9573299476485b30265996cdc405c229d91457e2
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 = $hookResults[1];
330 // If there are other types that shouldn't be formatted, add
333 if ($body_message->header
->type1
== 'html') {
334 if ( $show_html_default <> 1 ) {
335 $entity_conv = array(' ' => ' ',
338 $body = strip_tags( $body );
339 $body = strtr($body, $entity_conv);
341 translateText($body, $wrap_at,
342 $body_message->header
->getParameter('charset'));
344 $body = magicHTML( $body, $id, $message, $mailbox );
347 translateText($body, $wrap_at,
348 $body_message->header
->getParameter('charset'));
351 if ($has_unsafe_images) {
352 if ($view_unsafe_images) {
353 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&passed_ent_id=".$message->entity_id
."&mailbox=$urlmailbox&sort=$sort&startMessage=$startMessage&show_more=0\">". _("Hide Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
355 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$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";
363 function formatAttachments($message, $exclude_id, $mailbox, $id) {
364 global $where, $what, $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", '', $body);
495 } else if ($encoding == 'base64') {
496 $body = base64_decode($body);
499 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
500 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
501 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
504 // All other encodings are returned raw.
509 * This functions decode strings that is encoded according to
510 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
511 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
513 function decodeHeader ($string, $utfencode=true) {
514 global $languages, $squirrelmail_language;
515 if (is_array($string)) {
516 $string = implode("\n", $string);
519 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
520 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
521 $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
525 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
528 // Ignore white-space between consecutive encoded-words
529 if (strspn($res[2], " \t") != strlen($res[2])) {
533 if (ucfirst($res[4]) == 'B') {
534 $replace = base64_decode($res[5]);
536 $replace = str_replace('_', ' ', $res[5]);
537 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
539 /* Only encode into entities by default. Some places
540 don't need the encoding, like the compose form. */
542 $replace = charset_decode($res[3], $replace);
545 $string = $prefix . $replace . substr($string, strlen($res[0]));
546 $i = strlen($prefix) +
strlen($replace);
552 * Encode a string according to RFC 1522 for use in headers if it
553 * contains 8-bit characters or anything that looks like it should
556 function encodeHeader ($string) {
557 global $default_charset, $languages, $squirrelmail_language;
559 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
560 function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
561 return $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
564 // Encode only if the string contains 8-bit characters or =?
565 $j = strlen( $string );
566 $l = strstr($string, '=?'); // Must be encoded ?
568 for( $i=0; $i < $j; ++
$i) {
569 switch( $string{$i} ) {
583 $k = ord( $string{$i} );
585 $ret .= sprintf("=%02X", $k);
593 $string = "=?$default_charset?Q?$ret?=";
599 /* This function trys to locate the entity_id of a specific mime element */
601 function find_ent_id( $id, $message ) {
603 for ($i=0; $ret == '' && $i < count($message->entities
); $i++
) {
604 if ( $message->entities
[$i]->header
->type0
== 'multipart') {
605 $ret = find_ent_id( $id, $message->entities
[$i] );
607 if ( strcasecmp( $message->entities
[$i]->header
->id
, $id ) == 0 ) {
608 if (sq_check_save_extension($message->entities
[$i])) {
609 $ret = $message->entities
[$i]->entity_id
;
619 function sq_check_save_extension($message) {
620 $filename = $message->getFilename();
621 $ext = substr($filename, strrpos($filename,'.')+
1);
622 $save_extensions = array('jpg','jpeg','gif','png','bmp');
623 return (in_array($ext, $save_extensions));
628 ** HTMLFILTER ROUTINES
632 * This function returns the final tag out of the tag name, an array
633 * of attributes, and the type of the tag. This function is called by
634 * sq_sanitize internally.
636 * @param $tagname the name of the tag.
637 * @param $attary the array of attributes and their values
638 * @param $tagtype The type of the tag (see in comments).
639 * @return a string with the final tag representation.
641 function sq_tagprint($tagname, $attary, $tagtype){
646 $fulltag = '</' . $tagname . '>';
648 $fulltag = '<' . $tagname;
649 if (is_array($attary) && sizeof($attary)){
651 while (list($attname, $attvalue) = each($attary)){
652 array_push($atts, "$attname=$attvalue");
654 $fulltag .= ' ' . join(" ", $atts);
665 * A small helper function to use with array_walk. Modifies a by-ref
666 * value and makes it lowercase.
668 * @param $val a value passed by-ref.
669 * @return void since it modifies a by-ref value.
671 function sq_casenormalize(&$val){
672 $val = strtolower($val);
676 * This function skips any whitespace from the current position within
677 * a string and to the next non-whitespace value.
679 * @param $body the string
680 * @param $offset the offset within the string where we should start
681 * looking for the next non-whitespace character.
682 * @return the location within the $body where the next
683 * non-whitespace char is located.
685 function sq_skipspace($body, $offset){
686 $me = 'sq_skipspace';
687 preg_match("/^(\s*)/s", substr($body, $offset), $matches);
688 if (sizeof($matches{1})){
689 $count = strlen($matches{1});
696 * This function looks for the next character within a string. It's
697 * really just a glorified "strpos", except it catches if failures
700 * @param $body The string to look for needle in.
701 * @param $offset Start looking from this position.
702 * @param $needle The character/string to look for.
703 * @return location of the next occurance of the needle, or
704 * strlen($body) if needle wasn't found.
706 function sq_findnxstr($body, $offset, $needle){
707 $me = 'sq_findnxstr';
708 $pos = strpos($body, $needle, $offset);
710 $pos = strlen($body);
716 * This function takes a PCRE-style regexp and tries to match it
719 * @param $body The string to look for needle in.
720 * @param $offset Start looking from here.
721 * @param $reg A PCRE-style regex to match.
722 * @return Returns a false if no matches found, or an array
723 * with the following members:
724 * - integer with the location of the match within $body
725 * - string with whatever content between offset and the match
726 * - string with whatever it is we matched
728 function sq_findnxreg($body, $offset, $reg){
729 $me = 'sq_findnxreg';
732 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
736 $retarr{0} = $offset +
strlen($matches{1});
737 $retarr{1} = $matches{1};
738 $retarr{2} = $matches{2};
744 * This function looks for the next tag.
746 * @param $body String where to look for the next tag.
747 * @param $offset Start looking from here.
748 * @return false if no more tags exist in the body, or
749 * an array with the following members:
750 * - string with the name of the tag
751 * - array with attributes and their values
752 * - integer with tag type (1, 2, or 3)
753 * - integer where the tag starts (starting "<")
754 * - integer where the tag ends (ending ">")
755 * first three members will be false, if the tag is invalid.
757 function sq_getnxtag($body, $offset){
759 if ($offset > strlen($body)){
762 $lt = sq_findnxstr($body, $offset, "<");
763 if ($lt == strlen($body)){
768 * blah blah <tag attribute="value">
771 $pos = sq_skipspace($body, $lt+
1);
772 if ($pos >= strlen($body)){
773 return Array(false, false, false, $lt, strlen($body));
776 * There are 3 kinds of tags:
777 * 1. Opening tag, e.g.:
779 * 2. Closing tag, e.g.:
781 * 3. XHTML-style content-less tag, e.g.:
785 switch (substr($body, $pos, 1)){
792 * A comment or an SGML declaration.
794 if (substr($body, $pos+
1, 2) == "--"){
795 $gt = strpos($body, "-->", $pos);
801 return Array(false, false, false, $lt, $gt);
803 $gt = sq_findnxstr($body, $pos, ">");
804 return Array(false, false, false, $lt, $gt);
809 * Assume tagtype 1 for now. If it's type 3, we'll switch values
819 * Look for next [\W-_], which will indicate the end of the tag name.
821 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
822 if ($regary == false){
823 return Array(false, false, false, $lt, strlen($body));
825 list($pos, $tagname, $match) = $regary;
826 $tagname = strtolower($tagname);
829 * $match can be either of these:
830 * '>' indicating the end of the tag entirely.
831 * '\s' indicating the end of the tag name.
832 * '/' indicating that this is type-3 xhtml tag.
834 * Whatever else we find there indicates an invalid tag.
839 * This is an xhtml-style tag with a closing / at the
840 * end, like so: <img src="blah"/>. Check if it's followed
841 * by the closing bracket. If not, then this tag is invalid
843 if (substr($body, $pos, 2) == "/>"){
847 $gt = sq_findnxstr($body, $pos, ">");
848 $retary = Array(false, false, false, $lt, $gt);
852 return Array($tagname, false, $tagtype, $lt, $pos);
856 * Check if it's whitespace
858 if (preg_match("/\s/", $match)){
861 * This is an invalid tag! Look for the next closing ">".
863 $gt = sq_findnxstr($body, $offset, ">");
864 return Array(false, false, false, $lt, $gt);
869 * At this point we're here:
870 * <tagname attribute='blah'>
873 * At this point we loop in order to find all attributes.
879 while ($pos <= strlen($body)){
880 $pos = sq_skipspace($body, $pos);
881 if ($pos == strlen($body)){
885 return Array(false, false, false, $lt, $pos);
888 * See if we arrived at a ">" or "/>", which means that we reached
889 * the end of the tag.
892 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
896 $pos +
= strlen($matches{1});
897 if ($matches{2} == "/>"){
901 return Array($tagname, $attary, $tagtype, $lt, $pos);
905 * There are several types of attributes, with optional
906 * [:space:] between members.
908 * attrname[:space:]=[:space:]'CDATA'
910 * attrname[:space:]=[:space:]"CDATA"
912 * attr[:space:]=[:space:]CDATA
916 * We leave types 1 and 2 the same, type 3 we check for
917 * '"' and convert to """ if needed, then wrap in
918 * double quotes. Type 4 we convert into:
921 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
922 if ($regary == false){
924 * Looks like body ended before the end of tag.
926 return Array(false, false, false, $lt, strlen($body));
928 list($pos, $attname, $match) = $regary;
929 $attname = strtolower($attname);
931 * We arrived at the end of attribute name. Several things possible
933 * '>' means the end of the tag and this is attribute type 4
934 * '/' if followed by '>' means the same thing as above
935 * '\s' means a lot of things -- look what it's followed by.
936 * anything else means the attribute is invalid.
941 * This is an xhtml-style tag with a closing / at the
942 * end, like so: <img src="blah"/>. Check if it's followed
943 * by the closing bracket. If not, then this tag is invalid
945 if (substr($body, $pos, 2) == "/>"){
949 $gt = sq_findnxstr($body, $pos, ">");
950 $retary = Array(false, false, false, $lt, $gt);
954 $attary{$attname} = '"yes"';
955 return Array($tagname, $attary, $tagtype, $lt, $pos);
959 * Skip whitespace and see what we arrive at.
961 $pos = sq_skipspace($body, $pos);
962 $char = substr($body, $pos, 1);
964 * Two things are valid here:
965 * '=' means this is attribute type 1 2 or 3.
966 * \w means this was attribute type 4.
967 * anything else we ignore and re-loop. End of tag and
968 * invalid stuff will be caught by our checks at the beginning
973 $pos = sq_skipspace($body, $pos);
975 * Here are 3 possibilities:
976 * "'" attribute type 1
977 * '"' attribute type 2
978 * everything else is the content of tag type 3
980 $quot = substr($body, $pos, 1);
982 $regary = sq_findnxreg($body, $pos+
1, "\'");
983 if ($regary == false){
984 return Array(false, false, false, $lt, strlen($body));
986 list($pos, $attval, $match) = $regary;
988 $attary{$attname} = "'" . $attval . "'";
989 } else if ($quot == '"'){
990 $regary = sq_findnxreg($body, $pos+
1, '\"');
991 if ($regary == false){
992 return Array(false, false, false, $lt, strlen($body));
994 list($pos, $attval, $match) = $regary;
996 $attary{$attname} = '"' . $attval . '"';
999 * These are hateful. Look for \s, or >.
1001 $regary = sq_findnxreg($body, $pos, "[\s>]");
1002 if ($regary == false){
1003 return Array(false, false, false, $lt, strlen($body));
1005 list($pos, $attval, $match) = $regary;
1007 * If it's ">" it will be caught at the top.
1009 $attval = preg_replace("/\"/s", """, $attval);
1010 $attary{$attname} = '"' . $attval . '"';
1012 } else if (preg_match("|[\w/>]|", $char)) {
1014 * That was attribute type 4.
1016 $attary{$attname} = '"yes"';
1019 * An illegal character. Find next '>' and return.
1021 $gt = sq_findnxstr($body, $pos, ">");
1022 return Array(false, false, false, $lt, $gt);
1027 * The fact that we got here indicates that the tag end was never
1028 * found. Return invalid tag indication so it gets stripped.
1030 return Array(false, false, false, $lt, strlen($body));
1034 * This function checks attribute values for entity-encoded values
1035 * and returns them translated into 8-bit strings so we can run
1038 * @param $attvalue A string to run entity check against.
1039 * @return Translated value.
1041 function sq_deent($attvalue){
1044 * See if we have to run the checks first. All entities must start
1047 if (strpos($attvalue, "&") === false){
1051 * Check named entities first.
1053 $trans = get_html_translation_table(HTML_ENTITIES
);
1055 * Leave " in, as it can mess us up.
1057 $trans = array_flip($trans);
1058 unset($trans{"""});
1059 while (list($ent, $val) = each($trans)){
1060 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1063 * Now translate numbered entities from 1 to 255 if needed.
1065 if (strpos($attvalue, "#") !== false){
1066 $omit = Array(34, 39);
1067 for ($asc=1; $asc<256; $asc++
){
1068 if (!in_array($asc, $omit)){
1070 $attvalue = preg_replace("/\�*$asc;*(\D)/si", "$chr\\1",
1072 $attvalue = preg_replace("/\�*".dechex($asc).";*(\W)/si",
1073 "$chr\\1", $attvalue);
1081 * This function runs various checks against the attributes.
1083 * @param $tagname String with the name of the tag.
1084 * @param $attary Array with all tag attributes.
1085 * @param $rm_attnames See description for sq_sanitize
1086 * @param $bad_attvals See description for sq_sanitize
1087 * @param $add_attr_to_tag See description for sq_sanitize
1088 * @param $message message object
1089 * @param $id message id
1090 * @return Array with modified attributes.
1092 function sq_fixatts($tagname,
1102 while (list($attname, $attvalue) = each($attary)){
1104 * See if this attribute should be removed.
1106 foreach ($rm_attnames as $matchtag=>$matchattrs){
1107 if (preg_match($matchtag, $tagname)){
1108 foreach ($matchattrs as $matchattr){
1109 if (preg_match($matchattr, $attname)){
1110 unset($attary{$attname});
1117 * Remove any entities.
1119 $attvalue = sq_deent($attvalue);
1122 * Now let's run checks on the attvalues.
1123 * I don't expect anyone to comprehend this. If you do,
1124 * get in touch with me so I can drive to where you live and
1125 * shake your hand personally. :)
1127 foreach ($bad_attvals as $matchtag=>$matchattrs){
1128 if (preg_match($matchtag, $tagname)){
1129 foreach ($matchattrs as $matchattr=>$valary){
1130 if (preg_match($matchattr, $attname)){
1132 * There are two arrays in valary.
1134 * Second one is replacements
1136 list($valmatch, $valrepl) = $valary;
1138 preg_replace($valmatch, $valrepl, $attvalue);
1139 if ($newvalue != $attvalue){
1140 $attary{$attname} = $newvalue;
1147 * Turn cid: urls into http-friendly ones.
1149 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1150 $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1154 * See if we need to append any attributes to this tag.
1156 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1157 if (preg_match($matchtag, $tagname)){
1158 $attary = array_merge($attary, $addattary);
1165 * This function edits the style definition to make them friendly and
1166 * usable in squirrelmail.
1168 * @param $message the message object
1169 * @param $id the message id
1170 * @param $content a string with whatever is between <style> and </style>
1171 * @return a string with edited content.
1173 function sq_fixstyle($message, $id, $content){
1174 global $view_unsafe_images;
1175 $me = 'sq_fixstyle';
1177 * First look for general BODY style declaration, which would be
1179 * body {background: blah-blah}
1180 * and change it to .bodyclass so we can just assign it to a <div>
1182 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1183 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1185 * Fix url('blah') declarations.
1187 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1188 "url(\\1$secremoveimg\\2)", $content);
1190 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1193 if (!$view_unsafe_images){
1194 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1195 "url(\\1$secremoveimg\\2)", $content);
1199 * Fix urls that refer to cid:
1201 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1203 $cidurl = $matches{1};
1204 $httpurl = sq_cid2http($message, $id, $cidurl);
1205 $content = preg_replace("|url\($cidurl\)|si",
1206 "url($httpurl)", $content);
1210 * Fix stupid css declarations which lead to vulnerabilities
1213 $match = Array('/expression/si',
1216 $replace = Array('idiocy', 'idiocy', 'idiocy');
1217 $content = preg_replace($match, $replace, $content);
1222 * This function converts cid: url's into the ones that can be viewed in
1225 * @param $message the message object
1226 * @param $id the message id
1227 * @param $cidurl the cid: url.
1228 * @return a string with a http-friendly url
1230 function sq_cid2http($message, $id, $cidurl, $mailbox){
1232 * Get rid of quotes.
1234 $quotchar = substr($cidurl, 0, 1);
1235 $cidurl = str_replace($quotchar, "", $cidurl);
1236 $cidurl = substr(trim($cidurl), 4);
1237 $linkurl = find_ent_id($cidurl, $message);
1238 /* in case of non-save cid links $httpurl should be replaced by a sort of
1239 unsave link image */
1242 $httpurl = $quotchar . "../src/download.php?absolute_dl=true&" .
1243 "passed_id=$id&mailbox=" . urlencode($mailbox) .
1244 "&ent_id=" . $linkurl . $quotchar;
1250 * This function changes the <body> tag into a <div> tag since we
1251 * can't really have a body-within-body.
1253 * @param $attary an array of attributes and values of <body>
1254 * @return a modified array of attributes to be set for <div>
1256 function sq_body2div($attary){
1257 $me = 'sq_body2div';
1258 $divattary = Array( 'class' => "'bodyclass'" );
1259 $bgcolor = '#ffffff';
1262 if (is_array($attary) && sizeof($attary) > 0){
1263 foreach ($attary as $attname=>$attvalue){
1264 $quotchar = substr($attvalue, 0, 1);
1265 $attvalue = str_replace($quotchar, "", $attvalue);
1268 $styledef .= "background-image: url('$attvalue'); ";
1271 $styledef .= "background-color: $attvalue; ";
1274 $styledef .= "color: $attvalue; ";
1277 if (strlen($styledef) > 0){
1278 $divattary{"style"} = "\"$styledef\"";
1285 * This is the main function and the one you should actually be calling.
1286 * There are several variables you should be aware of an which need
1287 * special description.
1289 * Since the description is quite lengthy, see it here:
1290 * http://www.mricon.com/html/phpfilter.html
1292 * @param $body the string with HTML you wish to filter
1293 * @param $tag_list see description above
1294 * @param $rm_tags_with_content see description above
1295 * @param $self_closing_tags see description above
1296 * @param $force_tag_closing see description above
1297 * @param $rm_attnames see description above
1298 * @param $bad_attvals see description above
1299 * @param $add_attr_to_tag see description above
1300 * @param $message message object
1301 * @param $id message id
1302 * @return sanitized html safe to show on your pages.
1304 function sq_sanitize($body,
1306 $rm_tags_with_content,
1316 $me = 'sq_sanitize';
1318 * Normalize rm_tags and rm_tags_with_content.
1320 @array_walk
($rm_tags, 'sq_casenormalize');
1321 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
1322 @array_walk
($self_closing_tags, 'sq_casenormalize');
1324 * See if tag_list is of tags to remove or tags to allow.
1325 * false means remove these tags
1326 * true means allow these tags
1328 $rm_tags = array_shift($tag_list);
1330 $open_tags = Array();
1331 $trusted = "<!-- begin sanitized html -->\n";
1332 $skip_content = false;
1334 * Take care of netscape's stupid javascript entities like
1337 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
1339 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1340 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1341 $free_content = substr($body, $curpos, $lt-$curpos);
1343 * Take care of <style>
1345 if ($tagname == "style" && $tagtype == 2){
1347 * This is a closing </style>. Edit the
1348 * content before we apply it.
1350 $free_content = sq_fixstyle($message, $id, $free_content);
1352 if ($skip_content == false){
1353 $trusted .= $free_content;
1356 if ($tagname != FALSE){
1358 if ($skip_content == $tagname){
1360 * Got to the end of tag we needed to remove.
1363 $skip_content = false;
1365 if ($skip_content == false){
1366 if ($tagname == "body"){
1369 if (isset($open_tags{$tagname}) &&
1370 $open_tags{$tagname} > 0){
1371 $open_tags{$tagname}--;
1381 * $rm_tags_with_content
1383 if ($skip_content == false){
1385 * See if this is a self-closing type and change
1386 * tagtype appropriately.
1389 && in_array($tagname, $self_closing_tags)){
1393 * See if we should skip this tag and any content
1396 if ($tagtype == 1 &&
1397 in_array($tagname, $rm_tags_with_content)){
1398 $skip_content = $tagname;
1400 if (($rm_tags == false
1401 && in_array($tagname, $tag_list)) ||
1402 ($rm_tags == true &&
1403 !in_array($tagname, $tag_list))){
1407 if (isset($open_tags{$tagname})){
1408 $open_tags{$tagname}++
;
1410 $open_tags{$tagname}=1;
1414 * This is where we run other checks.
1416 if (is_array($attary) && sizeof($attary) > 0){
1417 $attary = sq_fixatts($tagname,
1428 * Convert body into div.
1430 if ($tagname == "body"){
1432 $attary = sq_body2div($attary, $message, $id);
1439 if ($tagname != false && $skip_content == false){
1440 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1446 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1447 if ($force_tag_closing == true){
1448 foreach ($open_tags as $tagname=>$opentimes){
1449 while ($opentimes > 0){
1450 $trusted .= '</' . $tagname . '>';
1456 $trusted .= "<!-- end sanitized html -->\n";
1461 * This is a wrapper function to call html sanitizing routines.
1463 * @param $body the body of the message
1464 * @param $id the id of the message
1465 * @return a string with html safe to display in the browser.
1467 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1468 global $attachment_common_show_images, $view_unsafe_images,
1471 * Don't display attached images in HTML mode.
1473 $attachment_common_show_images = false;
1484 $rm_tags_with_content = Array(
1491 $self_closing_tags = Array(
1498 $force_tag_closing = false;
1500 $rm_attnames = Array(
1510 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1511 $bad_attvals = Array(
1514 "/^src|background/i" =>
1517 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1518 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1519 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1520 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1523 "\\1$secremoveimg\\2",
1524 "\\1$secremoveimg\\2",
1525 "\\1$secremoveimg\\2",
1526 "\\1$secremoveimg\\2"
1529 "/^href|action/i" =>
1532 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1533 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1534 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1535 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1550 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1551 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1552 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1553 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1567 if (!$view_unsafe_images){
1569 * Remove any references to http/https if view_unsafe_images set
1572 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1573 '/^([\'\"])\s*https*:.*([\'\"])/si');
1574 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1575 "\\1$secremoveimg\\2");
1576 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1577 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1578 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1579 "url(\\1$secremoveimg\\2)");
1582 $add_attr_to_tag = Array(
1583 "/^a$/si" => Array('target'=>'"_new"')
1585 $trusted = sq_sanitize($body,
1587 $rm_tags_with_content,
1597 if (preg_match("|$secremoveimg|si", $trusted)){
1598 $has_unsafe_images = true;