92927d68f7a89cc5fa3f0a795397c9127694b5d2
[squirrelmail.git] / functions / mime.php
1 <?php
2
3 /**
4 * mime.php
5 *
6 * Copyright (c) 1999-2002 The SquirrelMail Project Team
7 * Licensed under the GNU GPL. For full terms see the file COPYING.
8 *
9 * This contains the functions necessary to detect and decode MIME
10 * messages.
11 *
12 * $Id$
13 */
14
15 require_once('../functions/imap.php');
16 require_once('../functions/attachment_common.php');
17
18 /** Setting up the objects that have the structure for the message **/
19 class msg_header {
20 /** msg_header contains generic variables for values that **/
21 /** could be in a header. **/
22
23 var $type0 = '', $type1 = '', $boundary = '', $charset = '',
24 $encoding = '', $size = 0, $to = array(), $from = '', $date = '',
25 $cc = array(), $bcc = array(), $reply_to = '', $subject = '',
26 $id = 0, $mailbox = '', $description = '', $filename = '',
27 $entity_id = 0, $message_id = 0, $name = '', $priority = 3;
28 }
29
30 class message {
31 /** message is the object that contains messages. It is a recursive
32 object in that through the $entities variable, it can contain
33 more objects of type message. See documentation in mime.txt for
34 a better description of how this works.
35 **/
36 var $header = '', $entities = array();
37
38 function addEntity ($msg) {
39 $this->entities[] = $msg;
40 }
41 }
42
43 /* --------------------------------------------------------------------------------- */
44 /* MIME DECODING */
45 /* --------------------------------------------------------------------------------- */
46
47 /* This function gets the structure of a message and stores it in the "message" class.
48 * It will return this object for use with all relevant header information and
49 * fully parsed into the standard "message" object format.
50 */
51 function mime_structure ($imap_stream, $header) {
52
53 sqimap_messages_flag ($imap_stream, $header->id, $header->id, 'Seen');
54 $ssid = sqimap_session_id();
55 $lsid = strlen( $ssid );
56 $id = $header->id;
57 fputs ($imap_stream, "$ssid FETCH $id BODYSTRUCTURE\r\n");
58 //
59 // This should use sqimap_read_data instead of reading it itself
60 //
61 $read = fgets ($imap_stream, 9216);
62 $bodystructure = '';
63 while ( substr($read, 0, $lsid) <> $ssid &&
64 !feof( $imap_stream ) ) {
65 $bodystructure .= $read;
66 $read = fgets ($imap_stream, 9216);
67 }
68 $read = $bodystructure;
69
70 // isolate the body structure and remove beginning and end parenthesis
71 $read = trim(substr ($read, strpos(strtolower($read), 'bodystructure') + 13));
72 $read = trim(substr ($read, 0, -1));
73 $end = mime_match_parenthesis(0, $read);
74 while ($end == strlen($read)-1) {
75 $read = trim(substr ($read, 0, -1));
76 $read = trim(substr ($read, 1));
77 $end = mime_match_parenthesis(0, $read);
78 }
79
80 $msg = mime_parse_structure ($read, 0);
81 $msg->header = $header;
82
83 return( $msg );
84 }
85
86 /* this starts the parsing of a particular structure. It is called recursively,
87 * so it can be passed different structures. It returns an object of type
88 * $message.
89 * First, it checks to see if it is a multipart message. If it is, then it
90 * handles that as it sees is necessary. If it is just a regular entity,
91 * then it parses it and adds the necessary header information (by calling out
92 * to mime_get_elements()
93 */
94 function mime_parse_structure ($structure, $ent_id) {
95
96 $msg = new message();
97 if ($structure{0} == '(') {
98 $ent_id = mime_new_element_level($ent_id);
99 $start = $end = -1;
100 do {
101 $start = $end+1;
102 $end = mime_match_parenthesis ($start, $structure);
103
104 $element = substr($structure, $start+1, ($end - $start)-1);
105 $ent_id = mime_increment_id ($ent_id);
106 $newmsg = mime_parse_structure ($element, $ent_id);
107 $msg->addEntity ($newmsg);
108 } while ($structure{$end+1} == '(');
109 } else {
110 // parse the elements
111 $msg = mime_get_element ($structure, $msg, $ent_id);
112 }
113 return $msg;
114 }
115
116 /* Increments the element ID. An element id can look like any of
117 * the following: 1, 1.2, 4.3.2.4.1, etc. This function increments
118 * the last number of the element id, changing 1.2 to 1.3.
119 */
120 function mime_increment_id ($id) {
121
122 if (strpos($id, '.')) {
123 $first = substr($id, 0, strrpos($id, '.'));
124 $last = substr($id, strrpos($id, '.')+1);
125 $last++;
126 $new = $first . '.' .$last;
127 } else {
128 $new = $id + 1;
129 }
130
131 return $new;
132 }
133
134 /*
135 * See comment for mime_increment_id().
136 * This adds another level on to the entity_id changing 1.3 to 1.3.0
137 * NOTE: 1.3.0 is not a valid element ID. It MUST be incremented
138 * before it can be used. I left it this way so as not to have
139 * to make a special case if it is the first entity_id. It
140 * always increments it, and that works fine.
141 */
142 function mime_new_element_level ($id) {
143
144 if (!$id) {
145 $id = 0;
146 } else {
147 $id = $id . '.0';
148 }
149
150 return( $id );
151 }
152
153 function mime_get_element (&$structure, $msg, $ent_id) {
154
155 $elem_num = 1;
156 $msg->header = new msg_header();
157 $msg->header->entity_id = $ent_id;
158 $properties = array();
159
160 while (strlen($structure) > 0) {
161 $structure = trim($structure);
162 $char = $structure{0};
163
164 if (strtolower(substr($structure, 0, 3)) == 'nil') {
165 $text = '';
166 $structure = substr($structure, 3);
167 } else if ($char == '"') {
168 // loop through until we find the matching quote, and return that as a string
169 $pos = 1;
170 $text = '';
171 while ( ($char = $structure{$pos} ) <> '"' && $pos < strlen($structure)) {
172 $text .= $char;
173 $pos++;
174 }
175 $structure = substr($structure, strlen($text) + 2);
176 } else if ($char == '(') {
177 // comment me
178 $end = mime_match_parenthesis (0, $structure);
179 $sub = substr($structure, 1, $end-1);
180 $properties = mime_get_props($properties, $sub);
181 $structure = substr($structure, strlen($sub) + 2);
182 } else {
183 // loop through until we find a space or an end parenthesis
184 $pos = 0;
185 $char = $structure{$pos};
186 $text = '';
187 while ($char != ' ' && $char != ')' && $pos < strlen($structure)) {
188 $text .= $char;
189 $pos++;
190 $char = $structure{$pos};
191 }
192 $structure = substr($structure, strlen($text));
193 }
194
195 // This is where all the text parts get put into the header
196 switch ($elem_num) {
197 case 1:
198 $msg->header->type0 = strtolower($text);
199 break;
200 case 2:
201 $msg->header->type1 = strtolower($text);
202 break;
203 case 4: // Id
204 // Invisimail enclose images with <>
205 $msg->header->id = str_replace( '<', '', str_replace( '>', '', $text ) );
206 break;
207 case 5:
208 $msg->header->description = $text;
209 break;
210 case 6:
211 $msg->header->encoding = strtolower($text);
212 break;
213 case 7:
214 $msg->header->size = $text;
215 break;
216 default:
217 if ($msg->header->type0 == 'text' && $elem_num == 8) {
218 // This is a plain text message, so lets get the number of lines
219 // that it contains.
220 $msg->header->num_lines = $text;
221
222 } else if ($msg->header->type0 == 'message' && $msg->header->type1 == 'rfc822' && $elem_num == 8) {
223 // This is an encapsulated message, so lets start all over again and
224 // parse this message adding it on to the existing one.
225 $structure = trim($structure);
226 if ( $structure{0} == '(' ) {
227 $e = mime_match_parenthesis (0, $structure);
228 $structure = substr($structure, 0, $e);
229 $structure = substr($structure, 1);
230 $m = mime_parse_structure($structure, $msg->header->entity_id);
231
232 // the following conditional is there to correct a bug that wasn't
233 // incrementing the entity IDs correctly because of the special case
234 // that message/rfc822 is. This fixes it fine.
235 if (substr($structure, 1, 1) != '(')
236 $m->header->entity_id = mime_increment_id(mime_new_element_level($ent_id));
237
238 // Now we'll go through and reformat the results.
239 if ($m->entities) {
240 for ($i=0; $i < count($m->entities); $i++) {
241 $msg->addEntity($m->entities[$i]);
242 }
243 } else {
244 $msg->addEntity($m);
245 }
246 $structure = "";
247 }
248 }
249 break;
250 }
251 $elem_num++;
252 $text = "";
253 }
254 // loop through the additional properties and put those in the various headers
255 if ($msg->header->type0 != 'message') {
256 for ($i=0; $i < count($properties); $i++) {
257 $msg->header->{$properties[$i]['name']} = $properties[$i]['value'];
258 }
259 }
260
261 return $msg;
262 }
263
264 /*
265 * I did most of the MIME stuff yesterday (June 20, 2000), but I couldn't
266 * figure out how to do this part, so I decided to go to bed. I woke up
267 * in the morning and had a flash of insight. I went to the white-board
268 * and scribbled it out, then spent a bit programming it, and this is the
269 * result. Nothing complicated, but I think my brain was fried yesterday.
270 * Funny how that happens some times.
271 *
272 * This gets properties in a nested parenthesisized list. For example,
273 * this would get passed something like: ("attachment" ("filename" "luke.tar.gz"))
274 * This returns an array called $props with all paired up properties.
275 * It ignores the "attachment" for now, maybe that should change later
276 * down the road. In this case, what is returned is:
277 * $props[0]["name"] = "filename";
278 * $props[0]["value"] = "luke.tar.gz";
279 */
280 function mime_get_props ($props, $structure) {
281
282 while (strlen($structure) > 0) {
283 $structure = trim($structure);
284 $char = $structure{0};
285
286 if ($char == '"') {
287 $pos = 1;
288 $tmp = '';
289 while ( ( $char = $structure{$pos} ) != '"' &&
290 $pos < strlen($structure)) {
291 $tmp .= $char;
292 $pos++;
293 }
294 $structure = trim(substr($structure, strlen($tmp) + 2));
295 $char = $structure{0};
296
297 if ($char == '"') {
298 $pos = 1;
299 $value = '';
300 while ( ( $char = $structure{$pos} ) != '"' &&
301 $pos < strlen($structure) ) {
302 $value .= $char;
303 $pos++;
304 }
305 $structure = trim(substr($structure, strlen($tmp) + 2));
306
307 $k = count($props);
308 $props[$k]['name'] = strtolower($tmp);
309 $props[$k]['value'] = $value;
310 } else if ($char == '(') {
311 $end = mime_match_parenthesis (0, $structure);
312 $sub = substr($structure, 1, $end-1);
313 if (! isset($props))
314 $props = array();
315 $props = mime_get_props($props, $sub);
316 $structure = substr($structure, strlen($sub) + 2);
317 }
318 return $props;
319 } else if ($char == '(') {
320 $end = mime_match_parenthesis (0, $structure);
321 $sub = substr($structure, 1, $end-1);
322 $props = mime_get_props($props, $sub);
323 $structure = substr($structure, strlen($sub) + 2);
324 return $props;
325 } else {
326 return $props;
327 }
328 }
329 }
330
331 /*
332 * Matches parenthesis. It will return the position of the matching
333 * parenthesis in $structure. For instance, if $structure was:
334 * ("text" "plain" ("val1name", "1") nil ... )
335 * x x
336 * then this would return 42 to match up those two.
337 */
338 function mime_match_parenthesis ($pos, $structure) {
339
340 $j = strlen( $structure );
341
342 // ignore all extra characters
343 // If inside of a string, skip string -- Boundary IDs and other
344 // things can have ) in them.
345 if ( $structure{$pos} != '(' ) {
346 return( $j );
347 }
348
349 while ( $pos < $j ) {
350 $pos++;
351 if ($structure{$pos} == ')') {
352 return $pos;
353 } elseif ($structure{$pos} == '"') {
354 $pos++;
355 while ( $structure{$pos} != '"' &&
356 $pos < $j ) {
357 if (substr($structure, $pos, 2) == '\\"') {
358 $pos++;
359 } elseif (substr($structure, $pos, 2) == '\\\\') {
360 $pos++;
361 }
362 $pos++;
363 }
364 } elseif ( $structure{$pos} == '(' ) {
365 $pos = mime_match_parenthesis ($pos, $structure);
366 }
367 }
368 echo _("Error decoding mime structure. Report this as a bug!") . '<br>';
369 return( $pos );
370 }
371
372 function mime_fetch_body($imap_stream, $id, $ent_id ) {
373
374 /*
375 * do a bit of error correction. If we couldn't find the entity id, just guess
376 * that it is the first one. That is usually the case anyway.
377 */
378 if (!$ent_id) {
379 $ent_id = 1;
380 }
381
382 $cmd = "FETCH $id BODY[$ent_id]";
383 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message);
384
385 do {
386 $topline = array_shift( $data );
387 } while( $topline && $topline == '*' && !preg_match( '/\\* [0-9] FETCH.*/i', $topline )) ;
388 $wholemessage = implode('', $data);
389 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
390
391 $ret = substr( $wholemessage, 0, $regs[1] );
392 /*
393 There is some information in the content info header that could be important
394 in order to parse html messages. Let's get them here.
395 */
396 if ( $ret{0} == '<' ) {
397 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message);
398 /* BASE within HTML documents is illegal (see w3 spec)
399 * $base = '';
400 * $k = 10;
401 * foreach( $data as $d ) {
402 * if ( substr( $d, 0, 13 ) == 'Content-Base:' ) {
403 * $j = strlen( $d );
404 * $i = 13;
405 * $base = '';
406 * while ( $i < $j &&
407 * ( !isNoSep( $d{$i} ) || $d{$i} == '"' ) )
408 * $i++;
409 * while ( $i < $j ) {
410 * if ( isNoSep( $d{$i} ) )
411 * $base .= $d{$i};
412 * $i++;
413 * }
414 * $k = 0;
415 * } elseif ( $k == 1 && !isnosep( $d{0} ) ) {
416 * $base .= substr( $d, 1 );
417 * }
418 * $k++;
419 * }
420 * if ( $base <> '' ) {
421 * $ret = "<base href=\"$base\">" . $ret;
422 * }
423 * */
424 }
425 } else if (ereg('"([^"]*)"', $topline, $regs)) {
426 $ret = $regs[1];
427 } else {
428 global $where, $what, $mailbox, $passed_id, $startMessage;
429 $par = 'mailbox=' . urlencode($mailbox) . "&amp;passed_id=$passed_id";
430 if (isset($where) && isset($what)) {
431 $par .= '&amp;where='. urlencode($where) . "&amp;what=" . urlencode($what);
432 } else {
433 $par .= "&amp;startMessage=$startMessage&amp;show_more=0";
434 }
435 $par .= '&amp;response=' . urlencode($response) .
436 '&amp;message=' . urlencode($message).
437 '&amp;topline=' . urlencode($topline);
438
439 echo '<tt><br>' .
440 '<table width="80%"><tr>' .
441 '<tr><td colspan=2>' .
442 _("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!") .
443 " <A HREF=\"../src/retrievalerror.php?$par\"><br>" .
444 _("Submit message") . '</A><BR>&nbsp;' .
445 '</td></tr>' .
446 '<td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
447 '<td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
448 '<td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
449 '<td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
450 "</table><BR></tt></font><hr>";
451
452 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message);
453 array_shift($data);
454 $wholemessage = implode('', $data);
455
456 $ret = $wholemessage;
457 }
458 return( $ret );
459 }
460
461 function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) {
462 // do a bit of error correction. If we couldn't find the entity id, just guess
463 // that it is the first one. That is usually the case anyway.
464 if (!$ent_id) {
465 $ent_id = 1;
466 }
467 $sid = sqimap_session_id();
468 // Don't kill the connection if the browser is over a dialup
469 // and it would take over 30 seconds to download it.
470
471 // donĀ“t call set_time_limit in safe mode.
472 if (!ini_get("safe_mode")) {
473 set_time_limit(0);
474 }
475
476 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
477 $cnt = 0;
478 $continue = true;
479 $read = fgets ($imap_stream,4096);
480 // This could be bad -- if the section has sqimap_session_id() . ' OK'
481 // or similar, it will kill the download.
482 while (!ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
483 if (trim($read) == ')==') {
484 $read1 = $read;
485 $read = fgets ($imap_stream,4096);
486 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
487 return;
488 } else {
489 echo decodeBody($read1, $encoding) .
490 decodeBody($read, $encoding);
491 }
492 } else if ($cnt) {
493 echo decodeBody($read, $encoding);
494 }
495 $read = fgets ($imap_stream,4096);
496 $cnt++;
497 }
498 }
499
500 /* -[ END MIME DECODING ]----------------------------------------------------------- */
501
502
503
504 /* This is the first function called. It decides if this is a multipart
505 message or if it should be handled as a single entity
506 */
507 function decodeMime ($imap_stream, &$header) {
508 global $username, $key, $imapServerAddress, $imapPort;
509 return mime_structure ($imap_stream, $header);
510 }
511
512 // This is here for debugging purposese. It will print out a list
513 // of all the entity IDs that are in the $message object.
514 /*
515 function listEntities ($message) {
516 if ($message) {
517 if ($message->header->entity_id)
518 echo "<tt>" . $message->header->entity_id . ' : ' . $message->header->type0 . '/' . $message->header->type1 . '<br>';
519 for ($i = 0; $message->entities[$i]; $i++) {
520 $msg = listEntities($message->entities[$i], $ent_id);
521 if ($msg)
522 return $msg;
523 }
524 }
525 }
526 */
527
528 /* returns a $message object for a particular entity id */
529 function getEntity ($message, $ent_id) {
530 if ($message) {
531 if ($message->header->entity_id == $ent_id && strlen($ent_id) == strlen($message->header->entity_id)) {
532 return $message;
533 } else {
534 for ($i = 0; isset($message->entities[$i]); $i++) {
535 $msg = getEntity ($message->entities[$i], $ent_id);
536 if ($msg) {
537 return $msg;
538 }
539 }
540 }
541 }
542 }
543
544 /*
545 * figures out what entity to display and returns the $message object
546 * for that entity.
547 */
548 function findDisplayEntity ($message, $textOnly = 1) {
549 global $show_html_default;
550
551 $entity = 0;
552
553 if ($message) {
554 if ( $message->header->type0 == 'multipart' &&
555 ( $message->header->type1 == 'alternative' ||
556 $message->header->type1 == 'mixed' ||
557 $message->header->type1 == 'related' ) &&
558 $show_html_default && ! $textOnly ) {
559 $entity = findDisplayEntityHTML($message);
560 }
561
562 // Show text/plain or text/html -- the first one we find.
563 if ( $entity == 0 &&
564 $message->header->type0 == 'text' &&
565 ( $message->header->type1 == 'plain' ||
566 $message->header->type1 == 'html' ) &&
567 isset($message->header->entity_id) ) {
568 $entity = $message->header->entity_id;
569 }
570
571 $i = 0;
572 while ($entity == 0 && isset($message->entities[$i]) ) {
573 $entity = findDisplayEntity($message->entities[$i], $textOnly);
574 $i++;
575 }
576 }
577
578 return( $entity );
579 }
580
581 /* Shows the HTML version */
582 function findDisplayEntityHTML ($message) {
583
584 if ( $message->header->type0 == 'text' &&
585 $message->header->type1 == 'html' &&
586 isset($message->header->entity_id)) {
587 return $message->header->entity_id;
588 }
589 for ($i = 0; isset($message->entities[$i]); $i ++) {
590 $entity = findDisplayEntityHTML($message->entities[$i]);
591 if ($entity != 0) {
592 return $entity;
593 }
594 }
595
596 return 0;
597 }
598
599 /* This returns a parsed string called $body. That string can then
600 be displayed as the actual message in the HTML. It contains
601 everything needed, including HTML Tags, Attachments at the
602 bottom, etc.
603 */
604 function formatBody($imap_stream, $message, $color, $wrap_at) {
605 // this if statement checks for the entity to show as the
606 // primary message. To add more of them, just put them in the
607 // order that is their priority.
608 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
609 $show_html_default, $has_unsafe_images, $view_unsafe_images, $sort;
610
611 $has_unsafe_images = 0;
612
613 $id = $message->header->id;
614 $urlmailbox = urlencode($message->header->mailbox);
615
616 // Get the right entity and redefine message to be this entity
617 // Pass the 0 to mean that we want the 'best' viewable one
618 $ent_num = findDisplayEntity ($message, 0);
619 $body_message = getEntity($message, $ent_num);
620 if (($body_message->header->type0 == 'text') ||
621 ($body_message->header->type0 == 'rfc822')) {
622
623 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
624 $body = decodeBody($body, $body_message->header->encoding);
625 $hookResults = do_hook("message_body", $body);
626 $body = $hookResults[1];
627
628 // If there are other types that shouldn't be formatted, add
629 // them here
630 if ($body_message->header->type1 == 'html') {
631 if ( $show_html_default <> 1 ) {
632 $body = strip_tags( $body );
633 translateText($body, $wrap_at, $body_message->header->charset);
634 } else {
635 $body = MagicHTML( $body, $id );
636 }
637 } else {
638 translateText($body, $wrap_at, $body_message->header->charset);
639 }
640
641 $body .= "<CENTER><SMALL><A HREF=\"../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;passed_ent_id=$ent_num&amp;mailbox=$urlmailbox&amp;showHeaders=1\">". _("Download this as a file") ."</A></SMALL></CENTER><BR>";
642 if ($has_unsafe_images) {
643 if ($view_unsafe_images) {
644 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&amp;mailbox=$urlmailbox&amp;sort=$sort&amp;startMessage=$startMessage&amp;show_more=0\">". _("Hide Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
645 } else {
646 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&amp;mailbox=$urlmailbox&amp;sort=$sort&amp;startMessage=$startMessage&amp;show_more=0&amp;view_unsafe_images=1\">". _("View Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
647 }
648 }
649
650 /** Display the ATTACHMENTS: message if there's more than one part **/
651 if (isset($message->entities[0])) {
652 $body .= formatAttachments ($message, $ent_num, $message->header->mailbox, $id);
653 }
654 } else {
655 $body = formatAttachments ($message, -1, $message->header->mailbox, $id);
656 }
657 return ($body);
658 }
659
660 /*
661 * A recursive function that returns a list of attachments with links
662 * to where to download these attachments
663 */
664 function formatAttachments($message, $ent_id, $mailbox, $id) {
665 global $where, $what;
666 global $startMessage, $color;
667 static $ShownHTML = 0;
668
669 $body = '';
670 if ($ShownHTML == 0) {
671
672 $ShownHTML = 1;
673 $body .= "<TABLE WIDTH=\"100%\" CELLSPACING=0 CELLPADDING=2 BORDER=0 BGCOLOR=\"$color[0]\"><TR>\n" .
674 "<TH ALIGN=\"left\" BGCOLOR=\"$color[9]\"><B>\n" .
675 _("Attachments") . ':' .
676 "</B></TH></TR><TR><TD>\n" .
677 "<TABLE CELLSPACING=0 CELLPADDING=1 BORDER=0>\n" .
678 formatAttachments($message, $ent_id, $mailbox, $id) .
679 "</TABLE></TD></TR></TABLE>";
680
681 } else if ($message) {
682
683 if (!$message->entities) {
684
685 $type0 = strtolower($message->header->type0);
686 $type1 = strtolower($message->header->type1);
687 $name = decodeHeader($message->header->name);
688
689 if ($message->header->entity_id != $ent_id) {
690 $filename = decodeHeader($message->header->filename);
691 if (trim($filename) == '') {
692 if (trim($name) == '') {
693 if ( trim( $message->header->id ) == '' )
694 $display_filename = 'untitled-[' . $message->header->entity_id . ']' ;
695 else
696 $display_filename = 'cid: ' . $message->header->id;
697 // $display_filename = 'untitled-[' . $message->header->entity_id . ']' ;
698 } else {
699 $display_filename = $name;
700 $filename = $name;
701 }
702 } else {
703 $display_filename = $filename;
704 }
705
706 $urlMailbox = urlencode($mailbox);
707 $ent = urlencode($message->header->entity_id);
708
709 $DefaultLink =
710 "../src/download.php?startMessage=$startMessage&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;passed_ent_id=$ent";
711 if ($where && $what) {
712 $DefaultLink .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
713 }
714 $Links['download link']['text'] = _("download");
715 $Links['download link']['href'] =
716 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;passed_ent_id=$ent";
717 $ImageURL = '';
718
719 /* this executes the attachment hook with a specific MIME-type.
720 * if that doens't have results, it tries if there's a rule
721 * for a more generic type. */
722 $HookResults = do_hook("attachment $type0/$type1", $Links,
723 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
724 $display_filename, $where, $what);
725 if(count($HookResults[1]) <= 1) {
726 $HookResults = do_hook("attachment $type0/*", $Links,
727 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
728 $display_filename, $where, $what);
729 }
730
731 $Links = $HookResults[1];
732 $DefaultLink = $HookResults[6];
733
734 $body .= '<TR><TD>&nbsp;&nbsp;</TD><TD>' .
735 "<A HREF=\"$DefaultLink\">$display_filename</A>&nbsp;</TD>" .
736 '<TD><SMALL><b>' . show_readable_size($message->header->size) .
737 '</b>&nbsp;&nbsp;</small></TD>' .
738 "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
739 '<TD><SMALL>';
740 if ($message->header->description) {
741 $body .= '<b>' . htmlspecialchars(_($message->header->description)) . '</b>';
742 }
743 $body .= '</SMALL></TD><TD><SMALL>&nbsp;';
744
745
746 $SkipSpaces = 1;
747 foreach ($Links as $Val) {
748 if ($SkipSpaces) {
749 $SkipSpaces = 0;
750 } else {
751 $body .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
752 }
753 $body .= '<a href="' . $Val['href'] . '">' . $Val['text'] . '</a>';
754 }
755
756 unset($Links);
757
758 $body .= "</SMALL></TD></TR>\n";
759 }
760 } else {
761 for ($i = 0; $i < count($message->entities); $i++) {
762 $body .= formatAttachments($message->entities[$i], $ent_id, $mailbox, $id);
763 }
764 }
765 }
766 return( $body );
767 }
768
769
770 /** this function decodes the body depending on the encoding type. **/
771 function decodeBody($body, $encoding) {
772 $body = str_replace("\r\n", "\n", $body);
773 $encoding = strtolower($encoding);
774
775 global $show_html_default;
776
777 if ($encoding == 'quoted-printable') {
778 $body = quoted_printable_decode($body);
779
780
781 while (ereg("=\n", $body))
782 $body = ereg_replace ("=\n", "", $body);
783
784 } else if ($encoding == 'base64') {
785 $body = base64_decode($body);
786 }
787
788 // All other encodings are returned raw.
789 return $body;
790 }
791
792 /*
793 * This functions decode strings that is encoded according to
794 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
795 */
796 function decodeHeader ($string, $utfencode=true) {
797
798 if ( is_array( $string ) ) {
799 $string = implode("\n", $string );
800 }
801
802 if (eregi('=\\?([^?]+)\\?(q|b)\\?([^?]+)\\?=',
803 $string, $res)) {
804 if (ucfirst($res[2]) == 'B') {
805 $replace = base64_decode($res[3]);
806 } else {
807 $replace = str_replace('_', ' ', $res[3]);
808 // Convert lowercase Quoted Printable to uppercase for
809 // quoted_printable_decode to understand it.
810 while (ereg("(=(([0-9][abcdef])|([abcdef][0-9])|([abcdef][abcdef])))",
811 $replace, $res)) {
812 $replace = str_replace($res[1], strtoupper($res[1]), $replace);
813 }
814 $replace = quoted_printable_decode($replace);
815 }
816 /* Only encode into entities by default. Some places
817 don't need the encoding, like the compose form. */
818 if ($utfencode){
819 $replace = charset_decode ($res[1], $replace);
820 }
821
822 // Remove the name of the character set.
823 $string = eregi_replace ('=\\?([^?]+)\\?(q|b)\\?([^?]+)\\?=',
824 $replace, $string);
825
826 // In case there should be more encoding in the string: recurse
827 $string = decodeHeader($string);
828 }
829
830 return ($string);
831 }
832
833 /*
834 * Encode a string according to RFC 1522 for use in headers if it
835 * contains 8-bit characters or anything that looks like it should
836 * be encoded.
837 */
838 function encodeHeader ($string) {
839 global $default_charset;
840
841 // Encode only if the string contains 8-bit characters or =?
842 $j = strlen( $string );
843 $l = strstr($string, '=?'); // Must be encoded ?
844 $ret = '';
845 for( $i=0; $i < $j; ++$i) {
846 switch( $string{$i} ) {
847 case '=':
848 $ret .= '=3D';
849 break;
850 case '?':
851 $ret .= '=3F';
852 break;
853 case '_':
854 $ret .= '=5F';
855 break;
856 case ' ':
857 $ret .= '_';
858 break;
859 default:
860 $k = ord( $string{$i} );
861 if ( $k > 126 ) {
862 $ret .= sprintf("=%02X", $k);
863 $l = TRUE;
864 } else
865 $ret .= $string{$i};
866 }
867 }
868
869 if ( $l ) {
870 $string = "=?$default_charset?Q?$ret?=";
871 }
872
873 return( $string );
874 }
875
876 /*
877 Strips dangerous tags from html messages.
878 */
879 function MagicHTML( $body, $id ) {
880
881 global $message, $HTTP_SERVER_VARS,
882 $attachment_common_show_images;
883
884 $attachment_common_show_images =
885 FALSE; // Don't display attached images in HTML mode
886 $j = strlen( $body ); // Legnth of the HTML
887 $ret = ''; // Returned string
888 $bgcolor = '#ffffff'; // Background style color (defaults to white)
889 $textcolor = '#000000'; // Foreground style color (defaults to black)
890 $leftmargin = ''; // Left margin style
891 $title = ''; // HTML title if any
892
893 $i = 0;
894 while ( $i < $j ) {
895 if ( $body{$i} == '<' ) {
896 $pos = $i + 1;
897 $tag = '';
898 while ($body{$pos} == ' ' || $body{$pos} == "\t" ||
899 $body{$pos} == "\n" ) {
900 $pos ++;
901 }
902 while (strlen($tag) < 4 && $body{$pos} != ' ' &&
903 $body{$pos} != "\t" && $body{$pos} != "\n" &&
904 $pos < $j ) {
905 $tag .= $body{$pos};
906 $pos ++;
907 }
908 switch( strtoupper( $tag ) ) {
909 // Strips the entire tag and contents
910 case 'APPL':
911 case 'EMBB':
912 case 'FRAM':
913 case 'SCRI':
914 case 'OBJE':
915 $etg = '/' . $tag;
916 while ( $body{$i+1}.$body{$i+2}.$body{$i+3}.$body{$i+4}.$body{$i+5} <> $etg &&
917 $i < $j ) $i++;
918 while ( $i < $j && $body{++$i} <> '>' );
919 // $ret .= "<!-- $tag removed -->";
920 break;
921 // Substitute Title
922 case 'TITL':
923 $i += 5;
924 while ( $body{$i} <> '>' && // </title>
925 $i < $j )
926 $i++;
927 $i++;
928 $title = '';
929 while ( $body{$i} <> '<' && // </title>
930 $i < $j ) {
931 $title .= $body{$i};
932 $i++;
933 }
934 $i += 7;
935 break;
936 // Destroy these tags
937 case 'HTML':
938 case 'HEAD':
939 case '/HTM':
940 case '/HEA':
941 case '!DOC':
942 case 'META':
943 //case 'DIV ':
944 //case '/DIV':
945 case '!-- ':
946 $i += 4;
947 while ( $body{$i} <> '>' &&
948 $i < $j )
949 $i++;
950 // $i++;
951 break;
952 case 'STYL':
953 $i += 5;
954 while ( $body{$i} <> '>' && // </title>
955 $i < $j )
956 $i++;
957 $i++;
958 // We parse the style to look for interesting stuff
959 $styleblk = '';
960 while ( $body{$i} <> '>' &&
961 $i < $j ) {
962 // First we get the name of the style
963 $style = '';
964 while ( $body{$i} <> '>' &&
965 $body{$i} <> '<' &&
966 $body{$i} <> '{' &&
967 $i < $j ) {
968 if ( isnoSep( $body{$i} ) )
969 $style .= $body{$i};
970 $i++;
971 }
972 stripComments( $i, $j, $body );
973 $style = strtoupper( trim( $style ) );
974 if ( $style == 'BODY' ) {
975 // Next we look into the definitions of the body style
976 while ( $body{$i} <> '>' &&
977 $body{$i} <> '}' &&
978 $i < $j ) {
979 // We look for the background color if any.
980 if ( substr( $body, $i, 17 ) == 'BACKGROUND-COLOR:' ) {
981 $i += 17;
982 $bgcolor = getStyleData( $i, $j, $body );
983 } elseif ( substr( $body, $i, 12 ) == 'MARGIN-LEFT:' ) {
984 $i += 12;
985 $leftmargin = getStyleData( $i, $j, $body );
986 }
987 $i++;
988 }
989 } else {
990 // Other style are mantained
991 $styleblk .= "$style ";
992 while ( $body{$i} <> '>' &&
993 $body{$i} <> '<' &&
994 $body{$i} <> '}' &&
995 $i < $j ) {
996 $styleblk .= $body{$i};
997 $i++;
998 }
999 $styleblk .= $body{$i};
1000 }
1001 stripComments( $i, $j, $body );
1002 if ( $body{$i} <> '>' )
1003 $i++;
1004 }
1005 if ( $styleblk <> '' )
1006 $ret .= "<style>$styleblk";
1007 break;
1008 case 'BODY':
1009 if ( $title <> '' )
1010 $ret .= '<b>' . _("Title:") . " </b>$title<br>\n";
1011 $ret .= "<TABLE";
1012 $i += 5;
1013 if (! isset($base)) {
1014 $base = '';
1015 }
1016 $ret .= stripEvent( $i, $j, $body, $id, $base );
1017 $ret .= " bgcolor=$bgcolor width=\"100%\"><tr>";
1018 if ( $leftmargin <> '' )
1019 $ret .= "<td width=$leftmargin>&nbsp;</td>";
1020 $ret .= '<td>';
1021 if (strtolower($bgcolor) == 'ffffff' ||
1022 strtolower($bgcolor) == '#ffffff')
1023 $ret .= '<font color=#000000>';
1024 break;
1025 case 'BASE':
1026 $i += 5;
1027 $base = '';
1028 while ( !isNoSep( $body{$i} ) &&
1029 $i < $j ) {
1030 $i++;
1031 }
1032 if ( strcasecmp( substr( $base, 0, 4 ), 'href' ) ) {
1033 $i += 5;
1034 while ( !isNoSep( $body{$i} ) &&
1035 $i < $j ) {
1036 $i++;
1037 }
1038 while ( $body{$i} <> '>' &&
1039 $i < $j ) {
1040 if ( $body{$i} <> '"' ) {
1041 $base .= $body{$i};
1042 }
1043 $i++;
1044 }
1045 // Debuging $ret .= "<!-- base == $base -->";
1046 if ( strcasecmp( substr( $base, 0, 4 ), 'file' ) <> 0 ) {
1047 $ret .= "\n<BASE HREF=\"$base\">\n";
1048 }
1049 }
1050 break;
1051 case '/BOD':
1052 $ret .= '</font></td></tr></TABLE>';
1053 $i += 6;
1054 break;
1055 default:
1056 // Following tags can contain some event handler, lets search it
1057 stripComments( $i, $j, $body );
1058 if (! isset($base)) {
1059 $base = '';
1060 }
1061 $ret .= stripEvent( $i, $j, $body, $id, $base ) . '>';
1062 // $ret .= "<!-- $tag detected -->";
1063 }
1064 } else {
1065 $ret .= $body{$i};
1066 }
1067 $i++;
1068 }
1069
1070 return( "\n\n<!-- HTML Output ahead -->\n" .
1071 $ret .
1072 /* Base is illegal within HTML
1073 "\n<!-- END of HTML Output --><base href=\"".
1074 get_location() . '/'.
1075 "\">\n\n" );
1076 */
1077 "\n<!-- END of HTML Output -->\n\n" );
1078 }
1079
1080 function isNoSep( $char ) {
1081
1082 switch( $char ) {
1083 case ' ':
1084 case "\n":
1085 case "\t":
1086 case "\r":
1087 case '>':
1088 case '"':
1089 return( FALSE );
1090 break;
1091 default:
1092 return( TRUE );
1093 }
1094
1095 }
1096
1097 /*
1098 The following function is usefull to remove extra data that can cause
1099 html not to display properly. Especialy with MS stuff.
1100 */
1101
1102 function stripComments( &$i, $j, &$body ) {
1103
1104 while ( $body{$i}.$body{$i+1}.$body{$i+2}.$body{$i+3} == '<!--' &&
1105 $i < $j ) {
1106 $i += 5;
1107 while ( $body{$i-2}.$body{$i-1}.$body{$i} <> '-->' &&
1108 $i < $j )
1109 $i++;
1110 $i++;
1111 }
1112
1113 return;
1114
1115 }
1116
1117 /* Gets the style data of a specific style */
1118
1119 function getStyleData( &$i, $j, &$body ) {
1120
1121 // We skip spaces
1122 while ( $body{$i} <> '>' && !isNoSep( $body{$i} ) &&
1123 $i < $j ) {
1124 $i++;
1125 }
1126 // And get the color
1127 $ret = '';
1128 while ( isNoSep( $body{$i} ) &&
1129 $i < $j ) {
1130 $ret .= $body{$i};
1131 $i++;
1132 }
1133
1134 return( $ret );
1135 }
1136
1137 /*
1138 Private function for strip_dangerous_tag. Look for event based coded and "remove" it
1139 change on with no (onload -> noload)
1140 */
1141
1142 function stripEvent( &$i, $j, &$body, $id, $base ) {
1143
1144 global $message, $base_uri, $has_unsafe_images, $view_unsafe_images;
1145
1146 $ret = '';
1147
1148 while ( $body{$i} <> '>' &&
1149 $i < $j ) {
1150 $etg = strtolower($body{$i}.$body{$i+1}.$body{$i+2});
1151 switch( $etg ) {
1152 case 'src':
1153 // This is probably a src specification
1154 $k = $i + 3;
1155 while( !isNoSep( $body{$k} )) {
1156 $k++;
1157 }
1158 if ( $body{$k} == '=' ) {
1159 /* It is indeed */
1160 $k++;
1161 while( !isNoSep( $body{$k} ) &&
1162 $k < $j ) {
1163 $k++;
1164 }
1165 $src = '';
1166 while ( $body{$k} <> '>' && isNoSep( $body{$k} ) &&
1167 $k < $j ) {
1168 $src .= $body{$k};
1169 $k++;
1170 }
1171 $k++;
1172 while( !isNoSep( $body{$k} ) &&
1173 $k < $j ) {
1174 $k++;
1175 }
1176 $k++;
1177 if ( strtolower( substr( $src, 0, 4 ) ) == 'cid:' ) {
1178 $src = substr( $src, 4 );
1179 $src = "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=" .
1180 urlencode( $message->header->mailbox ) .
1181 "&amp;passed_ent_id=" . find_ent_id( $src, $message );
1182 } else if ( strtolower( substr( $src, 0, 4 ) ) <> 'http' ||
1183 stristr( $src, $base_uri ) ) {
1184 /* Javascript and local urls goes out */
1185 if (!$view_unsafe_images) {
1186 $src = '../images/' . _("sec_remove_eng.png");
1187 }
1188 $has_unsafe_images = 1;
1189 }
1190 $ret .= 'src="' . $src . '" ';
1191 $i = $k - 2;
1192 } else {
1193 $ret .= 'src';
1194 $i = $i + 3;
1195 }
1196
1197 break;
1198 case '../':
1199 // Retrolinks are not allowed without a base because they mess with SM security
1200 if ( $base == '' ) {
1201 $i += 2;
1202 } else {
1203 $ret .= '.';
1204 }
1205 break;
1206 case 'cid':
1207 // Internal link
1208 $k = $i-1;
1209 if ( $body{$i+3} == ':') {
1210 $i +=4;
1211 $name = '';
1212 while ( isNoSep( $body{$i} ) &&
1213 $i < $j ) {
1214 $name .= $body{$i++};
1215 }
1216 if ( $name <> '' ) {
1217 $ret .= "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=" .
1218 urlencode( $message->header->mailbox ) .
1219 "&amp;passed_ent_id=" . find_ent_id( $name, $message );
1220 if ( $body{$k} == '"' )
1221 $ret .= '" ';
1222 else
1223 $ret .= ' ';
1224 }
1225 if ( $body{$i} == '>' )
1226 $i -= 1;
1227 }
1228 break;
1229 case ' on':
1230 case "\non":
1231 case "\ron":
1232 case "\ton":
1233 $ret .= ' no';
1234 $i += 2;
1235 break;
1236 case 'pt:':
1237 if ( strcasecmp( $body{$i-4}.$body{$i-3}.$body{$i-2}.$body{$i-1}.$body{$i}.$body{$i+1}.$body{$i+2}, 'script:') == 0 ) {
1238 $ret .= '_no/';
1239 } else {
1240 $ret .= $etg;
1241 }
1242 $i += 2;
1243 break;
1244 default:
1245 $ret .= $body{$i};
1246 }
1247 $i++;
1248 }
1249 return( $ret );
1250 }
1251
1252
1253 /* This function trys to locate the entity_id of a specific mime element */
1254
1255 function find_ent_id( $id, $message ) {
1256
1257 $ret = '';
1258 for ($i=0; $ret == '' && $i < count($message->entities); $i++) {
1259
1260 if ( $message->entities[$i]->header->entity_id == '' ) {
1261 $ret = find_ent_id( $id, $message->entities[$i] );
1262 } else {
1263 if ( strcasecmp( $message->entities[$i]->header->id, $id ) == 0 )
1264 $ret = $message->entities[$i]->header->entity_id;
1265 }
1266
1267 }
1268
1269 return( $ret );
1270
1271 }
1272 ?>