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, $view_unsafe_images, $sort;
328 $has_unsafe_images= 0;
330 $urlmailbox = urlencode($mailbox);
331 $body_message = getEntity($message, $ent_num);
332 if (($body_message->header
->type0
== 'text') ||
333 ($body_message->header
->type0
== 'rfc822')) {
334 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
335 $body = decodeBody($body, $body_message->header
->encoding
);
336 $hookResults = do_hook("message_body", $body);
337 $body = $hookResults[1];
339 /* If there are other types that shouldn't be formatted, add
343 if ($body_message->header
->type1
== 'html') {
344 if ($show_html_default <> 1) {
345 $entity_conv = array(' ' => ' ',
352 $body = strtr($body, $entity_conv);
353 $body = strip_tags($body);
355 translateText($body, $wrap_at,
356 $body_message->header
->getParameter('charset'));
358 $body = magicHTML($body, $id, $message, $mailbox);
361 translateText($body, $wrap_at,
362 $body_message->header
->getParameter('charset'));
365 if ($has_unsafe_images) {
366 if ($view_unsafe_images) {
367 $untext = '">' . _("Hide Unsafe Images");
369 $untext = '&view_unsafe_images=1">' . _("View Unsafe Images");
371 $body .= '<center><small><a href="read_body.php?passed_id=' . $id .
372 '&passed_ent_id=' . $message->entity_id
. '&mailbox=' . $urlmailbox .
373 '&sort=' . $sort . '&startMessage=' . $startMessage . '&show_more=0' .
374 $untext . '</a></small></center><br>' . "\n";
381 function formatAttachments($message, $exclude_id, $mailbox, $id) {
382 global $where, $what, $startMessage, $color;
383 static $ShownHTML = 0;
385 $att_ar = $message->getAttachments($exclude_id);
387 if (!count($att_ar)) return '';
391 $urlMailbox = urlencode($mailbox);
393 foreach ($att_ar as $att) {
394 $ent = urldecode($att->entity_id
);
395 $header = $att->header
;
396 $type0 = strtolower($header->type0
);
397 $type1 = strtolower($header->type1
);
399 $links['download link']['text'] = _("download");
400 $links['download link']['href'] =
401 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
403 if ($type0 =='message' && $type1 == 'rfc822') {
404 $default_page = '../src/read_body.php';
405 $rfc822_header = $att->rfc822_header
;
406 $filename = decodeHeader($rfc822_header->subject
);
408 $from_o = $rfc822_header->from
;
409 if (is_object($from_o)) {
410 $from_name = $from_o->getAddress(false);
412 $from_name = _("Unknown sender");
414 $from_name = decodeHeader(htmlspecialchars($from_name));
415 $description = $from_name;
417 $default_page = '../src/download.php';
418 if (is_object($header->disposition
)) {
419 $filename = decodeHeader($header->disposition
->getProperty('filename'));
420 if (trim($filename) == '') {
421 $name = decodeHeader($header->disposition
->getProperty('name'));
422 if (trim($name) == '') {
423 if (trim( $header->id
) == '') {
424 $filename = 'untitled-[' . $ent . ']' ;
426 $filename = 'cid: ' . $header->id
;
433 if (trim( $header->id
) == '') {
434 $filename = 'untitled-[' . $ent . ']' ;
436 $filename = 'cid: ' . $header->id
;
440 if ($header->description
) {
441 $description = htmlspecialchars($header->description
);
447 $display_filename = $filename;
448 if (isset($passed_ent_id)) {
449 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
451 $passed_ent_id_link = '';
453 $defaultlink = $default_page . "?startMessage=$startMessage"
454 . "&passed_id=$id&mailbox=$urlMailbox"
455 . '&ent_id='.$ent.$passed_ent_id_link;
456 if ($where && $what) {
457 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
459 /* This executes the attachment hook with a specific MIME-type.
460 * If that doesn't have results, it tries if there's a rule
461 * for a more generic type.
463 $hookresults = do_hook("attachment $type0/$type1", $links,
464 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
465 $display_filename, $where, $what);
466 if(count($hookresults[1]) <= 1) {
467 $hookresults = do_hook("attachment $type0/*", $links,
468 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
469 $display_filename, $where, $what);
472 $links = $hookresults[1];
473 $defaultlink = $hookresults[6];
475 $attachments .= '<TR><TD>' .
476 "<A HREF=\"$defaultlink\">$display_filename</A> </TD>" .
477 '<TD><SMALL><b>' . show_readable_size($header->size
) .
478 '</b> </small></TD>' .
479 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
481 $attachments .= '<b>' . $description . '</b>';
482 $attachments .= '</SMALL></TD><TD><SMALL> ';
485 foreach ($links as $val) {
489 $attachments .= ' | ';
491 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
494 $attachments .= "</TD></TR>\n";
499 /* This function decodes the body depending on the encoding type. */
500 function decodeBody($body, $encoding) {
501 global $languages, $squirrelmail_language;
502 global $show_html_default;
504 $body = str_replace("\r\n", "\n", $body);
505 $encoding = strtolower($encoding);
507 if ($encoding == 'quoted-printable' ||
508 $encoding == 'quoted_printable') {
509 $body = quoted_printable_decode($body);
511 while (ereg("=\n", $body)) {
512 $body = ereg_replace ("=\n", '', $body);
515 } else if ($encoding == 'base64') {
516 $body = base64_decode($body);
519 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
520 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
521 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
524 // All other encodings are returned raw.
529 * This functions decode strings that is encoded according to
530 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
531 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
533 function decodeHeader ($string, $utfencode=true) {
534 global $languages, $squirrelmail_language;
535 if (is_array($string)) {
536 $string = implode("\n", $string);
539 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
540 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
541 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
545 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
548 /* Ignore white-space between consecutive encoded-words. */
549 if (strspn($res[2], " \t") != strlen($res[2])) {
553 if (ucfirst($res[4]) == 'B') {
554 $replace = base64_decode($res[5]);
556 $replace = str_replace('_', ' ', $res[5]);
557 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
559 /* Only encode into entities by default. Some places
560 * don't need the encoding, like the compose form.
563 $replace = charset_decode($res[3], $replace);
566 $string = $prefix . $replace . substr($string, strlen($res[0]));
567 $i = strlen($prefix) +
strlen($replace);
573 * Encode a string according to RFC 1522 for use in headers if it
574 * contains 8-bit characters or anything that looks like it should
577 function encodeHeader ($string) {
578 global $default_charset, $languages, $squirrelmail_language;
580 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
581 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
582 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
585 // Encode only if the string contains 8-bit characters or =?
586 $j = strlen($string);
587 $l = strstr($string, '=?'); // Must be encoded ?
589 for($i = 0; $i < $j; ++
$i) {
590 switch($string{$i}) {
604 $k = ord($string{$i});
606 $ret .= sprintf("=%02X", $k);
616 $string = "=?$default_charset?Q?$ret?=";
622 /* This function trys to locate the entity_id of a specific mime element */
623 function find_ent_id($id, $message) {
624 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
625 if ($message->entities
[$i]->header
->type0
== 'multipart') {
626 $ret = find_ent_id($id, $message->entities
[$i]);
628 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
629 if (sq_check_save_extension($message->entities
[$i])) {
630 return $message->entities
[$i]->entity_id
;
638 function sq_check_save_extension($message) {
639 $filename = $message->getFilename();
640 $ext = substr($filename, strrpos($filename,'.')+
1);
641 $save_extensions = array('jpg','jpeg','gif','png','bmp');
642 return in_array($ext, $save_extensions);
647 ** HTMLFILTER ROUTINES
651 * This function returns the final tag out of the tag name, an array
652 * of attributes, and the type of the tag. This function is called by
653 * sq_sanitize internally.
655 * @param $tagname the name of the tag.
656 * @param $attary the array of attributes and their values
657 * @param $tagtype The type of the tag (see in comments).
658 * @return a string with the final tag representation.
660 function sq_tagprint($tagname, $attary, $tagtype){
664 $fulltag = '</' . $tagname . '>';
666 $fulltag = '<' . $tagname;
667 if (is_array($attary) && sizeof($attary)){
669 while (list($attname, $attvalue) = each($attary)){
670 array_push($atts, "$attname=$attvalue");
672 $fulltag .= ' ' . join(" ", $atts);
683 * A small helper function to use with array_walk. Modifies a by-ref
684 * value and makes it lowercase.
686 * @param $val a value passed by-ref.
687 * @return void since it modifies a by-ref value.
689 function sq_casenormalize(&$val){
690 $val = strtolower($val);
694 * This function skips any whitespace from the current position within
695 * a string and to the next non-whitespace value.
697 * @param $body the string
698 * @param $offset the offset within the string where we should start
699 * looking for the next non-whitespace character.
700 * @return the location within the $body where the next
701 * non-whitespace char is located.
703 function sq_skipspace($body, $offset){
704 $me = 'sq_skipspace';
705 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
706 if (sizeof($matches{1})){
707 $count = strlen($matches{1});
714 * This function looks for the next character within a string. It's
715 * really just a glorified "strpos", except it catches if failures
718 * @param $body The string to look for needle in.
719 * @param $offset Start looking from this position.
720 * @param $needle The character/string to look for.
721 * @return location of the next occurance of the needle, or
722 * strlen($body) if needle wasn't found.
724 function sq_findnxstr($body, $offset, $needle){
725 $me = 'sq_findnxstr';
726 $pos = strpos($body, $needle, $offset);
728 $pos = strlen($body);
734 * This function takes a PCRE-style regexp and tries to match it
737 * @param $body The string to look for needle in.
738 * @param $offset Start looking from here.
739 * @param $reg A PCRE-style regex to match.
740 * @return Returns a false if no matches found, or an array
741 * with the following members:
742 * - integer with the location of the match within $body
743 * - string with whatever content between offset and the match
744 * - string with whatever it is we matched
746 function sq_findnxreg($body, $offset, $reg){
747 $me = 'sq_findnxreg';
750 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
754 $retarr{0} = $offset +
strlen($matches{1});
755 $retarr{1} = $matches{1};
756 $retarr{2} = $matches{2};
762 * This function looks for the next tag.
764 * @param $body String where to look for the next tag.
765 * @param $offset Start looking from here.
766 * @return false if no more tags exist in the body, or
767 * an array with the following members:
768 * - string with the name of the tag
769 * - array with attributes and their values
770 * - integer with tag type (1, 2, or 3)
771 * - integer where the tag starts (starting "<")
772 * - integer where the tag ends (ending ">")
773 * first three members will be false, if the tag is invalid.
775 function sq_getnxtag($body, $offset){
777 if ($offset > strlen($body)){
780 $lt = sq_findnxstr($body, $offset, "<");
781 if ($lt == strlen($body)){
786 * blah blah <tag attribute="value">
789 $pos = sq_skipspace($body, $lt+
1);
790 if ($pos >= strlen($body)){
791 return Array(false, false, false, $lt, strlen($body));
794 * There are 3 kinds of tags:
795 * 1. Opening tag, e.g.:
797 * 2. Closing tag, e.g.:
799 * 3. XHTML-style content-less tag, e.g.:
803 switch (substr($body, $pos, 1)){
810 * A comment or an SGML declaration.
812 if (substr($body, $pos+
1, 2) == "--"){
813 $gt = strpos($body, "-->", $pos);
819 return Array(false, false, false, $lt, $gt);
821 $gt = sq_findnxstr($body, $pos, ">");
822 return Array(false, false, false, $lt, $gt);
827 * Assume tagtype 1 for now. If it's type 3, we'll switch values
837 * Look for next [\W-_], which will indicate the end of the tag name.
839 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
840 if ($regary == false){
841 return Array(false, false, false, $lt, strlen($body));
843 list($pos, $tagname, $match) = $regary;
844 $tagname = strtolower($tagname);
847 * $match can be either of these:
848 * '>' indicating the end of the tag entirely.
849 * '\s' indicating the end of the tag name.
850 * '/' indicating that this is type-3 xhtml tag.
852 * Whatever else we find there indicates an invalid tag.
857 * This is an xhtml-style tag with a closing / at the
858 * end, like so: <img src="blah"/>. Check if it's followed
859 * by the closing bracket. If not, then this tag is invalid
861 if (substr($body, $pos, 2) == "/>"){
865 $gt = sq_findnxstr($body, $pos, ">");
866 $retary = Array(false, false, false, $lt, $gt);
870 return Array($tagname, false, $tagtype, $lt, $pos);
874 * Check if it's whitespace
876 if (!preg_match('/\s/', $match)){
878 * This is an invalid tag! Look for the next closing ">".
880 $gt = sq_findnxstr($body, $offset, ">");
881 return Array(false, false, false, $lt, $gt);
887 * At this point we're here:
888 * <tagname attribute='blah'>
891 * At this point we loop in order to find all attributes.
897 while ($pos <= strlen($body)){
898 $pos = sq_skipspace($body, $pos);
899 if ($pos == strlen($body)){
903 return Array(false, false, false, $lt, $pos);
906 * See if we arrived at a ">" or "/>", which means that we reached
907 * the end of the tag.
910 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
914 $pos +
= strlen($matches{1});
915 if ($matches{2} == "/>"){
919 return Array($tagname, $attary, $tagtype, $lt, $pos);
923 * There are several types of attributes, with optional
924 * [:space:] between members.
926 * attrname[:space:]=[:space:]'CDATA'
928 * attrname[:space:]=[:space:]"CDATA"
930 * attr[:space:]=[:space:]CDATA
934 * We leave types 1 and 2 the same, type 3 we check for
935 * '"' and convert to """ if needed, then wrap in
936 * double quotes. Type 4 we convert into:
939 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
940 if ($regary == false){
942 * Looks like body ended before the end of tag.
944 return Array(false, false, false, $lt, strlen($body));
946 list($pos, $attname, $match) = $regary;
947 $attname = strtolower($attname);
949 * We arrived at the end of attribute name. Several things possible
951 * '>' means the end of the tag and this is attribute type 4
952 * '/' if followed by '>' means the same thing as above
953 * '\s' means a lot of things -- look what it's followed by.
954 * anything else means the attribute is invalid.
959 * This is an xhtml-style tag with a closing / at the
960 * end, like so: <img src="blah"/>. Check if it's followed
961 * by the closing bracket. If not, then this tag is invalid
963 if (substr($body, $pos, 2) == "/>"){
967 $gt = sq_findnxstr($body, $pos, ">");
968 $retary = Array(false, false, false, $lt, $gt);
972 $attary{$attname} = '"yes"';
973 return Array($tagname, $attary, $tagtype, $lt, $pos);
977 * Skip whitespace and see what we arrive at.
979 $pos = sq_skipspace($body, $pos);
980 $char = substr($body, $pos, 1);
982 * Two things are valid here:
983 * '=' means this is attribute type 1 2 or 3.
984 * \w means this was attribute type 4.
985 * anything else we ignore and re-loop. End of tag and
986 * invalid stuff will be caught by our checks at the beginning
991 $pos = sq_skipspace($body, $pos);
993 * Here are 3 possibilities:
994 * "'" attribute type 1
995 * '"' attribute type 2
996 * everything else is the content of tag type 3
998 $quot = substr($body, $pos, 1);
1000 $regary = sq_findnxreg($body, $pos+
1, "\'");
1001 if ($regary == false){
1002 return Array(false, false, false, $lt, strlen($body));
1004 list($pos, $attval, $match) = $regary;
1006 $attary{$attname} = "'" . $attval . "'";
1007 } else if ($quot == '"'){
1008 $regary = sq_findnxreg($body, $pos+
1, '\"');
1009 if ($regary == false){
1010 return Array(false, false, false, $lt, strlen($body));
1012 list($pos, $attval, $match) = $regary;
1014 $attary{$attname} = '"' . $attval . '"';
1017 * These are hateful. Look for \s, or >.
1019 $regary = sq_findnxreg($body, $pos, "[\s>]");
1020 if ($regary == false){
1021 return Array(false, false, false, $lt, strlen($body));
1023 list($pos, $attval, $match) = $regary;
1025 * If it's ">" it will be caught at the top.
1027 $attval = preg_replace("/\"/s", """, $attval);
1028 $attary{$attname} = '"' . $attval . '"';
1030 } else if (preg_match("|[\w/>]|", $char)) {
1032 * That was attribute type 4.
1034 $attary{$attname} = '"yes"';
1037 * An illegal character. Find next '>' and return.
1039 $gt = sq_findnxstr($body, $pos, ">");
1040 return Array(false, false, false, $lt, $gt);
1046 * The fact that we got here indicates that the tag end was never
1047 * found. Return invalid tag indication so it gets stripped.
1049 return Array(false, false, false, $lt, strlen($body));
1053 * This function checks attribute values for entity-encoded values
1054 * and returns them translated into 8-bit strings so we can run
1057 * @param $attvalue A string to run entity check against.
1058 * @return Translated value.
1060 function sq_deent($attvalue){
1063 * See if we have to run the checks first. All entities must start
1066 if (strpos($attvalue, "&") === false){
1070 * Check named entities first.
1072 $trans = get_html_translation_table(HTML_ENTITIES
);
1074 * Leave " in, as it can mess us up.
1076 $trans = array_flip($trans);
1077 unset($trans{"""});
1078 while (list($ent, $val) = each($trans)){
1079 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1082 * Now translate numbered entities from 1 to 255 if needed.
1084 if (strpos($attvalue, "#") !== false){
1085 $omit = Array(34, 39);
1086 for ($asc=1; $asc<256; $asc++
){
1087 if (!in_array($asc, $omit)){
1089 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1091 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1092 "$chr\\1", $attvalue);
1100 * This function runs various checks against the attributes.
1102 * @param $tagname String with the name of the tag.
1103 * @param $attary Array with all tag attributes.
1104 * @param $rm_attnames See description for sq_sanitize
1105 * @param $bad_attvals See description for sq_sanitize
1106 * @param $add_attr_to_tag See description for sq_sanitize
1107 * @param $message message object
1108 * @param $id message id
1109 * @return Array with modified attributes.
1111 function sq_fixatts($tagname,
1121 while (list($attname, $attvalue) = each($attary)){
1123 * See if this attribute should be removed.
1125 foreach ($rm_attnames as $matchtag=>$matchattrs){
1126 if (preg_match($matchtag, $tagname)){
1127 foreach ($matchattrs as $matchattr){
1128 if (preg_match($matchattr, $attname)){
1129 unset($attary{$attname});
1136 * Remove any entities.
1138 $attvalue = sq_deent($attvalue);
1141 * Now let's run checks on the attvalues.
1142 * I don't expect anyone to comprehend this. If you do,
1143 * get in touch with me so I can drive to where you live and
1144 * shake your hand personally. :)
1146 foreach ($bad_attvals as $matchtag=>$matchattrs){
1147 if (preg_match($matchtag, $tagname)){
1148 foreach ($matchattrs as $matchattr=>$valary){
1149 if (preg_match($matchattr, $attname)){
1151 * There are two arrays in valary.
1153 * Second one is replacements
1155 list($valmatch, $valrepl) = $valary;
1157 preg_replace($valmatch, $valrepl, $attvalue);
1158 if ($newvalue != $attvalue){
1159 $attary{$attname} = $newvalue;
1166 * Turn cid: urls into http-friendly ones.
1168 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1169 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1173 * See if we need to append any attributes to this tag.
1175 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1176 if (preg_match($matchtag, $tagname)){
1177 $attary = array_merge($attary, $addattary);
1184 * This function edits the style definition to make them friendly and
1185 * usable in squirrelmail.
1187 * @param $message the message object
1188 * @param $id the message id
1189 * @param $content a string with whatever is between <style> and </style>
1190 * @return a string with edited content.
1192 function sq_fixstyle($message, $id, $content){
1193 global $view_unsafe_images;
1194 $me = 'sq_fixstyle';
1196 * First look for general BODY style declaration, which would be
1198 * body {background: blah-blah}
1199 * and change it to .bodyclass so we can just assign it to a <div>
1201 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1202 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1204 * Fix url('blah') declarations.
1206 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1207 "url(\\1$secremoveimg\\2)", $content);
1209 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1212 if (!$view_unsafe_images){
1213 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1214 "url(\\1$secremoveimg\\2)", $content);
1218 * Fix urls that refer to cid:
1220 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1222 $cidurl = $matches{1};
1223 $httpurl = sq_cid2http($message, $id, $cidurl);
1224 $content = preg_replace("|url\($cidurl\)|si",
1225 "url($httpurl)", $content);
1229 * Fix stupid css declarations which lead to vulnerabilities
1232 $match = Array('/expression/si',
1235 $replace = Array('idiocy', 'idiocy', 'idiocy');
1236 $content = preg_replace($match, $replace, $content);
1241 * This function converts cid: url's into the ones that can be viewed in
1244 * @param $message the message object
1245 * @param $id the message id
1246 * @param $cidurl the cid: url.
1247 * @return a string with a http-friendly url
1249 function sq_cid2http($message, $id, $cidurl, $mailbox){
1251 * Get rid of quotes.
1253 $quotchar = substr($cidurl, 0, 1);
1254 $cidurl = str_replace($quotchar, "", $cidurl);
1255 $cidurl = substr(trim($cidurl), 4);
1256 $linkurl = find_ent_id($cidurl, $message);
1257 /* in case of non-save cid links $httpurl should be replaced by a sort of
1258 unsave link image */
1261 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&' .
1262 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1263 '&ent_id=' . $linkurl . $quotchar;
1269 * This function changes the <body> tag into a <div> tag since we
1270 * can't really have a body-within-body.
1272 * @param $attary an array of attributes and values of <body>
1273 * @return a modified array of attributes to be set for <div>
1275 function sq_body2div($attary){
1276 $me = 'sq_body2div';
1277 $divattary = Array('class' => "'bodyclass'");
1278 $bgcolor = '#ffffff';
1281 if (is_array($attary) && sizeof($attary) > 0){
1282 foreach ($attary as $attname=>$attvalue){
1283 $quotchar = substr($attvalue, 0, 1);
1284 $attvalue = str_replace($quotchar, "", $attvalue);
1287 $styledef .= "background-image: url('$attvalue'); ";
1290 $styledef .= "background-color: $attvalue; ";
1293 $styledef .= "color: $attvalue; ";
1297 if (strlen($styledef) > 0){
1298 $divattary{"style"} = "\"$styledef\"";
1305 * This is the main function and the one you should actually be calling.
1306 * There are several variables you should be aware of an which need
1307 * special description.
1309 * Since the description is quite lengthy, see it here:
1310 * http://www.mricon.com/html/phpfilter.html
1312 * @param $body the string with HTML you wish to filter
1313 * @param $tag_list see description above
1314 * @param $rm_tags_with_content see description above
1315 * @param $self_closing_tags see description above
1316 * @param $force_tag_closing see description above
1317 * @param $rm_attnames see description above
1318 * @param $bad_attvals see description above
1319 * @param $add_attr_to_tag see description above
1320 * @param $message message object
1321 * @param $id message id
1322 * @return sanitized html safe to show on your pages.
1324 function sq_sanitize($body,
1326 $rm_tags_with_content,
1336 $me = 'sq_sanitize';
1338 * Normalize rm_tags and rm_tags_with_content.
1340 @array_walk
($rm_tags, 'sq_casenormalize');
1341 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1342 @array_walk
($self_closing_tags, 'sq_casenormalize');
1344 * See if tag_list is of tags to remove or tags to allow.
1345 * false means remove these tags
1346 * true means allow these tags
1348 $rm_tags = array_shift($tag_list);
1350 $open_tags = Array();
1351 $trusted = "<!-- begin sanitized html -->\n";
1352 $skip_content = false;
1354 * Take care of netscape's stupid javascript entities like
1357 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1359 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1360 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1361 $free_content = substr($body, $curpos, $lt-$curpos);
1363 * Take care of <style>
1365 if ($tagname == "style" && $tagtype == 2){
1367 * This is a closing </style>. Edit the
1368 * content before we apply it.
1370 $free_content = sq_fixstyle($message, $id, $free_content);
1372 if ($skip_content == false){
1373 $trusted .= $free_content;
1375 if ($tagname != FALSE){
1377 if ($skip_content == $tagname){
1379 * Got to the end of tag we needed to remove.
1382 $skip_content = false;
1384 if ($skip_content == false){
1385 if ($tagname == "body"){
1388 if (isset($open_tags{$tagname}) &&
1389 $open_tags{$tagname} > 0){
1390 $open_tags{$tagname}--;
1399 * $rm_tags_with_content
1401 if ($skip_content == false){
1403 * See if this is a self-closing type and change
1404 * tagtype appropriately.
1407 && in_array($tagname, $self_closing_tags)){
1411 * See if we should skip this tag and any content
1414 if ($tagtype == 1 &&
1415 in_array($tagname, $rm_tags_with_content)){
1416 $skip_content = $tagname;
1418 if (($rm_tags == false
1419 && in_array($tagname, $tag_list)) ||
1420 ($rm_tags == true &&
1421 !in_array($tagname, $tag_list))){
1425 if (isset($open_tags{$tagname})){
1426 $open_tags{$tagname}++
;
1428 $open_tags{$tagname}=1;
1432 * This is where we run other checks.
1434 if (is_array($attary) && sizeof($attary) > 0){
1435 $attary = sq_fixatts($tagname,
1446 * Convert body into div.
1448 if ($tagname == "body"){
1450 $attary = sq_body2div($attary, $message, $id);
1456 if ($tagname != false && $skip_content == false){
1457 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1462 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1463 if ($force_tag_closing == true){
1464 foreach ($open_tags as $tagname=>$opentimes){
1465 while ($opentimes > 0){
1466 $trusted .= '</' . $tagname . '>';
1472 $trusted .= "<!-- end sanitized html -->\n";
1477 * This is a wrapper function to call html sanitizing routines.
1479 * @param $body the body of the message
1480 * @param $id the id of the message
1481 * @return a string with html safe to display in the browser.
1483 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1484 global $attachment_common_show_images, $view_unsafe_images,
1487 * Don't display attached images in HTML mode.
1489 $attachment_common_show_images = false;
1502 $rm_tags_with_content = Array(
1509 $self_closing_tags = Array(
1516 $force_tag_closing = false;
1518 $rm_attnames = Array(
1528 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1529 $bad_attvals = Array(
1532 "/^src|background/i" =>
1535 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1536 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1537 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1538 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1541 "\\1$secremoveimg\\2",
1542 "\\1$secremoveimg\\2",
1543 "\\1$secremoveimg\\2",
1544 "\\1$secremoveimg\\2"
1547 "/^href|action/i" =>
1550 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1551 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1552 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1553 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1568 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1569 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1570 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1571 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1585 if (!$view_unsafe_images){
1587 * Remove any references to http/https if view_unsafe_images set
1590 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1591 '/^([\'\"])\s*https*:.*([\'\"])/si');
1592 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1593 "\\1$secremoveimg\\2");
1594 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1595 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1596 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1597 "url(\\1$secremoveimg\\2)");
1600 $add_attr_to_tag = Array(
1601 "/^a$/si" => Array('target'=>'"_new"')
1603 $trusted = sq_sanitize($body,
1605 $rm_tags_with_content,
1615 if (preg_match("|$secremoveimg|si", $trusted)){
1616 $has_unsafe_images = true;