6 * Copyright (c) 1999-2003 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, $color );
44 echo '</body></html>';
48 foreach ($flags as $flag) {
49 $char = strtoupper($flag{1});
52 if (strtolower($flag) == '\\seen') {
57 if (strtolower($flag) == '\\answered') {
58 $msg->is_answered
= true;
62 if (strtolower($flag) == '\\deleted') {
63 $msg->is_deleted
= true;
67 if (strtolower($flag) == '\\flagged') {
68 $msg->is_flagged
= true;
72 if (strtolower($flag) == '$mdnsent') {
73 $msg->is_mdnsent
= true;
81 // listEntities($msg);
87 /* This starts the parsing of a particular structure. It is called recursively,
88 * so it can be passed different structures. It returns an object of type
90 * First, it checks to see if it is a multipart message. If it is, then it
91 * handles that as it sees is necessary. If it is just a regular entity,
92 * then it parses it and adds the necessary header information (by calling out
93 * to mime_get_elements()
96 function mime_fetch_body($imap_stream, $id, $ent_id=1) {
98 /* Do a bit of error correction. If we couldn't find the entity id, just guess
99 * that it is the first one. That is usually the case anyway.
102 $cmd = "FETCH $id BODY[]";
104 $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=1, $encoding) {
158 $sid = sqimap_session_id($uid_support);
159 /* Don't kill the connection if the browser is over a dialup
160 * and it would take over 30 seconds to download it.
161 * DonĀ“t call set_time_limit in safe mode.
164 if (!ini_get('safe_mode')) {
168 $sid_s = substr($sid,0,strpos($sid, ' '));
173 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
174 echo decodeBody($body, $encoding);
177 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
180 $read = fgets ($imap_stream,8192);
183 // This could be bad -- if the section has sqimap_session_id() . ' OK'
184 // or similar, it will kill the download.
185 while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
186 if (trim($read) == ')==') {
188 $read = fgets ($imap_stream,4096);
189 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
192 echo decodeBody($read1, $encoding) .
193 decodeBody($read, $encoding);
196 echo decodeBody($read, $encoding);
198 $read = fgets ($imap_stream,4096);
205 /* -[ END MIME DECODING ]----------------------------------------------------------- */
207 /* This is here for debugging purposes. It will print out a list
208 * of all the entity IDs that are in the $message object.
210 function listEntities ($message) {
212 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br>';
213 for ($i = 0; isset($message->entities
[$i]); $i++
) {
215 $msg = listEntities($message->entities
[$i]);
225 function getPriorityStr($priority) {
226 $priority_level = substr($priority,0,1);
228 switch($priority_level) {
229 /* Check for a higher then normal priority. */
232 $priority_string = _("High");
235 /* Check for a lower then normal priority. */
238 $priority_string = _("Low");
241 /* Check for a normal priority. */
244 $priority_level = '3';
245 $priority_string = _("Normal");
249 return $priority_string;
252 /* returns a $message object for a particular entity id */
253 function getEntity ($message, $ent_id) {
254 return $message->getEntity($ent_id);
258 * Extracted from strings.php 23/03/2002
261 function translateText(&$body, $wrap_at, $charset) {
262 global $where, $what; /* from searching */
263 global $color; /* color theme */
265 require_once(SM_PATH
. 'functions/url_parser.php');
267 $body_ary = explode("\n", $body);
268 for ($i=0; $i < count($body_ary); $i++
) {
269 $line = $body_ary[$i];
270 if (strlen($line) - 2 >= $wrap_at) {
271 sqWordWrap($line, $wrap_at);
273 $line = charset_decode($charset, $line);
274 $line = str_replace("\t", ' ', $line);
283 if ($line[$pos] == ' ') {
285 } else if (strpos($line, '>', $pos) === $pos) {
294 if (!isset($color[14])) {
295 $color[14] = '#FF0000';
297 $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
299 if (!isset($color[13])) {
300 $color[13] = '#800000';
302 $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
305 $body_ary[$i] = $line;
307 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
311 /* This returns a parsed string called $body. That string can then
312 * be displayed as the actual message in the HTML. It contains
313 * everything needed, including HTML Tags, Attachments at the
316 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
317 /* This if statement checks for the entity to show as the
318 * primary message. To add more of them, just put them in the
319 * order that is their priority.
321 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
322 $show_html_default, $sort, $has_unsafe_images;
324 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
325 $view_unsafe_images = false;
329 $urlmailbox = urlencode($mailbox);
330 $body_message = getEntity($message, $ent_num);
331 if (($body_message->header
->type0
== 'text') ||
332 ($body_message->header
->type0
== 'rfc822')) {
333 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
334 $body = decodeBody($body, $body_message->header
->encoding
);
335 $hookResults = do_hook("message_body", $body);
336 $body = $hookResults[1];
338 /* If there are other types that shouldn't be formatted, add
342 if ($body_message->header
->type1
== 'html') {
343 if ($show_html_default <> 1) {
344 $entity_conv = array(' ' => ' ',
351 $body = strtr($body, $entity_conv);
352 $body = strip_tags($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 $body_message->header
->getParameter('charset'));
364 if ($view_unsafe_images) {
365 $untext = '">' . _("Hide Unsafe Images");
367 if (isset($has_unsafe_images) && $has_unsafe_images) {
368 $untext = '&view_unsafe_images=1">' . _("View Unsafe Images");
373 $body .= '<center><small><a href="read_body.php?passed_id=' . $id .
374 '&passed_ent_id=' . $message->entity_id
. '&mailbox=' . $urlmailbox .
375 '&sort=' . $sort . '&startMessage=' . $startMessage . '&show_more=0' .
376 $untext . '</a></small></center><br>' . "\n";
383 function formatAttachments($message, $exclude_id, $mailbox, $id) {
384 global $where, $what, $startMessage, $color;
385 static $ShownHTML = 0;
387 $att_ar = $message->getAttachments($exclude_id);
389 if (!count($att_ar)) return '';
393 $urlMailbox = urlencode($mailbox);
395 foreach ($att_ar as $att) {
396 $ent = urldecode($att->entity_id
);
397 $header = $att->header
;
398 $type0 = strtolower($header->type0
);
399 $type1 = strtolower($header->type1
);
401 $links['download link']['text'] = _("download");
402 $links['download link']['href'] =
403 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
405 if ($type0 =='message' && $type1 == 'rfc822') {
406 $default_page = '../src/read_body.php';
407 $rfc822_header = $att->rfc822_header
;
408 $filename = decodeHeader($rfc822_header->subject
);
409 if (trim( $filename ) == '') {
410 $filename = 'untitled-[' . $ent . ']' ;
412 $from_o = $rfc822_header->from
;
413 if (is_object($from_o)) {
414 $from_name = $from_o->getAddress(false);
416 $from_name = _("Unknown sender");
418 $from_name = decodeHeader(htmlspecialchars($from_name));
419 $description = $from_name;
421 $default_page = '../src/download.php';
422 if (is_object($header->disposition
)) {
423 $filename = decodeHeader($header->disposition
->getProperty('filename'));
424 if (trim($filename) == '') {
425 $name = decodeHeader($header->disposition
->getProperty('name'));
426 if (trim($name) == '') {
427 $name = decodeHeader($header->getParameter('name'));
428 if(trim($name) == '') {
429 if (trim( $header->id
) == '') {
430 $filename = 'untitled-[' . $ent . ']' ;
432 $filename = 'cid: ' . $header->id
;
442 if (trim( $header->id
) == '') {
443 $filename = 'untitled-[' . $ent . ']' ;
445 $filename = 'cid: ' . $header->id
;
449 if ($header->description
) {
450 $description = htmlspecialchars($header->description
);
456 $display_filename = $filename;
457 if (isset($passed_ent_id)) {
458 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
460 $passed_ent_id_link = '';
462 $defaultlink = $default_page . "?startMessage=$startMessage"
463 . "&passed_id=$id&mailbox=$urlMailbox"
464 . '&ent_id='.$ent.$passed_ent_id_link.'&absolute_dl=true';
465 if ($where && $what) {
466 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
468 /* This executes the attachment hook with a specific MIME-type.
469 * If that doesn't have results, it tries if there's a rule
470 * for a more generic type.
472 $hookresults = do_hook("attachment $type0/$type1", $links,
473 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
474 $display_filename, $where, $what);
475 if(count($hookresults[1]) <= 1) {
476 $hookresults = do_hook("attachment $type0/*", $links,
477 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
478 $display_filename, $where, $what);
481 $links = $hookresults[1];
482 $defaultlink = $hookresults[6];
484 $attachments .= '<TR><TD>' .
485 "<A HREF=\"$defaultlink\">$display_filename</A> </TD>" .
486 '<TD><SMALL><b>' . show_readable_size($header->size
) .
487 '</b> </small></TD>' .
488 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
490 $attachments .= '<b>' . $description . '</b>';
491 $attachments .= '</SMALL></TD><TD><SMALL> ';
494 foreach ($links as $val) {
498 $attachments .= ' | ';
500 $attachments .= '<a href="' . $val['href'] . '">' . $val['text'] . '</a>';
503 $attachments .= "</TD></TR>\n";
508 /* This function decodes the body depending on the encoding type. */
509 function decodeBody($body, $encoding) {
510 global $languages, $squirrelmail_language;
511 global $show_html_default;
513 $body = str_replace("\r\n", "\n", $body);
514 $encoding = strtolower($encoding);
516 if ($encoding == 'quoted-printable' ||
517 $encoding == 'quoted_printable') {
518 $body = quoted_printable_decode($body);
520 while (ereg("=\n", $body)) {
521 $body = ereg_replace ("=\n", '', $body);
524 } else if ($encoding == 'base64') {
525 $body = base64_decode($body);
528 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
529 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
530 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
533 // All other encodings are returned raw.
538 * This functions decode strings that is encoded according to
539 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
540 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
542 function decodeHeader ($string, $utfencode=true) {
543 global $languages, $squirrelmail_language;
544 if (is_array($string)) {
545 $string = implode("\n", $string);
548 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
549 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
550 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
554 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
557 /* Ignore white-space between consecutive encoded-words. */
558 if (strspn($res[2], " \t") != strlen($res[2])) {
562 if (ucfirst($res[4]) == 'B') {
563 $replace = base64_decode($res[5]);
565 $replace = str_replace('_', ' ', $res[5]);
566 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
568 /* Only encode into entities by default. Some places
569 * don't need the encoding, like the compose form.
572 $replace = charset_decode($res[3], $replace);
575 $string = $prefix . $replace . substr($string, strlen($res[0]));
576 $i = strlen($prefix) +
strlen($replace);
582 * Encode a string according to RFC 1522 for use in headers if it
583 * contains 8-bit characters or anything that looks like it should
586 function encodeHeader ($string) {
587 global $default_charset, $languages, $squirrelmail_language;
589 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
590 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
591 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
594 // Encode only if the string contains 8-bit characters or =?
595 $j = strlen($string);
596 $l = strstr($string, '=?'); // Must be encoded ?
598 for($i = 0; $i < $j; ++
$i) {
599 switch($string{$i}) {
613 $k = ord($string{$i});
615 $ret .= sprintf("=%02X", $k);
625 $string = "=?$default_charset?Q?$ret?=";
631 /* This function trys to locate the entity_id of a specific mime element */
632 function find_ent_id($id, $message) {
633 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
634 if ($message->entities
[$i]->header
->type0
== 'multipart') {
635 $ret = find_ent_id($id, $message->entities
[$i]);
637 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
638 // if (sq_check_save_extension($message->entities[$i])) {
639 return $message->entities
[$i]->entity_id
;
647 function sq_check_save_extension($message) {
648 $filename = $message->getFilename();
649 $ext = substr($filename, strrpos($filename,'.')+
1);
650 $save_extensions = array('jpg','jpeg','gif','png','bmp');
651 return in_array($ext, $save_extensions);
656 ** HTMLFILTER ROUTINES
660 * This function returns the final tag out of the tag name, an array
661 * of attributes, and the type of the tag. This function is called by
662 * sq_sanitize internally.
664 * @param $tagname the name of the tag.
665 * @param $attary the array of attributes and their values
666 * @param $tagtype The type of the tag (see in comments).
667 * @return a string with the final tag representation.
669 function sq_tagprint($tagname, $attary, $tagtype){
673 $fulltag = '</' . $tagname . '>';
675 $fulltag = '<' . $tagname;
676 if (is_array($attary) && sizeof($attary)){
678 while (list($attname, $attvalue) = each($attary)){
679 array_push($atts, "$attname=$attvalue");
681 $fulltag .= ' ' . join(" ", $atts);
692 * A small helper function to use with array_walk. Modifies a by-ref
693 * value and makes it lowercase.
695 * @param $val a value passed by-ref.
696 * @return void since it modifies a by-ref value.
698 function sq_casenormalize(&$val){
699 $val = strtolower($val);
703 * This function skips any whitespace from the current position within
704 * a string and to the next non-whitespace value.
706 * @param $body the string
707 * @param $offset the offset within the string where we should start
708 * looking for the next non-whitespace character.
709 * @return the location within the $body where the next
710 * non-whitespace char is located.
712 function sq_skipspace($body, $offset){
713 $me = 'sq_skipspace';
714 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
715 if (sizeof($matches{1})){
716 $count = strlen($matches{1});
723 * This function looks for the next character within a string. It's
724 * really just a glorified "strpos", except it catches if failures
727 * @param $body The string to look for needle in.
728 * @param $offset Start looking from this position.
729 * @param $needle The character/string to look for.
730 * @return location of the next occurance of the needle, or
731 * strlen($body) if needle wasn't found.
733 function sq_findnxstr($body, $offset, $needle){
734 $me = 'sq_findnxstr';
735 $pos = strpos($body, $needle, $offset);
737 $pos = strlen($body);
743 * This function takes a PCRE-style regexp and tries to match it
746 * @param $body The string to look for needle in.
747 * @param $offset Start looking from here.
748 * @param $reg A PCRE-style regex to match.
749 * @return Returns a false if no matches found, or an array
750 * with the following members:
751 * - integer with the location of the match within $body
752 * - string with whatever content between offset and the match
753 * - string with whatever it is we matched
755 function sq_findnxreg($body, $offset, $reg){
756 $me = 'sq_findnxreg';
759 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
763 $retarr{0} = $offset +
strlen($matches{1});
764 $retarr{1} = $matches{1};
765 $retarr{2} = $matches{2};
771 * This function looks for the next tag.
773 * @param $body String where to look for the next tag.
774 * @param $offset Start looking from here.
775 * @return false if no more tags exist in the body, or
776 * an array with the following members:
777 * - string with the name of the tag
778 * - array with attributes and their values
779 * - integer with tag type (1, 2, or 3)
780 * - integer where the tag starts (starting "<")
781 * - integer where the tag ends (ending ">")
782 * first three members will be false, if the tag is invalid.
784 function sq_getnxtag($body, $offset){
786 if ($offset > strlen($body)){
789 $lt = sq_findnxstr($body, $offset, "<");
790 if ($lt == strlen($body)){
795 * blah blah <tag attribute="value">
798 $pos = sq_skipspace($body, $lt+
1);
799 if ($pos >= strlen($body)){
800 return Array(false, false, false, $lt, strlen($body));
803 * There are 3 kinds of tags:
804 * 1. Opening tag, e.g.:
806 * 2. Closing tag, e.g.:
808 * 3. XHTML-style content-less tag, e.g.:
812 switch (substr($body, $pos, 1)){
819 * A comment or an SGML declaration.
821 if (substr($body, $pos+
1, 2) == "--"){
822 $gt = strpos($body, "-->", $pos);
828 return Array(false, false, false, $lt, $gt);
830 $gt = sq_findnxstr($body, $pos, ">");
831 return Array(false, false, false, $lt, $gt);
836 * Assume tagtype 1 for now. If it's type 3, we'll switch values
846 * Look for next [\W-_], which will indicate the end of the tag name.
848 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
849 if ($regary == false){
850 return Array(false, false, false, $lt, strlen($body));
852 list($pos, $tagname, $match) = $regary;
853 $tagname = strtolower($tagname);
856 * $match can be either of these:
857 * '>' indicating the end of the tag entirely.
858 * '\s' indicating the end of the tag name.
859 * '/' indicating that this is type-3 xhtml tag.
861 * Whatever else we find there indicates an invalid tag.
866 * This is an xhtml-style tag with a closing / at the
867 * end, like so: <img src="blah"/>. Check if it's followed
868 * by the closing bracket. If not, then this tag is invalid
870 if (substr($body, $pos, 2) == "/>"){
874 $gt = sq_findnxstr($body, $pos, ">");
875 $retary = Array(false, false, false, $lt, $gt);
879 return Array($tagname, false, $tagtype, $lt, $pos);
883 * Check if it's whitespace
885 if (!preg_match('/\s/', $match)){
887 * This is an invalid tag! Look for the next closing ">".
889 $gt = sq_findnxstr($body, $offset, ">");
890 return Array(false, false, false, $lt, $gt);
896 * At this point we're here:
897 * <tagname attribute='blah'>
900 * At this point we loop in order to find all attributes.
906 while ($pos <= strlen($body)){
907 $pos = sq_skipspace($body, $pos);
908 if ($pos == strlen($body)){
912 return Array(false, false, false, $lt, $pos);
915 * See if we arrived at a ">" or "/>", which means that we reached
916 * the end of the tag.
919 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
923 $pos +
= strlen($matches{1});
924 if ($matches{2} == "/>"){
928 return Array($tagname, $attary, $tagtype, $lt, $pos);
932 * There are several types of attributes, with optional
933 * [:space:] between members.
935 * attrname[:space:]=[:space:]'CDATA'
937 * attrname[:space:]=[:space:]"CDATA"
939 * attr[:space:]=[:space:]CDATA
943 * We leave types 1 and 2 the same, type 3 we check for
944 * '"' and convert to """ if needed, then wrap in
945 * double quotes. Type 4 we convert into:
948 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
949 if ($regary == false){
951 * Looks like body ended before the end of tag.
953 return Array(false, false, false, $lt, strlen($body));
955 list($pos, $attname, $match) = $regary;
956 $attname = strtolower($attname);
958 * We arrived at the end of attribute name. Several things possible
960 * '>' means the end of the tag and this is attribute type 4
961 * '/' if followed by '>' means the same thing as above
962 * '\s' means a lot of things -- look what it's followed by.
963 * anything else means the attribute is invalid.
968 * This is an xhtml-style tag with a closing / at the
969 * end, like so: <img src="blah"/>. Check if it's followed
970 * by the closing bracket. If not, then this tag is invalid
972 if (substr($body, $pos, 2) == "/>"){
976 $gt = sq_findnxstr($body, $pos, ">");
977 $retary = Array(false, false, false, $lt, $gt);
981 $attary{$attname} = '"yes"';
982 return Array($tagname, $attary, $tagtype, $lt, $pos);
986 * Skip whitespace and see what we arrive at.
988 $pos = sq_skipspace($body, $pos);
989 $char = substr($body, $pos, 1);
991 * Two things are valid here:
992 * '=' means this is attribute type 1 2 or 3.
993 * \w means this was attribute type 4.
994 * anything else we ignore and re-loop. End of tag and
995 * invalid stuff will be caught by our checks at the beginning
1000 $pos = sq_skipspace($body, $pos);
1002 * Here are 3 possibilities:
1003 * "'" attribute type 1
1004 * '"' attribute type 2
1005 * everything else is the content of tag type 3
1007 $quot = substr($body, $pos, 1);
1009 $regary = sq_findnxreg($body, $pos+
1, "\'");
1010 if ($regary == false){
1011 return Array(false, false, false, $lt, strlen($body));
1013 list($pos, $attval, $match) = $regary;
1015 $attary{$attname} = "'" . $attval . "'";
1016 } else if ($quot == '"'){
1017 $regary = sq_findnxreg($body, $pos+
1, '\"');
1018 if ($regary == false){
1019 return Array(false, false, false, $lt, strlen($body));
1021 list($pos, $attval, $match) = $regary;
1023 $attary{$attname} = '"' . $attval . '"';
1026 * These are hateful. Look for \s, or >.
1028 $regary = sq_findnxreg($body, $pos, "[\s>]");
1029 if ($regary == false){
1030 return Array(false, false, false, $lt, strlen($body));
1032 list($pos, $attval, $match) = $regary;
1034 * If it's ">" it will be caught at the top.
1036 $attval = preg_replace("/\"/s", """, $attval);
1037 $attary{$attname} = '"' . $attval . '"';
1039 } else if (preg_match("|[\w/>]|", $char)) {
1041 * That was attribute type 4.
1043 $attary{$attname} = '"yes"';
1046 * An illegal character. Find next '>' and return.
1048 $gt = sq_findnxstr($body, $pos, ">");
1049 return Array(false, false, false, $lt, $gt);
1055 * The fact that we got here indicates that the tag end was never
1056 * found. Return invalid tag indication so it gets stripped.
1058 return Array(false, false, false, $lt, strlen($body));
1062 * This function checks attribute values for entity-encoded values
1063 * and returns them translated into 8-bit strings so we can run
1066 * @param $attvalue A string to run entity check against.
1067 * @return Translated value.
1069 function sq_deent($attvalue){
1072 * See if we have to run the checks first. All entities must start
1075 if (strpos($attvalue, "&") === false){
1079 * Check named entities first.
1081 $trans = get_html_translation_table(HTML_ENTITIES
);
1083 * Leave " in, as it can mess us up.
1085 $trans = array_flip($trans);
1086 unset($trans{"""});
1087 while (list($ent, $val) = each($trans)){
1088 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1091 * Now translate numbered entities from 1 to 255 if needed.
1093 if (strpos($attvalue, "#") !== false){
1094 $omit = Array(34, 39);
1095 for ($asc=1; $asc<256; $asc++
){
1096 if (!in_array($asc, $omit)){
1098 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1100 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1101 "$chr\\1", $attvalue);
1109 * This function runs various checks against the attributes.
1111 * @param $tagname String with the name of the tag.
1112 * @param $attary Array with all tag attributes.
1113 * @param $rm_attnames See description for sq_sanitize
1114 * @param $bad_attvals See description for sq_sanitize
1115 * @param $add_attr_to_tag See description for sq_sanitize
1116 * @param $message message object
1117 * @param $id message id
1118 * @return Array with modified attributes.
1120 function sq_fixatts($tagname,
1130 while (list($attname, $attvalue) = each($attary)){
1132 * See if this attribute should be removed.
1134 foreach ($rm_attnames as $matchtag=>$matchattrs){
1135 if (preg_match($matchtag, $tagname)){
1136 foreach ($matchattrs as $matchattr){
1137 if (preg_match($matchattr, $attname)){
1138 unset($attary{$attname});
1145 * Remove any entities.
1147 $attvalue = sq_deent($attvalue);
1150 * Now let's run checks on the attvalues.
1151 * I don't expect anyone to comprehend this. If you do,
1152 * get in touch with me so I can drive to where you live and
1153 * shake your hand personally. :)
1155 foreach ($bad_attvals as $matchtag=>$matchattrs){
1156 if (preg_match($matchtag, $tagname)){
1157 foreach ($matchattrs as $matchattr=>$valary){
1158 if (preg_match($matchattr, $attname)){
1160 * There are two arrays in valary.
1162 * Second one is replacements
1164 list($valmatch, $valrepl) = $valary;
1166 preg_replace($valmatch, $valrepl, $attvalue);
1167 if ($newvalue != $attvalue){
1168 $attary{$attname} = $newvalue;
1175 * Turn cid: urls into http-friendly ones.
1177 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1178 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1182 * See if we need to append any attributes to this tag.
1184 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1185 if (preg_match($matchtag, $tagname)){
1186 $attary = array_merge($attary, $addattary);
1193 * This function edits the style definition to make them friendly and
1194 * usable in squirrelmail.
1196 * @param $message the message object
1197 * @param $id the message id
1198 * @param $content a string with whatever is between <style> and </style>
1199 * @return a string with edited content.
1201 function sq_fixstyle($message, $id, $content){
1202 global $view_unsafe_images;
1203 $me = 'sq_fixstyle';
1205 * First look for general BODY style declaration, which would be
1207 * body {background: blah-blah}
1208 * and change it to .bodyclass so we can just assign it to a <div>
1210 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1211 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1213 * Fix url('blah') declarations.
1215 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1216 "url(\\1$secremoveimg\\2)", $content);
1218 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1221 if (!$view_unsafe_images){
1222 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1223 "url(\\1$secremoveimg\\2)", $content);
1227 * Fix urls that refer to cid:
1229 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1231 $cidurl = $matches{1};
1232 $httpurl = sq_cid2http($message, $id, $cidurl);
1233 $content = preg_replace("|url\($cidurl\)|si",
1234 "url($httpurl)", $content);
1238 * Fix stupid css declarations which lead to vulnerabilities
1241 $match = Array('/expression/si',
1244 $replace = Array('idiocy', 'idiocy', 'idiocy');
1245 $content = preg_replace($match, $replace, $content);
1250 * This function converts cid: url's into the ones that can be viewed in
1253 * @param $message the message object
1254 * @param $id the message id
1255 * @param $cidurl the cid: url.
1256 * @return a string with a http-friendly url
1258 function sq_cid2http($message, $id, $cidurl, $mailbox){
1260 * Get rid of quotes.
1262 $quotchar = substr($cidurl, 0, 1);
1263 $cidurl = str_replace($quotchar, "", $cidurl);
1264 $cidurl = substr(trim($cidurl), 4);
1265 $linkurl = find_ent_id($cidurl, $message);
1266 /* in case of non-save cid links $httpurl should be replaced by a sort of
1267 unsave link image */
1270 $httpurl = $quotchar . '../src/download.php?absolute_dl=true&' .
1271 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1272 '&ent_id=' . $linkurl . $quotchar;
1278 * This function changes the <body> tag into a <div> tag since we
1279 * can't really have a body-within-body.
1281 * @param $attary an array of attributes and values of <body>
1282 * @return a modified array of attributes to be set for <div>
1284 function sq_body2div($attary){
1285 $me = 'sq_body2div';
1286 $divattary = Array('class' => "'bodyclass'");
1287 $bgcolor = '#ffffff';
1290 if (is_array($attary) && sizeof($attary) > 0){
1291 foreach ($attary as $attname=>$attvalue){
1292 $quotchar = substr($attvalue, 0, 1);
1293 $attvalue = str_replace($quotchar, "", $attvalue);
1296 $styledef .= "background-image: url('$attvalue'); ";
1299 $styledef .= "background-color: $attvalue; ";
1302 $styledef .= "color: $attvalue; ";
1306 if (strlen($styledef) > 0){
1307 $divattary{"style"} = "\"$styledef\"";
1314 * This is the main function and the one you should actually be calling.
1315 * There are several variables you should be aware of an which need
1316 * special description.
1318 * Since the description is quite lengthy, see it here:
1319 * http://www.mricon.com/html/phpfilter.html
1321 * @param $body the string with HTML you wish to filter
1322 * @param $tag_list see description above
1323 * @param $rm_tags_with_content see description above
1324 * @param $self_closing_tags see description above
1325 * @param $force_tag_closing see description above
1326 * @param $rm_attnames see description above
1327 * @param $bad_attvals see description above
1328 * @param $add_attr_to_tag see description above
1329 * @param $message message object
1330 * @param $id message id
1331 * @return sanitized html safe to show on your pages.
1333 function sq_sanitize($body,
1335 $rm_tags_with_content,
1345 $me = 'sq_sanitize';
1347 * Normalize rm_tags and rm_tags_with_content.
1349 @array_walk
($rm_tags, 'sq_casenormalize');
1350 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1351 @array_walk
($self_closing_tags, 'sq_casenormalize');
1353 * See if tag_list is of tags to remove or tags to allow.
1354 * false means remove these tags
1355 * true means allow these tags
1357 $rm_tags = array_shift($tag_list);
1359 $open_tags = Array();
1360 $trusted = "<!-- begin sanitized html -->\n";
1361 $skip_content = false;
1363 * Take care of netscape's stupid javascript entities like
1366 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1368 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1369 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1370 $free_content = substr($body, $curpos, $lt-$curpos);
1372 * Take care of <style>
1374 if ($tagname == "style" && $tagtype == 2){
1376 * This is a closing </style>. Edit the
1377 * content before we apply it.
1379 $free_content = sq_fixstyle($message, $id, $free_content);
1381 if ($skip_content == false){
1382 $trusted .= $free_content;
1384 if ($tagname != FALSE){
1386 if ($skip_content == $tagname){
1388 * Got to the end of tag we needed to remove.
1391 $skip_content = false;
1393 if ($skip_content == false){
1394 if ($tagname == "body"){
1397 if (isset($open_tags{$tagname}) &&
1398 $open_tags{$tagname} > 0){
1399 $open_tags{$tagname}--;
1408 * $rm_tags_with_content
1410 if ($skip_content == false){
1412 * See if this is a self-closing type and change
1413 * tagtype appropriately.
1416 && in_array($tagname, $self_closing_tags)){
1420 * See if we should skip this tag and any content
1423 if ($tagtype == 1 &&
1424 in_array($tagname, $rm_tags_with_content)){
1425 $skip_content = $tagname;
1427 if (($rm_tags == false
1428 && in_array($tagname, $tag_list)) ||
1429 ($rm_tags == true &&
1430 !in_array($tagname, $tag_list))){
1434 if (isset($open_tags{$tagname})){
1435 $open_tags{$tagname}++
;
1437 $open_tags{$tagname}=1;
1441 * This is where we run other checks.
1443 if (is_array($attary) && sizeof($attary) > 0){
1444 $attary = sq_fixatts($tagname,
1455 * Convert body into div.
1457 if ($tagname == "body"){
1459 $attary = sq_body2div($attary, $message, $id);
1465 if ($tagname != false && $skip_content == false){
1466 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1471 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1472 if ($force_tag_closing == true){
1473 foreach ($open_tags as $tagname=>$opentimes){
1474 while ($opentimes > 0){
1475 $trusted .= '</' . $tagname . '>';
1481 $trusted .= "<!-- end sanitized html -->\n";
1486 * This is a wrapper function to call html sanitizing routines.
1488 * @param $body the body of the message
1489 * @param $id the id of the message
1490 * @return a string with html safe to display in the browser.
1492 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1493 global $attachment_common_show_images, $view_unsafe_images,
1496 * Don't display attached images in HTML mode.
1498 $attachment_common_show_images = false;
1511 $rm_tags_with_content = Array(
1518 $self_closing_tags = Array(
1525 $force_tag_closing = false;
1527 $rm_attnames = Array(
1537 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1538 $bad_attvals = Array(
1541 "/^src|background/i" =>
1544 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1545 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1546 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1547 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1550 "\\1$secremoveimg\\2",
1551 "\\1$secremoveimg\\2",
1552 "\\1$secremoveimg\\2",
1553 "\\1$secremoveimg\\2"
1556 "/^href|action/i" =>
1559 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1560 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1561 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1562 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1577 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1578 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1579 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1580 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1594 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1595 $view_unsafe_images = false;
1597 if (!$view_unsafe_images){
1599 * Remove any references to http/https if view_unsafe_images set
1602 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1603 '/^([\'\"])\s*https*:.*([\'\"])/si');
1604 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1605 "\\1$secremoveimg\\2");
1606 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1607 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1608 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1609 "url(\\1$secremoveimg\\2)");
1612 $add_attr_to_tag = Array(
1613 "/^a$/si" => Array('target'=>'"_new"')
1615 $trusted = sq_sanitize($body,
1617 $rm_tags_with_content,
1627 if (preg_match("|$secremoveimg|si", $trusted)){
1628 $has_unsafe_images = true;