6 * Copyright (c) 1999-2002 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));
32 $msg =& new Message();
33 $res = $msg->parseStructure($read);
35 if (!is_object($msg)) {
36 include_once(SM_PATH
. 'functions/display_messages.php');
37 global $color, $mailbox;
38 displayPageHeader( $color, urldecode($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, $color );
45 echo '</body></html>';
50 foreach ($flags as $flag) {
51 $char = strtoupper($flag{1});
54 if (strtolower($flag) == '\\seen') {
59 if (strtolower($flag) == '\\answered') {
60 $msg->is_answered
= true;
64 if (strtolower($flag) == '\\deleted') {
65 $msg->is_deleted
= true;
69 if (strtolower($flag) == '\\flagged') {
70 $msg->is_flagged
= true;
74 if (strtolower($flag) == '$mdnsent') {
75 $msg->is_mdnsent
= true;
83 // listEntities($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) {
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.
104 $cmd = "FETCH $id BODY[$ent_id]";
106 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
108 $topline = trim(array_shift($data));
109 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
111 $wholemessage = implode('', $data);
112 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
113 $ret = substr($wholemessage, 0, $regs[1]);
114 /* There is some information in the content info header that could be important
115 * in order to parse html messages. Let's get them here.
117 if ($ret{0} == '<') {
118 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
120 } else if (ereg('"([^"]*)"', $topline, $regs)) {
123 global $where, $what, $mailbox, $passed_id, $startMessage;
124 $par = 'mailbox=' . urlencode($mailbox) . '&passed_id=' . $passed_id;
125 if (isset($where) && isset($what)) {
126 $par .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
128 $par .= '&startMessage=' . $startMessage . '&show_more=0';
130 $par .= '&response=' . urlencode($response) .
131 '&message=' . urlencode($message) .
132 '&topline=' . urlencode($topline);
135 '<table width="80%"><tr>' .
136 '<tr><td colspan=2>' .
137 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
139 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
140 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
141 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
142 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
143 "</table><BR></tt></font><hr>";
145 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
147 $wholemessage = implode('', $data);
149 $ret = $wholemessage;
154 function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) {
156 /* Do a bit of error correction. If we couldn't find the entity id, just guess
157 * that it is the first one. That is usually the case anyway.
162 $sid = sqimap_session_id($uid_support);
163 /* Don't kill the connection if the browser is over a dialup
164 * and it would take over 30 seconds to download it.
165 * DonĀ“t call set_time_limit in safe mode.
168 if (!ini_get('safe_mode')) {
172 $sid_s = substr($sid,0,strpos($sid, ' '));
177 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
178 echo decodeBody($body, $encoding);
181 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
184 $read = fgets ($imap_stream,8192);
187 // This could be bad -- if the section has sqimap_session_id() . ' OK'
188 // or similar, it will kill the download.
189 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
190 if (trim($read) == ')==') {
192 $read = fgets ($imap_stream,4096);
193 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
196 echo decodeBody($read1, $encoding) .
197 decodeBody($read, $encoding);
200 echo decodeBody($read, $encoding);
202 $read = fgets ($imap_stream,4096);
209 /* -[ END MIME DECODING ]----------------------------------------------------------- */
211 /* This is here for debugging purposes. It will print out a list
212 * of all the entity IDs that are in the $message object.
214 function listEntities ($message) {
216 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
217 for ($i = 0; isset($message->entities
[$i]); $i++
) {
219 $msg = listEntities($message->entities
[$i]);
229 function getPriorityStr($priority) {
230 $priority_level = substr($priority,0,1);
232 switch($priority_level) {
233 /* Check for a higher then normal priority. */
236 $priority_string = _("High");
239 /* Check for a lower then normal priority. */
242 $priority_string = _("Low");
245 /* Check for a normal priority. */
248 $priority_level = '3';
249 $priority_string = _("Normal");
253 return $priority_string;
256 /* returns a $message object for a particular entity id */
257 function getEntity ($message, $ent_id) {
258 return $message->getEntity($ent_id);
262 * Extracted from strings.php 23/03/2002
265 function translateText(&$body, $wrap_at, $charset) {
266 global $where, $what; /* from searching */
267 global $color; /* color theme */
269 require_once(SM_PATH
. 'functions/url_parser.php');
271 $body_ary = explode("\n", $body);
272 for ($i=0; $i < count($body_ary); $i++
) {
273 $line = $body_ary[$i];
274 if (strlen($line) - 2 >= $wrap_at) {
275 sqWordWrap($line, $wrap_at);
277 $line = charset_decode($charset, $line);
278 $line = str_replace("\t", ' ', $line);
287 if ($line[$pos] == ' ') {
289 } else if (strpos($line, '>', $pos) === $pos) {
298 if (!isset($color[14])) {
299 $color[14] = '#FF0000';
301 $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
303 if (!isset($color[13])) {
304 $color[13] = '#800000';
306 $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
309 $body_ary[$i] = $line;
311 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
315 /* This returns a parsed string called $body. That string can then
316 * be displayed as the actual message in the HTML. It contains
317 * everything needed, including HTML Tags, Attachments at the
320 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
321 /* This if statement checks for the entity to show as the
322 * primary message. To add more of them, just put them in the
323 * order that is their priority.
325 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
326 $show_html_default, $has_unsafe_images, $sort;
328 if ( (float)substr(PHP_VERSION
,0,3) < 4.1 ) {
331 if(isset($_GET['view_unsafe_images'])) {
332 $view_unsafe_images = $_GET['view_unsafe_images'];
335 $has_unsafe_images= 0;
337 $urlmailbox = urlencode($mailbox);
338 $body_message = getEntity($message, $ent_num);
339 if (($body_message->header
->type0
== 'text') ||
340 ($body_message->header
->type0
== 'rfc822')) {
341 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
342 $body = decodeBody($body, $body_message->header
->encoding
);
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 $body_message->header
->getParameter('charset'));
372 if ($has_unsafe_images) {
373 if ($view_unsafe_images) {
374 $untext = '">' . _("Hide Unsafe Images");
376 $untext = '&view_unsafe_images=1">' . _("View Unsafe Images");
378 $body .= '<center><small><a href="read_body.php?passed_id=' . $id .
379 '&passed_ent_id=' . $message->entity_id
. '&mailbox=' . $urlmailbox .
380 '&sort=' . $sort . '&startMessage=' . $startMessage . '&show_more=0' .
381 $untext . '</a></small></center><br>' . "\n";
388 function formatAttachments($message, $exclude_id, $mailbox, $id) {
389 global $where, $what, $startMessage, $color;
390 static $ShownHTML = 0;
392 $att_ar = $message->getAttachments($exclude_id);
394 if (!count($att_ar)) return '';
398 $urlMailbox = urlencode($mailbox);
400 foreach ($att_ar as $att) {
401 $ent = urldecode($att->entity_id
);
402 $header = $att->header
;
403 $type0 = strtolower($header->type0
);
404 $type1 = strtolower($header->type1
);
406 $links['download link']['text'] = _("download");
407 $links['download link']['href'] =
408 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
410 if ($type0 =='message' && $type1 == 'rfc822') {
411 $default_page = '../src/read_body.php';
412 $rfc822_header = $att->rfc822_header
;
413 $filename = decodeHeader($rfc822_header->subject
);
415 $from_o = $rfc822_header->from
;
416 if (is_object($from_o)) {
417 $from_name = $from_o->getAddress(false);
419 $from_name = _("Unknown sender");
421 $from_name = decodeHeader(htmlspecialchars($from_name));
422 $description = $from_name;
424 $default_page = '../src/download.php';
425 if (is_object($header->disposition
)) {
426 $filename = decodeHeader($header->disposition
->getProperty('filename'));
427 if (trim($filename) == '') {
428 $name = decodeHeader($header->disposition
->getProperty('name'));
429 if (trim($name) == '') {
430 if (trim( $header->id
) == '') {
431 $filename = 'untitled-[' . $ent . ']' ;
433 $filename = 'cid: ' . $header->id
;
440 if (trim( $header->id
) == '') {
441 $filename = 'untitled-[' . $ent . ']' ;
443 $filename = 'cid: ' . $header->id
;
447 if ($header->description
) {
448 $description = htmlspecialchars($header->description
);
454 $display_filename = $filename;
455 if (isset($passed_ent_id)) {
456 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
458 $passed_ent_id_link = '';
460 $defaultlink = $default_page . "?startMessage=$startMessage"
461 . "&passed_id=$id&mailbox=$urlMailbox"
462 . '&ent_id='.$ent.$passed_ent_id_link;
463 if ($where && $what) {
464 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
466 /* This executes the attachment hook with a specific MIME-type.
467 * If that doesn't have results, it tries if there's a rule
468 * for a more generic type.
470 $hookresults = do_hook("attachment $type0/$type1", $links,
471 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
472 $display_filename, $where, $what);
473 if(count($hookresults[1]) <= 1) {
474 $hookresults = do_hook("attachment $type0/*", $links,
475 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
476 $display_filename, $where, $what);
479 $links = $hookresults[1];
480 $defaultlink = $hookresults[6];
482 $attachments .= '<TR><TD>' .
483 "<A HREF=\"$defaultlink\">$display_filename</A> </TD>" .
484 '<TD><SMALL><b>' . show_readable_size($header->size
) .
485 '</b> </small></TD>' .
486 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
488 $attachments .= '<b>' . $description . '</b>';
489 $attachments .= '</SMALL></TD><TD><SMALL> ';
492 foreach ($links as $val) {
496 $attachments .= ' | ';
498 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
501 $attachments .= "</TD></TR>\n";
506 /* This function decodes the body depending on the encoding type. */
507 function decodeBody($body, $encoding) {
508 global $languages, $squirrelmail_language;
509 global $show_html_default;
511 $body = str_replace("\r\n", "\n", $body);
512 $encoding = strtolower($encoding);
514 if ($encoding == 'quoted-printable' ||
515 $encoding == 'quoted_printable') {
516 $body = quoted_printable_decode($body);
518 while (ereg("=\n", $body)) {
519 $body = ereg_replace ("=\n", '', $body);
522 } else if ($encoding == 'base64') {
523 $body = base64_decode($body);
526 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
527 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
528 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
531 // All other encodings are returned raw.
536 * This functions decode strings that is encoded according to
537 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
538 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
540 function decodeHeader ($string, $utfencode=true) {
541 global $languages, $squirrelmail_language;
542 if (is_array($string)) {
543 $string = implode("\n", $string);
546 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
547 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
548 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
552 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
555 /* Ignore white-space between consecutive encoded-words. */
556 if (strspn($res[2], " \t") != strlen($res[2])) {
560 if (ucfirst($res[4]) == 'B') {
561 $replace = base64_decode($res[5]);
563 $replace = str_replace('_', ' ', $res[5]);
564 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
566 /* Only encode into entities by default. Some places
567 * don't need the encoding, like the compose form.
570 $replace = charset_decode($res[3], $replace);
573 $string = $prefix . $replace . substr($string, strlen($res[0]));
574 $i = strlen($prefix) +
strlen($replace);
580 * Encode a string according to RFC 1522 for use in headers if it
581 * contains 8-bit characters or anything that looks like it should
584 function encodeHeader ($string) {
585 global $default_charset, $languages, $squirrelmail_language;
587 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
588 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
589 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
592 // Encode only if the string contains 8-bit characters or =?
593 $j = strlen($string);
594 $l = strstr($string, '=?'); // Must be encoded ?
596 for($i = 0; $i < $j; ++
$i) {
597 switch($string{$i}) {
611 $k = ord($string{$i});
613 $ret .= sprintf("=%02X", $k);
623 $string = "=?$default_charset?Q?$ret?=";
629 /* This function trys to locate the entity_id of a specific mime element */
630 function find_ent_id($id, $message) {
631 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
632 if ($message->entities
[$i]->header
->type0
== 'multipart') {
633 $ret = find_ent_id($id, $message->entities
[$i]);
635 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
636 // if (sq_check_save_extension($message->entities[$i])) {
637 return $message->entities
[$i]->entity_id
;
645 function sq_check_save_extension($message) {
646 $filename = $message->getFilename();
647 $ext = substr($filename, strrpos($filename,'.')+
1);
648 $save_extensions = array('jpg','jpeg','gif','png','bmp');
649 return in_array($ext, $save_extensions);
654 ** HTMLFILTER ROUTINES
658 * This function returns the final tag out of the tag name, an array
659 * of attributes, and the type of the tag. This function is called by
660 * sq_sanitize internally.
662 * @param $tagname the name of the tag.
663 * @param $attary the array of attributes and their values
664 * @param $tagtype The type of the tag (see in comments).
665 * @return a string with the final tag representation.
667 function sq_tagprint($tagname, $attary, $tagtype){
671 $fulltag = '</' . $tagname . '>';
673 $fulltag = '<' . $tagname;
674 if (is_array($attary) && sizeof($attary)){
676 while (list($attname, $attvalue) = each($attary)){
677 array_push($atts, "$attname=$attvalue");
679 $fulltag .= ' ' . join(" ", $atts);
690 * A small helper function to use with array_walk. Modifies a by-ref
691 * value and makes it lowercase.
693 * @param $val a value passed by-ref.
694 * @return void since it modifies a by-ref value.
696 function sq_casenormalize(&$val){
697 $val = strtolower($val);
701 * This function skips any whitespace from the current position within
702 * a string and to the next non-whitespace value.
704 * @param $body the string
705 * @param $offset the offset within the string where we should start
706 * looking for the next non-whitespace character.
707 * @return the location within the $body where the next
708 * non-whitespace char is located.
710 function sq_skipspace($body, $offset){
711 $me = 'sq_skipspace';
712 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
713 if (sizeof($matches{1})){
714 $count = strlen($matches{1});
721 * This function looks for the next character within a string. It's
722 * really just a glorified "strpos", except it catches if failures
725 * @param $body The string to look for needle in.
726 * @param $offset Start looking from this position.
727 * @param $needle The character/string to look for.
728 * @return location of the next occurance of the needle, or
729 * strlen($body) if needle wasn't found.
731 function sq_findnxstr($body, $offset, $needle){
732 $me = 'sq_findnxstr';
733 $pos = strpos($body, $needle, $offset);
735 $pos = strlen($body);
741 * This function takes a PCRE-style regexp and tries to match it
744 * @param $body The string to look for needle in.
745 * @param $offset Start looking from here.
746 * @param $reg A PCRE-style regex to match.
747 * @return Returns a false if no matches found, or an array
748 * with the following members:
749 * - integer with the location of the match within $body
750 * - string with whatever content between offset and the match
751 * - string with whatever it is we matched
753 function sq_findnxreg($body, $offset, $reg){
754 $me = 'sq_findnxreg';
757 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
761 $retarr{0} = $offset +
strlen($matches{1});
762 $retarr{1} = $matches{1};
763 $retarr{2} = $matches{2};
769 * This function looks for the next tag.
771 * @param $body String where to look for the next tag.
772 * @param $offset Start looking from here.
773 * @return false if no more tags exist in the body, or
774 * an array with the following members:
775 * - string with the name of the tag
776 * - array with attributes and their values
777 * - integer with tag type (1, 2, or 3)
778 * - integer where the tag starts (starting "<")
779 * - integer where the tag ends (ending ">")
780 * first three members will be false, if the tag is invalid.
782 function sq_getnxtag($body, $offset){
784 if ($offset > strlen($body)){
787 $lt = sq_findnxstr($body, $offset, "<");
788 if ($lt == strlen($body)){
793 * blah blah <tag attribute="value">
796 $pos = sq_skipspace($body, $lt+
1);
797 if ($pos >= strlen($body)){
798 return Array(false, false, false, $lt, strlen($body));
801 * There are 3 kinds of tags:
802 * 1. Opening tag, e.g.:
804 * 2. Closing tag, e.g.:
806 * 3. XHTML-style content-less tag, e.g.:
810 switch (substr($body, $pos, 1)){
817 * A comment or an SGML declaration.
819 if (substr($body, $pos+
1, 2) == "--"){
820 $gt = strpos($body, "-->", $pos);
826 return Array(false, false, false, $lt, $gt);
828 $gt = sq_findnxstr($body, $pos, ">");
829 return Array(false, false, false, $lt, $gt);
834 * Assume tagtype 1 for now. If it's type 3, we'll switch values
844 * Look for next [\W-_], which will indicate the end of the tag name.
846 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
847 if ($regary == false){
848 return Array(false, false, false, $lt, strlen($body));
850 list($pos, $tagname, $match) = $regary;
851 $tagname = strtolower($tagname);
854 * $match can be either of these:
855 * '>' indicating the end of the tag entirely.
856 * '\s' indicating the end of the tag name.
857 * '/' indicating that this is type-3 xhtml tag.
859 * Whatever else we find there indicates an invalid tag.
864 * This is an xhtml-style tag with a closing / at the
865 * end, like so: <img src="blah"/>. Check if it's followed
866 * by the closing bracket. If not, then this tag is invalid
868 if (substr($body, $pos, 2) == "/>"){
872 $gt = sq_findnxstr($body, $pos, ">");
873 $retary = Array(false, false, false, $lt, $gt);
877 return Array($tagname, false, $tagtype, $lt, $pos);
881 * Check if it's whitespace
883 if (!preg_match('/\s/', $match)){
885 * This is an invalid tag! Look for the next closing ">".
887 $gt = sq_findnxstr($body, $offset, ">");
888 return Array(false, false, false, $lt, $gt);
894 * At this point we're here:
895 * <tagname attribute='blah'>
898 * At this point we loop in order to find all attributes.
904 while ($pos <= strlen($body)){
905 $pos = sq_skipspace($body, $pos);
906 if ($pos == strlen($body)){
910 return Array(false, false, false, $lt, $pos);
913 * See if we arrived at a ">" or "/>", which means that we reached
914 * the end of the tag.
917 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
921 $pos +
= strlen($matches{1});
922 if ($matches{2} == "/>"){
926 return Array($tagname, $attary, $tagtype, $lt, $pos);
930 * There are several types of attributes, with optional
931 * [:space:] between members.
933 * attrname[:space:]=[:space:]'CDATA'
935 * attrname[:space:]=[:space:]"CDATA"
937 * attr[:space:]=[:space:]CDATA
941 * We leave types 1 and 2 the same, type 3 we check for
942 * '"' and convert to """ if needed, then wrap in
943 * double quotes. Type 4 we convert into:
946 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
947 if ($regary == false){
949 * Looks like body ended before the end of tag.
951 return Array(false, false, false, $lt, strlen($body));
953 list($pos, $attname, $match) = $regary;
954 $attname = strtolower($attname);
956 * We arrived at the end of attribute name. Several things possible
958 * '>' means the end of the tag and this is attribute type 4
959 * '/' if followed by '>' means the same thing as above
960 * '\s' means a lot of things -- look what it's followed by.
961 * anything else means the attribute is invalid.
966 * This is an xhtml-style tag with a closing / at the
967 * end, like so: <img src="blah"/>. Check if it's followed
968 * by the closing bracket. If not, then this tag is invalid
970 if (substr($body, $pos, 2) == "/>"){
974 $gt = sq_findnxstr($body, $pos, ">");
975 $retary = Array(false, false, false, $lt, $gt);
979 $attary{$attname} = '"yes"';
980 return Array($tagname, $attary, $tagtype, $lt, $pos);
984 * Skip whitespace and see what we arrive at.
986 $pos = sq_skipspace($body, $pos);
987 $char = substr($body, $pos, 1);
989 * Two things are valid here:
990 * '=' means this is attribute type 1 2 or 3.
991 * \w means this was attribute type 4.
992 * anything else we ignore and re-loop. End of tag and
993 * invalid stuff will be caught by our checks at the beginning
998 $pos = sq_skipspace($body, $pos);
1000 * Here are 3 possibilities:
1001 * "'" attribute type 1
1002 * '"' attribute type 2
1003 * everything else is the content of tag type 3
1005 $quot = substr($body, $pos, 1);
1007 $regary = sq_findnxreg($body, $pos+
1, "\'");
1008 if ($regary == false){
1009 return Array(false, false, false, $lt, strlen($body));
1011 list($pos, $attval, $match) = $regary;
1013 $attary{$attname} = "'" . $attval . "'";
1014 } else if ($quot == '"'){
1015 $regary = sq_findnxreg($body, $pos+
1, '\"');
1016 if ($regary == false){
1017 return Array(false, false, false, $lt, strlen($body));
1019 list($pos, $attval, $match) = $regary;
1021 $attary{$attname} = '"' . $attval . '"';
1024 * These are hateful. Look for \s, or >.
1026 $regary = sq_findnxreg($body, $pos, "[\s>]");
1027 if ($regary == false){
1028 return Array(false, false, false, $lt, strlen($body));
1030 list($pos, $attval, $match) = $regary;
1032 * If it's ">" it will be caught at the top.
1034 $attval = preg_replace("/\"/s", """, $attval);
1035 $attary{$attname} = '"' . $attval . '"';
1037 } else if (preg_match("|[\w/>]|", $char)) {
1039 * That was attribute type 4.
1041 $attary{$attname} = '"yes"';
1044 * An illegal character. Find next '>' and return.
1046 $gt = sq_findnxstr($body, $pos, ">");
1047 return Array(false, false, false, $lt, $gt);
1053 * The fact that we got here indicates that the tag end was never
1054 * found. Return invalid tag indication so it gets stripped.
1056 return Array(false, false, false, $lt, strlen($body));
1060 * This function checks attribute values for entity-encoded values
1061 * and returns them translated into 8-bit strings so we can run
1064 * @param $attvalue A string to run entity check against.
1065 * @return Translated value.
1067 function sq_deent($attvalue){
1070 * See if we have to run the checks first. All entities must start
1073 if (strpos($attvalue, "&") === false){
1077 * Check named entities first.
1079 $trans = get_html_translation_table(HTML_ENTITIES
);
1081 * Leave " in, as it can mess us up.
1083 $trans = array_flip($trans);
1084 unset($trans{"""});
1085 while (list($ent, $val) = each($trans)){
1086 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1089 * Now translate numbered entities from 1 to 255 if needed.
1091 if (strpos($attvalue, "#") !== false){
1092 $omit = Array(34, 39);
1093 for ($asc=1; $asc<256; $asc++
){
1094 if (!in_array($asc, $omit)){
1096 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1098 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1099 "$chr\\1", $attvalue);
1107 * This function runs various checks against the attributes.
1109 * @param $tagname String with the name of the tag.
1110 * @param $attary Array with all tag attributes.
1111 * @param $rm_attnames See description for sq_sanitize
1112 * @param $bad_attvals See description for sq_sanitize
1113 * @param $add_attr_to_tag See description for sq_sanitize
1114 * @param $message message object
1115 * @param $id message id
1116 * @return Array with modified attributes.
1118 function sq_fixatts($tagname,
1128 while (list($attname, $attvalue) = each($attary)){
1130 * See if this attribute should be removed.
1132 foreach ($rm_attnames as $matchtag=>$matchattrs){
1133 if (preg_match($matchtag, $tagname)){
1134 foreach ($matchattrs as $matchattr){
1135 if (preg_match($matchattr, $attname)){
1136 unset($attary{$attname});
1143 * Remove any entities.
1145 $attvalue = sq_deent($attvalue);
1148 * Now let's run checks on the attvalues.
1149 * I don't expect anyone to comprehend this. If you do,
1150 * get in touch with me so I can drive to where you live and
1151 * shake your hand personally. :)
1153 foreach ($bad_attvals as $matchtag=>$matchattrs){
1154 if (preg_match($matchtag, $tagname)){
1155 foreach ($matchattrs as $matchattr=>$valary){
1156 if (preg_match($matchattr, $attname)){
1158 * There are two arrays in valary.
1160 * Second one is replacements
1162 list($valmatch, $valrepl) = $valary;
1164 preg_replace($valmatch, $valrepl, $attvalue);
1165 if ($newvalue != $attvalue){
1166 $attary{$attname} = $newvalue;
1173 * Turn cid: urls into http-friendly ones.
1175 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1176 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1180 * See if we need to append any attributes to this tag.
1182 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1183 if (preg_match($matchtag, $tagname)){
1184 $attary = array_merge($attary, $addattary);
1191 * This function edits the style definition to make them friendly and
1192 * usable in squirrelmail.
1194 * @param $message the message object
1195 * @param $id the message id
1196 * @param $content a string with whatever is between <style> and </style>
1197 * @return a string with edited content.
1199 function sq_fixstyle($message, $id, $content){
1200 global $view_unsafe_images;
1201 $me = 'sq_fixstyle';
1203 * First look for general BODY style declaration, which would be
1205 * body {background: blah-blah}
1206 * and change it to .bodyclass so we can just assign it to a <div>
1208 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1209 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1211 * Fix url('blah') declarations.
1213 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1214 "url(\\1$secremoveimg\\2)", $content);
1216 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1219 if (!$view_unsafe_images){
1220 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1221 "url(\\1$secremoveimg\\2)", $content);
1225 * Fix urls that refer to cid:
1227 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1229 $cidurl = $matches{1};
1230 $httpurl = sq_cid2http($message, $id, $cidurl);
1231 $content = preg_replace("|url\($cidurl\)|si",
1232 "url($httpurl)", $content);
1236 * Fix stupid css declarations which lead to vulnerabilities
1239 $match = Array('/expression/si',
1242 $replace = Array('idiocy', 'idiocy', 'idiocy');
1243 $content = preg_replace($match, $replace, $content);
1248 * This function converts cid: url's into the ones that can be viewed in
1251 * @param $message the message object
1252 * @param $id the message id
1253 * @param $cidurl the cid: url.
1254 * @return a string with a http-friendly url
1256 function sq_cid2http($message, $id, $cidurl, $mailbox){
1258 * Get rid of quotes.
1260 $quotchar = substr($cidurl, 0, 1);
1261 $cidurl = str_replace($quotchar, "", $cidurl);
1262 $cidurl = substr(trim($cidurl), 4);
1263 $linkurl = find_ent_id($cidurl, $message);
1264 /* in case of non-save cid links $httpurl should be replaced by a sort of
1265 unsave link image */
1268 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&' .
1269 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1270 '&ent_id=' . $linkurl . $quotchar;
1276 * This function changes the <body> tag into a <div> tag since we
1277 * can't really have a body-within-body.
1279 * @param $attary an array of attributes and values of <body>
1280 * @return a modified array of attributes to be set for <div>
1282 function sq_body2div($attary){
1283 $me = 'sq_body2div';
1284 $divattary = Array('class' => "'bodyclass'");
1285 $bgcolor = '#ffffff';
1288 if (is_array($attary) && sizeof($attary) > 0){
1289 foreach ($attary as $attname=>$attvalue){
1290 $quotchar = substr($attvalue, 0, 1);
1291 $attvalue = str_replace($quotchar, "", $attvalue);
1294 $styledef .= "background-image: url('$attvalue'); ";
1297 $styledef .= "background-color: $attvalue; ";
1300 $styledef .= "color: $attvalue; ";
1304 if (strlen($styledef) > 0){
1305 $divattary{"style"} = "\"$styledef\"";
1312 * This is the main function and the one you should actually be calling.
1313 * There are several variables you should be aware of an which need
1314 * special description.
1316 * Since the description is quite lengthy, see it here:
1317 * http://www.mricon.com/html/phpfilter.html
1319 * @param $body the string with HTML you wish to filter
1320 * @param $tag_list see description above
1321 * @param $rm_tags_with_content see description above
1322 * @param $self_closing_tags see description above
1323 * @param $force_tag_closing see description above
1324 * @param $rm_attnames see description above
1325 * @param $bad_attvals see description above
1326 * @param $add_attr_to_tag see description above
1327 * @param $message message object
1328 * @param $id message id
1329 * @return sanitized html safe to show on your pages.
1331 function sq_sanitize($body,
1333 $rm_tags_with_content,
1343 $me = 'sq_sanitize';
1345 * Normalize rm_tags and rm_tags_with_content.
1347 @array_walk
($rm_tags, 'sq_casenormalize');
1348 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1349 @array_walk
($self_closing_tags, 'sq_casenormalize');
1351 * See if tag_list is of tags to remove or tags to allow.
1352 * false means remove these tags
1353 * true means allow these tags
1355 $rm_tags = array_shift($tag_list);
1357 $open_tags = Array();
1358 $trusted = "<!-- begin sanitized html -->\n";
1359 $skip_content = false;
1361 * Take care of netscape's stupid javascript entities like
1364 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1366 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1367 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1368 $free_content = substr($body, $curpos, $lt-$curpos);
1370 * Take care of <style>
1372 if ($tagname == "style" && $tagtype == 2){
1374 * This is a closing </style>. Edit the
1375 * content before we apply it.
1377 $free_content = sq_fixstyle($message, $id, $free_content);
1379 if ($skip_content == false){
1380 $trusted .= $free_content;
1382 if ($tagname != FALSE){
1384 if ($skip_content == $tagname){
1386 * Got to the end of tag we needed to remove.
1389 $skip_content = false;
1391 if ($skip_content == false){
1392 if ($tagname == "body"){
1395 if (isset($open_tags{$tagname}) &&
1396 $open_tags{$tagname} > 0){
1397 $open_tags{$tagname}--;
1406 * $rm_tags_with_content
1408 if ($skip_content == false){
1410 * See if this is a self-closing type and change
1411 * tagtype appropriately.
1414 && in_array($tagname, $self_closing_tags)){
1418 * See if we should skip this tag and any content
1421 if ($tagtype == 1 &&
1422 in_array($tagname, $rm_tags_with_content)){
1423 $skip_content = $tagname;
1425 if (($rm_tags == false
1426 && in_array($tagname, $tag_list)) ||
1427 ($rm_tags == true &&
1428 !in_array($tagname, $tag_list))){
1432 if (isset($open_tags{$tagname})){
1433 $open_tags{$tagname}++
;
1435 $open_tags{$tagname}=1;
1439 * This is where we run other checks.
1441 if (is_array($attary) && sizeof($attary) > 0){
1442 $attary = sq_fixatts($tagname,
1453 * Convert body into div.
1455 if ($tagname == "body"){
1457 $attary = sq_body2div($attary, $message, $id);
1463 if ($tagname != false && $skip_content == false){
1464 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1469 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1470 if ($force_tag_closing == true){
1471 foreach ($open_tags as $tagname=>$opentimes){
1472 while ($opentimes > 0){
1473 $trusted .= '</' . $tagname . '>';
1479 $trusted .= "<!-- end sanitized html -->\n";
1484 * This is a wrapper function to call html sanitizing routines.
1486 * @param $body the body of the message
1487 * @param $id the id of the message
1488 * @return a string with html safe to display in the browser.
1490 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1491 global $attachment_common_show_images, $view_unsafe_images,
1494 * Don't display attached images in HTML mode.
1496 $attachment_common_show_images = false;
1509 $rm_tags_with_content = Array(
1516 $self_closing_tags = Array(
1523 $force_tag_closing = false;
1525 $rm_attnames = Array(
1535 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1536 $bad_attvals = Array(
1539 "/^src|background/i" =>
1542 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1543 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1544 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1545 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1548 "\\1$secremoveimg\\2",
1549 "\\1$secremoveimg\\2",
1550 "\\1$secremoveimg\\2",
1551 "\\1$secremoveimg\\2"
1554 "/^href|action/i" =>
1557 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1558 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1559 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1560 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1575 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1576 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1577 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1578 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1592 if (!$view_unsafe_images){
1594 * Remove any references to http/https if view_unsafe_images set
1597 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1598 '/^([\'\"])\s*https*:.*([\'\"])/si');
1599 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1600 "\\1$secremoveimg\\2");
1601 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1602 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1603 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1604 "url(\\1$secremoveimg\\2)");
1607 $add_attr_to_tag = Array(
1608 "/^a$/si" => Array('target'=>'"_new"')
1610 $trusted = sq_sanitize($body,
1612 $rm_tags_with_content,
1622 if (preg_match("|$secremoveimg|si", $trusted)){
1623 $has_unsafe_images = true;