f68d3717ba412cd63fec29986f9f4b3b2a1bce9f
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
);
416 $from_o = $rfc822_header->from
;
417 if (is_object($from_o)) {
418 $from_name = $from_o->getAddress(false);
420 $from_name = _("Unknown sender");
422 $from_name = decodeHeader(htmlspecialchars($from_name));
423 $description = $from_name;
425 $default_page = '../src/download.php';
426 if (is_object($header->disposition
)) {
427 $filename = decodeHeader($header->disposition
->getProperty('filename'));
428 if (trim($filename) == '') {
429 $name = decodeHeader($header->disposition
->getProperty('name'));
430 if (trim($name) == '') {
431 if (trim( $header->id
) == '') {
432 $filename = 'untitled-[' . $ent . ']' ;
434 $filename = 'cid: ' . $header->id
;
441 if (trim( $header->id
) == '') {
442 $filename = 'untitled-[' . $ent . ']' ;
444 $filename = 'cid: ' . $header->id
;
448 if ($header->description
) {
449 $description = htmlspecialchars($header->description
);
455 $display_filename = $filename;
456 if (isset($passed_ent_id)) {
457 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
459 $passed_ent_id_link = '';
461 $defaultlink = $default_page . "?startMessage=$startMessage"
462 . "&passed_id=$id&mailbox=$urlMailbox"
463 . '&ent_id='.$ent.$passed_ent_id_link.'&absolute_dl=true';
464 if ($where && $what) {
465 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
467 /* This executes the attachment hook with a specific MIME-type.
468 * If that doesn't have results, it tries if there's a rule
469 * for a more generic type.
471 $hookresults = do_hook("attachment $type0/$type1", $links,
472 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
473 $display_filename, $where, $what);
474 if(count($hookresults[1]) <= 1) {
475 $hookresults = do_hook("attachment $type0/*", $links,
476 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
477 $display_filename, $where, $what);
480 $links = $hookresults[1];
481 $defaultlink = $hookresults[6];
483 $attachments .= '<TR><TD>' .
484 "<A HREF=\"$defaultlink\">$display_filename</A> </TD>" .
485 '<TD><SMALL><b>' . show_readable_size($header->size
) .
486 '</b> </small></TD>' .
487 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
489 $attachments .= '<b>' . $description . '</b>';
490 $attachments .= '</SMALL></TD><TD><SMALL> ';
493 foreach ($links as $val) {
497 $attachments .= ' | ';
499 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
502 $attachments .= "</TD></TR>\n";
507 /* This function decodes the body depending on the encoding type. */
508 function decodeBody($body, $encoding) {
509 global $languages, $squirrelmail_language;
510 global $show_html_default;
512 $body = str_replace("\r\n", "\n", $body);
513 $encoding = strtolower($encoding);
515 if ($encoding == 'quoted-printable' ||
516 $encoding == 'quoted_printable') {
517 $body = quoted_printable_decode($body);
519 while (ereg("=\n", $body)) {
520 $body = ereg_replace ("=\n", '', $body);
523 } else if ($encoding == 'base64') {
524 $body = base64_decode($body);
527 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
528 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
529 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
532 // All other encodings are returned raw.
537 * This functions decode strings that is encoded according to
538 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
539 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
541 function decodeHeader ($string, $utfencode=true) {
542 global $languages, $squirrelmail_language;
543 if (is_array($string)) {
544 $string = implode("\n", $string);
547 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
548 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
549 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
553 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
556 /* Ignore white-space between consecutive encoded-words. */
557 if (strspn($res[2], " \t") != strlen($res[2])) {
561 if (ucfirst($res[4]) == 'B') {
562 $replace = base64_decode($res[5]);
564 $replace = str_replace('_', ' ', $res[5]);
565 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
567 /* Only encode into entities by default. Some places
568 * don't need the encoding, like the compose form.
571 $replace = charset_decode($res[3], $replace);
574 $string = $prefix . $replace . substr($string, strlen($res[0]));
575 $i = strlen($prefix) +
strlen($replace);
581 * Encode a string according to RFC 1522 for use in headers if it
582 * contains 8-bit characters or anything that looks like it should
585 function encodeHeader ($string) {
586 global $default_charset, $languages, $squirrelmail_language;
588 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
589 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
590 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
593 // Encode only if the string contains 8-bit characters or =?
594 $j = strlen($string);
595 $l = strstr($string, '=?'); // Must be encoded ?
597 for($i = 0; $i < $j; ++
$i) {
598 switch($string{$i}) {
612 $k = ord($string{$i});
614 $ret .= sprintf("=%02X", $k);
624 $string = "=?$default_charset?Q?$ret?=";
630 /* This function trys to locate the entity_id of a specific mime element */
631 function find_ent_id($id, $message) {
632 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
633 if ($message->entities
[$i]->header
->type0
== 'multipart') {
634 $ret = find_ent_id($id, $message->entities
[$i]);
636 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
637 // if (sq_check_save_extension($message->entities[$i])) {
638 return $message->entities
[$i]->entity_id
;
646 function sq_check_save_extension($message) {
647 $filename = $message->getFilename();
648 $ext = substr($filename, strrpos($filename,'.')+
1);
649 $save_extensions = array('jpg','jpeg','gif','png','bmp');
650 return in_array($ext, $save_extensions);
655 ** HTMLFILTER ROUTINES
659 * This function returns the final tag out of the tag name, an array
660 * of attributes, and the type of the tag. This function is called by
661 * sq_sanitize internally.
663 * @param $tagname the name of the tag.
664 * @param $attary the array of attributes and their values
665 * @param $tagtype The type of the tag (see in comments).
666 * @return a string with the final tag representation.
668 function sq_tagprint($tagname, $attary, $tagtype){
672 $fulltag = '</' . $tagname . '>';
674 $fulltag = '<' . $tagname;
675 if (is_array($attary) && sizeof($attary)){
677 while (list($attname, $attvalue) = each($attary)){
678 array_push($atts, "$attname=$attvalue");
680 $fulltag .= ' ' . join(" ", $atts);
691 * A small helper function to use with array_walk. Modifies a by-ref
692 * value and makes it lowercase.
694 * @param $val a value passed by-ref.
695 * @return void since it modifies a by-ref value.
697 function sq_casenormalize(&$val){
698 $val = strtolower($val);
702 * This function skips any whitespace from the current position within
703 * a string and to the next non-whitespace value.
705 * @param $body the string
706 * @param $offset the offset within the string where we should start
707 * looking for the next non-whitespace character.
708 * @return the location within the $body where the next
709 * non-whitespace char is located.
711 function sq_skipspace($body, $offset){
712 $me = 'sq_skipspace';
713 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
714 if (sizeof($matches{1})){
715 $count = strlen($matches{1});
722 * This function looks for the next character within a string. It's
723 * really just a glorified "strpos", except it catches if failures
726 * @param $body The string to look for needle in.
727 * @param $offset Start looking from this position.
728 * @param $needle The character/string to look for.
729 * @return location of the next occurance of the needle, or
730 * strlen($body) if needle wasn't found.
732 function sq_findnxstr($body, $offset, $needle){
733 $me = 'sq_findnxstr';
734 $pos = strpos($body, $needle, $offset);
736 $pos = strlen($body);
742 * This function takes a PCRE-style regexp and tries to match it
745 * @param $body The string to look for needle in.
746 * @param $offset Start looking from here.
747 * @param $reg A PCRE-style regex to match.
748 * @return Returns a false if no matches found, or an array
749 * with the following members:
750 * - integer with the location of the match within $body
751 * - string with whatever content between offset and the match
752 * - string with whatever it is we matched
754 function sq_findnxreg($body, $offset, $reg){
755 $me = 'sq_findnxreg';
758 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
762 $retarr{0} = $offset +
strlen($matches{1});
763 $retarr{1} = $matches{1};
764 $retarr{2} = $matches{2};
770 * This function looks for the next tag.
772 * @param $body String where to look for the next tag.
773 * @param $offset Start looking from here.
774 * @return false if no more tags exist in the body, or
775 * an array with the following members:
776 * - string with the name of the tag
777 * - array with attributes and their values
778 * - integer with tag type (1, 2, or 3)
779 * - integer where the tag starts (starting "<")
780 * - integer where the tag ends (ending ">")
781 * first three members will be false, if the tag is invalid.
783 function sq_getnxtag($body, $offset){
785 if ($offset > strlen($body)){
788 $lt = sq_findnxstr($body, $offset, "<");
789 if ($lt == strlen($body)){
794 * blah blah <tag attribute="value">
797 $pos = sq_skipspace($body, $lt+
1);
798 if ($pos >= strlen($body)){
799 return Array(false, false, false, $lt, strlen($body));
802 * There are 3 kinds of tags:
803 * 1. Opening tag, e.g.:
805 * 2. Closing tag, e.g.:
807 * 3. XHTML-style content-less tag, e.g.:
811 switch (substr($body, $pos, 1)){
818 * A comment or an SGML declaration.
820 if (substr($body, $pos+
1, 2) == "--"){
821 $gt = strpos($body, "-->", $pos);
827 return Array(false, false, false, $lt, $gt);
829 $gt = sq_findnxstr($body, $pos, ">");
830 return Array(false, false, false, $lt, $gt);
835 * Assume tagtype 1 for now. If it's type 3, we'll switch values
845 * Look for next [\W-_], which will indicate the end of the tag name.
847 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
848 if ($regary == false){
849 return Array(false, false, false, $lt, strlen($body));
851 list($pos, $tagname, $match) = $regary;
852 $tagname = strtolower($tagname);
855 * $match can be either of these:
856 * '>' indicating the end of the tag entirely.
857 * '\s' indicating the end of the tag name.
858 * '/' indicating that this is type-3 xhtml tag.
860 * Whatever else we find there indicates an invalid tag.
865 * This is an xhtml-style tag with a closing / at the
866 * end, like so: <img src="blah"/>. Check if it's followed
867 * by the closing bracket. If not, then this tag is invalid
869 if (substr($body, $pos, 2) == "/>"){
873 $gt = sq_findnxstr($body, $pos, ">");
874 $retary = Array(false, false, false, $lt, $gt);
878 return Array($tagname, false, $tagtype, $lt, $pos);
882 * Check if it's whitespace
884 if (!preg_match('/\s/', $match)){
886 * This is an invalid tag! Look for the next closing ">".
888 $gt = sq_findnxstr($body, $offset, ">");
889 return Array(false, false, false, $lt, $gt);
895 * At this point we're here:
896 * <tagname attribute='blah'>
899 * At this point we loop in order to find all attributes.
905 while ($pos <= strlen($body)){
906 $pos = sq_skipspace($body, $pos);
907 if ($pos == strlen($body)){
911 return Array(false, false, false, $lt, $pos);
914 * See if we arrived at a ">" or "/>", which means that we reached
915 * the end of the tag.
918 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
922 $pos +
= strlen($matches{1});
923 if ($matches{2} == "/>"){
927 return Array($tagname, $attary, $tagtype, $lt, $pos);
931 * There are several types of attributes, with optional
932 * [:space:] between members.
934 * attrname[:space:]=[:space:]'CDATA'
936 * attrname[:space:]=[:space:]"CDATA"
938 * attr[:space:]=[:space:]CDATA
942 * We leave types 1 and 2 the same, type 3 we check for
943 * '"' and convert to """ if needed, then wrap in
944 * double quotes. Type 4 we convert into:
947 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
948 if ($regary == false){
950 * Looks like body ended before the end of tag.
952 return Array(false, false, false, $lt, strlen($body));
954 list($pos, $attname, $match) = $regary;
955 $attname = strtolower($attname);
957 * We arrived at the end of attribute name. Several things possible
959 * '>' means the end of the tag and this is attribute type 4
960 * '/' if followed by '>' means the same thing as above
961 * '\s' means a lot of things -- look what it's followed by.
962 * anything else means the attribute is invalid.
967 * This is an xhtml-style tag with a closing / at the
968 * end, like so: <img src="blah"/>. Check if it's followed
969 * by the closing bracket. If not, then this tag is invalid
971 if (substr($body, $pos, 2) == "/>"){
975 $gt = sq_findnxstr($body, $pos, ">");
976 $retary = Array(false, false, false, $lt, $gt);
980 $attary{$attname} = '"yes"';
981 return Array($tagname, $attary, $tagtype, $lt, $pos);
985 * Skip whitespace and see what we arrive at.
987 $pos = sq_skipspace($body, $pos);
988 $char = substr($body, $pos, 1);
990 * Two things are valid here:
991 * '=' means this is attribute type 1 2 or 3.
992 * \w means this was attribute type 4.
993 * anything else we ignore and re-loop. End of tag and
994 * invalid stuff will be caught by our checks at the beginning
999 $pos = sq_skipspace($body, $pos);
1001 * Here are 3 possibilities:
1002 * "'" attribute type 1
1003 * '"' attribute type 2
1004 * everything else is the content of tag type 3
1006 $quot = substr($body, $pos, 1);
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} = "'" . $attval . "'";
1015 } else if ($quot == '"'){
1016 $regary = sq_findnxreg($body, $pos+
1, '\"');
1017 if ($regary == false){
1018 return Array(false, false, false, $lt, strlen($body));
1020 list($pos, $attval, $match) = $regary;
1022 $attary{$attname} = '"' . $attval . '"';
1025 * These are hateful. Look for \s, or >.
1027 $regary = sq_findnxreg($body, $pos, "[\s>]");
1028 if ($regary == false){
1029 return Array(false, false, false, $lt, strlen($body));
1031 list($pos, $attval, $match) = $regary;
1033 * If it's ">" it will be caught at the top.
1035 $attval = preg_replace("/\"/s", """, $attval);
1036 $attary{$attname} = '"' . $attval . '"';
1038 } else if (preg_match("|[\w/>]|", $char)) {
1040 * That was attribute type 4.
1042 $attary{$attname} = '"yes"';
1045 * An illegal character. Find next '>' and return.
1047 $gt = sq_findnxstr($body, $pos, ">");
1048 return Array(false, false, false, $lt, $gt);
1054 * The fact that we got here indicates that the tag end was never
1055 * found. Return invalid tag indication so it gets stripped.
1057 return Array(false, false, false, $lt, strlen($body));
1061 * This function checks attribute values for entity-encoded values
1062 * and returns them translated into 8-bit strings so we can run
1065 * @param $attvalue A string to run entity check against.
1066 * @return Translated value.
1068 function sq_deent($attvalue){
1071 * See if we have to run the checks first. All entities must start
1074 if (strpos($attvalue, "&") === false){
1078 * Check named entities first.
1080 $trans = get_html_translation_table(HTML_ENTITIES
);
1082 * Leave " in, as it can mess us up.
1084 $trans = array_flip($trans);
1085 unset($trans{"""});
1086 while (list($ent, $val) = each($trans)){
1087 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1090 * Now translate numbered entities from 1 to 255 if needed.
1092 if (strpos($attvalue, "#") !== false){
1093 $omit = Array(34, 39);
1094 for ($asc=1; $asc<256; $asc++
){
1095 if (!in_array($asc, $omit)){
1097 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1099 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1100 "$chr\\1", $attvalue);
1108 * This function runs various checks against the attributes.
1110 * @param $tagname String with the name of the tag.
1111 * @param $attary Array with all tag attributes.
1112 * @param $rm_attnames See description for sq_sanitize
1113 * @param $bad_attvals See description for sq_sanitize
1114 * @param $add_attr_to_tag See description for sq_sanitize
1115 * @param $message message object
1116 * @param $id message id
1117 * @return Array with modified attributes.
1119 function sq_fixatts($tagname,
1129 while (list($attname, $attvalue) = each($attary)){
1131 * See if this attribute should be removed.
1133 foreach ($rm_attnames as $matchtag=>$matchattrs){
1134 if (preg_match($matchtag, $tagname)){
1135 foreach ($matchattrs as $matchattr){
1136 if (preg_match($matchattr, $attname)){
1137 unset($attary{$attname});
1144 * Remove any entities.
1146 $attvalue = sq_deent($attvalue);
1149 * Now let's run checks on the attvalues.
1150 * I don't expect anyone to comprehend this. If you do,
1151 * get in touch with me so I can drive to where you live and
1152 * shake your hand personally. :)
1154 foreach ($bad_attvals as $matchtag=>$matchattrs){
1155 if (preg_match($matchtag, $tagname)){
1156 foreach ($matchattrs as $matchattr=>$valary){
1157 if (preg_match($matchattr, $attname)){
1159 * There are two arrays in valary.
1161 * Second one is replacements
1163 list($valmatch, $valrepl) = $valary;
1165 preg_replace($valmatch, $valrepl, $attvalue);
1166 if ($newvalue != $attvalue){
1167 $attary{$attname} = $newvalue;
1174 * Turn cid: urls into http-friendly ones.
1176 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1177 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1181 * See if we need to append any attributes to this tag.
1183 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1184 if (preg_match($matchtag, $tagname)){
1185 $attary = array_merge($attary, $addattary);
1192 * This function edits the style definition to make them friendly and
1193 * usable in squirrelmail.
1195 * @param $message the message object
1196 * @param $id the message id
1197 * @param $content a string with whatever is between <style> and </style>
1198 * @return a string with edited content.
1200 function sq_fixstyle($message, $id, $content){
1201 global $view_unsafe_images;
1202 $me = 'sq_fixstyle';
1204 * First look for general BODY style declaration, which would be
1206 * body {background: blah-blah}
1207 * and change it to .bodyclass so we can just assign it to a <div>
1209 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1210 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1212 * Fix url('blah') declarations.
1214 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1215 "url(\\1$secremoveimg\\2)", $content);
1217 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1220 if (!$view_unsafe_images){
1221 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1222 "url(\\1$secremoveimg\\2)", $content);
1226 * Fix urls that refer to cid:
1228 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1230 $cidurl = $matches{1};
1231 $httpurl = sq_cid2http($message, $id, $cidurl);
1232 $content = preg_replace("|url\($cidurl\)|si",
1233 "url($httpurl)", $content);
1237 * Fix stupid css declarations which lead to vulnerabilities
1240 $match = Array('/expression/si',
1243 $replace = Array('idiocy', 'idiocy', 'idiocy');
1244 $content = preg_replace($match, $replace, $content);
1249 * This function converts cid: url's into the ones that can be viewed in
1252 * @param $message the message object
1253 * @param $id the message id
1254 * @param $cidurl the cid: url.
1255 * @return a string with a http-friendly url
1257 function sq_cid2http($message, $id, $cidurl, $mailbox){
1259 * Get rid of quotes.
1261 $quotchar = substr($cidurl, 0, 1);
1262 $cidurl = str_replace($quotchar, "", $cidurl);
1263 $cidurl = substr(trim($cidurl), 4);
1264 $linkurl = find_ent_id($cidurl, $message);
1265 /* in case of non-save cid links $httpurl should be replaced by a sort of
1266 unsave link image */
1269 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&' .
1270 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1271 '&ent_id=' . $linkurl . $quotchar;
1277 * This function changes the <body> tag into a <div> tag since we
1278 * can't really have a body-within-body.
1280 * @param $attary an array of attributes and values of <body>
1281 * @return a modified array of attributes to be set for <div>
1283 function sq_body2div($attary){
1284 $me = 'sq_body2div';
1285 $divattary = Array('class' => "'bodyclass'");
1286 $bgcolor = '#ffffff';
1289 if (is_array($attary) && sizeof($attary) > 0){
1290 foreach ($attary as $attname=>$attvalue){
1291 $quotchar = substr($attvalue, 0, 1);
1292 $attvalue = str_replace($quotchar, "", $attvalue);
1295 $styledef .= "background-image: url('$attvalue'); ";
1298 $styledef .= "background-color: $attvalue; ";
1301 $styledef .= "color: $attvalue; ";
1305 if (strlen($styledef) > 0){
1306 $divattary{"style"} = "\"$styledef\"";
1313 * This is the main function and the one you should actually be calling.
1314 * There are several variables you should be aware of an which need
1315 * special description.
1317 * Since the description is quite lengthy, see it here:
1318 * http://www.mricon.com/html/phpfilter.html
1320 * @param $body the string with HTML you wish to filter
1321 * @param $tag_list see description above
1322 * @param $rm_tags_with_content see description above
1323 * @param $self_closing_tags see description above
1324 * @param $force_tag_closing see description above
1325 * @param $rm_attnames see description above
1326 * @param $bad_attvals see description above
1327 * @param $add_attr_to_tag see description above
1328 * @param $message message object
1329 * @param $id message id
1330 * @return sanitized html safe to show on your pages.
1332 function sq_sanitize($body,
1334 $rm_tags_with_content,
1344 $me = 'sq_sanitize';
1346 * Normalize rm_tags and rm_tags_with_content.
1348 @array_walk
($rm_tags, 'sq_casenormalize');
1349 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1350 @array_walk
($self_closing_tags, 'sq_casenormalize');
1352 * See if tag_list is of tags to remove or tags to allow.
1353 * false means remove these tags
1354 * true means allow these tags
1356 $rm_tags = array_shift($tag_list);
1358 $open_tags = Array();
1359 $trusted = "<!-- begin sanitized html -->\n";
1360 $skip_content = false;
1362 * Take care of netscape's stupid javascript entities like
1365 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1367 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1368 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1369 $free_content = substr($body, $curpos, $lt-$curpos);
1371 * Take care of <style>
1373 if ($tagname == "style" && $tagtype == 2){
1375 * This is a closing </style>. Edit the
1376 * content before we apply it.
1378 $free_content = sq_fixstyle($message, $id, $free_content);
1380 if ($skip_content == false){
1381 $trusted .= $free_content;
1383 if ($tagname != FALSE){
1385 if ($skip_content == $tagname){
1387 * Got to the end of tag we needed to remove.
1390 $skip_content = false;
1392 if ($skip_content == false){
1393 if ($tagname == "body"){
1396 if (isset($open_tags{$tagname}) &&
1397 $open_tags{$tagname} > 0){
1398 $open_tags{$tagname}--;
1407 * $rm_tags_with_content
1409 if ($skip_content == false){
1411 * See if this is a self-closing type and change
1412 * tagtype appropriately.
1415 && in_array($tagname, $self_closing_tags)){
1419 * See if we should skip this tag and any content
1422 if ($tagtype == 1 &&
1423 in_array($tagname, $rm_tags_with_content)){
1424 $skip_content = $tagname;
1426 if (($rm_tags == false
1427 && in_array($tagname, $tag_list)) ||
1428 ($rm_tags == true &&
1429 !in_array($tagname, $tag_list))){
1433 if (isset($open_tags{$tagname})){
1434 $open_tags{$tagname}++
;
1436 $open_tags{$tagname}=1;
1440 * This is where we run other checks.
1442 if (is_array($attary) && sizeof($attary) > 0){
1443 $attary = sq_fixatts($tagname,
1454 * Convert body into div.
1456 if ($tagname == "body"){
1458 $attary = sq_body2div($attary, $message, $id);
1464 if ($tagname != false && $skip_content == false){
1465 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1470 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1471 if ($force_tag_closing == true){
1472 foreach ($open_tags as $tagname=>$opentimes){
1473 while ($opentimes > 0){
1474 $trusted .= '</' . $tagname . '>';
1480 $trusted .= "<!-- end sanitized html -->\n";
1485 * This is a wrapper function to call html sanitizing routines.
1487 * @param $body the body of the message
1488 * @param $id the id of the message
1489 * @return a string with html safe to display in the browser.
1491 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1492 global $attachment_common_show_images, $view_unsafe_images,
1495 * Don't display attached images in HTML mode.
1497 $attachment_common_show_images = false;
1510 $rm_tags_with_content = Array(
1517 $self_closing_tags = Array(
1524 $force_tag_closing = false;
1526 $rm_attnames = Array(
1536 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1537 $bad_attvals = Array(
1540 "/^src|background/i" =>
1543 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1544 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1545 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1546 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1549 "\\1$secremoveimg\\2",
1550 "\\1$secremoveimg\\2",
1551 "\\1$secremoveimg\\2",
1552 "\\1$secremoveimg\\2"
1555 "/^href|action/i" =>
1558 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1559 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1560 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1561 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1576 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1577 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1578 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1579 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1593 if(isset($_GET['view_unsafe_images'])) {
1594 $view_unsafe_images = $_GET['view_unsafe_images'];
1596 if (!$view_unsafe_images){
1598 * Remove any references to http/https if view_unsafe_images set
1601 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1602 '/^([\'\"])\s*https*:.*([\'\"])/si');
1603 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1604 "\\1$secremoveimg\\2");
1605 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1606 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1607 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1608 "url(\\1$secremoveimg\\2)");
1611 $add_attr_to_tag = Array(
1612 "/^a$/si" => Array('target'=>'"_new"')
1614 $trusted = sq_sanitize($body,
1616 $rm_tags_with_content,
1626 if (preg_match("|$secremoveimg|si", $trusted)){
1627 $has_unsafe_images = true;