary == false){
1246 return Array(false, false, false, $lt, strlen($body));
1248 list($pos, $attval, $match) = $regary;
1250 * If it's ">" it will be caught at the top.
1252 $attval = preg_replace("/\"/s", """, $attval);
1253 $attary{$attname} = '"' . $attval . '"';
1255 } else if (preg_match("|[\w/>]|", $char)) {
1257 * That was attribute type 4.
1259 $attary{$attname} = '"yes"';
1262 * An illegal character. Find next '>' and return.
1264 $gt = sq_findnxstr($body, $pos, ">");
1265 return Array(false, false, false, $lt, $gt);
1271 * The fact that we got here indicates that the tag end was never
1272 * found. Return invalid tag indication so it gets stripped.
1274 return Array(false, false, false, $lt, strlen($body));
1278 * This function checks attribute values for entity-encoded values
1279 * and returns them translated into 8-bit strings so we can run
1282 * @param $attvalue A string to run entity check against.
1283 * @return Translated value.
1285 function sq_deent($attvalue){
1288 * See if we have to run the checks first. All entities must start
1291 if (strpos($attvalue, "&") === false){
1295 * Check named entities first.
1297 $trans = get_html_translation_table(HTML_ENTITIES
);
1299 * Leave " in, as it can mess us up.
1301 $trans = array_flip($trans);
1302 unset($trans{"""});
1303 while (list($ent, $val) = each($trans)){
1304 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1307 * Now translate numbered entities from 1 to 255 if needed.
1309 if (strpos($attvalue, "#") !== false){
1310 $omit = Array(34, 39);
1311 for ($asc=1; $asc<256; $asc++
){
1312 if (!in_array($asc, $omit)){
1314 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1316 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1317 "$chr\\1", $attvalue);
1325 * This function runs various checks against the attributes.
1327 * @param $tagname String with the name of the tag.
1328 * @param $attary Array with all tag attributes.
1329 * @param $rm_attnames See description for sq_sanitize
1330 * @param $bad_attvals See description for sq_sanitize
1331 * @param $add_attr_to_tag See description for sq_sanitize
1332 * @param $message message object
1333 * @param $id message id
1334 * @return Array with modified attributes.
1336 function sq_fixatts($tagname,
1346 while (list($attname, $attvalue) = each($attary)){
1348 * See if this attribute should be removed.
1350 foreach ($rm_attnames as $matchtag=>$matchattrs){
1351 if (preg_match($matchtag, $tagname)){
1352 foreach ($matchattrs as $matchattr){
1353 if (preg_match($matchattr, $attname)){
1354 unset($attary{$attname});
1361 * Remove any backslashes, entities, and extraneous whitespace.
1363 $attvalue = sq_unbackslash($attvalue);
1364 $attvalue = sq_deent($attvalue);
1365 $attvalue = sq_unspace($attvalue);
1368 * Now let's run checks on the attvalues.
1369 * I don't expect anyone to comprehend this. If you do,
1370 * get in touch with me so I can drive to where you live and
1371 * shake your hand personally. :)
1373 foreach ($bad_attvals as $matchtag=>$matchattrs){
1374 if (preg_match($matchtag, $tagname)){
1375 foreach ($matchattrs as $matchattr=>$valary){
1376 if (preg_match($matchattr, $attname)){
1378 * There are two arrays in valary.
1380 * Second one is replacements
1382 list($valmatch, $valrepl) = $valary;
1384 preg_replace($valmatch, $valrepl, $attvalue);
1385 if ($newvalue != $attvalue){
1386 $attary{$attname} = $newvalue;
1393 * Turn cid: urls into http-friendly ones.
1395 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1396 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1400 * See if we need to append any attributes to this tag.
1402 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1403 if (preg_match($matchtag, $tagname)){
1404 $attary = array_merge($attary, $addattary);
1411 * This function edits the style definition to make them friendly and
1412 * usable in squirrelmail.
1414 * @param $message the message object
1415 * @param $id the message id
1416 * @param $content a string with whatever is between <style> and </style>
1417 * @return a string with edited content.
1419 function sq_fixstyle($body, $pos, $message, $id){
1420 global $view_unsafe_images;
1421 $me = 'sq_fixstyle';
1422 $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1424 return array(FALSE, strlen($body));
1426 $newpos = $ret[0] +
strlen($ret[2]);
1429 * First look for general BODY style declaration, which would be
1431 * body {background: blah-blah}
1432 * and change it to .bodyclass so we can just assign it to a <div>
1434 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1435 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1437 * Fix url('blah') declarations.
1439 $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1440 "url(\\1$secremoveimg\\2)", $content);
1442 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1445 if (!$view_unsafe_images){
1446 $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1447 "url(\\1$secremoveimg\\2)", $content);
1451 * Fix urls that refer to cid:
1453 while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1454 $content, $matches)){
1455 $cidurl = $matches{1};
1456 $httpurl = sq_cid2http($message, $id, $cidurl);
1457 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1458 "url($httpurl)", $content);
1462 * Fix stupid css declarations which lead to vulnerabilities
1465 $match = Array('/expression/i',
1468 '/include-source/i');
1469 $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1470 $content = preg_replace($match, $replace, $content);
1471 return array($content, $newpos);
1475 * This function converts cid: url's into the ones that can be viewed in
1478 * @param $message the message object
1479 * @param $id the message id
1480 * @param $cidurl the cid: url.
1481 * @return a string with a http-friendly url
1483 function sq_cid2http($message, $id, $cidurl, $mailbox){
1485 * Get rid of quotes.
1487 $quotchar = substr($cidurl, 0, 1);
1488 if ($quotchar == '"' ||
$quotchar == "'"){
1489 $cidurl = str_replace($quotchar, "", $cidurl);
1493 $cidurl = substr(trim($cidurl), 4);
1494 $linkurl = find_ent_id($cidurl, $message);
1495 /* in case of non-save cid links $httpurl should be replaced by a sort of
1496 unsave link image */
1499 $httpurl = $quotchar . SM_PATH
. 'src/download.php?absolute_dl=true&' .
1500 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1501 '&ent_id=' . $linkurl . $quotchar;
1507 * This function changes the <body> tag into a <div> tag since we
1508 * can't really have a body-within-body.
1510 * @param $attary an array of attributes and values of <body>
1511 * @param $mailbox mailbox we're currently reading (for cid2http)
1512 * @param $message current message (for cid2http)
1513 * @param $id current message id (for cid2http)
1514 * @return a modified array of attributes to be set for <div>
1516 function sq_body2div($attary, $mailbox, $message, $id){
1517 $me = 'sq_body2div';
1518 $divattary = Array('class' => "'bodyclass'");
1519 $bgcolor = '#ffffff';
1522 if (is_array($attary) && sizeof($attary) > 0){
1523 foreach ($attary as $attname=>$attvalue){
1524 $quotchar = substr($attvalue, 0, 1);
1525 $attvalue = str_replace($quotchar, "", $attvalue);
1528 $attvalue = sq_cid2http($message, $id,
1529 $attvalue, $mailbox);
1530 $styledef .= "background-image: url('$attvalue'); ";
1533 $styledef .= "background-color: $attvalue; ";
1536 $styledef .= "color: $attvalue; ";
1540 if (strlen($styledef) > 0){
1541 $divattary{"style"} = "\"$styledef\"";
1548 * This is the main function and the one you should actually be calling.
1549 * There are several variables you should be aware of an which need
1550 * special description.
1552 * Since the description is quite lengthy, see it here:
1553 * http://www.mricon.com/html/phpfilter.html
1555 * @param $body the string with HTML you wish to filter
1556 * @param $tag_list see description above
1557 * @param $rm_tags_with_content see description above
1558 * @param $self_closing_tags see description above
1559 * @param $force_tag_closing see description above
1560 * @param $rm_attnames see description above
1561 * @param $bad_attvals see description above
1562 * @param $add_attr_to_tag see description above
1563 * @param $message message object
1564 * @param $id message id
1565 * @return sanitized html safe to show on your pages.
1567 function sq_sanitize($body,
1569 $rm_tags_with_content,
1579 $me = 'sq_sanitize';
1580 $rm_tags = array_shift($tag_list);
1582 * Normalize rm_tags and rm_tags_with_content.
1584 @array_walk
($tag_list, 'sq_casenormalize');
1585 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1586 @array_walk
($self_closing_tags, 'sq_casenormalize');
1588 * See if tag_list is of tags to remove or tags to allow.
1589 * false means remove these tags
1590 * true means allow these tags
1593 $open_tags = Array();
1594 $trusted = "\n<!-- begin sanitized html -->\n";
1595 $skip_content = false;
1597 * Take care of netscape's stupid javascript entities like
1600 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1602 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1603 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1604 $free_content = substr($body, $curpos, $lt-$curpos);
1606 * Take care of <style>
1608 if ($tagname == "style" && $tagtype == 1){
1609 list($free_content, $curpos) =
1610 sq_fixstyle($body, $gt+
1, $message, $id);
1611 if ($free_content != FALSE){
1612 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1613 $trusted .= $free_content;
1614 $trusted .= sq_tagprint($tagname, false, 2);
1618 if ($skip_content == false){
1619 $trusted .= $free_content;
1621 if ($tagname != FALSE){
1623 if ($skip_content == $tagname){
1625 * Got to the end of tag we needed to remove.
1628 $skip_content = false;
1630 if ($skip_content == false){
1631 if ($tagname == "body"){
1634 if (isset($open_tags{$tagname}) &&
1635 $open_tags{$tagname} > 0){
1636 $open_tags{$tagname}--;
1644 * $rm_tags_with_content
1646 if ($skip_content == false){
1648 * See if this is a self-closing type and change
1649 * tagtype appropriately.
1652 && in_array($tagname, $self_closing_tags)){
1656 * See if we should skip this tag and any content
1659 if ($tagtype == 1 &&
1660 in_array($tagname, $rm_tags_with_content)){
1661 $skip_content = $tagname;
1663 if (($rm_tags == false
1664 && in_array($tagname, $tag_list)) ||
1665 ($rm_tags == true &&
1666 !in_array($tagname, $tag_list))){
1670 * Convert body into div.
1672 if ($tagname == "body"){
1674 $attary = sq_body2div($attary, $mailbox,
1678 if (isset($open_tags{$tagname})){
1679 $open_tags{$tagname}++
;
1681 $open_tags{$tagname}=1;
1685 * This is where we run other checks.
1687 if (is_array($attary) && sizeof($attary) > 0){
1688 $attary = sq_fixatts($tagname,
1702 if ($tagname != false && $skip_content == false){
1703 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1708 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1709 if ($force_tag_closing == true){
1710 foreach ($open_tags as $tagname=>$opentimes){
1711 while ($opentimes > 0){
1712 $trusted .= '</' . $tagname . '>';
1718 $trusted .= "<!-- end sanitized html -->\n";
1723 * This is a wrapper function to call html sanitizing routines.
1725 * @param $body the body of the message
1726 * @param $id the id of the message
1727 * @return a string with html safe to display in the browser.
1729 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1730 global $attachment_common_show_images, $view_unsafe_images,
1733 * Don't display attached images in HTML mode.
1735 $attachment_common_show_images = false;
1750 $rm_tags_with_content = Array(
1759 $self_closing_tags = Array(
1766 $force_tag_closing = true;
1768 $rm_attnames = Array(
1779 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1780 $bad_attvals = Array(
1783 "/^src|background/i" =>
1786 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1787 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1788 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1791 "\\1$secremoveimg\\2",
1792 "\\1$secremoveimg\\2",
1793 "\\1$secremoveimg\\2",
1794 "\\1$secremoveimg\\2"
1797 "/^href|action/i" =>
1800 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1801 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1802 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1817 "/include-source/i",
1818 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1819 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1820 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si"
1835 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1836 $view_unsafe_images = false;
1838 if (!$view_unsafe_images){
1840 * Remove any references to http/https if view_unsafe_images set
1843 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1844 '/^([\'\"])\s*https*:.*([\'\"])/si');
1845 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1846 "\\1$secremoveimg\\1");
1847 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1848 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1849 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1850 "url(\\1$secremoveimg\\1)");
1853 $add_attr_to_tag = Array(
1855 Array('target'=>'"_new"',
1856 'title'=>'"'._("This external link will open in a new window").'"'
1859 $trusted = sq_sanitize($body,
1861 $rm_tags_with_content,
1871 if (preg_match("|$secremoveimg|i", $trusted)){
1872 $has_unsafe_images = true;