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 displayPageHeader( $color, urldecode($mailbox) );
38 echo "<BODY TEXT=\"$color[8]\" BGCOLOR=\"$color[4]\" LINK=\"$color[7]\" VLINK=\"$color[7]\" ALINK=\"$color[7]\">\n\n" .
40 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
41 $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
42 $errormessage .= '<table><tr><td>' . htmlspecialchars($read) . '</td></tr></table>';
43 plain_error_message( $errormessage, $color );
44 echo '</body></html>';
48 foreach ($flags as $flag) {
49 $char = strtoupper($flag{1});
52 if (strtolower($flag) == '\\seen') {
57 if (strtolower($flag) == '\\answered') {
58 $msg->is_answered
= true;
62 if (strtolower($flag) == '\\deleted') {
63 $msg->is_deleted
= true;
67 if (strtolower($flag) == '\\flagged') {
68 $msg->is_flagged
= true;
72 if (strtolower($flag) == '$mdnsent') {
73 $msg->is_mdnsent
= true;
81 // listEntities($msg);
87 /* This starts the parsing of a particular structure. It is called recursively,
88 * so it can be passed different structures. It returns an object of type
90 * First, it checks to see if it is a multipart message. If it is, then it
91 * handles that as it sees is necessary. If it is just a regular entity,
92 * then it parses it and adds the necessary header information (by calling out
93 * to mime_get_elements()
96 function mime_fetch_body($imap_stream, $id, $ent_id=1) {
98 /* Do a bit of error correction. If we couldn't find the entity id, just guess
99 * 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 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
109 $topline = trim(array_shift($data));
110 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
112 $wholemessage = implode('', $data);
113 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
114 $ret = substr($wholemessage, 0, $regs[1]);
115 /* There is some information in the content info header that could be important
116 * in order to parse html messages. Let's get them here.
118 if ($ret{0} == '<') {
119 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
121 } else if (ereg('"([^"]*)"', $topline, $regs)) {
124 global $where, $what, $mailbox, $passed_id, $startMessage;
125 $par = 'mailbox=' . urlencode($mailbox) . '&passed_id=' . $passed_id;
126 if (isset($where) && isset($what)) {
127 $par .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
129 $par .= '&startMessage=' . $startMessage . '&show_more=0';
131 $par .= '&response=' . urlencode($response) .
132 '&message=' . urlencode($message) .
133 '&topline=' . urlencode($topline);
136 '<table width="80%"><tr>' .
137 '<tr><td colspan=2>' .
138 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
140 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
141 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
142 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
143 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
144 "</table><BR></tt></font><hr>";
146 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
148 $wholemessage = implode('', $data);
150 $ret = $wholemessage;
155 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
158 $sid = sqimap_session_id($uid_support);
159 /* Don't kill the connection if the browser is over a dialup
160 * and it would take over 30 seconds to download it.
161 * DonĀ“t call set_time_limit in safe mode.
164 if (!ini_get('safe_mode')) {
168 $sid_s = substr($sid,0,strpos($sid, ' '));
173 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
174 echo decodeBody($body, $encoding);
177 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
180 $read = fgets ($imap_stream,8192);
183 // This could be bad -- if the section has sqimap_session_id() . ' OK'
184 // or similar, it will kill the download.
185 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
186 if (trim($read) == ')==') {
188 $read = fgets ($imap_stream,4096);
189 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
192 echo decodeBody($read1, $encoding) .
193 decodeBody($read, $encoding);
196 echo decodeBody($read, $encoding);
198 $read = fgets ($imap_stream,4096);
205 /* -[ END MIME DECODING ]----------------------------------------------------------- */
207 /* This is here for debugging purposes. It will print out a list
208 * of all the entity IDs that are in the $message object.
210 function listEntities ($message) {
212 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
213 for ($i = 0; isset($message->entities
[$i]); $i++
) {
215 $msg = listEntities($message->entities
[$i]);
225 function getPriorityStr($priority) {
226 $priority_level = substr($priority,0,1);
228 switch($priority_level) {
229 /* Check for a higher then normal priority. */
232 $priority_string = _("High");
235 /* Check for a lower then normal priority. */
238 $priority_string = _("Low");
241 /* Check for a normal priority. */
244 $priority_level = '3';
245 $priority_string = _("Normal");
249 return $priority_string;
252 /* returns a $message object for a particular entity id */
253 function getEntity ($message, $ent_id) {
254 return $message->getEntity($ent_id);
258 * Extracted from strings.php 23/03/2002
261 function translateText(&$body, $wrap_at, $charset) {
262 global $where, $what; /* from searching */
263 global $color; /* color theme */
265 require_once(SM_PATH
. 'functions/url_parser.php');
267 $body_ary = explode("\n", $body);
268 for ($i=0; $i < count($body_ary); $i++
) {
269 $line = $body_ary[$i];
270 if (strlen($line) - 2 >= $wrap_at) {
271 sqWordWrap($line, $wrap_at);
273 $line = charset_decode($charset, $line);
274 $line = str_replace("\t", ' ', $line);
283 if ($line[$pos] == ' ') {
285 } else if (strpos($line, '>', $pos) === $pos) {
294 if (!isset($color[13])) {
295 $color[13] = '#800000';
297 $line = '<font color="' . $color[13] . '">' . $line . '</font>';
299 if (!isset($color[14])) {
300 $color[14] = '#FF0000';
302 $line = '<font color="' . $color[14] . '">' . $line . '</font>';
305 $body_ary[$i] = $line;
307 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
310 /* This returns a parsed string called $body. That string can then
311 * be displayed as the actual message in the HTML. It contains
312 * everything needed, including HTML Tags, Attachments at the
315 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
316 /* This if statement checks for the entity to show as the
317 * primary message. To add more of them, just put them in the
318 * order that is their priority.
320 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
321 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
323 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
324 $view_unsafe_images = false;
328 $urlmailbox = urlencode($mailbox);
329 $body_message = getEntity($message, $ent_num);
330 if (($body_message->header
->type0
== 'text') ||
331 ($body_message->header
->type0
== 'rfc822')) {
332 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
333 $body = decodeBody($body, $body_message->header
->encoding
);
334 $hookResults = do_hook("message_body", $body);
335 $body = $hookResults[1];
337 /* If there are other types that shouldn't be formatted, add
341 if ($body_message->header
->type1
== 'html') {
342 if ($show_html_default <> 1) {
343 $entity_conv = array(' ' => ' ',
350 $body = strtr($body, $entity_conv);
351 $body = strip_tags($body);
353 translateText($body, $wrap_at,
354 $body_message->header
->getParameter('charset'));
356 $body = magicHTML($body, $id, $message, $mailbox);
359 translateText($body, $wrap_at,
360 $body_message->header
->getParameter('charset'));
362 $link = 'read_body.php?passed_id=' . $id . '&ent_id='.$ent_num.
363 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
364 '&startMessage=' . $startMessage . '&show_more=0';
365 if (isset($passed_ent_id)) {
366 $link .= '&passed_ent_id='.$passed_ent_id;
368 if ($view_unsafe_images) {
369 $text = _("Hide Unsafe Images");
371 if (isset($has_unsafe_images) && $has_unsafe_images) {
372 $link .= '&view_unsafe_images=1';
373 $text = _("View Unsafe Images");
378 $body .= '<center><small><a href="'.$link.'">'.$text.
379 '</a></small></center><br>' . "\n";
385 function formatAttachments($message, $exclude_id, $mailbox, $id) {
386 global $where, $what, $startMessage, $color;
387 static $ShownHTML = 0;
389 $att_ar = $message->getAttachments($exclude_id);
391 if (!count($att_ar)) return '';
395 $urlMailbox = urlencode($mailbox);
397 foreach ($att_ar as $att) {
398 $ent = urldecode($att->entity_id
);
399 $header = $att->header
;
400 $type0 = strtolower($header->type0
);
401 $type1 = strtolower($header->type1
);
403 $links['download link']['text'] = _("download");
404 $links['download link']['href'] =
405 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
407 if ($type0 =='message' && $type1 == 'rfc822') {
408 $default_page = '../src/read_body.php';
409 $rfc822_header = $att->rfc822_header
;
410 $filename = $rfc822_header->subject
;
411 if (trim( $filename ) == '') {
412 $filename = 'untitled-[' . $ent . ']' ;
414 $from_o = $rfc822_header->from
;
415 if (is_object($from_o)) {
416 $from_name = $from_o->getAddress(false);
418 $from_name = _("Unknown sender");
420 $from_name = decodeHeader(($from_name));
421 $description = $from_name;
423 $default_page = '../src/download.php';
424 if (is_object($header->disposition
)) {
425 $filename = $header->disposition
->getProperty('filename');
426 if (trim($filename) == '') {
427 $name = decodeHeader($header->disposition
->getProperty('name'));
428 if (trim($name) == '') {
429 $name = $header->getParameter('name');
430 if(trim($name) == '') {
431 if (trim( $header->id
) == '') {
432 $filename = 'untitled-[' . $ent . ']' ;
434 $filename = 'cid: ' . $header->id
;
444 $filename = $header->getParameter('name');
445 if (!trim($filename)) {
446 if (trim( $header->id
) == '') {
447 $filename = 'untitled-[' . $ent . ']' ;
449 $filename = 'cid: ' . $header->id
;
453 if ($header->description
) {
454 $description = decodeHeader($header->description
);
460 $display_filename = $filename;
461 if (isset($passed_ent_id)) {
462 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
464 $passed_ent_id_link = '';
466 $defaultlink = $default_page . "?startMessage=$startMessage"
467 . "&passed_id=$id&mailbox=$urlMailbox"
468 . '&ent_id='.$ent.$passed_ent_id_link.'&absolute_dl=true';
469 if ($where && $what) {
470 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
472 /* This executes the attachment hook with a specific MIME-type.
473 * If that doesn't have results, it tries if there's a rule
474 * for a more generic type.
476 $hookresults = do_hook("attachment $type0/$type1", $links,
477 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
478 $display_filename, $where, $what);
479 if(count($hookresults[1]) <= 1) {
480 $hookresults = do_hook("attachment $type0/*", $links,
481 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
482 $display_filename, $where, $what);
485 $links = $hookresults[1];
486 $defaultlink = $hookresults[6];
488 $attachments .= '<TR><TD>' .
489 '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A> </TD>' .
490 '<TD><SMALL><b>' . show_readable_size($header->size
) .
491 '</b> </small></TD>' .
492 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
494 $attachments .= '<b>' . $description . '</b>';
495 $attachments .= '</SMALL></TD><TD><SMALL> ';
498 foreach ($links as $val) {
502 $attachments .= ' | ';
504 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
507 $attachments .= "</TD></TR>\n";
512 /* This function decodes the body depending on the encoding type. */
513 function decodeBody($body, $encoding) {
514 global $languages, $squirrelmail_language;
515 global $show_html_default;
517 $body = str_replace("\r\n", "\n", $body);
518 $encoding = strtolower($encoding);
520 if ($encoding == 'quoted-printable' ||
521 $encoding == 'quoted_printable') {
522 $body = quoted_printable_decode($body);
524 while (ereg("=\n", $body)) {
525 $body = ereg_replace ("=\n", '', $body);
528 } else if ($encoding == 'base64') {
529 $body = base64_decode($body);
532 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
533 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
534 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
537 // All other encodings are returned raw.
542 * This functions decode strings that is encoded according to
543 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
544 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
546 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
547 global $languages, $squirrelmail_language;
548 if (is_array($string)) {
549 $string = implode("\n", $string);
552 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
553 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
554 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
558 $aString = explode(' ',$string);
559 foreach ($aString as $chunk) {
562 while (preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
563 //$aString[$i] .= $res[1];
564 //echo "match ". $res[5] . "<br>";
565 $encoding = ucfirst($res[3]);
569 $replace = base64_decode($res[4]);
570 $aString[$i] .= charset_decode($res[2],$replace);
573 $replace = str_replace('_', ' ', $res[4]);
574 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
576 /* Only encode into entities by default. Some places
577 * don't need the encoding, like the compose form.
580 $replace = charset_decode($res[2], $replace);
583 $replace = htmlspecialchars($res[4]);
586 $aString[$i] .= $replace;
594 if (!$encoded && $htmlsave) {
595 $aString[$i] = htmlspecialchars($chunk);
597 $aString[$i] .= $chunk;
601 return implode(' ',$aString);
605 * Encode a string according to RFC 1522 for use in headers if it
606 * contains 8-bit characters or anything that looks like it should
609 function encodeHeader ($string) {
610 global $default_charset, $languages, $squirrelmail_language;
612 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
613 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
614 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
617 // Encode only if the string contains 8-bit characters or =?
618 $j = strlen($string);
619 $l = strstr($string, '=?'); // Must be encoded ?
620 $max_l = 75 - strlen($default_charset) - 7;
624 $cur_l = $iOffset = 0;
625 for($i = 0; $i < $j; ++
$i) {
626 switch($string{$i}) {
629 if ($cur_l > $max_l) {
631 $aRet[] = "=?$default_charset?Q?$ret?=";
634 $aRet[] = substr($string,$iOffset,$i-$iOffset+
1);
644 if ($cur_l > $max_l) {
646 $aRet[] = "=?$default_charset?Q?$ret?=";
649 $aRet[] = substr($string,$iOffset,$i-$iOffset+
1);
659 if ($cur_l > $max_l) {
661 $aRet[] = "=?$default_charset?Q?$ret?=";
664 $aRet[] = substr($string,$iOffset,$i-$iOffset+
1);
674 if ($cur_l > $max_l) {
676 $aRet[] = "=?$default_charset?Q?$ret?=";
679 $aRet[] = substr($string,$iOffset,$i-$iOffset+
1);
688 $k = ord($string{$i});
691 $s = sprintf("=%02X", $k);
692 $cur_l +
= strlen($s);
693 if ($cur_l > $max_l) {
694 $aRet[] = "=?$default_charset?Q?$ret?=";
705 if ($cur_l > $max_l) {
707 $aRet[] = "=?$default_charset?Q?$ret?=";
710 $aRet[] = substr($ret,$iOffset,$i-$iOffset+
1);
724 $string = implode('',$aRet) . "=?$default_charset?Q?$ret?=";
730 /* This function trys to locate the entity_id of a specific mime element */
731 function find_ent_id($id, $message) {
732 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
733 if ($message->entities
[$i]->header
->type0
== 'multipart') {
734 $ret = find_ent_id($id, $message->entities
[$i]);
736 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
737 // if (sq_check_save_extension($message->entities[$i])) {
738 return $message->entities
[$i]->entity_id
;
746 function sq_check_save_extension($message) {
747 $filename = $message->getFilename();
748 $ext = substr($filename, strrpos($filename,'.')+
1);
749 $save_extensions = array('jpg','jpeg','gif','png','bmp');
750 return in_array($ext, $save_extensions);
755 ** HTMLFILTER ROUTINES
759 * This function returns the final tag out of the tag name, an array
760 * of attributes, and the type of the tag. This function is called by
761 * sq_sanitize internally.
763 * @param $tagname the name of the tag.
764 * @param $attary the array of attributes and their values
765 * @param $tagtype The type of the tag (see in comments).
766 * @return a string with the final tag representation.
768 function sq_tagprint($tagname, $attary, $tagtype){
772 $fulltag = '</' . $tagname . '>';
774 $fulltag = '<' . $tagname;
775 if (is_array($attary) && sizeof($attary)){
777 while (list($attname, $attvalue) = each($attary)){
778 array_push($atts, "$attname=$attvalue");
780 $fulltag .= ' ' . join(" ", $atts);
791 * A small helper function to use with array_walk. Modifies a by-ref
792 * value and makes it lowercase.
794 * @param $val a value passed by-ref.
795 * @return void since it modifies a by-ref value.
797 function sq_casenormalize(&$val){
798 $val = strtolower($val);
802 * This function skips any whitespace from the current position within
803 * a string and to the next non-whitespace value.
805 * @param $body the string
806 * @param $offset the offset within the string where we should start
807 * looking for the next non-whitespace character.
808 * @return the location within the $body where the next
809 * non-whitespace char is located.
811 function sq_skipspace($body, $offset){
812 $me = 'sq_skipspace';
813 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
814 if (sizeof($matches{1})){
815 $count = strlen($matches{1});
822 * This function looks for the next character within a string. It's
823 * really just a glorified "strpos", except it catches if failures
826 * @param $body The string to look for needle in.
827 * @param $offset Start looking from this position.
828 * @param $needle The character/string to look for.
829 * @return location of the next occurance of the needle, or
830 * strlen($body) if needle wasn't found.
832 function sq_findnxstr($body, $offset, $needle){
833 $me = 'sq_findnxstr';
834 $pos = strpos($body, $needle, $offset);
836 $pos = strlen($body);
842 * This function takes a PCRE-style regexp and tries to match it
845 * @param $body The string to look for needle in.
846 * @param $offset Start looking from here.
847 * @param $reg A PCRE-style regex to match.
848 * @return Returns a false if no matches found, or an array
849 * with the following members:
850 * - integer with the location of the match within $body
851 * - string with whatever content between offset and the match
852 * - string with whatever it is we matched
854 function sq_findnxreg($body, $offset, $reg){
855 $me = 'sq_findnxreg';
858 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
862 $retarr{0} = $offset +
strlen($matches{1});
863 $retarr{1} = $matches{1};
864 $retarr{2} = $matches{2};
870 * This function looks for the next tag.
872 * @param $body String where to look for the next tag.
873 * @param $offset Start looking from here.
874 * @return false if no more tags exist in the body, or
875 * an array with the following members:
876 * - string with the name of the tag
877 * - array with attributes and their values
878 * - integer with tag type (1, 2, or 3)
879 * - integer where the tag starts (starting "<")
880 * - integer where the tag ends (ending ">")
881 * first three members will be false, if the tag is invalid.
883 function sq_getnxtag($body, $offset){
885 if ($offset > strlen($body)){
888 $lt = sq_findnxstr($body, $offset, "<");
889 if ($lt == strlen($body)){
894 * blah blah <tag attribute="value">
897 $pos = sq_skipspace($body, $lt+
1);
898 if ($pos >= strlen($body)){
899 return Array(false, false, false, $lt, strlen($body));
902 * There are 3 kinds of tags:
903 * 1. Opening tag, e.g.:
905 * 2. Closing tag, e.g.:
907 * 3. XHTML-style content-less tag, e.g.:
911 switch (substr($body, $pos, 1)){
918 * A comment or an SGML declaration.
920 if (substr($body, $pos+
1, 2) == "--"){
921 $gt = strpos($body, "-->", $pos);
927 return Array(false, false, false, $lt, $gt);
929 $gt = sq_findnxstr($body, $pos, ">");
930 return Array(false, false, false, $lt, $gt);
935 * Assume tagtype 1 for now. If it's type 3, we'll switch values
945 * Look for next [\W-_], which will indicate the end of the tag name.
947 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
948 if ($regary == false){
949 return Array(false, false, false, $lt, strlen($body));
951 list($pos, $tagname, $match) = $regary;
952 $tagname = strtolower($tagname);
955 * $match can be either of these:
956 * '>' indicating the end of the tag entirely.
957 * '\s' indicating the end of the tag name.
958 * '/' indicating that this is type-3 xhtml tag.
960 * Whatever else we find there indicates an invalid tag.
965 * This is an xhtml-style tag with a closing / at the
966 * end, like so: <img src="blah"/>. Check if it's followed
967 * by the closing bracket. If not, then this tag is invalid
969 if (substr($body, $pos, 2) == "/>"){
973 $gt = sq_findnxstr($body, $pos, ">");
974 $retary = Array(false, false, false, $lt, $gt);
978 return Array($tagname, false, $tagtype, $lt, $pos);
982 * Check if it's whitespace
984 if (!preg_match('/\s/', $match)){
986 * This is an invalid tag! Look for the next closing ">".
988 $gt = sq_findnxstr($body, $offset, ">");
989 return Array(false, false, false, $lt, $gt);
995 * At this point we're here:
996 * <tagname attribute='blah'>
999 * At this point we loop in order to find all attributes.
1005 while ($pos <= strlen($body)){
1006 $pos = sq_skipspace($body, $pos);
1007 if ($pos == strlen($body)){
1011 return Array(false, false, false, $lt, $pos);
1014 * See if we arrived at a ">" or "/>", which means that we reached
1015 * the end of the tag.
1018 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1022 $pos +
= strlen($matches{1});
1023 if ($matches{2} == "/>"){
1027 return Array($tagname, $attary, $tagtype, $lt, $pos);
1031 * There are several types of attributes, with optional
1032 * [:space:] between members.
1034 * attrname[:space:]=[:space:]'CDATA'
1036 * attrname[:space:]=[:space:]"CDATA"
1038 * attr[:space:]=[:space:]CDATA
1042 * We leave types 1 and 2 the same, type 3 we check for
1043 * '"' and convert to """ if needed, then wrap in
1044 * double quotes. Type 4 we convert into:
1047 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1048 if ($regary == false){
1050 * Looks like body ended before the end of tag.
1052 return Array(false, false, false, $lt, strlen($body));
1054 list($pos, $attname, $match) = $regary;
1055 $attname = strtolower($attname);
1057 * We arrived at the end of attribute name. Several things possible
1059 * '>' means the end of the tag and this is attribute type 4
1060 * '/' if followed by '>' means the same thing as above
1061 * '\s' means a lot of things -- look what it's followed by.
1062 * anything else means the attribute is invalid.
1067 * This is an xhtml-style tag with a closing / at the
1068 * end, like so: <img src="blah"/>. Check if it's followed
1069 * by the closing bracket. If not, then this tag is invalid
1071 if (substr($body, $pos, 2) == "/>"){
1075 $gt = sq_findnxstr($body, $pos, ">");
1076 $retary = Array(false, false, false, $lt, $gt);
1080 $attary{$attname} = '"yes"';
1081 return Array($tagname, $attary, $tagtype, $lt, $pos);
1085 * Skip whitespace and see what we arrive at.
1087 $pos = sq_skipspace($body, $pos);
1088 $char = substr($body, $pos, 1);
1090 * Two things are valid here:
1091 * '=' means this is attribute type 1 2 or 3.
1092 * \w means this was attribute type 4.
1093 * anything else we ignore and re-loop. End of tag and
1094 * invalid stuff will be caught by our checks at the beginning
1099 $pos = sq_skipspace($body, $pos);
1101 * Here are 3 possibilities:
1102 * "'" attribute type 1
1103 * '"' attribute type 2
1104 * everything else is the content of tag type 3
1106 $quot = substr($body, $pos, 1);
1108 $regary = sq_findnxreg($body, $pos+
1, "\'");
1109 if ($regary == false){
1110 return Array(false, false, false, $lt, strlen($body));
1112 list($pos, $attval, $match) = $regary;
1114 $attary{$attname} = "'" . $attval . "'";
1115 } else if ($quot == '"'){
1116 $regary = sq_findnxreg($body, $pos+
1, '\"');
1117 if ($regary == false){
1118 return Array(false, false, false, $lt, strlen($body));
1120 list($pos, $attval, $match) = $regary;
1122 $attary{$attname} = '"' . $attval . '"';
1125 * These are hateful. Look for \s, or >.
1127 $regary = sq_findnxreg($body, $pos, "[\s>]");
1128 if ($regary == false){
1129 return Array(false, false, false, $lt, strlen($body));
1131 list($pos, $attval, $match) = $regary;
1133 * If it's ">" it will be caught at the top.
1135 $attval = preg_replace("/\"/s", """, $attval);
1136 $attary{$attname} = '"' . $attval . '"';
1138 } else if (preg_match("|[\w/>]|", $char)) {
1140 * That was attribute type 4.
1142 $attary{$attname} = '"yes"';
1145 * An illegal character. Find next '>' and return.
1147 $gt = sq_findnxstr($body, $pos, ">");
1148 return Array(false, false, false, $lt, $gt);
1154 * The fact that we got here indicates that the tag end was never
1155 * found. Return invalid tag indication so it gets stripped.
1157 return Array(false, false, false, $lt, strlen($body));
1161 * This function checks attribute values for entity-encoded values
1162 * and returns them translated into 8-bit strings so we can run
1165 * @param $attvalue A string to run entity check against.
1166 * @return Translated value.
1168 function sq_deent($attvalue){
1171 * See if we have to run the checks first. All entities must start
1174 if (strpos($attvalue, "&") === false){
1178 * Check named entities first.
1180 $trans = get_html_translation_table(HTML_ENTITIES
);
1182 * Leave " in, as it can mess us up.
1184 $trans = array_flip($trans);
1185 unset($trans{"""});
1186 while (list($ent, $val) = each($trans)){
1187 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1190 * Now translate numbered entities from 1 to 255 if needed.
1192 if (strpos($attvalue, "#") !== false){
1193 $omit = Array(34, 39);
1194 for ($asc=1; $asc<256; $asc++
){
1195 if (!in_array($asc, $omit)){
1197 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1199 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1200 "$chr\\1", $attvalue);
1208 * This function runs various checks against the attributes.
1210 * @param $tagname String with the name of the tag.
1211 * @param $attary Array with all tag attributes.
1212 * @param $rm_attnames See description for sq_sanitize
1213 * @param $bad_attvals See description for sq_sanitize
1214 * @param $add_attr_to_tag See description for sq_sanitize
1215 * @param $message message object
1216 * @param $id message id
1217 * @return Array with modified attributes.
1219 function sq_fixatts($tagname,
1229 while (list($attname, $attvalue) = each($attary)){
1231 * See if this attribute should be removed.
1233 foreach ($rm_attnames as $matchtag=>$matchattrs){
1234 if (preg_match($matchtag, $tagname)){
1235 foreach ($matchattrs as $matchattr){
1236 if (preg_match($matchattr, $attname)){
1237 unset($attary{$attname});
1244 * Remove any entities.
1246 $attvalue = sq_deent($attvalue);
1249 * Now let's run checks on the attvalues.
1250 * I don't expect anyone to comprehend this. If you do,
1251 * get in touch with me so I can drive to where you live and
1252 * shake your hand personally. :)
1254 foreach ($bad_attvals as $matchtag=>$matchattrs){
1255 if (preg_match($matchtag, $tagname)){
1256 foreach ($matchattrs as $matchattr=>$valary){
1257 if (preg_match($matchattr, $attname)){
1259 * There are two arrays in valary.
1261 * Second one is replacements
1263 list($valmatch, $valrepl) = $valary;
1265 preg_replace($valmatch, $valrepl, $attvalue);
1266 if ($newvalue != $attvalue){
1267 $attary{$attname} = $newvalue;
1274 * Turn cid: urls into http-friendly ones.
1276 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1277 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1281 * See if we need to append any attributes to this tag.
1283 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1284 if (preg_match($matchtag, $tagname)){
1285 $attary = array_merge($attary, $addattary);
1292 * This function edits the style definition to make them friendly and
1293 * usable in squirrelmail.
1295 * @param $message the message object
1296 * @param $id the message id
1297 * @param $content a string with whatever is between <style> and </style>
1298 * @return a string with edited content.
1300 function sq_fixstyle($message, $id, $content){
1301 global $view_unsafe_images;
1302 $me = 'sq_fixstyle';
1304 * First look for general BODY style declaration, which would be
1306 * body {background: blah-blah}
1307 * and change it to .bodyclass so we can just assign it to a <div>
1309 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1310 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1312 * Fix url('blah') declarations.
1314 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1315 "url(\\1$secremoveimg\\2)", $content);
1317 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1320 if (!$view_unsafe_images){
1321 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1322 "url(\\1$secremoveimg\\2)", $content);
1326 * Fix urls that refer to cid:
1328 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1330 $cidurl = $matches{1};
1331 $httpurl = sq_cid2http($message, $id, $cidurl);
1332 $content = preg_replace("|url\($cidurl\)|si",
1333 "url($httpurl)", $content);
1337 * Fix stupid css declarations which lead to vulnerabilities
1340 $match = Array('/expression/si',
1343 $replace = Array('idiocy', 'idiocy', 'idiocy');
1344 $content = preg_replace($match, $replace, $content);
1349 * This function converts cid: url's into the ones that can be viewed in
1352 * @param $message the message object
1353 * @param $id the message id
1354 * @param $cidurl the cid: url.
1355 * @return a string with a http-friendly url
1357 function sq_cid2http($message, $id, $cidurl, $mailbox){
1359 * Get rid of quotes.
1361 $quotchar = substr($cidurl, 0, 1);
1362 $cidurl = str_replace($quotchar, "", $cidurl);
1363 $cidurl = substr(trim($cidurl), 4);
1364 $linkurl = find_ent_id($cidurl, $message);
1365 /* in case of non-save cid links $httpurl should be replaced by a sort of
1366 unsave link image */
1369 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&' .
1370 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1371 '&ent_id=' . $linkurl . $quotchar;
1377 * This function changes the <body> tag into a <div> tag since we
1378 * can't really have a body-within-body.
1380 * @param $attary an array of attributes and values of <body>
1381 * @return a modified array of attributes to be set for <div>
1383 function sq_body2div($attary){
1384 $me = 'sq_body2div';
1385 $divattary = Array('class' => "'bodyclass'");
1386 $bgcolor = '#ffffff';
1389 if (is_array($attary) && sizeof($attary) > 0){
1390 foreach ($attary as $attname=>$attvalue){
1391 $quotchar = substr($attvalue, 0, 1);
1392 $attvalue = str_replace($quotchar, "", $attvalue);
1395 $styledef .= "background-image: url('$attvalue'); ";
1398 $styledef .= "background-color: $attvalue; ";
1401 $styledef .= "color: $attvalue; ";
1405 if (strlen($styledef) > 0){
1406 $divattary{"style"} = "\"$styledef\"";
1413 * This is the main function and the one you should actually be calling.
1414 * There are several variables you should be aware of an which need
1415 * special description.
1417 * Since the description is quite lengthy, see it here:
1418 * http://www.mricon.com/html/phpfilter.html
1420 * @param $body the string with HTML you wish to filter
1421 * @param $tag_list see description above
1422 * @param $rm_tags_with_content see description above
1423 * @param $self_closing_tags see description above
1424 * @param $force_tag_closing see description above
1425 * @param $rm_attnames see description above
1426 * @param $bad_attvals see description above
1427 * @param $add_attr_to_tag see description above
1428 * @param $message message object
1429 * @param $id message id
1430 * @return sanitized html safe to show on your pages.
1432 function sq_sanitize($body,
1434 $rm_tags_with_content,
1444 $me = 'sq_sanitize';
1446 * Normalize rm_tags and rm_tags_with_content.
1448 @array_walk
($rm_tags, 'sq_casenormalize');
1449 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1450 @array_walk
($self_closing_tags, 'sq_casenormalize');
1452 * See if tag_list is of tags to remove or tags to allow.
1453 * false means remove these tags
1454 * true means allow these tags
1456 $rm_tags = array_shift($tag_list);
1458 $open_tags = Array();
1459 $trusted = "<!-- begin sanitized html -->\n";
1460 $skip_content = false;
1462 * Take care of netscape's stupid javascript entities like
1465 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1467 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1468 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1469 $free_content = substr($body, $curpos, $lt-$curpos);
1471 * Take care of <style>
1473 if ($tagname == "style" && $tagtype == 2){
1475 * This is a closing </style>. Edit the
1476 * content before we apply it.
1478 $free_content = sq_fixstyle($message, $id, $free_content);
1480 if ($skip_content == false){
1481 $trusted .= $free_content;
1483 if ($tagname != FALSE){
1485 if ($skip_content == $tagname){
1487 * Got to the end of tag we needed to remove.
1490 $skip_content = false;
1492 if ($skip_content == false){
1493 if ($tagname == "body"){
1496 if (isset($open_tags{$tagname}) &&
1497 $open_tags{$tagname} > 0){
1498 $open_tags{$tagname}--;
1507 * $rm_tags_with_content
1509 if ($skip_content == false){
1511 * See if this is a self-closing type and change
1512 * tagtype appropriately.
1515 && in_array($tagname, $self_closing_tags)){
1519 * See if we should skip this tag and any content
1522 if ($tagtype == 1 &&
1523 in_array($tagname, $rm_tags_with_content)){
1524 $skip_content = $tagname;
1526 if (($rm_tags == false
1527 && in_array($tagname, $tag_list)) ||
1528 ($rm_tags == true &&
1529 !in_array($tagname, $tag_list))){
1533 if (isset($open_tags{$tagname})){
1534 $open_tags{$tagname}++
;
1536 $open_tags{$tagname}=1;
1540 * This is where we run other checks.
1542 if (is_array($attary) && sizeof($attary) > 0){
1543 $attary = sq_fixatts($tagname,
1554 * Convert body into div.
1556 if ($tagname == "body"){
1558 $attary = sq_body2div($attary, $message, $id);
1564 if ($tagname != false && $skip_content == false){
1565 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1570 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1571 if ($force_tag_closing == true){
1572 foreach ($open_tags as $tagname=>$opentimes){
1573 while ($opentimes > 0){
1574 $trusted .= '</' . $tagname . '>';
1580 $trusted .= "<!-- end sanitized html -->\n";
1585 * This is a wrapper function to call html sanitizing routines.
1587 * @param $body the body of the message
1588 * @param $id the id of the message
1589 * @return a string with html safe to display in the browser.
1591 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1592 global $attachment_common_show_images, $view_unsafe_images,
1595 * Don't display attached images in HTML mode.
1597 $attachment_common_show_images = false;
1610 $rm_tags_with_content = Array(
1617 $self_closing_tags = Array(
1624 $force_tag_closing = false;
1626 $rm_attnames = Array(
1636 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1637 $bad_attvals = Array(
1640 "/^src|background/i" =>
1643 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1644 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1645 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1646 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1649 "\\1$secremoveimg\\2",
1650 "\\1$secremoveimg\\2",
1651 "\\1$secremoveimg\\2",
1652 "\\1$secremoveimg\\2"
1655 "/^href|action/i" =>
1658 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1659 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1660 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1661 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1676 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1677 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1678 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1679 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1693 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1694 $view_unsafe_images = false;
1696 if (!$view_unsafe_images){
1698 * Remove any references to http/https if view_unsafe_images set
1701 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1702 '/^([\'\"])\s*https*:.*([\'\"])/si');
1703 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1704 "\\1$secremoveimg\\2");
1705 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1706 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1707 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1708 "url(\\1$secremoveimg\\2)");
1711 $add_attr_to_tag = Array(
1712 "/^a$/si" => Array('target'=>'"_new"')
1714 $trusted = sq_sanitize($body,
1716 $rm_tags_with_content,
1726 if (preg_match("|$secremoveimg|si", $trusted)){
1727 $has_unsafe_images = true;