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, $mailbox );
39 echo "<BODY TEXT=\"$color[8]\" BGCOLOR=\"$color[4]\" LINK=\"$color[7]\" VLINK=\"$color[7]\" ALINK=\"$color[7]\">\n\n" .
41 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
42 $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
43 $errormessage .= '<table><tr><td>' . htmlspecialchars($read) . '</td></tr></table>';
44 plain_error_message( $errormessage, $color );
45 echo '</body></html>';
49 foreach ($flags as $flag) {
50 $char = strtoupper($flag{1});
53 if (strtolower($flag) == '\\seen') {
58 if (strtolower($flag) == '\\answered') {
59 $msg->is_answered
= true;
63 if (strtolower($flag) == '\\deleted') {
64 $msg->is_deleted
= true;
68 if (strtolower($flag) == '\\flagged') {
69 $msg->is_flagged
= true;
73 if (strtolower($flag) == '$mdnsent') {
74 $msg->is_mdnsent
= true;
82 // listEntities($msg);
88 /* This starts the parsing of a particular structure. It is called recursively,
89 * so it can be passed different structures. It returns an object of type
91 * First, it checks to see if it is a multipart message. If it is, then it
92 * handles that as it sees is necessary. If it is just a regular entity,
93 * then it parses it and adds the necessary header information (by calling out
94 * to mime_get_elements()
97 function mime_fetch_body($imap_stream, $id, $ent_id=1) {
99 /* Do a bit of error correction. If we couldn't find the entity id, just guess
100 * that it is the first one. That is usually the case anyway.
104 $cmd = "FETCH $id BODY[]";
106 $cmd = "FETCH $id BODY[$ent_id]";
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 = $from_o->getAddress(false);
448 $from_name = _("Unknown sender");
450 $from_name = decodeHeader(($from_name));
451 $description = $from_name;
453 $default_page = SM_PATH
. 'src/download.php';
454 if (is_object($header->disposition
)) {
455 $filename = $header->disposition
->getProperty('filename');
456 if (trim($filename) == '') {
457 $name = decodeHeader($header->disposition
->getProperty('name'));
458 if (trim($name) == '') {
459 $name = $header->getParameter('name');
460 if(trim($name) == '') {
461 if (trim( $header->id
) == '') {
462 $filename = 'untitled-[' . $ent . ']' ;
464 $filename = 'cid: ' . $header->id
;
474 $filename = $header->getParameter('name');
475 if (!trim($filename)) {
476 if (trim( $header->id
) == '') {
477 $filename = 'untitled-[' . $ent . ']' ;
479 $filename = 'cid: ' . $header->id
;
483 if ($header->description
) {
484 $description = decodeHeader($header->description
);
490 $display_filename = $filename;
491 if (isset($passed_ent_id)) {
492 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
494 $passed_ent_id_link = '';
496 $defaultlink = $default_page . "?startMessage=$startMessage"
497 . "&passed_id=$id&mailbox=$urlMailbox"
498 . '&ent_id='.$ent.$passed_ent_id_link;
499 if ($where && $what) {
500 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
502 /* This executes the attachment hook with a specific MIME-type.
503 * If that doesn't have results, it tries if there's a rule
504 * for a more generic type.
506 $hookresults = do_hook("attachment $type0/$type1", $links,
507 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
508 $display_filename, $where, $what);
509 if(count($hookresults[1]) <= 1) {
510 $hookresults = do_hook("attachment $type0/*", $links,
511 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
512 $display_filename, $where, $what);
515 $links = $hookresults[1];
516 $defaultlink = $hookresults[6];
518 $attachments .= '<TR><TD>' .
519 '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A> </TD>' .
520 '<TD><SMALL><b>' . show_readable_size($header->size
) .
521 '</b> </small></TD>' .
522 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
524 $attachments .= '<b>' . $description . '</b>';
525 $attachments .= '</SMALL></TD><TD><SMALL> ';
528 foreach ($links as $val) {
532 $attachments .= ' | ';
534 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
537 $attachments .= "</TD></TR>\n";
542 function sqimap_base64_decode(&$string) {
543 $string = base64_decode($string);
546 /* This function decodes the body depending on the encoding type. */
547 function decodeBody($body, $encoding) {
548 global $show_html_default;
550 $body = str_replace("\r\n", "\n", $body);
551 $encoding = strtolower($encoding);
553 if ($encoding == 'quoted-printable' ||
554 $encoding == 'quoted_printable') {
555 $body = quoted_printable_decode($body);
557 while (ereg("=\n", $body)) {
558 $body = ereg_replace ("=\n", '', $body);
561 } else if ($encoding == 'base64') {
562 $body = base64_decode($body);
565 // All other encodings are returned raw.
570 * This functions decode strings that is encoded according to
571 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
572 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
574 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
575 global $languages, $squirrelmail_language;
576 if (is_array($string)) {
577 $string = implode("\n", $string);
580 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
581 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
582 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
583 // Do we need to return at this point?
590 $aString = explode(' ',$string);
592 foreach ($aString as $chunk) {
593 if ($encoded && !$chunk) {
600 /* if encoded words are not separated by a linear-space-white we still catch them */
602 if ($chunk{0} === '=') { /* performance, saves an unnessecarry preg call */
603 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
604 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
605 if ($iLastMatch !== $j) {
615 $encoding = ucfirst($res[3]);
619 $replace = base64_decode($res[4]);
620 $ret .= charset_decode($res[2],$replace);
623 $replace = str_replace('_', ' ', $res[4]);
624 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
626 /* Only encode into entities by default. Some places
627 * don't need the encoding, like the compose form.
630 $replace = charset_decode($res[2], $replace);
633 $replace = htmlspecialchars($replace);
653 if (!$encoded && $htmlsave) {
654 $ret .= htmlspecialchars($chunk);
660 /* remove the first added space */
663 $ret = substr($ret,6);
665 $ret = substr($ret,1);
673 * Encode a string according to RFC 1522 for use in headers if it
674 * contains 8-bit characters or anything that looks like it should
677 function encodeHeader ($string) {
678 global $default_charset, $languages, $squirrelmail_language;
680 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
681 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
682 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
684 if (strtolower($default_charset) == 'iso-8859-1') {
685 $string = str_replace("\240",' ',$string);
688 // Encode only if the string contains 8-bit characters or =?
689 $j = strlen($string);
690 $max_l = 75 - strlen($default_charset) - 7;
693 $iEncStart = $enc_init = false;
694 $cur_l = $iOffset = 0;
695 for($i = 0; $i < $j; ++
$i) {
704 if ($iEncStart === false) {
708 if ($cur_l > ($max_l-2)) {
709 /* if there is an stringpart that doesn't need encoding, add it */
710 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
711 $aRet[] = "=?$default_charset?Q?$ret?=";
717 $ret .= sprintf("=%02X",ord($string{$i}));
722 if ($iEncStart !== false) {
723 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
724 $aRet[] = "=?$default_charset?Q?$ret?=";
732 if ($iEncStart !== false) {
734 if ($cur_l > $max_l) {
735 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
736 $aRet[] = "=?$default_charset?Q?$ret?=";
747 $k = ord($string{$i});
749 if ($iEncStart === false) {
753 /* first we add the encoded string that reached it's max size */
754 if ($cur_l > ($max_l-2)) {
755 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
756 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
763 $ret .= sprintf("=%02X", $k);
765 if ($iEncStart !== false) {
767 if ($cur_l > $max_l) {
768 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
769 $aRet[] = "=?$default_charset?Q?$ret?=";
784 if ($iEncStart !== false) {
785 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
786 $aRet[] = "=?$default_charset?Q?$ret?=";
788 $aRet[] = substr($string,$iOffset);
790 $string = implode('',$aRet);
795 /* This function trys to locate the entity_id of a specific mime element */
796 function find_ent_id($id, $message) {
797 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
798 if ($message->entities
[$i]->header
->type0
== 'multipart') {
799 $ret = find_ent_id($id, $message->entities
[$i]);
801 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
802 // if (sq_check_save_extension($message->entities[$i])) {
803 return $message->entities
[$i]->entity_id
;
811 function sq_check_save_extension($message) {
812 $filename = $message->getFilename();
813 $ext = substr($filename, strrpos($filename,'.')+
1);
814 $save_extensions = array('jpg','jpeg','gif','png','bmp');
815 return in_array($ext, $save_extensions);
820 ** HTMLFILTER ROUTINES
824 * This function returns the final tag out of the tag name, an array
825 * of attributes, and the type of the tag. This function is called by
826 * sq_sanitize internally.
828 * @param $tagname the name of the tag.
829 * @param $attary the array of attributes and their values
830 * @param $tagtype The type of the tag (see in comments).
831 * @return a string with the final tag representation.
833 function sq_tagprint($tagname, $attary, $tagtype){
837 $fulltag = '</' . $tagname . '>';
839 $fulltag = '<' . $tagname;
840 if (is_array($attary) && sizeof($attary)){
842 while (list($attname, $attvalue) = each($attary)){
843 array_push($atts, "$attname=$attvalue");
845 $fulltag .= ' ' . join(" ", $atts);
856 * A small helper function to use with array_walk. Modifies a by-ref
857 * value and makes it lowercase.
859 * @param $val a value passed by-ref.
860 * @return void since it modifies a by-ref value.
862 function sq_casenormalize(&$val){
863 $val = strtolower($val);
867 * This function skips any whitespace from the current position within
868 * a string and to the next non-whitespace value.
870 * @param $body the string
871 * @param $offset the offset within the string where we should start
872 * looking for the next non-whitespace character.
873 * @return the location within the $body where the next
874 * non-whitespace char is located.
876 function sq_skipspace($body, $offset){
877 $me = 'sq_skipspace';
878 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
879 if (sizeof($matches{1})){
880 $count = strlen($matches{1});
887 * This function looks for the next character within a string. It's
888 * really just a glorified "strpos", except it catches if failures
891 * @param $body The string to look for needle in.
892 * @param $offset Start looking from this position.
893 * @param $needle The character/string to look for.
894 * @return location of the next occurance of the needle, or
895 * strlen($body) if needle wasn't found.
897 function sq_findnxstr($body, $offset, $needle){
898 $me = 'sq_findnxstr';
899 $pos = strpos($body, $needle, $offset);
901 $pos = strlen($body);
907 * This function takes a PCRE-style regexp and tries to match it
910 * @param $body The string to look for needle in.
911 * @param $offset Start looking from here.
912 * @param $reg A PCRE-style regex to match.
913 * @return Returns a false if no matches found, or an array
914 * with the following members:
915 * - integer with the location of the match within $body
916 * - string with whatever content between offset and the match
917 * - string with whatever it is we matched
919 function sq_findnxreg($body, $offset, $reg){
920 $me = 'sq_findnxreg';
923 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
927 $retarr{0} = $offset +
strlen($matches{1});
928 $retarr{1} = $matches{1};
929 $retarr{2} = $matches{2};
935 * This function looks for the next tag.
937 * @param $body String where to look for the next tag.
938 * @param $offset Start looking from here.
939 * @return false if no more tags exist in the body, or
940 * an array with the following members:
941 * - string with the name of the tag
942 * - array with attributes and their values
943 * - integer with tag type (1, 2, or 3)
944 * - integer where the tag starts (starting "<")
945 * - integer where the tag ends (ending ">")
946 * first three members will be false, if the tag is invalid.
948 function sq_getnxtag($body, $offset){
950 if ($offset > strlen($body)){
953 $lt = sq_findnxstr($body, $offset, "<");
954 if ($lt == strlen($body)){
959 * blah blah <tag attribute="value">
962 $pos = sq_skipspace($body, $lt+
1);
963 if ($pos >= strlen($body)){
964 return Array(false, false, false, $lt, strlen($body));
967 * There are 3 kinds of tags:
968 * 1. Opening tag, e.g.:
970 * 2. Closing tag, e.g.:
972 * 3. XHTML-style content-less tag, e.g.:
976 switch (substr($body, $pos, 1)){
983 * A comment or an SGML declaration.
985 if (substr($body, $pos+
1, 2) == "--"){
986 $gt = strpos($body, "-->", $pos);
992 return Array(false, false, false, $lt, $gt);
994 $gt = sq_findnxstr($body, $pos, ">");
995 return Array(false, false, false, $lt, $gt);
1000 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1010 * Look for next [\W-_], which will indicate the end of the tag name.
1012 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1013 if ($regary == false){
1014 return Array(false, false, false, $lt, strlen($body));
1016 list($pos, $tagname, $match) = $regary;
1017 $tagname = strtolower($tagname);
1020 * $match can be either of these:
1021 * '>' indicating the end of the tag entirely.
1022 * '\s' indicating the end of the tag name.
1023 * '/' indicating that this is type-3 xhtml tag.
1025 * Whatever else we find there indicates an invalid tag.
1030 * This is an xhtml-style tag with a closing / at the
1031 * end, like so: <img src="blah"/>. Check if it's followed
1032 * by the closing bracket. If not, then this tag is invalid
1034 if (substr($body, $pos, 2) == "/>"){
1038 $gt = sq_findnxstr($body, $pos, ">");
1039 $retary = Array(false, false, false, $lt, $gt);
1043 return Array($tagname, false, $tagtype, $lt, $pos);
1047 * Check if it's whitespace
1049 if (!preg_match('/\s/', $match)){
1051 * This is an invalid tag! Look for the next closing ">".
1053 $gt = sq_findnxstr($body, $offset, ">");
1054 return Array(false, false, false, $lt, $gt);
1060 * At this point we're here:
1061 * <tagname attribute='blah'>
1064 * At this point we loop in order to find all attributes.
1070 while ($pos <= strlen($body)){
1071 $pos = sq_skipspace($body, $pos);
1072 if ($pos == strlen($body)){
1076 return Array(false, false, false, $lt, $pos);
1079 * See if we arrived at a ">" or "/>", which means that we reached
1080 * the end of the tag.
1083 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1087 $pos +
= strlen($matches{1});
1088 if ($matches{2} == "/>"){
1092 return Array($tagname, $attary, $tagtype, $lt, $pos);
1096 * There are several types of attributes, with optional
1097 * [:space:] between members.
1099 * attrname[:space:]=[:space:]'CDATA'
1101 * attrname[:space:]=[:space:]"CDATA"
1103 * attr[:space:]=[:space:]CDATA
1107 * We leave types 1 and 2 the same, type 3 we check for
1108 * '"' and convert to """ if needed, then wrap in
1109 * double quotes. Type 4 we convert into:
1112 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1113 if ($regary == false){
1115 * Looks like body ended before the end of tag.
1117 return Array(false, false, false, $lt, strlen($body));
1119 list($pos, $attname, $match) = $regary;
1120 $attname = strtolower($attname);
1122 * We arrived at the end of attribute name. Several things possible
1124 * '>' means the end of the tag and this is attribute type 4
1125 * '/' if followed by '>' means the same thing as above
1126 * '\s' means a lot of things -- look what it's followed by.
1127 * anything else means the attribute is invalid.
1132 * This is an xhtml-style tag with a closing / at the
1133 * end, like so: <img src="blah"/>. Check if it's followed
1134 * by the closing bracket. If not, then this tag is invalid
1136 if (substr($body, $pos, 2) == "/>"){
1140 $gt = sq_findnxstr($body, $pos, ">");
1141 $retary = Array(false, false, false, $lt, $gt);
1145 $attary{$attname} = '"yes"';
1146 return Array($tagname, $attary, $tagtype, $lt, $pos);
1150 * Skip whitespace and see what we arrive at.
1152 $pos = sq_skipspace($body, $pos);
1153 $char = substr($body, $pos, 1);
1155 * Two things are valid here:
1156 * '=' means this is attribute type 1 2 or 3.
1157 * \w means this was attribute type 4.
1158 * anything else we ignore and re-loop. End of tag and
1159 * invalid stuff will be caught by our checks at the beginning
1164 $pos = sq_skipspace($body, $pos);
1166 * Here are 3 possibilities:
1167 * "'" attribute type 1
1168 * '"' attribute type 2
1169 * everything else is the content of tag type 3
1171 $quot = substr($body, $pos, 1);
1173 $regary = sq_findnxreg($body, $pos+
1, "\'");
1174 if ($regary == false){
1175 return Array(false, false, false, $lt, strlen($body));
1177 list($pos, $attval, $match) = $regary;
1179 $attary{$attname} = "'" . $attval . "'";
1180 } else if ($quot == '"'){
1181 $regary = sq_findnxreg($body, $pos+
1, '\"');
1182 if ($regary == false){
1183 return Array(false, false, false, $lt, strlen($body));
1185 list($pos, $attval, $match) = $regary;
1187 $attary{$attname} = '"' . $attval . '"';
1190 * These are hateful. Look for \s, or >.
1192 $regary = sq_findnxreg($body, $pos, "[\s>]");
1193 if ($regary == false){
1194 return Array(false, false, false, $lt, strlen($body));
1196 list($pos, $attval, $match) = $regary;
1198 * If it's ">" it will be caught at the top.
1200 $attval = preg_replace("/\"/s", """, $attval);
1201 $attary{$attname} = '"' . $attval . '"';
1203 } else if (preg_match("|[\w/>]|", $char)) {
1205 * That was attribute type 4.
1207 $attary{$attname} = '"yes"';
1210 * An illegal character. Find next '>' and return.
1212 $gt = sq_findnxstr($body, $pos, ">");
1213 return Array(false, false, false, $lt, $gt);
1219 * The fact that we got here indicates that the tag end was never
1220 * found. Return invalid tag indication so it gets stripped.
1222 return Array(false, false, false, $lt, strlen($body));
1226 * This function checks attribute values for entity-encoded values
1227 * and returns them translated into 8-bit strings so we can run
1230 * @param $attvalue A string to run entity check against.
1231 * @return Translated value.
1233 function sq_deent($attvalue){
1236 * See if we have to run the checks first. All entities must start
1239 if (strpos($attvalue, "&") === false){
1243 * Check named entities first.
1245 $trans = get_html_translation_table(HTML_ENTITIES
);
1247 * Leave " in, as it can mess us up.
1249 $trans = array_flip($trans);
1250 unset($trans{"""});
1251 while (list($ent, $val) = each($trans)){
1252 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1255 * Now translate numbered entities from 1 to 255 if needed.
1257 if (strpos($attvalue, "#") !== false){
1258 $omit = Array(34, 39);
1259 for ($asc=1; $asc<256; $asc++
){
1260 if (!in_array($asc, $omit)){
1262 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1264 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1265 "$chr\\1", $attvalue);
1273 * This function runs various checks against the attributes.
1275 * @param $tagname String with the name of the tag.
1276 * @param $attary Array with all tag attributes.
1277 * @param $rm_attnames See description for sq_sanitize
1278 * @param $bad_attvals See description for sq_sanitize
1279 * @param $add_attr_to_tag See description for sq_sanitize
1280 * @param $message message object
1281 * @param $id message id
1282 * @return Array with modified attributes.
1284 function sq_fixatts($tagname,
1294 while (list($attname, $attvalue) = each($attary)){
1296 * See if this attribute should be removed.
1298 foreach ($rm_attnames as $matchtag=>$matchattrs){
1299 if (preg_match($matchtag, $tagname)){
1300 foreach ($matchattrs as $matchattr){
1301 if (preg_match($matchattr, $attname)){
1302 unset($attary{$attname});
1309 * Remove any entities.
1311 $attvalue = sq_deent($attvalue);
1314 * Now let's run checks on the attvalues.
1315 * I don't expect anyone to comprehend this. If you do,
1316 * get in touch with me so I can drive to where you live and
1317 * shake your hand personally. :)
1319 foreach ($bad_attvals as $matchtag=>$matchattrs){
1320 if (preg_match($matchtag, $tagname)){
1321 foreach ($matchattrs as $matchattr=>$valary){
1322 if (preg_match($matchattr, $attname)){
1324 * There are two arrays in valary.
1326 * Second one is replacements
1328 list($valmatch, $valrepl) = $valary;
1330 preg_replace($valmatch, $valrepl, $attvalue);
1331 if ($newvalue != $attvalue){
1332 $attary{$attname} = $newvalue;
1339 * Turn cid: urls into http-friendly ones.
1341 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1342 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1346 * See if we need to append any attributes to this tag.
1348 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1349 if (preg_match($matchtag, $tagname)){
1350 $attary = array_merge($attary, $addattary);
1357 * This function edits the style definition to make them friendly and
1358 * usable in squirrelmail.
1360 * @param $message the message object
1361 * @param $id the message id
1362 * @param $content a string with whatever is between <style> and </style>
1363 * @return a string with edited content.
1365 function sq_fixstyle($message, $id, $content){
1366 global $view_unsafe_images;
1367 $me = 'sq_fixstyle';
1369 * First look for general BODY style declaration, which would be
1371 * body {background: blah-blah}
1372 * and change it to .bodyclass so we can just assign it to a <div>
1374 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1375 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1377 * Fix url('blah') declarations.
1379 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1380 "url(\\1$secremoveimg\\2)", $content);
1382 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1385 if (!$view_unsafe_images){
1386 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1387 "url(\\1$secremoveimg\\2)", $content);
1391 * Fix urls that refer to cid:
1393 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1395 $cidurl = $matches{1};
1396 $httpurl = sq_cid2http($message, $id, $cidurl);
1397 $content = preg_replace("|url\($cidurl\)|si",
1398 "url($httpurl)", $content);
1402 * Fix stupid css declarations which lead to vulnerabilities
1405 $match = Array('/expression/i',
1408 $replace = Array('idiocy', 'idiocy', 'idiocy');
1409 $content = preg_replace($match, $replace, $content);
1414 * This function converts cid: url's into the ones that can be viewed in
1417 * @param $message the message object
1418 * @param $id the message id
1419 * @param $cidurl the cid: url.
1420 * @return a string with a http-friendly url
1422 function sq_cid2http($message, $id, $cidurl, $mailbox){
1424 * Get rid of quotes.
1426 $quotchar = substr($cidurl, 0, 1);
1427 $cidurl = str_replace($quotchar, "", $cidurl);
1428 $cidurl = substr(trim($cidurl), 4);
1429 $linkurl = find_ent_id($cidurl, $message);
1430 /* in case of non-save cid links $httpurl should be replaced by a sort of
1431 unsave link image */
1434 $httpurl = $quotchar . SM_PATH
. 'src/download.php?absolute_dl=true&' .
1435 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1436 '&ent_id=' . $linkurl . $quotchar;
1442 * This function changes the <body> tag into a <div> tag since we
1443 * can't really have a body-within-body.
1445 * @param $attary an array of attributes and values of <body>
1446 * @return a modified array of attributes to be set for <div>
1448 function sq_body2div($attary){
1449 $me = 'sq_body2div';
1450 $divattary = Array('class' => "'bodyclass'");
1451 $bgcolor = '#ffffff';
1454 if (is_array($attary) && sizeof($attary) > 0){
1455 foreach ($attary as $attname=>$attvalue){
1456 $quotchar = substr($attvalue, 0, 1);
1457 $attvalue = str_replace($quotchar, "", $attvalue);
1460 $styledef .= "background-image: url('$attvalue'); ";
1463 $styledef .= "background-color: $attvalue; ";
1466 $styledef .= "color: $attvalue; ";
1470 if (strlen($styledef) > 0){
1471 $divattary{"style"} = "\"$styledef\"";
1478 * This is the main function and the one you should actually be calling.
1479 * There are several variables you should be aware of an which need
1480 * special description.
1482 * Since the description is quite lengthy, see it here:
1483 * http://www.mricon.com/html/phpfilter.html
1485 * @param $body the string with HTML you wish to filter
1486 * @param $tag_list see description above
1487 * @param $rm_tags_with_content see description above
1488 * @param $self_closing_tags see description above
1489 * @param $force_tag_closing see description above
1490 * @param $rm_attnames see description above
1491 * @param $bad_attvals see description above
1492 * @param $add_attr_to_tag see description above
1493 * @param $message message object
1494 * @param $id message id
1495 * @return sanitized html safe to show on your pages.
1497 function sq_sanitize($body,
1499 $rm_tags_with_content,
1509 $me = 'sq_sanitize';
1511 * Normalize rm_tags and rm_tags_with_content.
1513 @array_walk
($rm_tags, 'sq_casenormalize');
1514 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1515 @array_walk
($self_closing_tags, 'sq_casenormalize');
1517 * See if tag_list is of tags to remove or tags to allow.
1518 * false means remove these tags
1519 * true means allow these tags
1521 $rm_tags = array_shift($tag_list);
1523 $open_tags = Array();
1524 $trusted = "<!-- begin sanitized html -->\n";
1525 $skip_content = false;
1527 * Take care of netscape's stupid javascript entities like
1530 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1532 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1533 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1534 $free_content = substr($body, $curpos, $lt-$curpos);
1536 * Take care of <style>
1538 if ($tagname == "style" && $tagtype == 2){
1540 * This is a closing </style>. Edit the
1541 * content before we apply it.
1543 $free_content = sq_fixstyle($message, $id, $free_content);
1545 if ($skip_content == false){
1546 $trusted .= $free_content;
1548 if ($tagname != FALSE){
1550 if ($skip_content == $tagname){
1552 * Got to the end of tag we needed to remove.
1555 $skip_content = false;
1557 if ($skip_content == false){
1558 if ($tagname == "body"){
1561 if (isset($open_tags{$tagname}) &&
1562 $open_tags{$tagname} > 0){
1563 $open_tags{$tagname}--;
1572 * $rm_tags_with_content
1574 if ($skip_content == false){
1576 * See if this is a self-closing type and change
1577 * tagtype appropriately.
1580 && in_array($tagname, $self_closing_tags)){
1584 * See if we should skip this tag and any content
1587 if ($tagtype == 1 &&
1588 in_array($tagname, $rm_tags_with_content)){
1589 $skip_content = $tagname;
1591 if (($rm_tags == false
1592 && in_array($tagname, $tag_list)) ||
1593 ($rm_tags == true &&
1594 !in_array($tagname, $tag_list))){
1598 if (isset($open_tags{$tagname})){
1599 $open_tags{$tagname}++
;
1601 $open_tags{$tagname}=1;
1605 * This is where we run other checks.
1607 if (is_array($attary) && sizeof($attary) > 0){
1608 $attary = sq_fixatts($tagname,
1619 * Convert body into div.
1621 if ($tagname == "body"){
1623 $attary = sq_body2div($attary, $message, $id);
1629 if ($tagname != false && $skip_content == false){
1630 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1635 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1636 if ($force_tag_closing == true){
1637 foreach ($open_tags as $tagname=>$opentimes){
1638 while ($opentimes > 0){
1639 $trusted .= '</' . $tagname . '>';
1645 $trusted .= "<!-- end sanitized html -->\n";
1650 * This is a wrapper function to call html sanitizing routines.
1652 * @param $body the body of the message
1653 * @param $id the id of the message
1654 * @return a string with html safe to display in the browser.
1656 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1657 global $attachment_common_show_images, $view_unsafe_images,
1660 * Don't display attached images in HTML mode.
1662 $attachment_common_show_images = false;
1675 $rm_tags_with_content = Array(
1682 $self_closing_tags = Array(
1689 $force_tag_closing = false;
1691 $rm_attnames = Array(
1702 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1703 $bad_attvals = Array(
1706 "/^src|background/i" =>
1709 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1710 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1711 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1712 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1715 "\\1$secremoveimg\\2",
1716 "\\1$secremoveimg\\2",
1717 "\\1$secremoveimg\\2",
1718 "\\1$secremoveimg\\2"
1721 "/^href|action/i" =>
1724 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1725 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1726 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1727 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1742 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1743 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1744 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1745 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1759 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1760 $view_unsafe_images = false;
1762 if (!$view_unsafe_images){
1764 * Remove any references to http/https if view_unsafe_images set
1767 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1768 '/^([\'\"])\s*https*:.*([\'\"])/si');
1769 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1770 "\\1$secremoveimg\\2");
1771 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1772 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1773 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1774 "url(\\1$secremoveimg\\2)");
1777 $add_attr_to_tag = Array(
1778 "/^a$/i" => Array('target'=>'"_new"')
1780 $trusted = sq_sanitize($body,
1782 $rm_tags_with_content,
1792 if (preg_match("|$secremoveimg|i", $trusted)){
1793 $has_unsafe_images = true;