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 /** Setting up the objects that have the structure for the message **/
20 /** msg_header contains generic variables for values that **/
21 /** could be in a header. **/
23 var $type0 = '', $type1 = '', $boundary = '', $charset = '';
24 var $encoding = '', $size = 0, $to = array(), $from = '', $date = '';
25 var $cc = array(), $bcc = array(), $reply_to = '', $subject = '';
26 var $id = 0, $mailbox = '', $description = '', $filename = '';
27 var $entity_id = 0, $message_id = 0, $name = '';
28 // var $priority = "";
32 /** message is the object that contains messages. It is a recursive
33 object in that through the $entities variable, it can contain
34 more objects of type message. See documentation in mime.txt for
35 a better description of how this works.
38 var $entities = array();
40 function addEntity ($msg) {
41 $this->entities
[] = $msg;
45 /* --------------------------------------------------------------------------------- */
47 /* --------------------------------------------------------------------------------- */
49 /* This function gets the structure of a message and stores it in the "message" class.
50 * It will return this object for use with all relevant header information and
51 * fully parsed into the standard "message" object format.
53 function mime_structure ($imap_stream, $header) {
55 sqimap_messages_flag ($imap_stream, $header->id
, $header->id
, 'Seen');
56 $ssid = sqimap_session_id();
57 $lsid = strlen( $ssid );
59 fputs ($imap_stream, "$ssid FETCH $id BODYSTRUCTURE\r\n");
61 // This should use sqimap_read_data instead of reading it itself
63 $read = fgets ($imap_stream, 10000);
65 while ( substr($read, 0, $lsid) <> $ssid &&
66 !feof( $imap_stream ) ) {
67 $bodystructure .= $read;
68 $read = fgets ($imap_stream, 10000);
70 $read = $bodystructure;
72 // isolate the body structure and remove beginning and end parenthesis
73 $read = trim(substr ($read, strpos(strtolower($read), 'bodystructure') +
13));
74 $read = trim(substr ($read, 0, -1));
75 $end = mime_match_parenthesis(0, $read);
76 while ($end == strlen($read)-1) {
77 $read = trim(substr ($read, 0, -1));
78 $read = trim(substr ($read, 1));
79 $end = mime_match_parenthesis(0, $read);
82 $msg = mime_parse_structure ($read, 0);
83 $msg->header
= $header;
88 /* this starts the parsing of a particular structure. It is called recursively,
89 * so it can be passed different structures. It returns an object of type
91 * First, it checks to see if it is a multipart message. If it is, then it
92 * handles that as it sees is necessary. If it is just a regular entity,
93 * then it parses it and adds the necessary header information (by calling out
94 * to mime_get_elements()
96 function mime_parse_structure ($structure, $ent_id) {
99 if ($structure{0} == '(') {
100 $ent_id = mime_new_element_level($ent_id);
104 $end = mime_match_parenthesis ($start, $structure);
106 $element = substr($structure, $start+
1, ($end - $start)-1);
107 $ent_id = mime_increment_id ($ent_id);
108 $newmsg = mime_parse_structure ($element, $ent_id);
109 $msg->addEntity ($newmsg);
110 } while ($structure{$end+
1} == '(');
112 // parse the elements
113 $msg = mime_get_element ($structure, $msg, $ent_id);
118 /* Increments the element ID. An element id can look like any of
119 * the following: 1, 1.2, 4.3.2.4.1, etc. This function increments
120 * the last number of the element id, changing 1.2 to 1.3.
122 function mime_increment_id ($id) {
124 if (strpos($id, ".")) {
125 $first = substr($id, 0, strrpos($id, "."));
126 $last = substr($id, strrpos($id, ".")+
1);
128 $new = $first . "." .$last;
137 * See comment for mime_increment_id().
138 * This adds another level on to the entity_id changing 1.3 to 1.3.0
139 * NOTE: 1.3.0 is not a valid element ID. It MUST be incremented
140 * before it can be used. I left it this way so as not to have
141 * to make a special case if it is the first entity_id. It
142 * always increments it, and that works fine.
144 function mime_new_element_level ($id) {
155 function mime_get_element (&$structure, $msg, $ent_id) {
158 $msg->header
= new msg_header();
159 $msg->header
->entity_id
= $ent_id;
160 $properties = array();
162 while (strlen($structure) > 0) {
163 $structure = trim($structure);
164 $char = $structure{0};
166 if (strtolower(substr($structure, 0, 3)) == 'nil') {
168 $structure = substr($structure, 3);
169 } else if ($char == '"') {
170 // loop through until we find the matching quote, and return that as a string
173 while ( ($char = $structure{$pos} ) <> '"' && $pos < strlen($structure)) {
177 $structure = substr($structure, strlen($text) +
2);
178 } else if ($char == '(') {
180 $end = mime_match_parenthesis (0, $structure);
181 $sub = substr($structure, 1, $end-1);
182 $properties = mime_get_props($properties, $sub);
183 $structure = substr($structure, strlen($sub) +
2);
185 // loop through until we find a space or an end parenthesis
187 $char = $structure{$pos};
189 while ($char != ' ' && $char != ')' && $pos < strlen($structure)) {
192 $char = $structure{$pos};
194 $structure = substr($structure, strlen($text));
197 // This is where all the text parts get put into the header
200 $msg->header
->type0
= strtolower($text);
203 $msg->header
->type1
= strtolower($text);
206 // Invisimail enclose images with <>
207 $msg->header
->id
= str_replace( '<', '', str_replace( '>', '', $text ) );
210 $msg->header
->description
= $text;
213 $msg->header
->encoding
= strtolower($text);
216 $msg->header
->size
= $text;
219 if ($msg->header
->type0
== 'text' && $elem_num == 8) {
220 // This is a plain text message, so lets get the number of lines
222 $msg->header
->num_lines
= $text;
224 } else if ($msg->header
->type0
== 'message' && $msg->header
->type1
== 'rfc822' && $elem_num == 8) {
225 // This is an encapsulated message, so lets start all over again and
226 // parse this message adding it on to the existing one.
227 $structure = trim($structure);
228 if ( $structure{0} == '(' ) {
229 $e = mime_match_parenthesis (0, $structure);
230 $structure = substr($structure, 0, $e);
231 $structure = substr($structure, 1);
232 $m = mime_parse_structure($structure, $msg->header
->entity_id
);
234 // the following conditional is there to correct a bug that wasn't
235 // incrementing the entity IDs correctly because of the special case
236 // that message/rfc822 is. This fixes it fine.
237 if (substr($structure, 1, 1) != '(')
238 $m->header
->entity_id
= mime_increment_id(mime_new_element_level($ent_id));
240 // Now we'll go through and reformat the results.
242 for ($i=0; $i < count($m->entities
); $i++
) {
243 $msg->addEntity($m->entities
[$i]);
256 // loop through the additional properties and put those in the various headers
257 if ($msg->header
->type0
!= 'message') {
258 for ($i=0; $i < count($properties); $i++
) {
259 $msg->header
->{$properties[$i]['name']} = $properties[$i]['value'];
267 * I did most of the MIME stuff yesterday (June 20, 2000), but I couldn't
268 * figure out how to do this part, so I decided to go to bed. I woke up
269 * in the morning and had a flash of insight. I went to the white-board
270 * and scribbled it out, then spent a bit programming it, and this is the
271 * result. Nothing complicated, but I think my brain was fried yesterday.
272 * Funny how that happens some times.
274 * This gets properties in a nested parenthesisized list. For example,
275 * this would get passed something like: ("attachment" ("filename" "luke.tar.gz"))
276 * This returns an array called $props with all paired up properties.
277 * It ignores the "attachment" for now, maybe that should change later
278 * down the road. In this case, what is returned is:
279 * $props[0]["name"] = "filename";
280 * $props[0]["value"] = "luke.tar.gz";
282 function mime_get_props ($props, $structure) {
284 while (strlen($structure) > 0) {
285 $structure = trim($structure);
286 $char = $structure{0};
291 while ( ( $char = $structure{$pos} ) != '"' &&
292 $pos < strlen($structure)) {
296 $structure = trim(substr($structure, strlen($tmp) +
2));
297 $char = $structure{0};
302 while ( ( $char = $structure{$pos} ) != '"' &&
303 $pos < strlen($structure) ) {
307 $structure = trim(substr($structure, strlen($tmp) +
2));
310 $props[$k]['name'] = strtolower($tmp);
311 $props[$k]['value'] = $value;
312 } else if ($char == '(') {
313 $end = mime_match_parenthesis (0, $structure);
314 $sub = substr($structure, 1, $end-1);
317 $props = mime_get_props($props, $sub);
318 $structure = substr($structure, strlen($sub) +
2);
321 } else if ($char == '(') {
322 $end = mime_match_parenthesis (0, $structure);
323 $sub = substr($structure, 1, $end-1);
324 $props = mime_get_props($props, $sub);
325 $structure = substr($structure, strlen($sub) +
2);
334 * Matches parenthesis. It will return the position of the matching
335 * parenthesis in $structure. For instance, if $structure was:
336 * ("text" "plain" ("val1name", "1") nil ... )
338 * then this would return 42 to match up those two.
340 function mime_match_parenthesis ($pos, $structure) {
342 $j = strlen( $structure );
344 // ignore all extra characters
345 // If inside of a string, skip string -- Boundary IDs and other
346 // things can have ) in them.
347 if ( $structure{$pos} != '(' ) {
351 while ( $pos < $j ) {
353 if ($structure{$pos} == ')') {
355 } elseif ($structure{$pos} == '"') {
357 while ( $structure{$pos} != '"' &&
359 if (substr($structure, $pos, 2) == '\\"') {
361 } elseif (substr($structure, $pos, 2) == '\\\\') {
366 } elseif ( $structure{$pos} == '(' ) {
367 $pos = mime_match_parenthesis ($pos, $structure);
370 echo _("Error decoding mime structure. Report this as a bug!") . '<br>';
374 function mime_fetch_body ($imap_stream, $id, $ent_id ) {
375 // do a bit of error correction. If we couldn't find the entity id, just guess
376 // that it is the first one. That is usually the case anyway.
379 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id]", true, $response, $message);
380 $topline = array_shift($data);
381 while (! ereg('\\* [0-9]+ FETCH ', $topline) && $data)
382 $topline = array_shift($data);
383 $wholemessage = implode('', $data);
384 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
385 $ret = substr( $wholemessage, 0, $regs[1] );
387 There is some information in the content info header that could be important
388 in order to parse html messages. Let's get them here.
390 if ( $ret{0} == '<' ) {
391 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message);
394 foreach( $data as $d ) {
395 if ( substr( $d, 0, 13 ) == 'Content-Base:' ) {
400 ( !isNoSep( $d{$i} ) ||
$d{$i} == '"' ) )
403 if ( isNoSep( $d{$i} ) )
408 } elseif ( $k == 1 && !isnosep( $d{0} ) ) {
409 $base .= substr( $d, 1 );
415 $ret = "<base href=\"$base\">" . $ret;
417 } else if (ereg('"([^"]*)"', $topline, $regs)) {
420 global $where, $what, $mailbox, $passed_id, $startMessage;
421 $par = "mailbox=".urlencode($mailbox)."&passed_id=$passed_id";
422 if (isset($where) && isset($what)) {
423 $par .= "&where=".urlencode($where)."&what=".urlencode($what);
425 $par .= "&startMessage=$startMessage&show_more=0";
427 $par .= '&response='.urlencode($response).'&message='.urlencode($message).
428 '&topline='.urlencode($topline);
430 echo '<b><font color=$color[2]>' .
431 _("Body retrieval error. The reason for this is most probably that the message is malformed. Please help us making future versions better by submitting this message to the developers knowledgebase!") .
432 "<A HREF=\"../src/retrievalerror.php?$par\">Submit message</A><BR>" .
433 '<tt>' . _("Response:") . "$response<BR>" .
434 _("Message:") . " $message<BR>" .
435 _("FETCH line:") . " $topline<BR></tt></font></b>";
437 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message);
439 $wholemessage = implode('', $data);
441 $ret = "---------------\n$wholemessage";
447 function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) {
448 // do a bit of error correction. If we couldn't find the entity id, just guess
449 // that it is the first one. That is usually the case anyway.
453 $sid = sqimap_session_id();
454 // Don't kill the connection if the browser is over a dialup
455 // and it would take over 30 seconds to download it.
457 // don´t call set_time_limit in safe mode.
458 if (!ini_get("safe_mode")) {
462 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
465 $read = fgets ($imap_stream,4096);
466 // This could be bad -- if the section has sqimap_session_id() . ' OK'
467 // or similar, it will kill the download.
468 while (!ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
469 if (trim($read) == ')==') {
471 $read = fgets ($imap_stream,4096);
472 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
475 echo decodeBody($read1, $encoding) .
476 decodeBody($read, $encoding);
479 echo decodeBody($read, $encoding);
481 $read = fgets ($imap_stream,4096);
486 /* -[ END MIME DECODING ]----------------------------------------------------------- */
490 /* This is the first function called. It decides if this is a multipart
491 message or if it should be handled as a single entity
493 function decodeMime ($imap_stream, &$header) {
494 global $username, $key, $imapServerAddress, $imapPort;
495 return mime_structure ($imap_stream, $header);
498 // This is here for debugging purposese. It will print out a list
499 // of all the entity IDs that are in the $message object.
501 function listEntities ($message) {
503 if ($message->header->entity_id)
504 echo "<tt>" . $message->header->entity_id . ' : ' . $message->header->type0 . '/' . $message->header->type1 . '<br>';
505 for ($i = 0; $message->entities[$i]; $i++) {
506 $msg = listEntities($message->entities[$i], $ent_id);
514 /* returns a $message object for a particular entity id */
515 function getEntity ($message, $ent_id) {
517 if ($message->header
->entity_id
== $ent_id && strlen($ent_id) == strlen($message->header
->entity_id
)) {
520 for ($i = 0; isset($message->entities
[$i]); $i++
) {
521 $msg = getEntity ($message->entities
[$i], $ent_id);
531 * figures out what entity to display and returns the $message object
534 function findDisplayEntity ($message, $textOnly = 1) {
535 global $show_html_default;
540 if ( $message->header
->type0
== 'multipart' &&
541 ( $message->header
->type1
== 'alternative' ||
542 $message->header
->type1
== 'related' ) &&
543 $show_html_default && ! $textOnly ) {
544 $entity = findDisplayEntityHTML($message);
547 // Show text/plain or text/html -- the first one we find.
549 $message->header
->type0
== 'text' &&
550 ( $message->header
->type1
== 'plain' ||
551 $message->header
->type1
== 'html' ) &&
552 isset($message->header
->entity_id
) ) {
553 $entity = $message->header
->entity_id
;
557 while ($entity == 0 && isset($message->entities
[$i]) ) {
558 $entity = findDisplayEntity($message->entities
[$i], $textOnly);
566 /* Shows the HTML version */
567 function findDisplayEntityHTML ($message) {
569 if ( $message->header
->type0
== 'text' &&
570 $message->header
->type1
== 'html' &&
571 isset($message->header
->entity_id
)) {
572 return $message->header
->entity_id
;
574 for ($i = 0; isset($message->entities
[$i]); $i ++
) {
575 $entity = findDisplayEntityHTML($message->entities
[$i]);
584 /* This returns a parsed string called $body. That string can then
585 be displayed as the actual message in the HTML. It contains
586 everything needed, including HTML Tags, Attachments at the
589 function formatBody($imap_stream, $message, $color, $wrap_at) {
590 // this if statement checks for the entity to show as the
591 // primary message. To add more of them, just put them in the
592 // order that is their priority.
593 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
596 $id = $message->header
->id
;
597 $urlmailbox = urlencode($message->header
->mailbox
);
599 // Get the right entity and redefine message to be this entity
600 // Pass the 0 to mean that we want the 'best' viewable one
601 $ent_num = findDisplayEntity ($message, 0);
602 $body_message = getEntity($message, $ent_num);
603 if (($body_message->header
->type0
== 'text') ||
604 ($body_message->header
->type0
== 'rfc822')) {
606 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
607 $body = decodeBody($body, $body_message->header
->encoding
);
608 $hookResults = do_hook("message_body", $body);
609 $body = $hookResults[1];
611 // If there are other types that shouldn't be formatted, add
613 if ($body_message->header
->type1
== 'html') {
614 if ( $show_html_default <> 1 ) {
615 $body = strip_tags( $body );
616 translateText($body, $wrap_at, $body_message->header
->charset
);
618 $body = MagicHTML( $body, $id );
621 translateText($body, $wrap_at, $body_message->header
->charset
);
624 $body .= "<SMALL><CENTER><A HREF=\"../src/download.php?absolute_dl=true&passed_id=$id&passed_ent_id=$ent_num&mailbox=$urlmailbox&showHeaders=1\">". _("Download this as a file") ."</A></CENTER><BR></SMALL>";
626 /** Display the ATTACHMENTS: message if there's more than one part **/
627 $body .= "</TD></TR></TABLE>";
628 if (isset($message->entities
[0])) {
629 $body .= formatAttachments ($message, $ent_num, $message->header
->mailbox
, $id);
631 $body .= "</TD></TR></TABLE>";
633 $body = formatAttachments ($message, -1, $message->header
->mailbox
, $id);
639 * A recursive function that returns a list of attachments with links
640 * to where to download these attachments
642 function formatAttachments ($message, $ent_id, $mailbox, $id) {
643 global $where, $what;
644 global $startMessage, $color;
645 static $ShownHTML = 0;
648 if ($ShownHTML == 0) {
651 $body .= "<TABLE WIDTH=\"100%\" CELLSPACING=0 CELLPADDING=2 BORDER=0 BGCOLOR=\"$color[0]\"><TR>\n" .
652 "<TH ALIGN=\"left\" BGCOLOR=\"$color[9]\"><B>\n" .
653 _("Attachments") . ':' .
654 "</B></TH></TR><TR><TD>\n" .
655 "<TABLE CELLSPACING=0 CELLPADDING=1 BORDER=0>\n" .
656 formatAttachments ($message, $ent_id, $mailbox, $id) .
657 "</TABLE></TD></TR></TABLE>";
663 if (!$message->entities
) {
664 $type0 = strtolower($message->header
->type0
);
665 $type1 = strtolower($message->header
->type1
);
666 $name = decodeHeader($message->header
->name
);
668 if ($message->header
->entity_id
!= $ent_id) {
669 $filename = decodeHeader($message->header
->filename
);
670 if (trim($filename) == '') {
671 if (trim($name) == '') {
672 if ( trim( $message->header
->id
) == '' )
673 $display_filename = 'untitled-[' . $message->header
->entity_id
. ']' ;
675 $display_filename = 'cid: ' . $message->header
->id
;
676 // $display_filename = 'untitled-[' . $message->header->entity_id . ']' ;
678 $display_filename = $name;
682 $display_filename = $filename;
685 $urlMailbox = urlencode($mailbox);
686 $ent = urlencode($message->header
->entity_id
);
689 "../src/download.php?startMessage=$startMessage&passed_id=$id&mailbox=$urlMailbox&passed_ent_id=$ent";
691 $DefaultLink .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
692 $Links['download link']['text'] = _("download");
693 $Links['download link']['href'] =
694 "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&passed_ent_id=$ent";
697 /* this executes the attachment hook with a specific MIME-type.
698 * if that doens't have results, it tries if there's a rule
699 * for a more generic type. */
700 $HookResults = do_hook("attachment $type0/$type1", $Links,
701 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
702 $display_filename, $where, $what);
703 if(count($HookResults[1]) <= 1) {
704 $HookResults = do_hook("attachment $type0/*", $Links,
705 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
706 $display_filename, $where, $what);
709 $Links = $HookResults[1];
710 $DefaultLink = $HookResults[6];
712 $body .= '<TR><TD> </TD><TD>' .
713 "<A HREF=\"$DefaultLink\">$display_filename</A> </TD>" .
714 '<TD><SMALL><b>' . show_readable_size($message->header
->size
) .
715 '</b> </small></TD>' .
716 "<TD><SMALL>[ $type0/$type1 ] </SMALL></TD>" .
718 if ($message->header
->description
)
719 $body .= '<b>' . htmlspecialchars($message->header
->description
) . '</b>';
720 $body .= '</SMALL></TD><TD><SMALL> ';
724 foreach ($Links as $Val) {
728 $body .= ' | ';
730 $body .= '<a href="' . $Val['href'] . '">' . $Val['text'] . '</a>';
735 $body .= "</SMALL></TD></TR>\n";
738 for ($i = 0; $i < count($message->entities
); $i++
) {
739 $body .= formatAttachments ($message->entities
[$i], $ent_id, $mailbox, $id);
747 /** this function decodes the body depending on the encoding type. **/
748 function decodeBody($body, $encoding) {
749 $body = str_replace("\r\n", "\n", $body);
750 $encoding = strtolower($encoding);
752 global $show_html_default;
754 if ($encoding == 'quoted-printable') {
755 $body = quoted_printable_decode($body);
758 while (ereg("=\n", $body))
759 $body = ereg_replace ("=\n", "", $body);
761 } else if ($encoding == 'base64') {
762 $body = base64_decode($body);
765 // All other encodings are returned raw.
770 * This functions decode strings that is encoded according to
771 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
773 function decodeHeader ($string, $utfencode=true) {
774 if (eregi('=\\?([^?]+)\\?(q|b)\\?([^?]+)\\?=',
776 if (ucfirst($res[2]) == 'B') {
777 $replace = base64_decode($res[3]);
779 $replace = str_replace('_', ' ', $res[3]);
780 // Convert lowercase Quoted Printable to uppercase for
781 // quoted_printable_decode to understand it.
782 while (ereg("(=(([0-9][abcdef])|([abcdef][0-9])|([abcdef][abcdef])))", $replace, $res)) {
783 $replace = str_replace($res[1], strtoupper($res[1]), $replace);
785 $replace = quoted_printable_decode($replace);
787 /* Only encode into entities by default. Some places
788 don't need the encoding, like the compose form. */
790 $replace = charset_decode ($res[1], $replace);
793 // Remove the name of the character set.
794 $string = eregi_replace ('=\\?([^?]+)\\?(q|b)\\?([^?]+)\\?=',
797 // In case there should be more encoding in the string: recurse
798 return (decodeHeader($string));
804 * Encode a string according to RFC 1522 for use in headers if it
805 * contains 8-bit characters or anything that looks like it should
808 function encodeHeader ($string) {
809 global $default_charset;
811 // Encode only if the string contains 8-bit characters or =?
812 $j = strlen( $string );
813 $l = strstr($string, '=?'); // Must be encoded ?
815 for( $i=0; $i < $j; ++
$i) {
816 switch( $string{$i} ) {
830 $k = ord( $string{$i} );
832 $ret .= sprintf("=%02X", $k);
840 $string = "=?$default_charset?Q?$ret?=";
847 Strips dangerous tags from html messages.
849 function MagicHTML( $body, $id ) {
851 global $message, $HTTP_SERVER_VARS,
852 $attachment_common_show_images;
854 $attachment_common_show_images =
855 FALSE; // Don't display attached images in HTML mode
856 $j = strlen( $body ); // Legnth of the HTML
857 $ret = ''; // Returned string
858 $bgcolor = '#ffffff'; // Background style color (defaults to white)
859 $textcolor = '#000000'; // Foreground style color (defaults to black)
860 $leftmargin = ''; // Left margin style
861 $title = ''; // HTML title if any
865 if ( $body{$i} == '<' ) {
868 while ($body{$pos} == ' ' ||
$body{$pos} == "\t" ||
869 $body{$pos} == "\n") {
872 while (strlen($tag) < 4 && $body{$pos} != ' ' &&
873 $body{$pos} != "\t" && $body{$pos} != "\n") {
877 switch( strtoupper( $tag ) ) {
878 // Strips the entire tag and contents
885 while ( $body{$i+
1}.$body{$i+
2}.$body{$i+
3}.$body{$i+
4}.$body{$i+
5} <> $etg &&
887 while ( $i < $j && $body{++
$i} <> '>' );
888 // $ret .= "<!-- $tag removed -->";
893 while ( $body{$i} <> '>' && // </title>
898 while ( $body{$i} <> '<' && // </title>
905 // Destroy these tags
916 while ( $body{$i} <> '>' &&
923 while ( $body{$i} <> '>' && // </title>
927 // We parse the style to look for interesting stuff
929 while ( $body{$i} <> '>' &&
931 // First we get the name of the style
933 while ( $body{$i} <> '>' &&
937 if ( isnoSep( $body{$i} ) )
941 stripComments( $i, $j, $body );
942 $style = strtoupper( trim( $style ) );
943 if ( $style == 'BODY' ) {
944 // Next we look into the definitions of the body style
945 while ( $body{$i} <> '>' &&
948 // We look for the background color if any.
949 if ( substr( $body, $i, 17 ) == 'BACKGROUND-COLOR:' ) {
951 $bgcolor = getStyleData( $i, $j, $body );
952 } elseif ( substr( $body, $i, 12 ) == 'MARGIN-LEFT:' ) {
954 $leftmargin = getStyleData( $i, $j, $body );
959 // Other style are mantained
960 $styleblk .= "$style ";
961 while ( $body{$i} <> '>' &&
965 $styleblk .= $body{$i};
968 $styleblk .= $body{$i};
970 stripComments( $i, $j, $body );
971 if ( $body{$i} <> '>' )
974 if ( $styleblk <> '' )
975 $ret .= "<style>$styleblk";
979 $ret .= '<b>' . _("Title:") . " </b>$title<br>\n";
982 if (! isset($base)) {
985 $ret .= stripEvent( $i, $j, $body, $id, $base );
986 $ret .= " bgcolor=$bgcolor width=\"100%\"><tr>";
987 if ( $leftmargin <> '' )
988 $ret .= "<td width=$leftmargin> </td>";
990 if (strtolower($bgcolor) == 'ffffff' ||
991 strtolower($bgcolor) == '#ffffff')
992 $ret .= '<font color=#000000>';
997 while ( !isNoSep( $body{$i} ) &&
1001 if ( strcasecmp( substr( $base, 0, 4 ), 'href' ) ) {
1003 while ( !isNoSep( $body{$i} ) &&
1007 while ( $body{$i} <> '>' &&
1009 if ( $body{$i} <> '"' ) {
1014 // Debuging $ret .= "<!-- base == $base -->";
1015 if ( strcasecmp( substr( $base, 0, 4 ), 'file' ) <> 0 ) {
1016 $ret .= "\n<BASE HREF=\"$base\">\n";
1021 $ret .= '</font></td></tr></TABLE>';
1025 // Following tags can contain some event handler, lets search it
1026 stripComments( $i, $j, $body );
1027 if (! isset($base)) {
1030 $ret .= stripEvent( $i, $j, $body, $id, $base ) . '>';
1031 // $ret .= "<!-- $tag detected -->";
1039 return( "\n\n<!-- HTML Output ahead -->\n" .
1041 "\n<!-- END of HTML Output --><base href=\"".
1042 get_location() . '/'.
1046 function isNoSep( $char ) {
1064 The following function is usefull to remove extra data that can cause
1065 html not to display properly. Especialy with MS stuff.
1068 function stripComments( &$i, $j, &$body ) {
1070 while ( $body{$i}.$body{$i+
1}.$body{$i+
2}.$body{$i+
3} == '<!--' &&
1073 while ( $body{$i-2}.$body{$i-1}.$body{$i} <> '-->' &&
1083 /* Gets the style data of a specific style */
1085 function getStyleData( &$i, $j, &$body ) {
1088 while ( $body{$i} <> '>' && !isNoSep( $body{$i} ) &&
1092 // And get the color
1094 while ( isNoSep( $body{$i} ) &&
1104 Private function for strip_dangerous_tag. Look for event based coded and "remove" it
1105 change on with no (onload -> noload)
1108 function stripEvent( &$i, $j, &$body, $id, $base ) {
1110 global $message, $base_uri;
1114 while ( $body{$i} <> '>' &&
1116 $etg = strtolower($body{$i}.$body{$i+
1}.$body{$i+
2});
1119 // This is probably a src specification
1121 while( !isNoSep( $body{$k} )) {
1124 if ( $body{$k} == '=' ) {
1127 while( !isNoSep( $body{$k} ) &&
1132 while ( $body{$k} <> '>' && isNoSep( $body{$k} ) &&
1137 while( !isNoSep( $body{$k} ) &&
1141 if ( strtolower( substr( $src, 0, 4 ) ) == 'cid:' ) {
1142 $src = substr( $src, 4 );
1143 $src = "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=" .
1144 urlencode( $message->header
->mailbox
) .
1145 "&passed_ent_id=" . find_ent_id( $src, $message );
1146 } else if ( strtolower( substr( $src, 0, 4 ) ) <> 'http' ||
1147 stristr( $src, $base_uri ) ) {
1148 /* Javascript and local urls goes out */
1149 $src = '../images/' . _("sec_remove_eng.png");
1151 $ret .= 'src="' . $src . '" ';
1160 // Retrolinks are not allowed without a base because they mess with SM security
1161 if ( $base == '' ) {
1170 if ( $body{$i+
3} == ':') {
1173 while ( isNoSep( $body{$i} ) &&
1175 $name .= $body{$i++
};
1177 if ( $name <> '' ) {
1178 $ret .= "../src/download.php?absolute_dl=true&passed_id=$id&mailbox=" .
1179 urlencode( $message->header
->mailbox
) .
1180 "&passed_ent_id=" . find_ent_id( $name, $message );
1181 if ( $body{$k} == '"' )
1186 if ( $body{$i} == '>' )
1198 if ( strcasecmp( $body{$i-4}.$body{$i-3}.$body{$i-2}.$body{$i-1}.$body{$i}.$body{$i+
1}.$body{$i+
2}, 'script:') == 0 ) {
1214 /* This function trys to locate the entity_id of a specific mime element */
1216 function find_ent_id( $id, $message ) {
1219 for ($i=0; $ret == '' && $i < count($message->entities
); $i++
) {
1221 if ( $message->entities
[$i]->header
->entity_id
== '' ) {
1222 $ret = find_ent_id( $id, $message->entities
[$i] );
1224 if ( strcasecmp( $message->entities
[$i]->header
->id
, $id ) == 0 )
1225 $ret = $message->entities
[$i]->header
->entity_id
;