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 // don\8e´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 ]----------------------------------------------------------- */
202 * Checks to see if $message contains content of type $type0/$type1
203 * returns the first entity number it finds of that type, or NULL if
204 * none is found. Takes optional argument $start to allow the caller
205 * to continue where they left off
207 function findDisplayEntity($message, $type0, $type1, $start=0) {
209 for ($i = $start;isset($message->entities
[$i]); $i++
) {
210 $entity = $message->entities
[$i];
211 if ($entity->type0
== $type0 && $entity->type1
== $type1) {
219 // This is here for debugging purposese. It will print out a list
220 // of all the entity IDs that are in the $message object.
222 function listEntities ($message) {
224 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
225 for ($i = 0;isset($message->entities
[$i]); $i++
) {
227 $msg = listEntities($message->entities
[$i]);
237 function getPriorityStr($priority) {
238 $priority_level = substr($priority,0,1);
240 switch($priority_level) {
241 /* check for a higher then normal priority. */
244 $priority_string = _("High");
247 /* check for a lower then normal priority. */
250 $priority_string = _("Low");
253 /* check for a normal priority. */
256 $priority_level = '3';
257 $priority_string = _("Normal");
261 return $priority_string;
264 /* returns a $message object for a particular entity id */
265 function getEntity ($message, $ent_id) {
266 return $message->getEntity($ent_id);
271 * Extracted from strings.php 23/03/2002
274 function translateText(&$body, $wrap_at, $charset) {
275 global $where, $what; /* from searching */
276 global $color; /* color theme */
278 require_once('../functions/url_parser.php');
280 $body_ary = explode("\n", $body);
282 for ($i=0; $i < count($body_ary); $i++
) {
283 $line = $body_ary[$i];
284 if (strlen($line) - 2 >= $wrap_at) {
285 sqWordWrap($line, $wrap_at);
287 $line = charset_decode($charset, $line);
288 $line = str_replace("\t", ' ', $line);
294 $j = strlen( $line );
296 while ( $pos < $j ) {
297 if ($line[$pos] == ' ') {
299 } else if (strpos($line, '>', $pos) === $pos) {
308 if (! isset($color[14])) {
309 $color[14] = '#FF0000';
311 $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
313 if (! isset($color[13])) {
314 $color[13] = '#800000';
316 $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
319 $body_ary[$i] = $line;
321 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
325 /* This returns a parsed string called $body. That string can then
326 be displayed as the actual message in the HTML. It contains
327 everything needed, including HTML Tags, Attachments at the
330 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
331 // this if statement checks for the entity to show as the
332 // primary message. To add more of them, just put them in the
333 // order that is their priority.
334 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
335 $show_html_default, $has_unsafe_images, $view_unsafe_images, $sort;
337 $has_unsafe_images= 0;
339 $urlmailbox = urlencode($mailbox);
340 $body_message = getEntity($message, $ent_num);
341 if (($body_message->header
->type0
== 'text') ||
342 ($body_message->header
->type0
== 'rfc822')) {
343 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
344 $body = decodeBody($body, $body_message->header
->encoding
);
345 $hookResults = do_hook("message_body", $body);
346 $body = $hookResults[1];
348 // If there are other types that shouldn't be formatted, add
350 if ($body_message->header
->type1
== 'html') {
351 $body = charset_decode_japanese($body);
352 if ( $show_html_default <> 1 ) {
353 $body = strip_tags( $body );
354 translateText($body, $wrap_at,
355 $body_message->header
->getParameter
['charset']);
357 $body = magicHTML( $body, $id, $message, $mailbox );
360 translateText($body, $wrap_at,
361 $body_message->header
->getParameter('charset'));
364 if ($has_unsafe_images) {
365 if ($view_unsafe_images) {
366 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&passed_ent_id=".$message->entity_id
."&mailbox=$urlmailbox&sort=$sort&startMessage=$startMessage&show_more=0\">". _("Hide Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
368 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$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";
376 function formatAttachments($message, $exclude_id, $mailbox, $id) {
377 global $where, $what;
378 global $startMessage, $color;
379 static $ShownHTML = 0;
381 $att_ar = $message->getAttachments($exclude_id);
383 if (!count($att_ar)) return '';
387 $urlMailbox = urlencode($mailbox);
389 foreach ($att_ar as $att) {
390 $ent = urldecode($att->entity_id
);
391 $header = $att->header
;
392 $type0 = strtolower($header->type0
);
393 $type1 = strtolower($header->type1
);
395 $Links['download link']['text'] = _("download");
396 $Links['download link']['href'] =
397 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
399 if ($type0 =='message' && $type1 == 'rfc822') {
400 $default_page = '../src/read_body.php';
401 $rfc822_header = $att->rfc822_header
;
402 $filename = decodeHeader($rfc822_header->subject
);
404 $from_o = $rfc822_header->from
;
405 if (is_object($from_o)) {
406 $from_name = $from_o->getAddress(false);
408 $from_name = _("Unknown sender");
410 $from_name = decodeHeader(htmlspecialchars($from_name));
411 $description = $from_name;
413 $default_page = '../src/download.php';
414 if (is_object($header->disposition
)) {
415 $filename = decodeHeader($header->disposition
->getProperty('filename'));
416 if (trim($filename) == '') {
417 $name = decodeHeader($header->disposition
->getProperty('name'));
418 if (trim($name) == '') {
419 if ( trim( $header->id
) == '' )
420 $filename = 'untitled-[' . $ent . ']' ;
422 $filename = 'cid: ' . $header->id
;
428 if ( trim( $header->id
) == '' )
429 $filename = 'untitled-[' . $ent . ']' ;
431 $filename = 'cid: ' . $header->id
;
434 if ($header->description
) {
435 $description = htmlspecialchars($header->description
);
441 $display_filename = $filename;
442 if (isset($passed_ent_id)) {
443 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
445 $passed_ent_id_link = '';
447 $DefaultLink = $default_page . "?startMessage=$startMessage"
448 . "&passed_id=$id&mailbox=$urlMailbox"
449 . '&ent_id='.$ent.$passed_ent_id_link;
450 if ($where && $what) {
451 $DefaultLink = '&where='. urlencode($where).'&what='.urlencode($what);
453 /* this executes the attachment hook with a specific MIME-type.
454 * if that doens't have results, it tries if there's a rule
455 * for a more generic type. */
456 $HookResults = do_hook("attachment $type0/$type1", $Links,
457 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
458 $display_filename, $where, $what);
459 if(count($HookResults[1]) <= 1) {
460 $HookResults = do_hook("attachment $type0/*", $Links,
461 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
462 $display_filename, $where, $what);
465 $Links = $HookResults[1];
466 $DefaultLink = $HookResults[6];
468 $attachments .= '<TR><TD>' .
469 "<A HREF=\"$DefaultLink\">$display_filename</A> </TD>" .
470 '<TD><SMALL><b>' . show_readable_size($header->size
) .
471 '</b> </small></TD>' .
472 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
474 $attachments .= '<b>' . $description . '</b>';
475 $attachments .= '</SMALL></TD><TD><SMALL> ';
478 foreach ($Links as $Val) {
482 $attachments .= ' | ';
484 $attachments .= '<a href="' . $Val['href'] . '">' . $Val['text'] . '</a>';
487 $attachments .= "</TD></TR>\n";
492 /** this function decodes the body depending on the encoding type. **/
493 function decodeBody($body, $encoding) {
495 $body = str_replace("\r\n", "\n", $body);
496 $encoding = strtolower($encoding);
498 global $show_html_default;
500 if ($encoding == 'quoted-printable' ||
501 $encoding == 'quoted_printable') {
502 $body = quoted_printable_decode($body);
504 while (ereg("=\n", $body))
505 $body = ereg_replace ("=\n", "", $body);
507 } else if ($encoding == 'base64') {
508 $body = base64_decode($body);
511 // All other encodings are returned raw.
516 * This functions decode strings that is encoded according to
517 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
518 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
520 function decodeHeader ($string, $utfencode=true) {
521 global $squirrelmail_language;
522 if (is_array($string)) {
523 $string = implode("\n", $string);
526 if ($squirrelmail_language == 'ja_JP' && function_exists('mb_convert_encoding')) {
527 $string = str_replace("\t", "", $string);
528 if (eregi('=\\?([^?]+)\\?(q|b)\\?([^?]+)\\?=', $string))
529 $string = mb_decode_mimeheader($string);
530 return @mb_convert_encoding
($string, 'EUC-JP', 'AUTO');
534 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
537 // Ignore white-space between consecutive encoded-words
538 if (strspn($res[2], " \t") != strlen($res[2])) {
542 if (ucfirst($res[4]) == 'B') {
543 $replace = base64_decode($res[5]);
545 $replace = str_replace('_', ' ', $res[5]);
546 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
548 /* Only encode into entities by default. Some places
549 don't need the encoding, like the compose form. */
551 $replace = charset_decode($res[3], $replace);
554 $string = $prefix . $replace . substr($string, strlen($res[0]));
555 $i = strlen($prefix) +
strlen($replace);
561 * Encode a string according to RFC 1522 for use in headers if it
562 * contains 8-bit characters or anything that looks like it should
565 function encodeHeader ($string) {
566 global $default_charset, $squirrelmail_language;
568 if ($squirrelmail_language == 'ja_JP' && function_exists('mb_encode_mimeheader')) {
569 return mb_encode_mimeheader($string);
572 // Encode only if the string contains 8-bit characters or =?
573 $j = strlen( $string );
574 $l = strstr($string, '=?'); // Must be encoded ?
576 for( $i=0; $i < $j; ++
$i) {
577 switch( $string{$i} ) {
591 $k = ord( $string{$i} );
593 $ret .= sprintf("=%02X", $k);
601 $string = "=?$default_charset?Q?$ret?=";
607 /* This function trys to locate the entity_id of a specific mime element */
609 function find_ent_id( $id, $message ) {
611 for ($i=0; $ret == '' && $i < count($message->entities
); $i++
) {
612 if ( $message->entities
[$i]->header
->type0
== 'multipart') {
613 $ret = find_ent_id( $id, $message->entities
[$i] );
615 if ( strcasecmp( $message->entities
[$i]->header
->id
, $id ) == 0 )
616 $ret = $message->entities
[$i]->entity_id
;
623 ** HTMLFILTER ROUTINES
627 * This function returns the final tag out of the tag name, an array
628 * of attributes, and the type of the tag. This function is called by
629 * sq_sanitize internally.
631 * @param $tagname the name of the tag.
632 * @param $attary the array of attributes and their values
633 * @param $tagtype The type of the tag (see in comments).
634 * @return a string with the final tag representation.
636 function sq_tagprint($tagname, $attary, $tagtype){
639 $fulltag = '</' . $tagname . '>';
641 $fulltag = '<' . $tagname;
642 if (is_array($attary) && sizeof($attary)){
644 while (list($attname, $attvalue) = each($attary)){
645 array_push($atts, "$attname=$attvalue");
647 $fulltag .= ' ' . join(" ", $atts);
658 * A small helper function to use with array_walk. Modifies a by-ref
659 * value and makes it lowercase.
661 * @param $val a value passed by-ref.
662 * @return void since it modifies a by-ref value.
664 function sq_casenormalize(&$val){
665 $val = strtolower($val);
669 * This function skips any whitespace from the current position within
670 * a string and to the next non-whitespace value.
672 * @param $body the string
673 * @param $offset the offset within the string where we should start
674 * looking for the next non-whitespace character.
675 * @return the location within the $body where the next
676 * non-whitespace char is located.
678 function sq_skipspace($body, $offset){
679 $me = "sq_skipspace";
680 preg_match("/^(\s*)/s", substr($body, $offset), $matches);
681 if (sizeof($matches{1})){
682 $count = strlen($matches{1});
689 * This function looks for the next character within a string. It's
690 * really just a glorified "strpos", except it catches if failures
693 * @param $body The string to look for needle in.
694 * @param $offset Start looking from this position.
695 * @param $needle The character/string to look for.
696 * @return location of the next occurance of the needle, or
697 * strlen($body) if needle wasn't found.
699 function sq_findnxstr($body, $offset, $needle){
700 $me = "sq_findnxstr";
701 $pos = strpos($body, $needle, $offset);
703 $pos = strlen($body);
709 * This function takes a PCRE-style regexp and tries to match it
712 * @param $body The string to look for needle in.
713 * @param $offset Start looking from here.
714 * @param $reg A PCRE-style regex to match.
715 * @return Returns a false if no matches found, or an array
716 * with the following members:
717 * - integer with the location of the match within $body
718 * - string with whatever content between offset and the match
719 * - string with whatever it is we matched
721 function sq_findnxreg($body, $offset, $reg){
722 $me = "sq_findnxreg";
725 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
729 $retarr{0} = $offset +
strlen($matches{1});
730 $retarr{1} = $matches{1};
731 $retarr{2} = $matches{2};
737 * This function looks for the next tag.
739 * @param $body String where to look for the next tag.
740 * @param $offset Start looking from here.
741 * @return false if no more tags exist in the body, or
742 * an array with the following members:
743 * - string with the name of the tag
744 * - array with attributes and their values
745 * - integer with tag type (1, 2, or 3)
746 * - integer where the tag starts (starting "<")
747 * - integer where the tag ends (ending ">")
748 * first three members will be false, if the tag is invalid.
750 function sq_getnxtag($body, $offset){
752 if ($offset > strlen($body)){
755 $lt = sq_findnxstr($body, $offset, "<");
756 if ($lt == strlen($body)){
761 * blah blah <tag attribute="value">
764 $pos = sq_skipspace($body, $lt+
1);
765 if ($pos >= strlen($body)){
766 return Array(false, false, false, $lt, strlen($body));
769 * There are 3 kinds of tags:
770 * 1. Opening tag, e.g.:
772 * 2. Closing tag, e.g.:
774 * 3. XHTML-style content-less tag, e.g.:
778 switch (substr($body, $pos, 1)){
785 * A comment or an SGML declaration.
787 if (substr($body, $pos+
1, 2) == "--"){
788 $gt = strpos($body, "-->", $pos);
794 return Array(false, false, false, $lt, $gt);
796 $gt = sq_findnxstr($body, $pos, ">");
797 return Array(false, false, false, $lt, $gt);
802 * Assume tagtype 1 for now. If it's type 3, we'll switch values
812 * Look for next [\W-_], which will indicate the end of the tag name.
814 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
815 if ($regary == false){
816 return Array(false, false, false, $lt, strlen($body));
818 list($pos, $tagname, $match) = $regary;
819 $tagname = strtolower($tagname);
822 * $match can be either of these:
823 * '>' indicating the end of the tag entirely.
824 * '\s' indicating the end of the tag name.
825 * '/' indicating that this is type-3 xhtml tag.
827 * Whatever else we find there indicates an invalid tag.
832 * This is an xhtml-style tag with a closing / at the
833 * end, like so: <img src="blah"/>. Check if it's followed
834 * by the closing bracket. If not, then this tag is invalid
836 if (substr($body, $pos, 2) == "/>"){
840 $gt = sq_findnxstr($body, $pos, ">");
841 $retary = Array(false, false, false, $lt, $gt);
845 return Array($tagname, false, $tagtype, $lt, $pos);
849 * Check if it's whitespace
851 if (preg_match("/\s/", $match)){
854 * This is an invalid tag! Look for the next closing ">".
856 $gt = sq_findnxstr($body, $offset, ">");
857 return Array(false, false, false, $lt, $gt);
862 * At this point we're here:
863 * <tagname attribute='blah'>
866 * At this point we loop in order to find all attributes.
872 while ($pos <= strlen($body)){
873 $pos = sq_skipspace($body, $pos);
874 if ($pos == strlen($body)){
878 return Array(false, false, false, $lt, $pos);
881 * See if we arrived at a ">" or "/>", which means that we reached
882 * the end of the tag.
885 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
889 $pos +
= strlen($matches{1});
890 if ($matches{2} == "/>"){
894 return Array($tagname, $attary, $tagtype, $lt, $pos);
898 * There are several types of attributes, with optional
899 * [:space:] between members.
901 * attrname[:space:]=[:space:]'CDATA'
903 * attrname[:space:]=[:space:]"CDATA"
905 * attr[:space:]=[:space:]CDATA
909 * We leave types 1 and 2 the same, type 3 we check for
910 * '"' and convert to """ if needed, then wrap in
911 * double quotes. Type 4 we convert into:
914 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
915 if ($regary == false){
917 * Looks like body ended before the end of tag.
919 return Array(false, false, false, $lt, strlen($body));
921 list($pos, $attname, $match) = $regary;
922 $attname = strtolower($attname);
924 * We arrived at the end of attribute name. Several things possible
926 * '>' means the end of the tag and this is attribute type 4
927 * '/' if followed by '>' means the same thing as above
928 * '\s' means a lot of things -- look what it's followed by.
929 * anything else means the attribute is invalid.
934 * This is an xhtml-style tag with a closing / at the
935 * end, like so: <img src="blah"/>. Check if it's followed
936 * by the closing bracket. If not, then this tag is invalid
938 if (substr($body, $pos, 2) == "/>"){
942 $gt = sq_findnxstr($body, $pos, ">");
943 $retary = Array(false, false, false, $lt, $gt);
947 $attary{$attname} = '"yes"';
948 return Array($tagname, $attary, $tagtype, $lt, $pos);
952 * Skip whitespace and see what we arrive at.
954 $pos = sq_skipspace($body, $pos);
955 $char = substr($body, $pos, 1);
957 * Two things are valid here:
958 * '=' means this is attribute type 1 2 or 3.
959 * \w means this was attribute type 4.
960 * anything else we ignore and re-loop. End of tag and
961 * invalid stuff will be caught by our checks at the beginning
966 $pos = sq_skipspace($body, $pos);
968 * Here are 3 possibilities:
969 * "'" attribute type 1
970 * '"' attribute type 2
971 * everything else is the content of tag type 3
973 $quot = substr($body, $pos, 1);
975 $regary = sq_findnxreg($body, $pos+
1, "\'");
976 if ($regary == false){
977 return Array(false, false, false, $lt, strlen($body));
979 list($pos, $attval, $match) = $regary;
981 $attary{$attname} = "'" . $attval . "'";
982 } else if ($quot == '"'){
983 $regary = sq_findnxreg($body, $pos+
1, '\"');
984 if ($regary == false){
985 return Array(false, false, false, $lt, strlen($body));
987 list($pos, $attval, $match) = $regary;
989 $attary{$attname} = '"' . $attval . '"';
992 * These are hateful. Look for \s, or >.
994 $regary = sq_findnxreg($body, $pos, "[\s>]");
995 if ($regary == false){
996 return Array(false, false, false, $lt, strlen($body));
998 list($pos, $attval, $match) = $regary;
1000 * If it's ">" it will be caught at the top.
1002 $attval = preg_replace("/\"/s", """, $attval);
1003 $attary{$attname} = '"' . $attval . '"';
1005 } else if (preg_match("|[\w/>]|", $char)) {
1007 * That was attribute type 4.
1009 $attary{$attname} = '"yes"';
1012 * An illegal character. Find next '>' and return.
1014 $gt = sq_findnxstr($body, $pos, ">");
1015 return Array(false, false, false, $lt, $gt);
1020 * The fact that we got here indicates that the tag end was never
1021 * found. Return invalid tag indication so it gets stripped.
1023 return Array(false, false, false, $lt, strlen($body));
1027 * This function checks attribute values for entity-encoded values
1028 * and returns them translated into 8-bit strings so we can run
1031 * @param $attvalue A string to run entity check against.
1032 * @return Translated value.
1034 function sq_deent($attvalue){
1037 * See if we have to run the checks first. All entities must start
1040 if (strpos($attvalue, "&") === false){
1044 * Check named entities first.
1046 $trans = get_html_translation_table(HTML_ENTITIES
);
1048 * Leave " in, as it can mess us up.
1050 $trans = array_flip($trans);
1051 unset($trans{"""});
1052 while (list($ent, $val) = each($trans)){
1053 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1056 * Now translate numbered entities from 1 to 255 if needed.
1058 if (strpos($attvalue, "#") !== false){
1059 $omit = Array(34, 39);
1060 for ($asc=1; $asc<256; $asc++
){
1061 if (!in_array($asc, $omit)){
1063 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1065 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1066 "$chr\\1", $attvalue);
1074 * This function runs various checks against the attributes.
1076 * @param $tagname String with the name of the tag.
1077 * @param $attary Array with all tag attributes.
1078 * @param $rm_attnames See description for sq_sanitize
1079 * @param $bad_attvals See description for sq_sanitize
1080 * @param $add_attr_to_tag See description for sq_sanitize
1081 * @param $message message object
1082 * @param $id message id
1083 * @return Array with modified attributes.
1085 function sq_fixatts($tagname,
1095 while (list($attname, $attvalue) = each($attary)){
1097 * See if this attribute should be removed.
1099 foreach ($rm_attnames as $matchtag=>$matchattrs){
1100 if (preg_match($matchtag, $tagname)){
1101 foreach ($matchattrs as $matchattr){
1102 if (preg_match($matchattr, $attname)){
1103 unset($attary{$attname});
1110 * Remove any entities.
1112 $attvalue = sq_deent($attvalue);
1115 * Now let's run checks on the attvalues.
1116 * I don't expect anyone to comprehend this. If you do,
1117 * get in touch with me so I can drive to where you live and
1118 * shake your hand personally. :)
1120 foreach ($bad_attvals as $matchtag=>$matchattrs){
1121 if (preg_match($matchtag, $tagname)){
1122 foreach ($matchattrs as $matchattr=>$valary){
1123 if (preg_match($matchattr, $attname)){
1125 * There are two arrays in valary.
1127 * Second one is replacements
1129 list($valmatch, $valrepl) = $valary;
1131 preg_replace($valmatch, $valrepl, $attvalue);
1132 if ($newvalue != $attvalue){
1133 $attary{$attname} = $newvalue;
1140 * Turn cid: urls into http-friendly ones.
1142 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1143 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1147 * See if we need to append any attributes to this tag.
1149 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1150 if (preg_match($matchtag, $tagname)){
1151 $attary = array_merge($attary, $addattary);
1158 * This function edits the style definition to make them friendly and
1159 * usable in squirrelmail.
1161 * @param $message the message object
1162 * @param $id the message id
1163 * @param $content a string with whatever is between <style> and </style>
1164 * @return a string with edited content.
1166 function sq_fixstyle($message, $id, $content){
1167 global $view_unsafe_images;
1168 $me = "sq_fixstyle";
1170 * First look for general BODY style declaration, which would be
1172 * body {background: blah-blah}
1173 * and change it to .bodyclass so we can just assign it to a <div>
1175 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1176 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1178 * Fix url('blah') declarations.
1180 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1181 "url(\\1$secremoveimg\\2)", $content);
1183 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1186 if (!$view_unsafe_images){
1187 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1188 "url(\\1$secremoveimg\\2)", $content);
1192 * Fix urls that refer to cid:
1194 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1196 $cidurl = $matches{1};
1197 $httpurl = sq_cid2http($message, $id, $cidurl);
1198 $content = preg_replace("|url\($cidurl\)|si",
1199 "url($httpurl)", $content);
1203 * Fix stupid css declarations which lead to vulnerabilities
1206 $match = Array('/expression/si',
1209 $replace = Array('idiocy', 'idiocy', 'idiocy');
1210 $content = preg_replace($match, $replace, $content);
1215 * This function converts cid: url's into the ones that can be viewed in
1218 * @param $message the message object
1219 * @param $id the message id
1220 * @param $cidurl the cid: url.
1221 * @return a string with a http-friendly url
1223 function sq_cid2http($message, $id, $cidurl, $mailbox){
1225 * Get rid of quotes.
1227 $quotchar = substr($cidurl, 0, 1);
1228 $cidurl = str_replace($quotchar, "", $cidurl);
1229 $cidurl = substr(trim($cidurl), 4);
1230 $httpurl = $quotchar . "../src/download.php?absolute_dl=true&" .
1231 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1232 "&ent_id=" . find_ent_id($cidurl, $message) . $quotchar;
1237 * This function changes the <body> tag into a <div> tag since we
1238 * can't really have a body-within-body.
1240 * @param $attary an array of attributes and values of <body>
1241 * @return a modified array of attributes to be set for <div>
1243 function sq_body2div($attary){
1244 $me = "sq_body2div";
1245 $divattary = Array("class"=>"'bodyclass'");
1249 if (is_array($attary) && sizeof($attary) > 0){
1250 foreach ($attary as $attname=>$attvalue){
1251 $quotchar = substr($attvalue, 0, 1);
1252 $attvalue = str_replace($quotchar, "", $attvalue);
1255 $styledef .= "background-image: url('$attvalue'); ";
1258 $styledef .= "background-color: $attvalue; ";
1261 $styledef .= "color: $attvalue; ";
1264 if (strlen($styledef) > 0){
1265 $divattary{"style"} = "\"$styledef\"";
1272 * This is the main function and the one you should actually be calling.
1273 * There are several variables you should be aware of an which need
1274 * special description.
1276 * Since the description is quite lengthy, see it here:
1277 * http://www.mricon.com/html/phpfilter.html
1279 * @param $body the string with HTML you wish to filter
1280 * @param $tag_list see description above
1281 * @param $rm_tags_with_content see description above
1282 * @param $self_closing_tags see description above
1283 * @param $force_tag_closing see description above
1284 * @param $rm_attnames see description above
1285 * @param $bad_attvals see description above
1286 * @param $add_attr_to_tag see description above
1287 * @param $message message object
1288 * @param $id message id
1289 * @return sanitized html safe to show on your pages.
1291 function sq_sanitize($body,
1293 $rm_tags_with_content,
1303 $me = "sq_sanitize";
1305 * Normalize rm_tags and rm_tags_with_content.
1307 @array_walk
($rm_tags, 'sq_casenormalize');
1308 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1309 @array_walk
($self_closing_tags, 'sq_casenormalize');
1311 * See if tag_list is of tags to remove or tags to allow.
1312 * false means remove these tags
1313 * true means allow these tags
1315 $rm_tags = array_shift($tag_list);
1317 $open_tags = Array();
1318 $trusted = "<!-- begin sanitized html -->\n";
1319 $skip_content = false;
1321 * Take care of netscape's stupid javascript entities like
1324 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1326 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1327 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1328 $free_content = substr($body, $curpos, $lt-$curpos);
1330 * Take care of <style>
1332 if ($tagname == "style" && $tagtype == 2){
1334 * This is a closing </style>. Edit the
1335 * content before we apply it.
1337 $free_content = sq_fixstyle($message, $id, $free_content);
1339 if ($skip_content == false){
1340 $trusted .= $free_content;
1343 if ($tagname != FALSE){
1345 if ($skip_content == $tagname){
1347 * Got to the end of tag we needed to remove.
1350 $skip_content = false;
1352 if ($skip_content == false){
1353 if ($tagname == "body"){
1356 if (isset($open_tags{$tagname}) &&
1357 $open_tags{$tagname} > 0){
1358 $open_tags{$tagname}--;
1368 * $rm_tags_with_content
1370 if ($skip_content == false){
1372 * See if this is a self-closing type and change
1373 * tagtype appropriately.
1376 && in_array($tagname, $self_closing_tags)){
1380 * See if we should skip this tag and any content
1383 if ($tagtype == 1 &&
1384 in_array($tagname, $rm_tags_with_content)){
1385 $skip_content = $tagname;
1387 if (($rm_tags == false
1388 && in_array($tagname, $tag_list)) ||
1389 ($rm_tags == true &&
1390 !in_array($tagname, $tag_list))){
1394 if (isset($open_tags{$tagname})){
1395 $open_tags{$tagname}++
;
1397 $open_tags{$tagname}=1;
1401 * This is where we run other checks.
1403 if (is_array($attary) && sizeof($attary) > 0){
1404 $attary = sq_fixatts($tagname,
1415 * Convert body into div.
1417 if ($tagname == "body"){
1419 $attary = sq_body2div($attary, $message, $id);
1426 if ($tagname != false && $skip_content == false){
1427 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1433 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1434 if ($force_tag_closing == true){
1435 foreach ($open_tags as $tagname=>$opentimes){
1436 while ($opentimes > 0){
1437 $trusted .= '</' . $tagname . '>';
1443 $trusted .= "<!-- end sanitized html -->\n";
1448 * This is a wrapper function to call html sanitizing routines.
1450 * @param $body the body of the message
1451 * @param $id the id of the message
1452 * @return a string with html safe to display in the browser.
1454 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1455 global $attachment_common_show_images, $view_unsafe_images,
1458 * Don't display attached images in HTML mode.
1460 $attachment_common_show_images = false;
1471 $rm_tags_with_content = Array(
1478 $self_closing_tags = Array(
1485 $force_tag_closing = false;
1487 $rm_attnames = Array(
1497 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1498 $bad_attvals = Array(
1501 "/^src|background/i" =>
1504 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1505 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1506 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1507 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1510 "\\1$secremoveimg\\2",
1511 "\\1$secremoveimg\\2",
1512 "\\1$secremoveimg\\2",
1513 "\\1$secremoveimg\\2"
1516 "/^href|action/i" =>
1519 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1520 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1521 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1522 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1537 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1538 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1539 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1540 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1554 if (!$view_unsafe_images){
1556 * Remove any references to http/https if view_unsafe_images set
1559 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1560 '/^([\'\"])\s*https*:.*([\'\"])/si');
1561 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1562 "\\1$secremoveimg\\2");
1563 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1564 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1565 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1566 "url(\\1$secremoveimg\\2)");
1569 $add_attr_to_tag = Array(
1570 "/^a$/si" => Array('target'=>'"_new"')
1572 $trusted = sq_sanitize($body,
1574 $rm_tags_with_content,
1584 if (preg_match("|$secremoveimg|si", $trusted)){
1585 $has_unsafe_images = true;