f56f54d86db02e3c159750b50852ad48444ea982
[squirrelmail.git] / functions / mime.php
1 <?php
2 /** mime.php
3 **
4 ** This contains the functions necessary to detect and decode MIME
5 ** messages.
6 **
7 **/
8
9 $debug_mime = false;
10 $mime_php = true;
11
12 if (!isset($i18n_php))
13 include "../functions/i18n.php";
14 if (!isset($imap_php))
15 include "../functions/imap.php";
16 if (!isset($config_php))
17 include "../config/config.php";
18
19
20 /** Setting up the objects that have the structure for the message **/
21
22 class msg_header {
23 /** msg_header contains generic variables for values that **/
24 /** could be in a header. **/
25
26 var $type0, $type1, $boundary, $charset, $encoding;
27 var $to, $from, $date, $cc, $bcc, $reply_to, $subject;
28 var $id, $mailbox, $description;
29 var $entity_id, $message_id, $charset;
30 }
31
32 class message {
33 /** message is the object that contains messages. It is a recursive
34 object in that through the $entities variable, it can contain
35 more objects of type message. See documentation in mime.txt for
36 a better description of how this works.
37 **/
38 var $header;
39 var $entities;
40
41 function addEntity ($msg) {
42 $this->entities[count($this->entities)] = $msg;
43 }
44 }
45
46
47
48 /* --------------------------------------------------------------------------------- */
49 /* MIME DECODING */
50 /* --------------------------------------------------------------------------------- */
51
52 // This function gets the structure of a message and stores it in the "message" class.
53 // It will return this object for use with all relevant header information and
54 // fully parsed into the standard "message" object format.
55 function mime_structure ($imap_stream, $header) {
56 global $debug_mime;
57 sqimap_messages_flag ($imap_stream, $header->id, $header->id, "Seen");
58
59 $id = $header->id;
60 fputs ($imap_stream, "a001 FETCH $id BODYSTRUCTURE\r\n");
61 $read = fgets ($imap_stream, 10000);
62 $response = substr($read, 0, 4);
63 while ($response != "a001") {
64 $bodystructure .= $read;
65 $read = fgets ($imap_stream, 10000);
66 $response = substr($read, 0, 4);
67 }
68 // $read = strtolower($bodystructure);
69 $read = $bodystructure;
70
71 if ($debug_mime) echo "<tt>$read</tt><br><br>\n";
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, $read);
80 }
81
82 if ($debug_mime) echo "<tt>$read</tt><br><br>\n";
83
84 $msg = mime_parse_structure ($read, 0);
85 $msg->header = $header;
86 return $msg;
87 }
88
89 // this starts the parsing of a particular structure. It is called recursively,
90 // so it can be passed different structures. It returns an object of type
91 // $message.
92 // First, it checks to see if it is a multipart message. If it is, then it
93 // handles that as it sees is necessary. If it is just a regular entity,
94 // then it parses it and adds the necessary header information (by calling out
95 // to mime_get_elements()
96 function mime_parse_structure ($structure, $ent_id) {
97 global $debug_mime;
98 if ($debug_mime) echo "<font color=008800><tt>START: mime_parse_structure()</tt></font><br>\n";
99 $msg = new message();
100 if (substr($structure, 0, 1) == "(") {
101 $ent_id = mime_new_element_level($ent_id);
102 $start = $end = -1;
103 if ($debug_mime) echo "<br><font color=0000aa><tt>$structure</tt></font><br>";
104 do {
105 if ($debug_mime) echo "<font color=008800><tt>Found entity...</tt></font><br>";
106 $start = $end+1;
107 $end = mime_match_parenthesis ($start, $structure);
108
109 $element = substr($structure, $start+1, ($end - $start)-1);
110 $ent_id = mime_increment_id ($ent_id);
111 $newmsg = mime_parse_structure ($element, $ent_id);
112 $msg->addEntity ($newmsg);
113 } while (substr($structure, $end+1, 1) == "(");
114 } else {
115 // parse the elements
116 if ($debug_mime) echo "<br><font color=0000aa><tt>$structure</tt></font><br>";
117 $msg = mime_get_element (&$structure, $msg, $ent_id);
118 if ($debug_mime) echo "<br>";
119 }
120 return $msg;
121 if ($debug_mime) echo "<font color=008800><tt>&nbsp;&nbsp;END: mime_parse_structure()</tt></font><br>";
122 }
123
124 // Increments the element ID. An element id can look like any of
125 // the following: 1, 1.2, 4.3.2.4.1, etc. This function increments
126 // the last number of the element id, changing 1.2 to 1.3.
127 function mime_increment_id ($id) {
128 global $debug_mime;
129 if (strpos($id, ".")) {
130 $first = substr($id, 0, strrpos($id, "."));
131 $last = substr($id, strrpos($id, ".")+1);
132 $last++;
133 $new = $first . "." .$last;
134 } else {
135 $new = $id + 1;
136 }
137 if ($debug_mime) echo "<b>INCREMENT: $new</b><br>";
138 return $new;
139 }
140
141 // See comment for mime_increment_id().
142 // This adds another level on to the entity_id changing 1.3 to 1.3.0
143 // NOTE: 1.3.0 is not a valid element ID. It MUST be incremented
144 // before it can be used. I left it this way so as not to have
145 // to make a special case if it is the first entity_id. It
146 // always increments it, and that works fine.
147 function mime_new_element_level ($id) {
148 if (!$id) $id = 0;
149 else $id = $id . ".0";
150
151 return $id;
152 }
153
154 function mime_get_element (&$structure, $msg, $ent_id) {
155 global $debug_mime;
156 $elem_num = 1;
157 $msg->header = new msg_header();
158 $msg->header->entity_id = $ent_id;
159
160 while (strlen($structure) > 0) {
161 $structure = trim($structure);
162 $char = substr($structure, 0, 1);
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 $char = substr($structure, $pos, 1);
171 while ($char != "\"" && $pos < strlen($structure)) {
172 $text .= $char;
173 $pos++;
174 $char = substr($structure, $pos, 1);
175 }
176 $structure = substr($structure, strlen($text) + 2);
177 } else if ($char == "(") {
178 // comment me
179 $end = mime_match_parenthesis (0, $structure);
180 $sub = substr($structure, 1, $end-1);
181 $properties = mime_get_props($properties, $sub);
182 $structure = substr($structure, strlen($sub) + 2);
183 } else {
184 // loop through until we find a space or an end parenthesis
185 $pos = 0;
186 $char = substr($structure, $pos, 1);
187 while ($char != " " && $char != ")" && $pos < strlen($structure)) {
188 $text .= $char;
189 $pos++;
190 $char = substr($structure, $pos, 1);
191 }
192 $structure = substr($structure, strlen($text));
193 }
194 if ($debug_mime) echo "<tt>$elem_num : $text</tt><br>";
195
196 // This is where all the text parts get put into the header
197 switch ($elem_num) {
198 case 1:
199 $msg->header->type0 = strtolower($text);
200 if ($debug_mime) echo "<tt>type0 = ".strtolower($text)."</tt><br>";
201 break;
202 case 2:
203 $msg->header->type1 = strtolower($text);
204 if ($debug_mime) echo "<tt>type1 = ".strtolower($text)."</tt><br>";
205 break;
206 case 5:
207 $msg->header->description = $text;
208 if ($debug_mime) echo "<tt>description = $text</tt><br>";
209 break;
210 case 6:
211 $msg->header->encoding = strtolower($text);
212 if ($debug_mime) echo "<tt>encoding = ".strtolower($text)."</tt><br>";
213 break;
214 case 7:
215 $msg->header->size = $text;
216 if ($debug_mime) echo "<tt>size = $text</tt><br>";
217 break;
218 default:
219 if ($msg->header->type0 == "text" && $elem_num == 8) {
220 // This is a plain text message, so lets get the number of lines
221 // that it contains.
222 $msg->header->num_lines = $text;
223 if ($debug_mime) echo "<tt>num_lines = $text</tt><br>";
224
225 } else if ($msg->header->type0 == "message" && $msg->header->type1 == "rfc822" && $elem_num == 8) {
226 // This is an encapsulated message, so lets start all over again and
227 // parse this message adding it on to the existing one.
228 $structure = trim($structure);
229 if (substr($structure, 0, 1) == "(") {
230 $e = mime_match_parenthesis (0, $structure);
231 $structure = substr($structure, 0, $e);
232 $structure = substr($structure, 1);
233 $m = mime_parse_structure($structure, $msg->header->entity_id);
234
235 // the following conditional is there to correct a bug that wasn't
236 // incrementing the entity IDs correctly because of the special case
237 // that message/rfc822 is. This fixes it fine.
238 if (substr($structure, 1, 1) != "(")
239 $m->header->entity_id = mime_increment_id(mime_new_element_level($ent_id));
240
241 // Now we'll go through and reformat the results.
242 if ($m->entities) {
243 for ($i=0; $i < count($m->entities); $i++) {
244 $msg->addEntity($m->entities[$i]);
245 }
246 } else {
247 $msg->addEntity($m);
248 }
249 $structure = "";
250 }
251 }
252 break;
253 }
254 $elem_num++;
255 $text = "";
256 }
257 // loop through the additional properties and put those in the various headers
258 if ($msg->header->type0 != "message") {
259 for ($i=0; $i < count($properties); $i++) {
260 $msg->header->{$properties[$i]["name"]} = $properties[$i]["value"];
261 if ($debug_mime) echo "<tt>".$properties[$i]["name"]." = " . $properties[$i]["value"] . "</tt><br>";
262 }
263 }
264
265 return $msg;
266 }
267
268 // I did most of the MIME stuff yesterday (June 20, 2000), but I couldn't
269 // figure out how to do this part, so I decided to go to bed. I woke up
270 // in the morning and had a flash of insight. I went to the white-board
271 // and scribbled it out, then spent a bit programming it, and this is the
272 // result. Nothing complicated, but I think my brain was fried yesterday.
273 // Funny how that happens some times.
274 //
275 // This gets properties in a nested parenthesisized list. For example,
276 // this would get passed something like: ("attachment" ("filename" "luke.tar.gz"))
277 // This returns an array called $props with all paired up properties.
278 // It ignores the "attachment" for now, maybe that should change later
279 // down the road. In this case, what is returned is:
280 // $props[0]["name"] = "filename";
281 // $props[0]["value"] = "luke.tar.gz";
282 function mime_get_props ($props, $structure) {
283 global $debug_mime;
284 while (strlen($structure) > 0) {
285 $structure = trim($structure);
286 $char = substr($structure, 0, 1);
287
288 if ($char == "\"") {
289 $pos = 1;
290 $char = substr($structure, $pos, 1);
291 while ($char != "\"" && $pos < strlen($structure)) {
292 $tmp .= $char;
293 $pos++;
294 $char = substr($structure, $pos, 1);
295 }
296 $structure = trim(substr($structure, strlen($tmp) + 2));
297 $char = substr($structure, 0, 1);
298
299 if ($char == "\"") {
300 $pos = 1;
301 $char = substr($structure, $pos, 1);
302 while ($char != "\"" && $pos < strlen($structure)) {
303 $value .= $char;
304 $pos++;
305 $char = substr($structure, $pos, 1);
306 }
307 $structure = trim(substr($structure, strlen($tmp) + 2));
308
309 $k = count($props);
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);
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 // Matches parenthesis. It will return the position of the matching
332 // parenthesis in $structure. For instance, if $structure was:
333 // ("text" "plain" ("val1name", "1") nil ... )
334 // x x
335 // then this would return 42 to match up those two.
336 function mime_match_parenthesis ($pos, $structure) {
337 $char = substr($structure, $pos, 1);
338
339 // ignore all extra characters
340 while ($pos < strlen($structure)) {
341 $pos++;
342 $char = substr($structure, $pos, 1);
343 if ($char == ")") {
344 return $pos;
345 } else if ($char == "(") {
346 $pos = mime_match_parenthesis ($pos, $structure);
347 }
348 }
349 }
350
351 function mime_fetch_body ($imap_stream, $id, $ent_id) {
352 // do a bit of error correction. If we couldn't find the entity id, just guess
353 // that it is the first one. That is usually the case anyway.
354 if (!$ent_id) $ent_id = 1;
355
356 fputs ($imap_stream, "a001 FETCH $id BODY[$ent_id]\r\n");
357 $topline = fgets ($imap_stream, 1024);
358 $size = substr ($topline, strpos($topline, "{")+1);
359 $size = substr ($size, 0, strpos($size, "}"));
360 $read = fread ($imap_stream, $size);
361 return $read;
362 }
363
364 /* -[ END MIME DECODING ]----------------------------------------------------------- */
365
366
367
368 /** This is the first function called. It decides if this is a multipart
369 message or if it should be handled as a single entity
370 **/
371 function decodeMime ($imap_stream, $body, $header) {
372 global $username, $key, $imapServerAddress, $imapPort;
373 return mime_structure ($imap_stream, $header);
374 }
375
376 // This is here for debugging purposese. It will print out a list
377 // of all the entity IDs that are in the $message object.
378 function listEntities ($message) {
379 if ($message) {
380 if ($message->header->entity_id)
381 echo "<tt>" . $message->header->entity_id . " : " . $message->header->type0 . "/" . $message->header->type1 . "<br>";
382 for ($i = 0; $message->entities[$i]; $i++) {
383 $msg = listEntities($message->entities[$i], $ent_id);
384 if ($msg)
385 return $msg;
386 }
387 }
388 }
389
390 // returns a $message object for a particular entity id
391 function getEntity ($message, $ent_id) {
392 if ($message) {
393 if ($message->header->entity_id == $ent_id && strlen($ent_id) == strlen($message->header->entity_id)) {
394 return $message;
395 } else {
396 for ($i = 0; $message->entities[$i]; $i++) {
397 $msg = getEntity ($message->entities[$i], $ent_id);
398 if ($msg)
399 return $msg;
400 }
401 }
402 }
403 }
404
405 // figures out what entity to display and returns the $message object
406 // for that entity.
407 function findDisplayEntity ($message) {
408 if ($message) {
409 if ($message->header->type0 == "text") {
410 if ($message->header->type1 == "plain" ||
411 $message->header->type1 == "html") {
412 return $message->header->entity_id;
413 }
414 } else {
415 for ($i=0; $message->entities[$i]; $i++) {
416 return findDisplayEntity($message->entities[$i]);
417 }
418 }
419 }
420 }
421
422 /** This returns a parsed string called $body. That string can then
423 be displayed as the actual message in the HTML. It contains
424 everything needed, including HTML Tags, Attachments at the
425 bottom, etc.
426 **/
427 function formatBody($imap_stream, $message, $color, $wrap_at) {
428 // this if statement checks for the entity to show as the
429 // primary message. To add more of them, just put them in the
430 // order that is their priority.
431 global $startMessage, $username, $key, $imapServerAddress, $imapPort;
432
433 $id = $message->header->id;
434 $urlmailbox = urlencode($message->header->mailbox);
435
436 // Get the right entity and redefine message to be this entity
437 $ent_num = findDisplayEntity ($message);
438 $body_message = getEntity($message, $ent_num);
439
440 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
441 $body = decodeBody($body, $body_message->header->encoding);
442
443 // If there are other types that shouldn't be formatted, add
444 // them here
445 if ($message->header->type1 != "html") {
446 $body = translateText($body, $wrap_at, $body_message->header->charset);
447 }
448
449 $body .= "<SMALL><CENTER><A HREF=\"../src/download.php?absolute_dl=true&passed_id=$id&passed_ent_id=$ent_num&mailbox=$urlmailbox\">". _("Download this as a file") ."</A></CENTER><BR></SMALL>";
450
451 /** Display the ATTACHMENTS: message if there's more than one part **/
452 if ($message->entities) {
453 $body .= "</TD></TR></TABLE>";
454 $body .= "<TABLE WIDTH=100% CELLSPACING=0 CELLPADDING=4 BORDER=0><TR><TD BGCOLOR=\"$color[0]\">";
455 $body .= "<TT><B>ATTACHMENTS:</B></TT>";
456 $body .= "</TD></TR><TR><TD BGCOLOR=\"$color[0]\">";
457 $num = 0;
458
459 $body .= formatAttachments ($message, $ent_num, $message->header->mailbox, $id);
460 $body .= "</TD></TR></TABLE>";
461 } else {
462 $body .= "</TD></TR></TABLE>";
463 }
464 return $body;
465 }
466
467 // A recursive function that returns a list of attachments with links
468 // to where to download these attachments
469 function formatAttachments ($message, $ent_id, $mailbox, $id) {
470 global $where, $what;
471 global $startMessage;
472 if ($message) {
473 if (!$message->entities) {
474 $type0 = strtolower($message->header->type0);
475 $type1 = strtolower($message->header->type1);
476
477 if ($message->header->entity_id != $ent_id) {
478 $filename = decodeHeader($message->header->filename);
479 if (trim($filename) == "") {
480 $display_filename = "untitled-".$message->header->entity_id;
481 } else {
482 $display_filename = $filename;
483 }
484
485 $urlMailbox = urlencode($mailbox);
486 $ent = urlencode($message->header->entity_id);
487 if ($where && $what) {
488 // from a search
489 $body .= "<TT>&nbsp;&nbsp;&nbsp;<A HREF=\"../src/download.php?startMessage=$startMessage&where=".urlencode($where)."&what=".urlencode($what)."&passed_id=$id&mailbox=$urlMailbox&passed_ent_id=$ent\">" . $display_filename . "</A>&nbsp;&nbsp;(TYPE: $type0/$type1)";
490 } else {
491 $body .= "<TT>&nbsp;&nbsp;&nbsp;<A HREF=\"../src/download.php?startMessage=$startMessage&passed_id=$id&mailbox=$urlMailbox&passed_ent_id=$ent\">" . $display_filename . "</A>&nbsp;&nbsp;(TYPE: $type0/$type1)";
492 }
493 if ($message->header->description)
494 $body .= "&nbsp;&nbsp;<b>" . htmlspecialchars($message->header->description)."</b>";
495 $body .= "&nbsp;(<a href=\"../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&passed_ent_id=$ent\">"._("download")."</a>)\n";
496 do_hook("attachment $type0/$type1");
497 $body .= "</TT><BR>";
498 $num++;
499 }
500 return $body;
501 } else {
502 for ($i = 0; $i < count($message->entities); $i++) {
503 $body .= formatAttachments ($message->entities[$i], $ent_id, $mailbox, $id);
504 }
505 return $body;
506 }
507 }
508 }
509
510
511 /** this function decodes the body depending on the encoding type. **/
512 function decodeBody($body, $encoding) {
513 $body = str_replace("\r\n", "\n", $body);
514 $encoding = strtolower($encoding);
515
516 if ($encoding == "quoted-printable") {
517 $body = quoted_printable_decode($body);
518
519 while (ereg("=\n", $body))
520 $body = ereg_replace ("=\n", "", $body);
521 } else if ($encoding == "base64") {
522 $body = base64_decode($body);
523 }
524
525 // All other encodings are returned raw.
526 return $body;
527 }
528
529
530 // This functions decode strings that is encoded according to
531 // RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
532 function decodeHeader ($string) {
533 if (eregi('=\?([^?]+)\?(q|b)\?([^?]+)\?=',
534 $string, $res)) {
535 if (ucfirst($res[2]) == "B") {
536 $replace = base64_decode($res[3]);
537 } else {
538 $replace = ereg_replace("_", " ", $res[3]);
539 // Convert lowercase Quoted Printable to uppercase for
540 // quoted_printable_decode to understand it.
541 while (ereg("(=([0-9][a-f])|([a-f][0-9])|([a-f][0-9]))", $replace, $res)) {
542 $replace = str_replace($res[1], strtoupper($res[1]), $replace);
543 }
544 $replace = quoted_printable_decode($replace);
545 }
546
547 $replace = charset_decode ($res[1], $replace);
548
549 $string = eregi_replace
550 ('=\?([^?]+)\?(q|b)\?([^?]+)\?=',
551 $replace, $string);
552 // In case there should be more encoding in the string: recurse
553 return (decodeHeader($string));
554 } else
555 return ($string);
556 }
557
558 // Encode a string according to RFC 1522 for use in headers if it
559 // contains 8-bit characters or anything that looks like it should
560 // be encoded.
561 function encodeHeader ($string) {
562 global $default_charset;
563
564 // Encode only if the string contains 8-bit characters or =?
565 if (ereg("([\200-\377])|=\\?", $string)) {
566 $newstring = "=?$default_charset?Q?";
567
568 // First the special characters
569 $string = str_replace("=", "=3D", $string);
570 $string = str_replace("?", "=3F", $string);
571 $string = str_replace("_", "=5F", $string);
572 $string = str_replace(" ", "_", $string);
573
574
575 while (ereg("([\200-\377])", $string, $regs)) {
576 $replace = $regs[1];
577 $insert = "=" . strtoupper(bin2hex($replace));
578 $string = str_replace($replace, $insert, $string);
579 }
580
581 $newstring = "=?$default_charset?Q?".$string."?=";
582
583 return $newstring;
584 }
585
586 return $string;
587 }
588
589 ?>