19fa5a83adde004d75fe329cae722b3a0134b4b2
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));
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, $color );
44 echo '</body></html>';
49 foreach ($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) {
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.
105 $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, $encoding) {
157 /* Do a bit of error correction. If we couldn't find the entity id, just guess
158 * that it is the first one. That is usually the case anyway.
163 $sid = sqimap_session_id($uid_support);
164 /* Don't kill the connection if the browser is over a dialup
165 * and it would take over 30 seconds to download it.
166 * DonĀ“t call set_time_limit in safe mode.
169 if (!ini_get('safe_mode')) {
173 $sid_s = substr($sid,0,strpos($sid, ' '));
178 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
179 echo decodeBody($body, $encoding);
182 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
185 $read = fgets ($imap_stream,8192);
188 // This could be bad -- if the section has sqimap_session_id() . ' OK'
189 // or similar, it will kill the download.
190 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
191 if (trim($read) == ')==') {
193 $read = fgets ($imap_stream,4096);
194 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
197 echo decodeBody($read1, $encoding) .
198 decodeBody($read, $encoding);
201 echo decodeBody($read, $encoding);
203 $read = fgets ($imap_stream,4096);
210 /* -[ END MIME DECODING ]----------------------------------------------------------- */
212 /* This is here for debugging purposes. It will print out a list
213 * of all the entity IDs that are in the $message object.
215 function listEntities ($message) {
217 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
218 for ($i = 0; isset($message->entities
[$i]); $i++
) {
220 $msg = listEntities($message->entities
[$i]);
230 function getPriorityStr($priority) {
231 $priority_level = substr($priority,0,1);
233 switch($priority_level) {
234 /* Check for a higher then normal priority. */
237 $priority_string = _("High");
240 /* Check for a lower then normal priority. */
243 $priority_string = _("Low");
246 /* Check for a normal priority. */
249 $priority_level = '3';
250 $priority_string = _("Normal");
254 return $priority_string;
257 /* returns a $message object for a particular entity id */
258 function getEntity ($message, $ent_id) {
259 return $message->getEntity($ent_id);
263 * Extracted from strings.php 23/03/2002
266 function translateText(&$body, $wrap_at, $charset) {
267 global $where, $what; /* from searching */
268 global $color; /* color theme */
270 require_once(SM_PATH
. 'functions/url_parser.php');
272 $body_ary = explode("\n", $body);
273 for ($i=0; $i < count($body_ary); $i++
) {
274 $line = $body_ary[$i];
275 if (strlen($line) - 2 >= $wrap_at) {
276 sqWordWrap($line, $wrap_at);
278 $line = charset_decode($charset, $line);
279 $line = str_replace("\t", ' ', $line);
288 if ($line[$pos] == ' ') {
290 } else if (strpos($line, '>', $pos) === $pos) {
299 if (!isset($color[14])) {
300 $color[14] = '#FF0000';
302 $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
304 if (!isset($color[13])) {
305 $color[13] = '#800000';
307 $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
310 $body_ary[$i] = $line;
312 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
316 /* This returns a parsed string called $body. That string can then
317 * be displayed as the actual message in the HTML. It contains
318 * everything needed, including HTML Tags, Attachments at the
321 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
322 /* This if statement checks for the entity to show as the
323 * primary message. To add more of them, just put them in the
324 * order that is their priority.
326 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
327 $show_html_default, $has_unsafe_images, $sort;
329 if ( !check_php_version(4,1) ) {
332 if(isset($_GET['view_unsafe_images'])) {
333 $view_unsafe_images = $_GET['view_unsafe_images'];
336 $has_unsafe_images= 0;
338 $urlmailbox = urlencode($mailbox);
339 $body_message = getEntity($message, $ent_num);
340 if (($body_message->header
->type0
== 'text') ||
341 ($body_message->header
->type0
== 'rfc822')) {
342 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
343 $body = decodeBody($body, $body_message->header
->encoding
);
344 $hookResults = do_hook("message_body", $body);
345 $body = $hookResults[1];
347 /* If there are other types that shouldn't be formatted, add
351 if ($body_message->header
->type1
== 'html') {
352 if ($show_html_default <> 1) {
353 $entity_conv = array(' ' => ' ',
360 $body = strtr($body, $entity_conv);
361 $body = strip_tags($body);
363 translateText($body, $wrap_at,
364 $body_message->header
->getParameter('charset'));
366 $body = magicHTML($body, $id, $message, $mailbox);
369 translateText($body, $wrap_at,
370 $body_message->header
->getParameter('charset'));
373 if ($has_unsafe_images) {
374 if ($view_unsafe_images) {
375 $untext = '">' . _("Hide Unsafe Images");
377 $untext = '&view_unsafe_images=1">' . _("View Unsafe Images");
379 $body .= '<center><small><a href="read_body.php?passed_id=' . $id .
380 '&passed_ent_id=' . $message->entity_id
. '&mailbox=' . $urlmailbox .
381 '&sort=' . $sort . '&startMessage=' . $startMessage . '&show_more=0' .
382 $untext . '</a></small></center><br>' . "\n";
389 function formatAttachments($message, $exclude_id, $mailbox, $id) {
390 global $where, $what, $startMessage, $color;
391 static $ShownHTML = 0;
393 $att_ar = $message->getAttachments($exclude_id);
395 if (!count($att_ar)) return '';
399 $urlMailbox = urlencode($mailbox);
401 foreach ($att_ar as $att) {
402 $ent = urldecode($att->entity_id
);
403 $header = $att->header
;
404 $type0 = strtolower($header->type0
);
405 $type1 = strtolower($header->type1
);
407 $links['download link']['text'] = _("download");
408 $links['download link']['href'] =
409 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
411 if ($type0 =='message' && $type1 == 'rfc822') {
412 $default_page = '../src/read_body.php';
413 $rfc822_header = $att->rfc822_header
;
414 $filename = decodeHeader($rfc822_header->subject
);
415 if (trim( $filename ) == '') {
416 $filename = 'untitled-[' . $ent . ']' ;
418 $from_o = $rfc822_header->from
;
419 if (is_object($from_o)) {
420 $from_name = $from_o->getAddress(false);
422 $from_name = _("Unknown sender");
424 $from_name = decodeHeader(htmlspecialchars($from_name));
425 $description = $from_name;
427 $default_page = '../src/download.php';
428 if (is_object($header->disposition
)) {
429 $filename = decodeHeader($header->disposition
->getProperty('filename'));
430 if (trim($filename) == '') {
431 $name = decodeHeader($header->disposition
->getProperty('name'));
432 if (trim($name) == '') {
433 if (trim( $header->id
) == '') {
434 $filename = 'untitled-[' . $ent . ']' ;
436 $filename = 'cid: ' . $header->id
;
443 if (trim( $header->id
) == '') {
444 $filename = 'untitled-[' . $ent . ']' ;
446 $filename = 'cid: ' . $header->id
;
450 if ($header->description
) {
451 $description = htmlspecialchars($header->description
);
457 $display_filename = $filename;
458 if (isset($passed_ent_id)) {
459 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
461 $passed_ent_id_link = '';
463 $defaultlink = $default_page . "?startMessage=$startMessage"
464 . "&passed_id=$id&mailbox=$urlMailbox"
465 . '&ent_id='.$ent.$passed_ent_id_link.'&absolute_dl=true';
466 if ($where && $what) {
467 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
469 /* This executes the attachment hook with a specific MIME-type.
470 * If that doesn't have results, it tries if there's a rule
471 * for a more generic type.
473 $hookresults = do_hook("attachment $type0/$type1", $links,
474 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
475 $display_filename, $where, $what);
476 if(count($hookresults[1]) <= 1) {
477 $hookresults = do_hook("attachment $type0/*", $links,
478 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
479 $display_filename, $where, $what);
482 $links = $hookresults[1];
483 $defaultlink = $hookresults[6];
485 $attachments .= '<TR><TD>' .
486 "<A HREF=\"$defaultlink\">$display_filename</A> </TD>" .
487 '<TD><SMALL><b>' . show_readable_size($header->size
) .
488 '</b> </small></TD>' .
489 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
491 $attachments .= '<b>' . $description . '</b>';
492 $attachments .= '</SMALL></TD><TD><SMALL> ';
495 foreach ($links as $val) {
499 $attachments .= ' | ';
501 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
504 $attachments .= "</TD></TR>\n";
509 /* This function decodes the body depending on the encoding type. */
510 function decodeBody($body, $encoding) {
511 global $languages, $squirrelmail_language;
512 global $show_html_default;
514 $body = str_replace("\r\n", "\n", $body);
515 $encoding = strtolower($encoding);
517 if ($encoding == 'quoted-printable' ||
518 $encoding == 'quoted_printable') {
519 $body = quoted_printable_decode($body);
521 while (ereg("=\n", $body)) {
522 $body = ereg_replace ("=\n", '', $body);
525 } else if ($encoding == 'base64') {
526 $body = base64_decode($body);
529 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
530 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
531 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
534 // All other encodings are returned raw.
539 * This functions decode strings that is encoded according to
540 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
541 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
543 function decodeHeader ($string, $utfencode=true) {
544 global $languages, $squirrelmail_language;
545 if (is_array($string)) {
546 $string = implode("\n", $string);
549 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
550 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
551 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
555 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
558 /* Ignore white-space between consecutive encoded-words. */
559 if (strspn($res[2], " \t") != strlen($res[2])) {
563 if (ucfirst($res[4]) == 'B') {
564 $replace = base64_decode($res[5]);
566 $replace = str_replace('_', ' ', $res[5]);
567 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
569 /* Only encode into entities by default. Some places
570 * don't need the encoding, like the compose form.
573 $replace = charset_decode($res[3], $replace);
576 $string = $prefix . $replace . substr($string, strlen($res[0]));
577 $i = strlen($prefix) +
strlen($replace);
583 * Encode a string according to RFC 1522 for use in headers if it
584 * contains 8-bit characters or anything that looks like it should
587 function encodeHeader ($string) {
588 global $default_charset, $languages, $squirrelmail_language;
590 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
591 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
592 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
595 // Encode only if the string contains 8-bit characters or =?
596 $j = strlen($string);
597 $l = strstr($string, '=?'); // Must be encoded ?
599 for($i = 0; $i < $j; ++
$i) {
600 switch($string{$i}) {
614 $k = ord($string{$i});
616 $ret .= sprintf("=%02X", $k);
626 $string = "=?$default_charset?Q?$ret?=";
632 /* This function trys to locate the entity_id of a specific mime element */
633 function find_ent_id($id, $message) {
634 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
635 if ($message->entities
[$i]->header
->type0
== 'multipart') {
636 $ret = find_ent_id($id, $message->entities
[$i]);
638 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
639 // if (sq_check_save_extension($message->entities[$i])) {
640 return $message->entities
[$i]->entity_id
;
648 function sq_check_save_extension($message) {
649 $filename = $message->getFilename();
650 $ext = substr($filename, strrpos($filename,'.')+
1);
651 $save_extensions = array('jpg','jpeg','gif','png','bmp');
652 return in_array($ext, $save_extensions);
657 ** HTMLFILTER ROUTINES
661 * This function returns the final tag out of the tag name, an array
662 * of attributes, and the type of the tag. This function is called by
663 * sq_sanitize internally.
665 * @param $tagname the name of the tag.
666 * @param $attary the array of attributes and their values
667 * @param $tagtype The type of the tag (see in comments).
668 * @return a string with the final tag representation.
670 function sq_tagprint($tagname, $attary, $tagtype){
674 $fulltag = '</' . $tagname . '>';
676 $fulltag = '<' . $tagname;
677 if (is_array($attary) && sizeof($attary)){
679 while (list($attname, $attvalue) = each($attary)){
680 array_push($atts, "$attname=$attvalue");
682 $fulltag .= ' ' . join(" ", $atts);
693 * A small helper function to use with array_walk. Modifies a by-ref
694 * value and makes it lowercase.
696 * @param $val a value passed by-ref.
697 * @return void since it modifies a by-ref value.
699 function sq_casenormalize(&$val){
700 $val = strtolower($val);
704 * This function skips any whitespace from the current position within
705 * a string and to the next non-whitespace value.
707 * @param $body the string
708 * @param $offset the offset within the string where we should start
709 * looking for the next non-whitespace character.
710 * @return the location within the $body where the next
711 * non-whitespace char is located.
713 function sq_skipspace($body, $offset){
714 $me = 'sq_skipspace';
715 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
716 if (sizeof($matches{1})){
717 $count = strlen($matches{1});
724 * This function looks for the next character within a string. It's
725 * really just a glorified "strpos", except it catches if failures
728 * @param $body The string to look for needle in.
729 * @param $offset Start looking from this position.
730 * @param $needle The character/string to look for.
731 * @return location of the next occurance of the needle, or
732 * strlen($body) if needle wasn't found.
734 function sq_findnxstr($body, $offset, $needle){
735 $me = 'sq_findnxstr';
736 $pos = strpos($body, $needle, $offset);
738 $pos = strlen($body);
744 * This function takes a PCRE-style regexp and tries to match it
747 * @param $body The string to look for needle in.
748 * @param $offset Start looking from here.
749 * @param $reg A PCRE-style regex to match.
750 * @return Returns a false if no matches found, or an array
751 * with the following members:
752 * - integer with the location of the match within $body
753 * - string with whatever content between offset and the match
754 * - string with whatever it is we matched
756 function sq_findnxreg($body, $offset, $reg){
757 $me = 'sq_findnxreg';
760 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
764 $retarr{0} = $offset +
strlen($matches{1});
765 $retarr{1} = $matches{1};
766 $retarr{2} = $matches{2};
772 * This function looks for the next tag.
774 * @param $body String where to look for the next tag.
775 * @param $offset Start looking from here.
776 * @return false if no more tags exist in the body, or
777 * an array with the following members:
778 * - string with the name of the tag
779 * - array with attributes and their values
780 * - integer with tag type (1, 2, or 3)
781 * - integer where the tag starts (starting "<")
782 * - integer where the tag ends (ending ">")
783 * first three members will be false, if the tag is invalid.
785 function sq_getnxtag($body, $offset){
787 if ($offset > strlen($body)){
790 $lt = sq_findnxstr($body, $offset, "<");
791 if ($lt == strlen($body)){
796 * blah blah <tag attribute="value">
799 $pos = sq_skipspace($body, $lt+
1);
800 if ($pos >= strlen($body)){
801 return Array(false, false, false, $lt, strlen($body));
804 * There are 3 kinds of tags:
805 * 1. Opening tag, e.g.:
807 * 2. Closing tag, e.g.:
809 * 3. XHTML-style content-less tag, e.g.:
813 switch (substr($body, $pos, 1)){
820 * A comment or an SGML declaration.
822 if (substr($body, $pos+
1, 2) == "--"){
823 $gt = strpos($body, "-->", $pos);
829 return Array(false, false, false, $lt, $gt);
831 $gt = sq_findnxstr($body, $pos, ">");
832 return Array(false, false, false, $lt, $gt);
837 * Assume tagtype 1 for now. If it's type 3, we'll switch values
847 * Look for next [\W-_], which will indicate the end of the tag name.
849 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
850 if ($regary == false){
851 return Array(false, false, false, $lt, strlen($body));
853 list($pos, $tagname, $match) = $regary;
854 $tagname = strtolower($tagname);
857 * $match can be either of these:
858 * '>' indicating the end of the tag entirely.
859 * '\s' indicating the end of the tag name.
860 * '/' indicating that this is type-3 xhtml tag.
862 * Whatever else we find there indicates an invalid tag.
867 * This is an xhtml-style tag with a closing / at the
868 * end, like so: <img src="blah"/>. Check if it's followed
869 * by the closing bracket. If not, then this tag is invalid
871 if (substr($body, $pos, 2) == "/>"){
875 $gt = sq_findnxstr($body, $pos, ">");
876 $retary = Array(false, false, false, $lt, $gt);
880 return Array($tagname, false, $tagtype, $lt, $pos);
884 * Check if it's whitespace
886 if (!preg_match('/\s/', $match)){
888 * This is an invalid tag! Look for the next closing ">".
890 $gt = sq_findnxstr($body, $offset, ">");
891 return Array(false, false, false, $lt, $gt);
897 * At this point we're here:
898 * <tagname attribute='blah'>
901 * At this point we loop in order to find all attributes.
907 while ($pos <= strlen($body)){
908 $pos = sq_skipspace($body, $pos);
909 if ($pos == strlen($body)){
913 return Array(false, false, false, $lt, $pos);
916 * See if we arrived at a ">" or "/>", which means that we reached
917 * the end of the tag.
920 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
924 $pos +
= strlen($matches{1});
925 if ($matches{2} == "/>"){
929 return Array($tagname, $attary, $tagtype, $lt, $pos);
933 * There are several types of attributes, with optional
934 * [:space:] between members.
936 * attrname[:space:]=[:space:]'CDATA'
938 * attrname[:space:]=[:space:]"CDATA"
940 * attr[:space:]=[:space:]CDATA
944 * We leave types 1 and 2 the same, type 3 we check for
945 * '"' and convert to """ if needed, then wrap in
946 * double quotes. Type 4 we convert into:
949 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
950 if ($regary == false){
952 * Looks like body ended before the end of tag.
954 return Array(false, false, false, $lt, strlen($body));
956 list($pos, $attname, $match) = $regary;
957 $attname = strtolower($attname);
959 * We arrived at the end of attribute name. Several things possible
961 * '>' means the end of the tag and this is attribute type 4
962 * '/' if followed by '>' means the same thing as above
963 * '\s' means a lot of things -- look what it's followed by.
964 * anything else means the attribute is invalid.
969 * This is an xhtml-style tag with a closing / at the
970 * end, like so: <img src="blah"/>. Check if it's followed
971 * by the closing bracket. If not, then this tag is invalid
973 if (substr($body, $pos, 2) == "/>"){
977 $gt = sq_findnxstr($body, $pos, ">");
978 $retary = Array(false, false, false, $lt, $gt);
982 $attary{$attname} = '"yes"';
983 return Array($tagname, $attary, $tagtype, $lt, $pos);
987 * Skip whitespace and see what we arrive at.
989 $pos = sq_skipspace($body, $pos);
990 $char = substr($body, $pos, 1);
992 * Two things are valid here:
993 * '=' means this is attribute type 1 2 or 3.
994 * \w means this was attribute type 4.
995 * anything else we ignore and re-loop. End of tag and
996 * invalid stuff will be caught by our checks at the beginning
1001 $pos = sq_skipspace($body, $pos);
1003 * Here are 3 possibilities:
1004 * "'" attribute type 1
1005 * '"' attribute type 2
1006 * everything else is the content of tag type 3
1008 $quot = substr($body, $pos, 1);
1010 $regary = sq_findnxreg($body, $pos+
1, "\'");
1011 if ($regary == false){
1012 return Array(false, false, false, $lt, strlen($body));
1014 list($pos, $attval, $match) = $regary;
1016 $attary{$attname} = "'" . $attval . "'";
1017 } else if ($quot == '"'){
1018 $regary = sq_findnxreg($body, $pos+
1, '\"');
1019 if ($regary == false){
1020 return Array(false, false, false, $lt, strlen($body));
1022 list($pos, $attval, $match) = $regary;
1024 $attary{$attname} = '"' . $attval . '"';
1027 * These are hateful. Look for \s, or >.
1029 $regary = sq_findnxreg($body, $pos, "[\s>]");
1030 if ($regary == false){
1031 return Array(false, false, false, $lt, strlen($body));
1033 list($pos, $attval, $match) = $regary;
1035 * If it's ">" it will be caught at the top.
1037 $attval = preg_replace("/\"/s", """, $attval);
1038 $attary{$attname} = '"' . $attval . '"';
1040 } else if (preg_match("|[\w/>]|", $char)) {
1042 * That was attribute type 4.
1044 $attary{$attname} = '"yes"';
1047 * An illegal character. Find next '>' and return.
1049 $gt = sq_findnxstr($body, $pos, ">");
1050 return Array(false, false, false, $lt, $gt);
1056 * The fact that we got here indicates that the tag end was never
1057 * found. Return invalid tag indication so it gets stripped.
1059 return Array(false, false, false, $lt, strlen($body));
1063 * This function checks attribute values for entity-encoded values
1064 * and returns them translated into 8-bit strings so we can run
1067 * @param $attvalue A string to run entity check against.
1068 * @return Translated value.
1070 function sq_deent($attvalue){
1073 * See if we have to run the checks first. All entities must start
1076 if (strpos($attvalue, "&") === false){
1080 * Check named entities first.
1082 $trans = get_html_translation_table(HTML_ENTITIES
);
1084 * Leave " in, as it can mess us up.
1086 $trans = array_flip($trans);
1087 unset($trans{"""});
1088 while (list($ent, $val) = each($trans)){
1089 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1092 * Now translate numbered entities from 1 to 255 if needed.
1094 if (strpos($attvalue, "#") !== false){
1095 $omit = Array(34, 39);
1096 for ($asc=1; $asc<256; $asc++
){
1097 if (!in_array($asc, $omit)){
1099 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1101 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1102 "$chr\\1", $attvalue);
1110 * This function runs various checks against the attributes.
1112 * @param $tagname String with the name of the tag.
1113 * @param $attary Array with all tag attributes.
1114 * @param $rm_attnames See description for sq_sanitize
1115 * @param $bad_attvals See description for sq_sanitize
1116 * @param $add_attr_to_tag See description for sq_sanitize
1117 * @param $message message object
1118 * @param $id message id
1119 * @return Array with modified attributes.
1121 function sq_fixatts($tagname,
1131 while (list($attname, $attvalue) = each($attary)){
1133 * See if this attribute should be removed.
1135 foreach ($rm_attnames as $matchtag=>$matchattrs){
1136 if (preg_match($matchtag, $tagname)){
1137 foreach ($matchattrs as $matchattr){
1138 if (preg_match($matchattr, $attname)){
1139 unset($attary{$attname});
1146 * Remove any entities.
1148 $attvalue = sq_deent($attvalue);
1151 * Now let's run checks on the attvalues.
1152 * I don't expect anyone to comprehend this. If you do,
1153 * get in touch with me so I can drive to where you live and
1154 * shake your hand personally. :)
1156 foreach ($bad_attvals as $matchtag=>$matchattrs){
1157 if (preg_match($matchtag, $tagname)){
1158 foreach ($matchattrs as $matchattr=>$valary){
1159 if (preg_match($matchattr, $attname)){
1161 * There are two arrays in valary.
1163 * Second one is replacements
1165 list($valmatch, $valrepl) = $valary;
1167 preg_replace($valmatch, $valrepl, $attvalue);
1168 if ($newvalue != $attvalue){
1169 $attary{$attname} = $newvalue;
1176 * Turn cid: urls into http-friendly ones.
1178 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1179 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1183 * See if we need to append any attributes to this tag.
1185 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1186 if (preg_match($matchtag, $tagname)){
1187 $attary = array_merge($attary, $addattary);
1194 * This function edits the style definition to make them friendly and
1195 * usable in squirrelmail.
1197 * @param $message the message object
1198 * @param $id the message id
1199 * @param $content a string with whatever is between <style> and </style>
1200 * @return a string with edited content.
1202 function sq_fixstyle($message, $id, $content){
1203 global $view_unsafe_images;
1204 $me = 'sq_fixstyle';
1206 * First look for general BODY style declaration, which would be
1208 * body {background: blah-blah}
1209 * and change it to .bodyclass so we can just assign it to a <div>
1211 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1212 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1214 * Fix url('blah') declarations.
1216 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1217 "url(\\1$secremoveimg\\2)", $content);
1219 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1222 if (!$view_unsafe_images){
1223 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1224 "url(\\1$secremoveimg\\2)", $content);
1228 * Fix urls that refer to cid:
1230 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1232 $cidurl = $matches{1};
1233 $httpurl = sq_cid2http($message, $id, $cidurl);
1234 $content = preg_replace("|url\($cidurl\)|si",
1235 "url($httpurl)", $content);
1239 * Fix stupid css declarations which lead to vulnerabilities
1242 $match = Array('/expression/si',
1245 $replace = Array('idiocy', 'idiocy', 'idiocy');
1246 $content = preg_replace($match, $replace, $content);
1251 * This function converts cid: url's into the ones that can be viewed in
1254 * @param $message the message object
1255 * @param $id the message id
1256 * @param $cidurl the cid: url.
1257 * @return a string with a http-friendly url
1259 function sq_cid2http($message, $id, $cidurl, $mailbox){
1261 * Get rid of quotes.
1263 $quotchar = substr($cidurl, 0, 1);
1264 $cidurl = str_replace($quotchar, "", $cidurl);
1265 $cidurl = substr(trim($cidurl), 4);
1266 $linkurl = find_ent_id($cidurl, $message);
1267 /* in case of non-save cid links $httpurl should be replaced by a sort of
1268 unsave link image */
1271 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&' .
1272 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1273 '&ent_id=' . $linkurl . $quotchar;
1279 * This function changes the <body> tag into a <div> tag since we
1280 * can't really have a body-within-body.
1282 * @param $attary an array of attributes and values of <body>
1283 * @return a modified array of attributes to be set for <div>
1285 function sq_body2div($attary){
1286 $me = 'sq_body2div';
1287 $divattary = Array('class' => "'bodyclass'");
1288 $bgcolor = '#ffffff';
1291 if (is_array($attary) && sizeof($attary) > 0){
1292 foreach ($attary as $attname=>$attvalue){
1293 $quotchar = substr($attvalue, 0, 1);
1294 $attvalue = str_replace($quotchar, "", $attvalue);
1297 $styledef .= "background-image: url('$attvalue'); ";
1300 $styledef .= "background-color: $attvalue; ";
1303 $styledef .= "color: $attvalue; ";
1307 if (strlen($styledef) > 0){
1308 $divattary{"style"} = "\"$styledef\"";
1315 * This is the main function and the one you should actually be calling.
1316 * There are several variables you should be aware of an which need
1317 * special description.
1319 * Since the description is quite lengthy, see it here:
1320 * http://www.mricon.com/html/phpfilter.html
1322 * @param $body the string with HTML you wish to filter
1323 * @param $tag_list see description above
1324 * @param $rm_tags_with_content see description above
1325 * @param $self_closing_tags see description above
1326 * @param $force_tag_closing see description above
1327 * @param $rm_attnames see description above
1328 * @param $bad_attvals see description above
1329 * @param $add_attr_to_tag see description above
1330 * @param $message message object
1331 * @param $id message id
1332 * @return sanitized html safe to show on your pages.
1334 function sq_sanitize($body,
1336 $rm_tags_with_content,
1346 $me = 'sq_sanitize';
1348 * Normalize rm_tags and rm_tags_with_content.
1350 @array_walk
($rm_tags, 'sq_casenormalize');
1351 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1352 @array_walk
($self_closing_tags, 'sq_casenormalize');
1354 * See if tag_list is of tags to remove or tags to allow.
1355 * false means remove these tags
1356 * true means allow these tags
1358 $rm_tags = array_shift($tag_list);
1360 $open_tags = Array();
1361 $trusted = "<!-- begin sanitized html -->\n";
1362 $skip_content = false;
1364 * Take care of netscape's stupid javascript entities like
1367 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1369 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1370 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1371 $free_content = substr($body, $curpos, $lt-$curpos);
1373 * Take care of <style>
1375 if ($tagname == "style" && $tagtype == 2){
1377 * This is a closing </style>. Edit the
1378 * content before we apply it.
1380 $free_content = sq_fixstyle($message, $id, $free_content);
1382 if ($skip_content == false){
1383 $trusted .= $free_content;
1385 if ($tagname != FALSE){
1387 if ($skip_content == $tagname){
1389 * Got to the end of tag we needed to remove.
1392 $skip_content = false;
1394 if ($skip_content == false){
1395 if ($tagname == "body"){
1398 if (isset($open_tags{$tagname}) &&
1399 $open_tags{$tagname} > 0){
1400 $open_tags{$tagname}--;
1409 * $rm_tags_with_content
1411 if ($skip_content == false){
1413 * See if this is a self-closing type and change
1414 * tagtype appropriately.
1417 && in_array($tagname, $self_closing_tags)){
1421 * See if we should skip this tag and any content
1424 if ($tagtype == 1 &&
1425 in_array($tagname, $rm_tags_with_content)){
1426 $skip_content = $tagname;
1428 if (($rm_tags == false
1429 && in_array($tagname, $tag_list)) ||
1430 ($rm_tags == true &&
1431 !in_array($tagname, $tag_list))){
1435 if (isset($open_tags{$tagname})){
1436 $open_tags{$tagname}++
;
1438 $open_tags{$tagname}=1;
1442 * This is where we run other checks.
1444 if (is_array($attary) && sizeof($attary) > 0){
1445 $attary = sq_fixatts($tagname,
1456 * Convert body into div.
1458 if ($tagname == "body"){
1460 $attary = sq_body2div($attary, $message, $id);
1466 if ($tagname != false && $skip_content == false){
1467 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1472 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1473 if ($force_tag_closing == true){
1474 foreach ($open_tags as $tagname=>$opentimes){
1475 while ($opentimes > 0){
1476 $trusted .= '</' . $tagname . '>';
1482 $trusted .= "<!-- end sanitized html -->\n";
1487 * This is a wrapper function to call html sanitizing routines.
1489 * @param $body the body of the message
1490 * @param $id the id of the message
1491 * @return a string with html safe to display in the browser.
1493 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1494 global $attachment_common_show_images, $view_unsafe_images,
1497 * Don't display attached images in HTML mode.
1499 $attachment_common_show_images = false;
1512 $rm_tags_with_content = Array(
1519 $self_closing_tags = Array(
1526 $force_tag_closing = false;
1528 $rm_attnames = Array(
1538 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1539 $bad_attvals = Array(
1542 "/^src|background/i" =>
1545 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1546 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1547 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1548 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1551 "\\1$secremoveimg\\2",
1552 "\\1$secremoveimg\\2",
1553 "\\1$secremoveimg\\2",
1554 "\\1$secremoveimg\\2"
1557 "/^href|action/i" =>
1560 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1561 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1562 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1563 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1578 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1579 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1580 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1581 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1595 if(isset($_GET['view_unsafe_images'])) {
1596 $view_unsafe_images = $_GET['view_unsafe_images'];
1598 if (!$view_unsafe_images){
1600 * Remove any references to http/https if view_unsafe_images set
1603 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1604 '/^([\'\"])\s*https*:.*([\'\"])/si');
1605 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1606 "\\1$secremoveimg\\2");
1607 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1608 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1609 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1610 "url(\\1$secremoveimg\\2)");
1613 $add_attr_to_tag = Array(
1614 "/^a$/si" => Array('target'=>'"_new"')
1616 $trusted = sq_sanitize($body,
1618 $rm_tags_with_content,
1628 if (preg_match("|$secremoveimg|si", $trusted)){
1629 $has_unsafe_images = true;