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.
103 $cmd = "FETCH $id BODY[]";
105 $cmd = "FETCH $id BODY[$ent_id]";
108 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
110 $topline = trim(array_shift($data));
111 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
113 $wholemessage = implode('', $data);
114 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
115 $ret = substr($wholemessage, 0, $regs[1]);
116 /* There is some information in the content info header that could be important
117 * in order to parse html messages. Let's get them here.
119 // if ($ret{0} == '<') {
120 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
122 } else if (ereg('"([^"]*)"', $topline, $regs)) {
125 global $where, $what, $mailbox, $passed_id, $startMessage;
126 $par = 'mailbox=' . urlencode($mailbox) . '&passed_id=' . $passed_id;
127 if (isset($where) && isset($what)) {
128 $par .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
130 $par .= '&startMessage=' . $startMessage . '&show_more=0';
132 $par .= '&response=' . urlencode($response) .
133 '&message=' . urlencode($message) .
134 '&topline=' . urlencode($topline);
137 '<table width="80%"><tr>' .
138 '<tr><td colspan=2>' .
139 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
141 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
142 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
143 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
144 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
145 "</table><BR></tt></font><hr>";
147 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
149 $wholemessage = implode('', $data);
151 $ret = $wholemessage;
156 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
159 $sid = sqimap_session_id($uid_support);
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')) {
169 $sid_s = substr($sid,0,strpos($sid, ' '));
174 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
175 echo decodeBody($body, $encoding);
178 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
181 $read = fgets ($imap_stream,8192);
184 // This could be bad -- if the section has sqimap_session_id() . ' OK'
185 // or similar, it will kill the download.
186 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
187 if (trim($read) == ')==') {
189 $read = fgets ($imap_stream,4096);
190 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
193 echo decodeBody($read1, $encoding) .
194 decodeBody($read, $encoding);
197 echo decodeBody($read, $encoding);
199 $read = fgets ($imap_stream,4096);
206 /* -[ END MIME DECODING ]----------------------------------------------------------- */
208 /* This is here for debugging purposes. It will print out a list
209 * of all the entity IDs that are in the $message object.
211 function listEntities ($message) {
213 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
214 for ($i = 0; isset($message->entities
[$i]); $i++
) {
216 $msg = listEntities($message->entities
[$i]);
226 function getPriorityStr($priority) {
227 $priority_level = substr($priority,0,1);
229 switch($priority_level) {
230 /* Check for a higher then normal priority. */
233 $priority_string = _("High");
236 /* Check for a lower then normal priority. */
239 $priority_string = _("Low");
242 /* Check for a normal priority. */
245 $priority_level = '3';
246 $priority_string = _("Normal");
250 return $priority_string;
253 /* returns a $message object for a particular entity id */
254 function getEntity ($message, $ent_id) {
255 return $message->getEntity($ent_id);
259 * Extracted from strings.php 23/03/2002
262 function translateText(&$body, $wrap_at, $charset) {
263 global $where, $what; /* from searching */
264 global $color; /* color theme */
266 require_once(SM_PATH
. 'functions/url_parser.php');
268 $body_ary = explode("\n", $body);
269 for ($i=0; $i < count($body_ary); $i++
) {
270 $line = $body_ary[$i];
271 if (strlen($line) - 2 >= $wrap_at) {
272 sqWordWrap($line, $wrap_at);
274 $line = charset_decode($charset, $line);
275 $line = str_replace("\t", ' ', $line);
284 if ($line[$pos] == ' ') {
286 } else if (strpos($line, '>', $pos) === $pos) {
295 if (!isset($color[13])) {
296 $color[13] = '#800000';
298 $line = '<font color="' . $color[13] . '">' . $line . '</font>';
300 if (!isset($color[14])) {
301 $color[14] = '#FF0000';
303 $line = '<font color="' . $color[14] . '">' . $line . '</font>';
306 $body_ary[$i] = $line;
308 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
311 /* This returns a parsed string called $body. That string can then
312 * be displayed as the actual message in the HTML. It contains
313 * everything needed, including HTML Tags, Attachments at the
316 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
317 /* This if statement checks for the entity to show as the
318 * primary message. To add more of them, just put them in the
319 * order that is their priority.
321 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
322 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
323 global $languages, $squirrelmail_language;
325 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
326 $view_unsafe_images = false;
330 $urlmailbox = urlencode($mailbox);
331 $body_message = getEntity($message, $ent_num);
332 if (($body_message->header
->type0
== 'text') ||
333 ($body_message->header
->type0
== 'rfc822')) {
334 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
335 $body = decodeBody($body, $body_message->header
->encoding
);
337 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
338 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
339 if (mb_detect_encoding($body) != 'ASCII') {
340 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
343 $hookResults = do_hook("message_body", $body);
344 $body = $hookResults[1];
346 /* If there are other types that shouldn't be formatted, add
350 if ($body_message->header
->type1
== 'html') {
351 if ($show_html_default <> 1) {
352 $entity_conv = array(' ' => ' ',
359 $body = strtr($body, $entity_conv);
360 $body = strip_tags($body);
362 translateText($body, $wrap_at,
363 $body_message->header
->getParameter('charset'));
365 $body = magicHTML($body, $id, $message, $mailbox);
368 translateText($body, $wrap_at,
369 $body_message->header
->getParameter('charset'));
371 $link = 'read_body.php?passed_id=' . $id . '&ent_id='.$ent_num.
372 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
373 '&startMessage=' . $startMessage . '&show_more=0';
374 if (isset($passed_ent_id)) {
375 $link .= '&passed_ent_id='.$passed_ent_id;
377 if ($view_unsafe_images) {
378 $text = _("Hide Unsafe Images");
380 if (isset($has_unsafe_images) && $has_unsafe_images) {
381 $link .= '&view_unsafe_images=1';
382 $text = _("View Unsafe Images");
387 $body .= '<center><small><a href="'.$link.'">'.$text.
388 '</a></small></center><br>' . "\n";
394 function formatAttachments($message, $exclude_id, $mailbox, $id) {
395 global $where, $what, $startMessage, $color;
396 static $ShownHTML = 0;
398 $att_ar = $message->getAttachments($exclude_id);
400 if (!count($att_ar)) return '';
404 $urlMailbox = urlencode($mailbox);
406 foreach ($att_ar as $att) {
407 $ent = $att->entity_id
;
408 $header = $att->header
;
409 $type0 = strtolower($header->type0
);
410 $type1 = strtolower($header->type1
);
412 $links['download link']['text'] = _("download");
413 $links['download link']['href'] =
414 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
416 if ($type0 =='message' && $type1 == 'rfc822') {
417 $default_page = '../src/read_body.php';
418 $rfc822_header = $att->rfc822_header
;
419 $filename = $rfc822_header->subject
;
420 if (trim( $filename ) == '') {
421 $filename = 'untitled-[' . $ent . ']' ;
423 $from_o = $rfc822_header->from
;
424 if (is_object($from_o)) {
425 $from_name = $from_o->getAddress(false);
427 $from_name = _("Unknown sender");
429 $from_name = decodeHeader(($from_name));
430 $description = $from_name;
432 $default_page = '../src/download.php';
433 if (is_object($header->disposition
)) {
434 $filename = $header->disposition
->getProperty('filename');
435 if (trim($filename) == '') {
436 $name = decodeHeader($header->disposition
->getProperty('name'));
437 if (trim($name) == '') {
438 $name = $header->getParameter('name');
439 if(trim($name) == '') {
440 if (trim( $header->id
) == '') {
441 $filename = 'untitled-[' . $ent . ']' ;
443 $filename = 'cid: ' . $header->id
;
453 $filename = $header->getParameter('name');
454 if (!trim($filename)) {
455 if (trim( $header->id
) == '') {
456 $filename = 'untitled-[' . $ent . ']' ;
458 $filename = 'cid: ' . $header->id
;
462 if ($header->description
) {
463 $description = decodeHeader($header->description
);
469 $display_filename = $filename;
470 if (isset($passed_ent_id)) {
471 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
473 $passed_ent_id_link = '';
475 $defaultlink = $default_page . "?startMessage=$startMessage"
476 . "&passed_id=$id&mailbox=$urlMailbox"
477 . '&ent_id='.$ent.$passed_ent_id_link.'&absolute_dl=true';
478 if ($where && $what) {
479 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
481 /* This executes the attachment hook with a specific MIME-type.
482 * If that doesn't have results, it tries if there's a rule
483 * for a more generic type.
485 $hookresults = do_hook("attachment $type0/$type1", $links,
486 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
487 $display_filename, $where, $what);
488 if(count($hookresults[1]) <= 1) {
489 $hookresults = do_hook("attachment $type0/*", $links,
490 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
491 $display_filename, $where, $what);
494 $links = $hookresults[1];
495 $defaultlink = $hookresults[6];
497 $attachments .= '<TR><TD>' .
498 '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A> </TD>' .
499 '<TD><SMALL><b>' . show_readable_size($header->size
) .
500 '</b> </small></TD>' .
501 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
503 $attachments .= '<b>' . $description . '</b>';
504 $attachments .= '</SMALL></TD><TD><SMALL> ';
507 foreach ($links as $val) {
511 $attachments .= ' | ';
513 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
516 $attachments .= "</TD></TR>\n";
521 /* This function decodes the body depending on the encoding type. */
522 function decodeBody($body, $encoding) {
523 global $show_html_default;
525 $body = str_replace("\r\n", "\n", $body);
526 $encoding = strtolower($encoding);
528 if ($encoding == 'quoted-printable' ||
529 $encoding == 'quoted_printable') {
530 $body = quoted_printable_decode($body);
532 while (ereg("=\n", $body)) {
533 $body = ereg_replace ("=\n", '', $body);
536 } else if ($encoding == 'base64') {
537 $body = base64_decode($body);
540 // All other encodings are returned raw.
545 * This functions decode strings that is encoded according to
546 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
547 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
549 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
550 global $languages, $squirrelmail_language;
551 if (is_array($string)) {
552 $string = implode("\n", $string);
555 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
556 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
557 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
558 // Do we need to return at this point?
562 $aString = explode(' ',$string);
563 foreach ($aString as $chunk) {
566 while (preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
567 $aString[$i] .= $res[1];
568 //echo "$chunk match ". $res[5] . "<br>";
569 $encoding = ucfirst($res[3]);
573 $replace = base64_decode($res[4]);
574 $aString[$i] .= charset_decode($res[2],$replace);
577 $replace = str_replace('_', ' ', $res[4]);
578 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
580 /* Only encode into entities by default. Some places
581 * don't need the encoding, like the compose form.
584 $replace = charset_decode($res[2], $replace);
587 $replace = htmlspecialchars($replace);
590 $aString[$i] .= $replace;
598 if (!$encoded && $htmlsave) {
599 $aString[$i] = htmlspecialchars($chunk);
601 $aString[$i] .= $chunk;
605 return implode (' ',$aString);
610 * Encode a string according to RFC 1522 for use in headers if it
611 * contains 8-bit characters or anything that looks like it should
614 function encodeHeader ($string) {
615 global $default_charset, $languages, $squirrelmail_language;
617 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
618 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
619 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
621 if (strtolower($default_charset) == 'iso-8859-1') {
622 $string = str_replace("\240",' ',$string);
625 // Encode only if the string contains 8-bit characters or =?
626 $j = strlen($string);
627 $max_l = 75 - strlen($default_charset) - 7;
630 $iEncStart = $enc_init = false;
631 $cur_l = $iOffset = 0;
632 for($i = 0; $i < $j; ++
$i) {
641 if ($iEncStart === false) {
645 if ($cur_l > ($max_l-2)) {
646 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
647 $aRet[] = "=?$default_charset?Q?$ret?=";
653 $ret .= sprintf("=%02X",ord($string{$i}));
658 if ($iEncStart !== false) {
659 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
660 $aRet[] = "=?$default_charset?Q?$ret?=";
668 if ($iEncStart !== false) {
670 if ($cur_l > $max_l) {
671 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
672 $aRet[] = "=?$default_charset?Q?$ret?=";
683 $k = ord($string{$i});
685 if ($iEncStart === false) {
689 if ($cur_l > ($max_l-2)) {
690 if ($iEncStart !== false) {
691 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
692 $aRet[] = "=?$default_charset?Q?$ret?=";
694 $aRet[] = substr($string,$iOffset,$i-$iOffset);
701 $ret .= sprintf("=%02X", $k);
703 if ($iEncStart !== false) {
705 if ($cur_l > $max_l) {
706 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
707 $aRet[] = "=?$default_charset?Q?$ret?=";
722 if ($iEncStart !== false) {
723 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
724 $aRet[] = "=?$default_charset?Q?$ret?=";
726 $aRet[] = substr($string,$iOffset);
728 $string = implode('',$aRet);
733 /* This function trys to locate the entity_id of a specific mime element */
734 function find_ent_id($id, $message) {
735 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
736 if ($message->entities
[$i]->header
->type0
== 'multipart') {
737 $ret = find_ent_id($id, $message->entities
[$i]);
739 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
740 // if (sq_check_save_extension($message->entities[$i])) {
741 return $message->entities
[$i]->entity_id
;
749 function sq_check_save_extension($message) {
750 $filename = $message->getFilename();
751 $ext = substr($filename, strrpos($filename,'.')+
1);
752 $save_extensions = array('jpg','jpeg','gif','png','bmp');
753 return in_array($ext, $save_extensions);
758 ** HTMLFILTER ROUTINES
762 * This function returns the final tag out of the tag name, an array
763 * of attributes, and the type of the tag. This function is called by
764 * sq_sanitize internally.
766 * @param $tagname the name of the tag.
767 * @param $attary the array of attributes and their values
768 * @param $tagtype The type of the tag (see in comments).
769 * @return a string with the final tag representation.
771 function sq_tagprint($tagname, $attary, $tagtype){
775 $fulltag = '</' . $tagname . '>';
777 $fulltag = '<' . $tagname;
778 if (is_array($attary) && sizeof($attary)){
780 while (list($attname, $attvalue) = each($attary)){
781 array_push($atts, "$attname=$attvalue");
783 $fulltag .= ' ' . join(" ", $atts);
794 * A small helper function to use with array_walk. Modifies a by-ref
795 * value and makes it lowercase.
797 * @param $val a value passed by-ref.
798 * @return void since it modifies a by-ref value.
800 function sq_casenormalize(&$val){
801 $val = strtolower($val);
805 * This function skips any whitespace from the current position within
806 * a string and to the next non-whitespace value.
808 * @param $body the string
809 * @param $offset the offset within the string where we should start
810 * looking for the next non-whitespace character.
811 * @return the location within the $body where the next
812 * non-whitespace char is located.
814 function sq_skipspace($body, $offset){
815 $me = 'sq_skipspace';
816 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
817 if (sizeof($matches{1})){
818 $count = strlen($matches{1});
825 * This function looks for the next character within a string. It's
826 * really just a glorified "strpos", except it catches if failures
829 * @param $body The string to look for needle in.
830 * @param $offset Start looking from this position.
831 * @param $needle The character/string to look for.
832 * @return location of the next occurance of the needle, or
833 * strlen($body) if needle wasn't found.
835 function sq_findnxstr($body, $offset, $needle){
836 $me = 'sq_findnxstr';
837 $pos = strpos($body, $needle, $offset);
839 $pos = strlen($body);
845 * This function takes a PCRE-style regexp and tries to match it
848 * @param $body The string to look for needle in.
849 * @param $offset Start looking from here.
850 * @param $reg A PCRE-style regex to match.
851 * @return Returns a false if no matches found, or an array
852 * with the following members:
853 * - integer with the location of the match within $body
854 * - string with whatever content between offset and the match
855 * - string with whatever it is we matched
857 function sq_findnxreg($body, $offset, $reg){
858 $me = 'sq_findnxreg';
861 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
865 $retarr{0} = $offset +
strlen($matches{1});
866 $retarr{1} = $matches{1};
867 $retarr{2} = $matches{2};
873 * This function looks for the next tag.
875 * @param $body String where to look for the next tag.
876 * @param $offset Start looking from here.
877 * @return false if no more tags exist in the body, or
878 * an array with the following members:
879 * - string with the name of the tag
880 * - array with attributes and their values
881 * - integer with tag type (1, 2, or 3)
882 * - integer where the tag starts (starting "<")
883 * - integer where the tag ends (ending ">")
884 * first three members will be false, if the tag is invalid.
886 function sq_getnxtag($body, $offset){
888 if ($offset > strlen($body)){
891 $lt = sq_findnxstr($body, $offset, "<");
892 if ($lt == strlen($body)){
897 * blah blah <tag attribute="value">
900 $pos = sq_skipspace($body, $lt+
1);
901 if ($pos >= strlen($body)){
902 return Array(false, false, false, $lt, strlen($body));
905 * There are 3 kinds of tags:
906 * 1. Opening tag, e.g.:
908 * 2. Closing tag, e.g.:
910 * 3. XHTML-style content-less tag, e.g.:
914 switch (substr($body, $pos, 1)){
921 * A comment or an SGML declaration.
923 if (substr($body, $pos+
1, 2) == "--"){
924 $gt = strpos($body, "-->", $pos);
930 return Array(false, false, false, $lt, $gt);
932 $gt = sq_findnxstr($body, $pos, ">");
933 return Array(false, false, false, $lt, $gt);
938 * Assume tagtype 1 for now. If it's type 3, we'll switch values
948 * Look for next [\W-_], which will indicate the end of the tag name.
950 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
951 if ($regary == false){
952 return Array(false, false, false, $lt, strlen($body));
954 list($pos, $tagname, $match) = $regary;
955 $tagname = strtolower($tagname);
958 * $match can be either of these:
959 * '>' indicating the end of the tag entirely.
960 * '\s' indicating the end of the tag name.
961 * '/' indicating that this is type-3 xhtml tag.
963 * Whatever else we find there indicates an invalid tag.
968 * This is an xhtml-style tag with a closing / at the
969 * end, like so: <img src="blah"/>. Check if it's followed
970 * by the closing bracket. If not, then this tag is invalid
972 if (substr($body, $pos, 2) == "/>"){
976 $gt = sq_findnxstr($body, $pos, ">");
977 $retary = Array(false, false, false, $lt, $gt);
981 return Array($tagname, false, $tagtype, $lt, $pos);
985 * Check if it's whitespace
987 if (!preg_match('/\s/', $match)){
989 * This is an invalid tag! Look for the next closing ">".
991 $gt = sq_findnxstr($body, $offset, ">");
992 return Array(false, false, false, $lt, $gt);
998 * At this point we're here:
999 * <tagname attribute='blah'>
1002 * At this point we loop in order to find all attributes.
1008 while ($pos <= strlen($body)){
1009 $pos = sq_skipspace($body, $pos);
1010 if ($pos == strlen($body)){
1014 return Array(false, false, false, $lt, $pos);
1017 * See if we arrived at a ">" or "/>", which means that we reached
1018 * the end of the tag.
1021 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1025 $pos +
= strlen($matches{1});
1026 if ($matches{2} == "/>"){
1030 return Array($tagname, $attary, $tagtype, $lt, $pos);
1034 * There are several types of attributes, with optional
1035 * [:space:] between members.
1037 * attrname[:space:]=[:space:]'CDATA'
1039 * attrname[:space:]=[:space:]"CDATA"
1041 * attr[:space:]=[:space:]CDATA
1045 * We leave types 1 and 2 the same, type 3 we check for
1046 * '"' and convert to """ if needed, then wrap in
1047 * double quotes. Type 4 we convert into:
1050 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1051 if ($regary == false){
1053 * Looks like body ended before the end of tag.
1055 return Array(false, false, false, $lt, strlen($body));
1057 list($pos, $attname, $match) = $regary;
1058 $attname = strtolower($attname);
1060 * We arrived at the end of attribute name. Several things possible
1062 * '>' means the end of the tag and this is attribute type 4
1063 * '/' if followed by '>' means the same thing as above
1064 * '\s' means a lot of things -- look what it's followed by.
1065 * anything else means the attribute is invalid.
1070 * This is an xhtml-style tag with a closing / at the
1071 * end, like so: <img src="blah"/>. Check if it's followed
1072 * by the closing bracket. If not, then this tag is invalid
1074 if (substr($body, $pos, 2) == "/>"){
1078 $gt = sq_findnxstr($body, $pos, ">");
1079 $retary = Array(false, false, false, $lt, $gt);
1083 $attary{$attname} = '"yes"';
1084 return Array($tagname, $attary, $tagtype, $lt, $pos);
1088 * Skip whitespace and see what we arrive at.
1090 $pos = sq_skipspace($body, $pos);
1091 $char = substr($body, $pos, 1);
1093 * Two things are valid here:
1094 * '=' means this is attribute type 1 2 or 3.
1095 * \w means this was attribute type 4.
1096 * anything else we ignore and re-loop. End of tag and
1097 * invalid stuff will be caught by our checks at the beginning
1102 $pos = sq_skipspace($body, $pos);
1104 * Here are 3 possibilities:
1105 * "'" attribute type 1
1106 * '"' attribute type 2
1107 * everything else is the content of tag type 3
1109 $quot = substr($body, $pos, 1);
1111 $regary = sq_findnxreg($body, $pos+
1, "\'");
1112 if ($regary == false){
1113 return Array(false, false, false, $lt, strlen($body));
1115 list($pos, $attval, $match) = $regary;
1117 $attary{$attname} = "'" . $attval . "'";
1118 } else if ($quot == '"'){
1119 $regary = sq_findnxreg($body, $pos+
1, '\"');
1120 if ($regary == false){
1121 return Array(false, false, false, $lt, strlen($body));
1123 list($pos, $attval, $match) = $regary;
1125 $attary{$attname} = '"' . $attval . '"';
1128 * These are hateful. Look for \s, or >.
1130 $regary = sq_findnxreg($body, $pos, "[\s>]");
1131 if ($regary == false){
1132 return Array(false, false, false, $lt, strlen($body));
1134 list($pos, $attval, $match) = $regary;
1136 * If it's ">" it will be caught at the top.
1138 $attval = preg_replace("/\"/s", """, $attval);
1139 $attary{$attname} = '"' . $attval . '"';
1141 } else if (preg_match("|[\w/>]|", $char)) {
1143 * That was attribute type 4.
1145 $attary{$attname} = '"yes"';
1148 * An illegal character. Find next '>' and return.
1150 $gt = sq_findnxstr($body, $pos, ">");
1151 return Array(false, false, false, $lt, $gt);
1157 * The fact that we got here indicates that the tag end was never
1158 * found. Return invalid tag indication so it gets stripped.
1160 return Array(false, false, false, $lt, strlen($body));
1164 * This function checks attribute values for entity-encoded values
1165 * and returns them translated into 8-bit strings so we can run
1168 * @param $attvalue A string to run entity check against.
1169 * @return Translated value.
1171 function sq_deent($attvalue){
1174 * See if we have to run the checks first. All entities must start
1177 if (strpos($attvalue, "&") === false){
1181 * Check named entities first.
1183 $trans = get_html_translation_table(HTML_ENTITIES
);
1185 * Leave " in, as it can mess us up.
1187 $trans = array_flip($trans);
1188 unset($trans{"""});
1189 while (list($ent, $val) = each($trans)){
1190 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1193 * Now translate numbered entities from 1 to 255 if needed.
1195 if (strpos($attvalue, "#") !== false){
1196 $omit = Array(34, 39);
1197 for ($asc=1; $asc<256; $asc++
){
1198 if (!in_array($asc, $omit)){
1200 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1202 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1203 "$chr\\1", $attvalue);
1211 * This function runs various checks against the attributes.
1213 * @param $tagname String with the name of the tag.
1214 * @param $attary Array with all tag attributes.
1215 * @param $rm_attnames See description for sq_sanitize
1216 * @param $bad_attvals See description for sq_sanitize
1217 * @param $add_attr_to_tag See description for sq_sanitize
1218 * @param $message message object
1219 * @param $id message id
1220 * @return Array with modified attributes.
1222 function sq_fixatts($tagname,
1232 while (list($attname, $attvalue) = each($attary)){
1234 * See if this attribute should be removed.
1236 foreach ($rm_attnames as $matchtag=>$matchattrs){
1237 if (preg_match($matchtag, $tagname)){
1238 foreach ($matchattrs as $matchattr){
1239 if (preg_match($matchattr, $attname)){
1240 unset($attary{$attname});
1247 * Remove any entities.
1249 $attvalue = sq_deent($attvalue);
1252 * Now let's run checks on the attvalues.
1253 * I don't expect anyone to comprehend this. If you do,
1254 * get in touch with me so I can drive to where you live and
1255 * shake your hand personally. :)
1257 foreach ($bad_attvals as $matchtag=>$matchattrs){
1258 if (preg_match($matchtag, $tagname)){
1259 foreach ($matchattrs as $matchattr=>$valary){
1260 if (preg_match($matchattr, $attname)){
1262 * There are two arrays in valary.
1264 * Second one is replacements
1266 list($valmatch, $valrepl) = $valary;
1268 preg_replace($valmatch, $valrepl, $attvalue);
1269 if ($newvalue != $attvalue){
1270 $attary{$attname} = $newvalue;
1277 * Turn cid: urls into http-friendly ones.
1279 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1280 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1284 * See if we need to append any attributes to this tag.
1286 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1287 if (preg_match($matchtag, $tagname)){
1288 $attary = array_merge($attary, $addattary);
1295 * This function edits the style definition to make them friendly and
1296 * usable in squirrelmail.
1298 * @param $message the message object
1299 * @param $id the message id
1300 * @param $content a string with whatever is between <style> and </style>
1301 * @return a string with edited content.
1303 function sq_fixstyle($message, $id, $content){
1304 global $view_unsafe_images;
1305 $me = 'sq_fixstyle';
1307 * First look for general BODY style declaration, which would be
1309 * body {background: blah-blah}
1310 * and change it to .bodyclass so we can just assign it to a <div>
1312 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1313 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1315 * Fix url('blah') declarations.
1317 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1318 "url(\\1$secremoveimg\\2)", $content);
1320 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1323 if (!$view_unsafe_images){
1324 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1325 "url(\\1$secremoveimg\\2)", $content);
1329 * Fix urls that refer to cid:
1331 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1333 $cidurl = $matches{1};
1334 $httpurl = sq_cid2http($message, $id, $cidurl);
1335 $content = preg_replace("|url\($cidurl\)|si",
1336 "url($httpurl)", $content);
1340 * Fix stupid css declarations which lead to vulnerabilities
1343 $match = Array('/expression/si',
1346 $replace = Array('idiocy', 'idiocy', 'idiocy');
1347 $content = preg_replace($match, $replace, $content);
1352 * This function converts cid: url's into the ones that can be viewed in
1355 * @param $message the message object
1356 * @param $id the message id
1357 * @param $cidurl the cid: url.
1358 * @return a string with a http-friendly url
1360 function sq_cid2http($message, $id, $cidurl, $mailbox){
1362 * Get rid of quotes.
1364 $quotchar = substr($cidurl, 0, 1);
1365 $cidurl = str_replace($quotchar, "", $cidurl);
1366 $cidurl = substr(trim($cidurl), 4);
1367 $linkurl = find_ent_id($cidurl, $message);
1368 /* in case of non-save cid links $httpurl should be replaced by a sort of
1369 unsave link image */
1372 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&' .
1373 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1374 '&ent_id=' . $linkurl . $quotchar;
1380 * This function changes the <body> tag into a <div> tag since we
1381 * can't really have a body-within-body.
1383 * @param $attary an array of attributes and values of <body>
1384 * @return a modified array of attributes to be set for <div>
1386 function sq_body2div($attary){
1387 $me = 'sq_body2div';
1388 $divattary = Array('class' => "'bodyclass'");
1389 $bgcolor = '#ffffff';
1392 if (is_array($attary) && sizeof($attary) > 0){
1393 foreach ($attary as $attname=>$attvalue){
1394 $quotchar = substr($attvalue, 0, 1);
1395 $attvalue = str_replace($quotchar, "", $attvalue);
1398 $styledef .= "background-image: url('$attvalue'); ";
1401 $styledef .= "background-color: $attvalue; ";
1404 $styledef .= "color: $attvalue; ";
1408 if (strlen($styledef) > 0){
1409 $divattary{"style"} = "\"$styledef\"";
1416 * This is the main function and the one you should actually be calling.
1417 * There are several variables you should be aware of an which need
1418 * special description.
1420 * Since the description is quite lengthy, see it here:
1421 * http://www.mricon.com/html/phpfilter.html
1423 * @param $body the string with HTML you wish to filter
1424 * @param $tag_list see description above
1425 * @param $rm_tags_with_content see description above
1426 * @param $self_closing_tags see description above
1427 * @param $force_tag_closing see description above
1428 * @param $rm_attnames see description above
1429 * @param $bad_attvals see description above
1430 * @param $add_attr_to_tag see description above
1431 * @param $message message object
1432 * @param $id message id
1433 * @return sanitized html safe to show on your pages.
1435 function sq_sanitize($body,
1437 $rm_tags_with_content,
1447 $me = 'sq_sanitize';
1449 * Normalize rm_tags and rm_tags_with_content.
1451 @array_walk
($rm_tags, 'sq_casenormalize');
1452 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1453 @array_walk
($self_closing_tags, 'sq_casenormalize');
1455 * See if tag_list is of tags to remove or tags to allow.
1456 * false means remove these tags
1457 * true means allow these tags
1459 $rm_tags = array_shift($tag_list);
1461 $open_tags = Array();
1462 $trusted = "<!-- begin sanitized html -->\n";
1463 $skip_content = false;
1465 * Take care of netscape's stupid javascript entities like
1468 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1470 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1471 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1472 $free_content = substr($body, $curpos, $lt-$curpos);
1474 * Take care of <style>
1476 if ($tagname == "style" && $tagtype == 2){
1478 * This is a closing </style>. Edit the
1479 * content before we apply it.
1481 $free_content = sq_fixstyle($message, $id, $free_content);
1483 if ($skip_content == false){
1484 $trusted .= $free_content;
1486 if ($tagname != FALSE){
1488 if ($skip_content == $tagname){
1490 * Got to the end of tag we needed to remove.
1493 $skip_content = false;
1495 if ($skip_content == false){
1496 if ($tagname == "body"){
1499 if (isset($open_tags{$tagname}) &&
1500 $open_tags{$tagname} > 0){
1501 $open_tags{$tagname}--;
1510 * $rm_tags_with_content
1512 if ($skip_content == false){
1514 * See if this is a self-closing type and change
1515 * tagtype appropriately.
1518 && in_array($tagname, $self_closing_tags)){
1522 * See if we should skip this tag and any content
1525 if ($tagtype == 1 &&
1526 in_array($tagname, $rm_tags_with_content)){
1527 $skip_content = $tagname;
1529 if (($rm_tags == false
1530 && in_array($tagname, $tag_list)) ||
1531 ($rm_tags == true &&
1532 !in_array($tagname, $tag_list))){
1536 if (isset($open_tags{$tagname})){
1537 $open_tags{$tagname}++
;
1539 $open_tags{$tagname}=1;
1543 * This is where we run other checks.
1545 if (is_array($attary) && sizeof($attary) > 0){
1546 $attary = sq_fixatts($tagname,
1557 * Convert body into div.
1559 if ($tagname == "body"){
1561 $attary = sq_body2div($attary, $message, $id);
1567 if ($tagname != false && $skip_content == false){
1568 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1573 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1574 if ($force_tag_closing == true){
1575 foreach ($open_tags as $tagname=>$opentimes){
1576 while ($opentimes > 0){
1577 $trusted .= '</' . $tagname . '>';
1583 $trusted .= "<!-- end sanitized html -->\n";
1588 * This is a wrapper function to call html sanitizing routines.
1590 * @param $body the body of the message
1591 * @param $id the id of the message
1592 * @return a string with html safe to display in the browser.
1594 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1595 global $attachment_common_show_images, $view_unsafe_images,
1598 * Don't display attached images in HTML mode.
1600 $attachment_common_show_images = false;
1613 $rm_tags_with_content = Array(
1620 $self_closing_tags = Array(
1627 $force_tag_closing = false;
1629 $rm_attnames = Array(
1640 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1641 $bad_attvals = Array(
1644 "/^src|background/i" =>
1647 "|^([\'\"])\s*\.\./.*([\'\"])|i",
1648 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/i",
1649 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/i",
1650 "/^([\'\"])\s*about\s*:.*([\'\"])/i"
1653 "\\1$secremoveimg\\2",
1654 "\\1$secremoveimg\\2",
1655 "\\1$secremoveimg\\2",
1656 "\\1$secremoveimg\\2"
1659 "/^href|action/i" =>
1662 "|^([\'\"])\s*\.\./.*([\'\"])|i",
1663 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/i",
1664 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/i",
1665 "/^([\'\"])\s*about\s*:.*([\'\"])/i"
1680 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|i",
1681 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/i",
1682 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/i",
1683 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/i"
1697 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1698 $view_unsafe_images = false;
1700 if (!$view_unsafe_images){
1702 * Remove any references to http/https if view_unsafe_images set
1705 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1706 '/^([\'\"])\s*https*:.*([\'\"])/i');
1707 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1708 "\\1$secremoveimg\\2");
1709 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1710 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1711 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1712 "url(\\1$secremoveimg\\2)");
1715 $add_attr_to_tag = Array(
1716 "/^a$/si" => Array('target'=>'"_new"')
1718 $trusted = sq_sanitize($body,
1720 $rm_tags_with_content,
1730 if (preg_match("|$secremoveimg|si", $trusted)){
1731 $has_unsafe_images = true;