0ae940ed8bac8aa748437bdb8d92b26851a5d049
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('../functions/imap.php');
16 require_once('../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 $msg = &new message();
32 $read = trim(substr ($read, 0, -1));
33 $res = $msg->parseStructure($read);
37 foreach ($flags as $flag) {
38 $char = strtoupper($flag{1});
41 if (strtolower($flag) == '\\seen') {
46 if (strtolower($flag) == '\\answered') {
47 $msg->is_answered
= true;
51 if (strtolower($flag) == '\\deleted') {
52 $msg->is_deleted
= true;
56 if (strtolower($flag) == '\\flagged') {
57 $msg->is_flagged
= true;
61 if (strtolower($flag) == '$mdnsent') {
62 $msg->is_mdnsent
= true;
70 // listEntities($msg);
74 /* this starts the parsing of a particular structure. It is called recursively,
75 * so it can be passed different structures. It returns an object of type
77 * First, it checks to see if it is a multipart message. If it is, then it
78 * handles that as it sees is necessary. If it is just a regular entity,
79 * then it parses it and adds the necessary header information (by calling out
80 * to mime_get_elements()
83 function mime_fetch_body($imap_stream, $id, $ent_id) {
86 * do a bit of error correction. If we couldn't find the entity id, just guess
87 * that it is the first one. That is usually the case anyway.
92 $cmd = "FETCH $id BODY[$ent_id]";
94 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
96 $topline = trim(array_shift( $data ));
97 } while( $topline && $topline[0] == '*' && !preg_match( '/\* [0-9]+ FETCH.*/i', $topline )) ;
99 $wholemessage = implode('', $data);
100 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
102 $ret = substr( $wholemessage, 0, $regs[1] );
104 There is some information in the content info header that could be important
105 in order to parse html messages. Let's get them here.
107 if ( $ret{0} == '<' ) {
108 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
110 } else if (ereg('"([^"]*)"', $topline, $regs)) {
113 global $where, $what, $mailbox, $passed_id, $startMessage;
114 $par = 'mailbox=' . urlencode($mailbox) . "&passed_id=$passed_id";
115 if (isset($where) && isset($what)) {
116 $par .= '&where='. urlencode($where) . "&what=" . urlencode($what);
118 $par .= "&startMessage=$startMessage&show_more=0";
120 $par .= '&response=' . urlencode($response) .
121 '&message=' . urlencode($message).
122 '&topline=' . urlencode($topline);
125 '<table width="80%"><tr>' .
126 '<tr><td colspan=2>' .
127 _("Body retrieval error. The reason for this is most probably that the message is malformed. Please help us making future versions better by submitting this message to the developers knowledgebase!") .
128 " <A HREF=\"../src/retrievalerror.php?$par\"><br>" .
129 _("Submit message") . '</A><BR> ' .
131 '<td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
132 '<td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
133 '<td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
134 '<td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
135 "</table><BR></tt></font><hr>";
137 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
139 $wholemessage = implode('', $data);
141 $ret = $wholemessage;
146 function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) {
148 // do a bit of error correction. If we couldn't find the entity id, just guess
149 // that it is the first one. That is usually the case anyway.
153 $sid = sqimap_session_id($uid_support);
154 // Don't kill the connection if the browser is over a dialup
155 // and it would take over 30 seconds to download it.
157 // donĀ“t call set_time_limit in safe mode.
158 if (!ini_get("safe_mode")) {
162 $sid_s = substr($sid,0,strpos($sid, ' '));
167 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
168 echo decodeBody($body, $encoding);
171 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
174 $read = fgets ($imap_stream,8192);
177 // This could be bad -- if the section has sqimap_session_id() . ' OK'
178 // or similar, it will kill the download.
179 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
180 if (trim($read) == ')==') {
182 $read = fgets ($imap_stream,4096);
183 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
186 echo decodeBody($read1, $encoding) .
187 decodeBody($read, $encoding);
190 echo decodeBody($read, $encoding);
192 $read = fgets ($imap_stream,4096);
199 /* -[ END MIME DECODING ]----------------------------------------------------------- */
201 // This is here for debugging purposese. It will print out a list
202 // of all the entity IDs that are in the $message object.
204 function listEntities ($message) {
206 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
207 for ($i = 0;isset($message->entities
[$i]); $i++
) {
209 $msg = listEntities($message->entities
[$i]);
219 function getPriorityStr($priority) {
220 $priority_level = substr($priority,0,1);
222 switch($priority_level) {
223 /* check for a higher then normal priority. */
226 $priority_string = _("High");
229 /* check for a lower then normal priority. */
232 $priority_string = _("Low");
235 /* check for a normal priority. */
238 $priority_level = '3';
239 $priority_string = _("Normal");
243 return $priority_string;
246 /* returns a $message object for a particular entity id */
247 function getEntity ($message, $ent_id) {
248 return $message->getEntity($ent_id);
253 * Extracted from strings.php 23/03/2002
256 function translateText(&$body, $wrap_at, $charset) {
257 global $where, $what; /* from searching */
258 global $color; /* color theme */
260 require_once('../functions/url_parser.php');
262 $body_ary = explode("\n", $body);
264 for ($i=0; $i < count($body_ary); $i++
) {
265 $line = $body_ary[$i];
266 if (strlen($line) - 2 >= $wrap_at) {
267 sqWordWrap($line, $wrap_at);
269 $line = charset_decode($charset, $line);
270 $line = str_replace("\t", ' ', $line);
276 $j = strlen( $line );
278 while ( $pos < $j ) {
279 if ($line[$pos] == ' ') {
281 } else if (strpos($line, '>', $pos) === $pos) {
290 if (! isset($color[14])) {
291 $color[14] = '#FF0000';
293 $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
295 if (! isset($color[13])) {
296 $color[13] = '#800000';
298 $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
301 $body_ary[$i] = $line;
303 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
307 /* This returns a parsed string called $body. That string can then
308 be displayed as the actual message in the HTML. It contains
309 everything needed, including HTML Tags, Attachments at the
312 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
313 // this if statement checks for the entity to show as the
314 // primary message. To add more of them, just put them in the
315 // order that is their priority.
316 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
317 $show_html_default, $has_unsafe_images, $view_unsafe_images, $sort;
319 $has_unsafe_images= 0;
321 $urlmailbox = urlencode($mailbox);
322 $body_message = getEntity($message, $ent_num);
323 if (($body_message->header
->type0
== 'text') ||
324 ($body_message->header
->type0
== 'rfc822')) {
325 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
326 $body = decodeBody($body, $body_message->header
->encoding
);
327 $hookResults = do_hook("message_body", $body);
328 $body = $hookResults[1];
330 // If there are other types that shouldn't be formatted, add
332 if ($body_message->header
->type1
== 'html') {
333 if ( $show_html_default <> 1 ) {
334 $entity_conv = array(' ' => ' ',
337 $body = strip_tags( $body );
338 $body = strtr($body, $entity_conv);
340 translateText($body, $wrap_at,
341 $body_message->header
->getParameter
['charset']);
343 $body = magicHTML( $body, $id, $message, $mailbox );
346 translateText($body, $wrap_at,
347 $body_message->header
->getParameter('charset'));
350 if ($has_unsafe_images) {
351 if ($view_unsafe_images) {
352 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&passed_ent_id=".$message->entity_id
."&mailbox=$urlmailbox&sort=$sort&startMessage=$startMessage&show_more=0\">". _("Hide Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
354 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&passed_ent_id=".$message->entity_id
."&mailbox=$urlmailbox&sort=$sort&startMessage=$startMessage&show_more=0&view_unsafe_images=1\">". _("View Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
362 function formatAttachments($message, $exclude_id, $mailbox, $id) {
363 global $where, $what;
364 global $startMessage, $color;
365 static $ShownHTML = 0;
367 $att_ar = $message->getAttachments($exclude_id);
369 if (!count($att_ar)) return '';
373 $urlMailbox = urlencode($mailbox);
375 foreach ($att_ar as $att) {
376 $ent = urldecode($att->entity_id
);
377 $header = $att->header
;
378 $type0 = strtolower($header->type0
);
379 $type1 = strtolower($header->type1
);
381 $Links['download link']['text'] = _("download");
382 $Links['download link']['href'] =
383 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
385 if ($type0 =='message' && $type1 == 'rfc822') {
386 $default_page = '../src/read_body.php';
387 $rfc822_header = $att->rfc822_header
;
388 $filename = decodeHeader($rfc822_header->subject
);
390 $from_o = $rfc822_header->from
;
391 if (is_object($from_o)) {
392 $from_name = $from_o->getAddress(false);
394 $from_name = _("Unknown sender");
396 $from_name = decodeHeader(htmlspecialchars($from_name));
397 $description = $from_name;
399 $default_page = '../src/download.php';
400 if (is_object($header->disposition
)) {
401 $filename = decodeHeader($header->disposition
->getProperty('filename'));
402 if (trim($filename) == '') {
403 $name = decodeHeader($header->disposition
->getProperty('name'));
404 if (trim($name) == '') {
405 if ( trim( $header->id
) == '' )
406 $filename = 'untitled-[' . $ent . ']' ;
408 $filename = 'cid: ' . $header->id
;
414 if ( trim( $header->id
) == '' )
415 $filename = 'untitled-[' . $ent . ']' ;
417 $filename = 'cid: ' . $header->id
;
420 if ($header->description
) {
421 $description = htmlspecialchars($header->description
);
427 $display_filename = $filename;
428 if (isset($passed_ent_id)) {
429 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
431 $passed_ent_id_link = '';
433 $DefaultLink = $default_page . "?startMessage=$startMessage"
434 . "&passed_id=$id&mailbox=$urlMailbox"
435 . '&ent_id='.$ent.$passed_ent_id_link;
436 if ($where && $what) {
437 $DefaultLink = '&where='. urlencode($where).'&what='.urlencode($what);
439 /* this executes the attachment hook with a specific MIME-type.
440 * if that doens't have results, it tries if there's a rule
441 * for a more generic type. */
442 $HookResults = do_hook("attachment $type0/$type1", $Links,
443 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
444 $display_filename, $where, $what);
445 if(count($HookResults[1]) <= 1) {
446 $HookResults = do_hook("attachment $type0/*", $Links,
447 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
448 $display_filename, $where, $what);
451 $Links = $HookResults[1];
452 $DefaultLink = $HookResults[6];
454 $attachments .= '<TR><TD>' .
455 "<A HREF=\"$DefaultLink\">$display_filename</A> </TD>" .
456 '<TD><SMALL><b>' . show_readable_size($header->size
) .
457 '</b> </small></TD>' .
458 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
460 $attachments .= '<b>' . $description . '</b>';
461 $attachments .= '</SMALL></TD><TD><SMALL> ';
464 foreach ($Links as $Val) {
468 $attachments .= ' | ';
470 $attachments .= '<a href="' . $Val['href'] . '">' . $Val['text'] . '</a>';
473 $attachments .= "</TD></TR>\n";
478 /** this function decodes the body depending on the encoding type. **/
479 function decodeBody($body, $encoding) {
480 global $languages, $squirrelmail_language;
482 $body = str_replace("\r\n", "\n", $body);
483 $encoding = strtolower($encoding);
485 global $show_html_default;
487 if ($encoding == 'quoted-printable' ||
488 $encoding == 'quoted_printable') {
489 $body = quoted_printable_decode($body);
491 while (ereg("=\n", $body))
492 $body = ereg_replace ("=\n", "", $body);
494 } else if ($encoding == 'base64') {
495 $body = base64_decode($body);
498 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
499 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
500 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
503 // All other encodings are returned raw.
508 * This functions decode strings that is encoded according to
509 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
510 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
512 function decodeHeader ($string, $utfencode=true) {
513 global $languages, $squirrelmail_language;
514 if (is_array($string)) {
515 $string = implode("\n", $string);
518 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
519 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
520 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
524 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
527 // Ignore white-space between consecutive encoded-words
528 if (strspn($res[2], " \t") != strlen($res[2])) {
532 if (ucfirst($res[4]) == 'B') {
533 $replace = base64_decode($res[5]);
535 $replace = str_replace('_', ' ', $res[5]);
536 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
538 /* Only encode into entities by default. Some places
539 don't need the encoding, like the compose form. */
541 $replace = charset_decode($res[3], $replace);
544 $string = $prefix . $replace . substr($string, strlen($res[0]));
545 $i = strlen($prefix) +
strlen($replace);
551 * Encode a string according to RFC 1522 for use in headers if it
552 * contains 8-bit characters or anything that looks like it should
555 function encodeHeader ($string) {
556 global $default_charset, $languages, $squirrelmail_language;
558 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
559 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
560 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
563 // Encode only if the string contains 8-bit characters or =?
564 $j = strlen( $string );
565 $l = strstr($string, '=?'); // Must be encoded ?
567 for( $i=0; $i < $j; ++
$i) {
568 switch( $string{$i} ) {
582 $k = ord( $string{$i} );
584 $ret .= sprintf("=%02X", $k);
592 $string = "=?$default_charset?Q?$ret?=";
598 /* This function trys to locate the entity_id of a specific mime element */
600 function find_ent_id( $id, $message ) {
602 for ($i=0; $ret == '' && $i < count($message->entities
); $i++
) {
603 if ( $message->entities
[$i]->header
->type0
== 'multipart') {
604 $ret = find_ent_id( $id, $message->entities
[$i] );
606 if ( strcasecmp( $message->entities
[$i]->header
->id
, $id ) == 0 )
607 $ret = $message->entities
[$i]->entity_id
;
614 ** HTMLFILTER ROUTINES
618 * This function returns the final tag out of the tag name, an array
619 * of attributes, and the type of the tag. This function is called by
620 * sq_sanitize internally.
622 * @param $tagname the name of the tag.
623 * @param $attary the array of attributes and their values
624 * @param $tagtype The type of the tag (see in comments).
625 * @return a string with the final tag representation.
627 function sq_tagprint($tagname, $attary, $tagtype){
630 $fulltag = '</' . $tagname . '>';
632 $fulltag = '<' . $tagname;
633 if (is_array($attary) && sizeof($attary)){
635 while (list($attname, $attvalue) = each($attary)){
636 array_push($atts, "$attname=$attvalue");
638 $fulltag .= ' ' . join(" ", $atts);
649 * A small helper function to use with array_walk. Modifies a by-ref
650 * value and makes it lowercase.
652 * @param $val a value passed by-ref.
653 * @return void since it modifies a by-ref value.
655 function sq_casenormalize(&$val){
656 $val = strtolower($val);
660 * This function skips any whitespace from the current position within
661 * a string and to the next non-whitespace value.
663 * @param $body the string
664 * @param $offset the offset within the string where we should start
665 * looking for the next non-whitespace character.
666 * @return the location within the $body where the next
667 * non-whitespace char is located.
669 function sq_skipspace($body, $offset){
670 $me = "sq_skipspace";
671 preg_match("/^(\s*)/s", substr($body, $offset), $matches);
672 if (sizeof($matches{1})){
673 $count = strlen($matches{1});
680 * This function looks for the next character within a string. It's
681 * really just a glorified "strpos", except it catches if failures
684 * @param $body The string to look for needle in.
685 * @param $offset Start looking from this position.
686 * @param $needle The character/string to look for.
687 * @return location of the next occurance of the needle, or
688 * strlen($body) if needle wasn't found.
690 function sq_findnxstr($body, $offset, $needle){
691 $me = "sq_findnxstr";
692 $pos = strpos($body, $needle, $offset);
694 $pos = strlen($body);
700 * This function takes a PCRE-style regexp and tries to match it
703 * @param $body The string to look for needle in.
704 * @param $offset Start looking from here.
705 * @param $reg A PCRE-style regex to match.
706 * @return Returns a false if no matches found, or an array
707 * with the following members:
708 * - integer with the location of the match within $body
709 * - string with whatever content between offset and the match
710 * - string with whatever it is we matched
712 function sq_findnxreg($body, $offset, $reg){
713 $me = "sq_findnxreg";
716 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
720 $retarr{0} = $offset +
strlen($matches{1});
721 $retarr{1} = $matches{1};
722 $retarr{2} = $matches{2};
728 * This function looks for the next tag.
730 * @param $body String where to look for the next tag.
731 * @param $offset Start looking from here.
732 * @return false if no more tags exist in the body, or
733 * an array with the following members:
734 * - string with the name of the tag
735 * - array with attributes and their values
736 * - integer with tag type (1, 2, or 3)
737 * - integer where the tag starts (starting "<")
738 * - integer where the tag ends (ending ">")
739 * first three members will be false, if the tag is invalid.
741 function sq_getnxtag($body, $offset){
743 if ($offset > strlen($body)){
746 $lt = sq_findnxstr($body, $offset, "<");
747 if ($lt == strlen($body)){
752 * blah blah <tag attribute="value">
755 $pos = sq_skipspace($body, $lt+
1);
756 if ($pos >= strlen($body)){
757 return Array(false, false, false, $lt, strlen($body));
760 * There are 3 kinds of tags:
761 * 1. Opening tag, e.g.:
763 * 2. Closing tag, e.g.:
765 * 3. XHTML-style content-less tag, e.g.:
769 switch (substr($body, $pos, 1)){
776 * A comment or an SGML declaration.
778 if (substr($body, $pos+
1, 2) == "--"){
779 $gt = strpos($body, "-->", $pos);
785 return Array(false, false, false, $lt, $gt);
787 $gt = sq_findnxstr($body, $pos, ">");
788 return Array(false, false, false, $lt, $gt);
793 * Assume tagtype 1 for now. If it's type 3, we'll switch values
803 * Look for next [\W-_], which will indicate the end of the tag name.
805 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
806 if ($regary == false){
807 return Array(false, false, false, $lt, strlen($body));
809 list($pos, $tagname, $match) = $regary;
810 $tagname = strtolower($tagname);
813 * $match can be either of these:
814 * '>' indicating the end of the tag entirely.
815 * '\s' indicating the end of the tag name.
816 * '/' indicating that this is type-3 xhtml tag.
818 * Whatever else we find there indicates an invalid tag.
823 * This is an xhtml-style tag with a closing / at the
824 * end, like so: <img src="blah"/>. Check if it's followed
825 * by the closing bracket. If not, then this tag is invalid
827 if (substr($body, $pos, 2) == "/>"){
831 $gt = sq_findnxstr($body, $pos, ">");
832 $retary = Array(false, false, false, $lt, $gt);
836 return Array($tagname, false, $tagtype, $lt, $pos);
840 * Check if it's whitespace
842 if (preg_match("/\s/", $match)){
845 * This is an invalid tag! Look for the next closing ">".
847 $gt = sq_findnxstr($body, $offset, ">");
848 return Array(false, false, false, $lt, $gt);
853 * At this point we're here:
854 * <tagname attribute='blah'>
857 * At this point we loop in order to find all attributes.
863 while ($pos <= strlen($body)){
864 $pos = sq_skipspace($body, $pos);
865 if ($pos == strlen($body)){
869 return Array(false, false, false, $lt, $pos);
872 * See if we arrived at a ">" or "/>", which means that we reached
873 * the end of the tag.
876 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
880 $pos +
= strlen($matches{1});
881 if ($matches{2} == "/>"){
885 return Array($tagname, $attary, $tagtype, $lt, $pos);
889 * There are several types of attributes, with optional
890 * [:space:] between members.
892 * attrname[:space:]=[:space:]'CDATA'
894 * attrname[:space:]=[:space:]"CDATA"
896 * attr[:space:]=[:space:]CDATA
900 * We leave types 1 and 2 the same, type 3 we check for
901 * '"' and convert to """ if needed, then wrap in
902 * double quotes. Type 4 we convert into:
905 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
906 if ($regary == false){
908 * Looks like body ended before the end of tag.
910 return Array(false, false, false, $lt, strlen($body));
912 list($pos, $attname, $match) = $regary;
913 $attname = strtolower($attname);
915 * We arrived at the end of attribute name. Several things possible
917 * '>' means the end of the tag and this is attribute type 4
918 * '/' if followed by '>' means the same thing as above
919 * '\s' means a lot of things -- look what it's followed by.
920 * anything else means the attribute is invalid.
925 * This is an xhtml-style tag with a closing / at the
926 * end, like so: <img src="blah"/>. Check if it's followed
927 * by the closing bracket. If not, then this tag is invalid
929 if (substr($body, $pos, 2) == "/>"){
933 $gt = sq_findnxstr($body, $pos, ">");
934 $retary = Array(false, false, false, $lt, $gt);
938 $attary{$attname} = '"yes"';
939 return Array($tagname, $attary, $tagtype, $lt, $pos);
943 * Skip whitespace and see what we arrive at.
945 $pos = sq_skipspace($body, $pos);
946 $char = substr($body, $pos, 1);
948 * Two things are valid here:
949 * '=' means this is attribute type 1 2 or 3.
950 * \w means this was attribute type 4.
951 * anything else we ignore and re-loop. End of tag and
952 * invalid stuff will be caught by our checks at the beginning
957 $pos = sq_skipspace($body, $pos);
959 * Here are 3 possibilities:
960 * "'" attribute type 1
961 * '"' attribute type 2
962 * everything else is the content of tag type 3
964 $quot = substr($body, $pos, 1);
966 $regary = sq_findnxreg($body, $pos+
1, "\'");
967 if ($regary == false){
968 return Array(false, false, false, $lt, strlen($body));
970 list($pos, $attval, $match) = $regary;
972 $attary{$attname} = "'" . $attval . "'";
973 } else if ($quot == '"'){
974 $regary = sq_findnxreg($body, $pos+
1, '\"');
975 if ($regary == false){
976 return Array(false, false, false, $lt, strlen($body));
978 list($pos, $attval, $match) = $regary;
980 $attary{$attname} = '"' . $attval . '"';
983 * These are hateful. Look for \s, or >.
985 $regary = sq_findnxreg($body, $pos, "[\s>]");
986 if ($regary == false){
987 return Array(false, false, false, $lt, strlen($body));
989 list($pos, $attval, $match) = $regary;
991 * If it's ">" it will be caught at the top.
993 $attval = preg_replace("/\"/s", """, $attval);
994 $attary{$attname} = '"' . $attval . '"';
996 } else if (preg_match("|[\w/>]|", $char)) {
998 * That was attribute type 4.
1000 $attary{$attname} = '"yes"';
1003 * An illegal character. Find next '>' and return.
1005 $gt = sq_findnxstr($body, $pos, ">");
1006 return Array(false, false, false, $lt, $gt);
1011 * The fact that we got here indicates that the tag end was never
1012 * found. Return invalid tag indication so it gets stripped.
1014 return Array(false, false, false, $lt, strlen($body));
1018 * This function checks attribute values for entity-encoded values
1019 * and returns them translated into 8-bit strings so we can run
1022 * @param $attvalue A string to run entity check against.
1023 * @return Translated value.
1025 function sq_deent($attvalue){
1028 * See if we have to run the checks first. All entities must start
1031 if (strpos($attvalue, "&") === false){
1035 * Check named entities first.
1037 $trans = get_html_translation_table(HTML_ENTITIES
);
1039 * Leave " in, as it can mess us up.
1041 $trans = array_flip($trans);
1042 unset($trans{"""});
1043 while (list($ent, $val) = each($trans)){
1044 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1047 * Now translate numbered entities from 1 to 255 if needed.
1049 if (strpos($attvalue, "#") !== false){
1050 $omit = Array(34, 39);
1051 for ($asc=1; $asc<256; $asc++
){
1052 if (!in_array($asc, $omit)){
1054 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1056 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1057 "$chr\\1", $attvalue);
1065 * This function runs various checks against the attributes.
1067 * @param $tagname String with the name of the tag.
1068 * @param $attary Array with all tag attributes.
1069 * @param $rm_attnames See description for sq_sanitize
1070 * @param $bad_attvals See description for sq_sanitize
1071 * @param $add_attr_to_tag See description for sq_sanitize
1072 * @param $message message object
1073 * @param $id message id
1074 * @return Array with modified attributes.
1076 function sq_fixatts($tagname,
1086 while (list($attname, $attvalue) = each($attary)){
1088 * See if this attribute should be removed.
1090 foreach ($rm_attnames as $matchtag=>$matchattrs){
1091 if (preg_match($matchtag, $tagname)){
1092 foreach ($matchattrs as $matchattr){
1093 if (preg_match($matchattr, $attname)){
1094 unset($attary{$attname});
1101 * Remove any entities.
1103 $attvalue = sq_deent($attvalue);
1106 * Now let's run checks on the attvalues.
1107 * I don't expect anyone to comprehend this. If you do,
1108 * get in touch with me so I can drive to where you live and
1109 * shake your hand personally. :)
1111 foreach ($bad_attvals as $matchtag=>$matchattrs){
1112 if (preg_match($matchtag, $tagname)){
1113 foreach ($matchattrs as $matchattr=>$valary){
1114 if (preg_match($matchattr, $attname)){
1116 * There are two arrays in valary.
1118 * Second one is replacements
1120 list($valmatch, $valrepl) = $valary;
1122 preg_replace($valmatch, $valrepl, $attvalue);
1123 if ($newvalue != $attvalue){
1124 $attary{$attname} = $newvalue;
1131 * Turn cid: urls into http-friendly ones.
1133 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1134 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1138 * See if we need to append any attributes to this tag.
1140 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1141 if (preg_match($matchtag, $tagname)){
1142 $attary = array_merge($attary, $addattary);
1149 * This function edits the style definition to make them friendly and
1150 * usable in squirrelmail.
1152 * @param $message the message object
1153 * @param $id the message id
1154 * @param $content a string with whatever is between <style> and </style>
1155 * @return a string with edited content.
1157 function sq_fixstyle($message, $id, $content){
1158 global $view_unsafe_images;
1159 $me = "sq_fixstyle";
1161 * First look for general BODY style declaration, which would be
1163 * body {background: blah-blah}
1164 * and change it to .bodyclass so we can just assign it to a <div>
1166 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1167 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1169 * Fix url('blah') declarations.
1171 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1172 "url(\\1$secremoveimg\\2)", $content);
1174 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1177 if (!$view_unsafe_images){
1178 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1179 "url(\\1$secremoveimg\\2)", $content);
1183 * Fix urls that refer to cid:
1185 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1187 $cidurl = $matches{1};
1188 $httpurl = sq_cid2http($message, $id, $cidurl);
1189 $content = preg_replace("|url\($cidurl\)|si",
1190 "url($httpurl)", $content);
1194 * Fix stupid css declarations which lead to vulnerabilities
1197 $match = Array('/expression/si',
1200 $replace = Array('idiocy', 'idiocy', 'idiocy');
1201 $content = preg_replace($match, $replace, $content);
1206 * This function converts cid: url's into the ones that can be viewed in
1209 * @param $message the message object
1210 * @param $id the message id
1211 * @param $cidurl the cid: url.
1212 * @return a string with a http-friendly url
1214 function sq_cid2http($message, $id, $cidurl, $mailbox){
1216 * Get rid of quotes.
1218 $quotchar = substr($cidurl, 0, 1);
1219 $cidurl = str_replace($quotchar, "", $cidurl);
1220 $cidurl = substr(trim($cidurl), 4);
1221 $httpurl = $quotchar . "../src/download.php?absolute_dl=true&" .
1222 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1223 "&ent_id=" . find_ent_id($cidurl, $message) . $quotchar;
1228 * This function changes the <body> tag into a <div> tag since we
1229 * can't really have a body-within-body.
1231 * @param $attary an array of attributes and values of <body>
1232 * @return a modified array of attributes to be set for <div>
1234 function sq_body2div($attary){
1235 $me = "sq_body2div";
1236 $divattary = Array("class"=>"'bodyclass'");
1240 if (is_array($attary) && sizeof($attary) > 0){
1241 foreach ($attary as $attname=>$attvalue){
1242 $quotchar = substr($attvalue, 0, 1);
1243 $attvalue = str_replace($quotchar, "", $attvalue);
1246 $styledef .= "background-image: url('$attvalue'); ";
1249 $styledef .= "background-color: $attvalue; ";
1252 $styledef .= "color: $attvalue; ";
1255 if (strlen($styledef) > 0){
1256 $divattary{"style"} = "\"$styledef\"";
1263 * This is the main function and the one you should actually be calling.
1264 * There are several variables you should be aware of an which need
1265 * special description.
1267 * Since the description is quite lengthy, see it here:
1268 * http://www.mricon.com/html/phpfilter.html
1270 * @param $body the string with HTML you wish to filter
1271 * @param $tag_list see description above
1272 * @param $rm_tags_with_content see description above
1273 * @param $self_closing_tags see description above
1274 * @param $force_tag_closing see description above
1275 * @param $rm_attnames see description above
1276 * @param $bad_attvals see description above
1277 * @param $add_attr_to_tag see description above
1278 * @param $message message object
1279 * @param $id message id
1280 * @return sanitized html safe to show on your pages.
1282 function sq_sanitize($body,
1284 $rm_tags_with_content,
1294 $me = "sq_sanitize";
1296 * Normalize rm_tags and rm_tags_with_content.
1298 @array_walk
($rm_tags, 'sq_casenormalize');
1299 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1300 @array_walk
($self_closing_tags, 'sq_casenormalize');
1302 * See if tag_list is of tags to remove or tags to allow.
1303 * false means remove these tags
1304 * true means allow these tags
1306 $rm_tags = array_shift($tag_list);
1308 $open_tags = Array();
1309 $trusted = "<!-- begin sanitized html -->\n";
1310 $skip_content = false;
1312 * Take care of netscape's stupid javascript entities like
1315 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1317 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1318 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1319 $free_content = substr($body, $curpos, $lt-$curpos);
1321 * Take care of <style>
1323 if ($tagname == "style" && $tagtype == 2){
1325 * This is a closing </style>. Edit the
1326 * content before we apply it.
1328 $free_content = sq_fixstyle($message, $id, $free_content);
1330 if ($skip_content == false){
1331 $trusted .= $free_content;
1334 if ($tagname != FALSE){
1336 if ($skip_content == $tagname){
1338 * Got to the end of tag we needed to remove.
1341 $skip_content = false;
1343 if ($skip_content == false){
1344 if ($tagname == "body"){
1347 if (isset($open_tags{$tagname}) &&
1348 $open_tags{$tagname} > 0){
1349 $open_tags{$tagname}--;
1359 * $rm_tags_with_content
1361 if ($skip_content == false){
1363 * See if this is a self-closing type and change
1364 * tagtype appropriately.
1367 && in_array($tagname, $self_closing_tags)){
1371 * See if we should skip this tag and any content
1374 if ($tagtype == 1 &&
1375 in_array($tagname, $rm_tags_with_content)){
1376 $skip_content = $tagname;
1378 if (($rm_tags == false
1379 && in_array($tagname, $tag_list)) ||
1380 ($rm_tags == true &&
1381 !in_array($tagname, $tag_list))){
1385 if (isset($open_tags{$tagname})){
1386 $open_tags{$tagname}++
;
1388 $open_tags{$tagname}=1;
1392 * This is where we run other checks.
1394 if (is_array($attary) && sizeof($attary) > 0){
1395 $attary = sq_fixatts($tagname,
1406 * Convert body into div.
1408 if ($tagname == "body"){
1410 $attary = sq_body2div($attary, $message, $id);
1417 if ($tagname != false && $skip_content == false){
1418 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1424 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1425 if ($force_tag_closing == true){
1426 foreach ($open_tags as $tagname=>$opentimes){
1427 while ($opentimes > 0){
1428 $trusted .= '</' . $tagname . '>';
1434 $trusted .= "<!-- end sanitized html -->\n";
1439 * This is a wrapper function to call html sanitizing routines.
1441 * @param $body the body of the message
1442 * @param $id the id of the message
1443 * @return a string with html safe to display in the browser.
1445 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1446 global $attachment_common_show_images, $view_unsafe_images,
1449 * Don't display attached images in HTML mode.
1451 $attachment_common_show_images = false;
1462 $rm_tags_with_content = Array(
1469 $self_closing_tags = Array(
1476 $force_tag_closing = false;
1478 $rm_attnames = Array(
1488 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1489 $bad_attvals = Array(
1492 "/^src|background/i" =>
1495 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1496 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1497 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1498 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1501 "\\1$secremoveimg\\2",
1502 "\\1$secremoveimg\\2",
1503 "\\1$secremoveimg\\2",
1504 "\\1$secremoveimg\\2"
1507 "/^href|action/i" =>
1510 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1511 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1512 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1513 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1528 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1529 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1530 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1531 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1545 if (!$view_unsafe_images){
1547 * Remove any references to http/https if view_unsafe_images set
1550 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1551 '/^([\'\"])\s*https*:.*([\'\"])/si');
1552 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1553 "\\1$secremoveimg\\2");
1554 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1555 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1556 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1557 "url(\\1$secremoveimg\\2)");
1560 $add_attr_to_tag = Array(
1561 "/^a$/si" => Array('target'=>'"_new"')
1563 $trusted = sq_sanitize($body,
1565 $rm_tags_with_content,
1575 if (preg_match("|$secremoveimg|si", $trusted)){
1576 $has_unsafe_images = true;