updated todo, trying to fix this stupid multiple log email problem,
[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;
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 $read = strtolower($read);
63
64 if ($debug_mime) echo "<tt>$read</tt><br><br>";
65 // isolate the body structure and remove beginning and end parenthesis
66 $read = trim(substr ($read, strpos($read, "bodystructure") + 13));
67 $read = trim(substr ($read, 0, -1));
68 $end = mime_match_parenthesis(0, $read);
69 while ($end == strlen($read)-1) {
70 $read = trim(substr ($read, 0, -1));
71 $read = trim(substr ($read, 1));
72 $end = mime_match_parenthesis(0, $read);
73 }
74
75 if ($debug_mime) echo "<tt>$read</tt><br><br>";
76
77 $msg = mime_parse_structure ($read);
78 $msg->header = $header;
79 return $msg;
80 }
81
82 // this starts the parsing of a particular structure. It is called recursively,
83 // so it can be passed different structures. It returns an object of type
84 // $message.
85 // First, it checks to see if it is a multipart message. If it is, then it
86 // handles that as it sees is necessary. If it is just a regular entity,
87 // then it parses it and adds the necessary header information (by calling out
88 // to mime_get_elements()
89 function mime_parse_structure ($structure, $ent_id) {
90 global $debug_mime;
91 if ($debug_mime) echo "<font color=008800><tt>START: mime_parse_structure()</tt></font><br>";
92 $msg = new message();
93 if (substr($structure, 0, 1) == "(") {
94 $ent_id = mime_new_element_level($ent_id);
95 $start = $end = -1;
96 if ($debug_mime) echo "<br><font color=0000aa><tt>$structure</tt></font><br>";
97 do {
98 if ($debug_mime) echo "<font color=008800><tt>Found entity...</tt></font><br>";
99 $start = $end+1;
100 $end = mime_match_parenthesis ($start, $structure);
101
102 $element = substr($structure, $start+1, ($end - $start)-1);
103 $ent_id = mime_increment_id ($ent_id);
104 $newmsg = mime_parse_structure ($element, $ent_id);
105 $msg->addEntity ($newmsg);
106 } while (substr($structure, $end+1, 1) == "(");
107 } else {
108 // parse the elements
109 if ($debug_mime) echo "<br><font color=0000aa><tt>$structure</tt></font><br>";
110 $msg = mime_get_element (&$structure, $msg, $ent_id);
111 if ($debug_mime) echo "<br>";
112 }
113 return $msg;
114 if ($debug_mime) echo "<font color=008800><tt>&nbsp;&nbsp;END: mime_parse_structure()</tt></font><br>";
115 }
116
117 // Increments the element ID. An element id can look like any of
118 // the following: 1, 1.2, 4.3.2.4.1, etc. This function increments
119 // the last number of the element id, changing 1.2 to 1.3.
120 function mime_increment_id ($id) {
121 global $debug_mime;
122 if (strpos($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 = $id + 1;
129 }
130 if ($debug_mime) echo "<b>INCREMENT: $new</b><br>";
131 return $new;
132 }
133
134 // See comment for mime_increment_id().
135 // This adds another level on to the entity_id changing 1.3 to 1.3.0
136 // NOTE: 1.3.0 is not a valid element ID. It MUST be incremented
137 // before it can be used. I left it this way so as not to have
138 // to make a special case if it is the first entity_id. It
139 // always increments it, and that works fine.
140 function mime_new_element_level ($id) {
141 if (!$id) $id = 0;
142 else $id = $id . ".0";
143
144 return $id;
145 }
146
147 function mime_get_element (&$structure, $msg, $ent_id) {
148 global $debug_mime;
149 $elem_num = 1;
150 $msg->header = new msg_header();
151 $msg->header->entity_id = $ent_id;
152
153 while (strlen($structure) > 0) {
154 $structure = trim($structure);
155 $char = substr($structure, 0, 1);
156
157 if (substr($structure, 0, 3) == "nil") {
158 $text = "";
159 $structure = substr($structure, 3);
160 } else if ($char == "\"") {
161 // loop through until we find the matching quote, and return that as a string
162 $pos = 1;
163 $char = substr($structure, $pos, 1);
164 while ($char != "\"" && $pos < strlen($structure)) {
165 $text .= $char;
166 $pos++;
167 $char = substr($structure, $pos, 1);
168 }
169 $structure = substr($structure, strlen($text) + 2);
170 } else if ($char == "(") {
171 // comment me
172 $end = mime_match_parenthesis (0, $structure);
173 $sub = substr($structure, 1, $end-1);
174 $properties = mime_get_props($properties, $sub);
175 $structure = substr($structure, strlen($sub) + 2);
176 } else {
177 // loop through until we find a space or an end parenthesis
178 $pos = 0;
179 $char = substr($structure, $pos, 1);
180 while ($char != " " && $char != ")" && $pos < strlen($structure)) {
181 $text .= $char;
182 $pos++;
183 $char = substr($structure, $pos, 1);
184 }
185 $structure = substr($structure, strlen($text));
186 }
187 if ($debug_mime) echo "<tt>$elem_num : $text</tt><br>";
188
189 // This is where all the text parts get put into the header
190 switch ($elem_num) {
191 case 1:
192 $msg->header->type0 = $text;
193 if ($debug_mime) echo "<tt>type0 = $text</tt><br>";
194 break;
195 case 2:
196 $msg->header->type1 = $text;
197 if ($debug_mime) echo "<tt>type1 = $text</tt><br>";
198 break;
199 case 5:
200 $msg->header->description = $text;
201 if ($debug_mime) echo "<tt>description = $text</tt><br>";
202 break;
203 case 6:
204 $msg->header->encoding = $text;
205 if ($debug_mime) echo "<tt>encoding = $text</tt><br>";
206 break;
207 case 7:
208 $msg->header->size = $text;
209 if ($debug_mime) echo "<tt>size = $text</tt><br>";
210 break;
211 default:
212 if ($msg->header->type0 == "text" && $elem_num == 8) {
213 // This is a plain text message, so lets get the number of lines
214 // that it contains.
215 $msg->header->num_lines = $text;
216 if ($debug_mime) echo "<tt>num_lines = $text</tt><br>";
217
218 } else if ($msg->header->type0 == "message" && $msg->header->type1 == "rfc822" && $elem_num == 8) {
219 // This is an encapsulated message, so lets start all over again and
220 // parse this message adding it on to the existing one.
221 $structure = trim($structure);
222 if (substr($structure, 0, 1) == "(") {
223 $e = mime_match_parenthesis (0, $structure);
224 $structure = substr($structure, 0, $e);
225 $structure = substr($structure, 1);
226 $m = mime_parse_structure($structure, $msg->header->entity_id);
227
228 // the following conditional is there to correct a bug that wasn't
229 // incrementing the entity IDs correctly because of the special case
230 // that message/rfc822 is. This fixes it fine.
231 if (substr($structure, 1, 1) != "(")
232 $m->header->entity_id = mime_increment_id(mime_new_element_level($ent_id));
233
234 // Now we'll go through and reformat the results.
235 if ($m->entities) {
236 for ($i=0; $i < count($m->entities); $i++) {
237 $msg->addEntity($m->entities[$i]);
238 }
239 } else {
240 $msg->addEntity($m);
241 }
242 $structure = "";
243 }
244 }
245 break;
246 }
247 $elem_num++;
248 $text = "";
249 }
250 // loop through the additional properties and put those in the various headers
251 if ($msg->header->type0 != "message") {
252 for ($i=0; $i < count($properties); $i++) {
253 $msg->header->{$properties[$i]["name"]} = $properties[$i]["value"];
254 if ($debug_mime) echo "<tt>".$properties[$i]["name"]." = " . $properties[$i]["value"] . "</tt><br>";
255 }
256 }
257 return $msg;
258 }
259
260 // I did most of the MIME stuff yesterday (June 20, 2000), but I couldn't
261 // figure out how to do this part, so I decided to go to bed. I woke up
262 // in the morning and had a flash of insight. I went to the white-board
263 // and scribbled it out, then spent a bit programming it, and this is the
264 // result. Nothing complicated, but I think my brain was fried yesterday.
265 // Funny how that happens some times.
266 //
267 // This gets properties in a nested parenthesisized list. For example,
268 // this would get passed something like: ("attachment" ("filename" "luke.tar.gz"))
269 // This returns an array called $props with all paired up properties.
270 // It ignores the "attachment" for now, maybe that should change later
271 // down the road. In this case, what is returned is:
272 // $props[0]["name"] = "filename";
273 // $props[0]["value"] = "luke.tar.gz";
274 function mime_get_props ($props, $structure) {
275 global $debug_mime;
276 while (strlen($structure) > 0) {
277 $structure = trim($structure);
278 $char = substr($structure, 0, 1);
279
280 if ($char == "\"") {
281 $pos = 1;
282 $char = substr($structure, $pos, 1);
283 while ($char != "\"" && $pos < strlen($structure)) {
284 $tmp .= $char;
285 $pos++;
286 $char = substr($structure, $pos, 1);
287 }
288 $structure = trim(substr($structure, strlen($tmp) + 2));
289 $char = substr($structure, 0, 1);
290
291 if ($char == "\"") {
292 $pos = 1;
293 $char = substr($structure, $pos, 1);
294 while ($char != "\"" && $pos < strlen($structure)) {
295 $value .= $char;
296 $pos++;
297 $char = substr($structure, $pos, 1);
298 }
299 $structure = trim(substr($structure, strlen($tmp) + 2));
300
301 $k = count($props);
302 $props[$k]["name"] = $tmp;
303 $props[$k]["value"] = $value;
304 } else if ($char == "(") {
305 $end = mime_match_parenthesis (0, $structure);
306 $sub = substr($structure, 1, $end-1);
307 $props = mime_get_props($props, $sub);
308 $structure = substr($structure, strlen($sub) + 2);
309 }
310 return $props;
311 } else if ($char == "(") {
312 $end = mime_match_parenthesis (0, $structure);
313 $sub = substr($structure, 1, $end-1);
314 $props = mime_get_props($props, $sub);
315 $structure = substr($structure, strlen($sub) + 2);
316 return $props;
317 } else {
318 return $props;
319 }
320 }
321 }
322
323 // Matches parenthesis. It will return the position of the matching
324 // parenthesis in $structure. For instance, if $structure was:
325 // ("text" "plain" ("val1name", "1") nil ... )
326 // x x
327 // then this would return 42 to match up those two.
328 function mime_match_parenthesis ($pos, $structure) {
329 $char = substr($structure, $pos, 1);
330
331 // ignore all extra characters
332 while ($pos < strlen($structure)) {
333 $pos++;
334 $char = substr($structure, $pos, 1);
335 if ($char == ")") {
336 return $pos;
337 } else if ($char == "(") {
338 $pos = mime_match_parenthesis ($pos, $structure);
339 }
340 }
341 }
342
343 function mime_fetch_body ($imap_stream, $id, $ent_id) {
344 // do a bit of error correction. If we couldn't find the entity id, just guess
345 // that it is the first one. That is usually the case anyway.
346 if (!$ent_id) $ent_id = 1;
347
348 fputs ($imap_stream, "a001 FETCH $id BODY[$ent_id]\r\n");
349 $topline = fgets ($imap_stream, 1024);
350 $size = substr ($topline, strpos($topline, "{")+1);
351 $size = substr ($size, 0, strpos($size, "}"));
352 $read = fread ($imap_stream, $size);
353 return $read;
354 }
355
356 /* -[ END MIME DECODING ]----------------------------------------------------------- */
357
358
359
360 /** This is the first function called. It decides if this is a multipart
361 message or if it should be handled as a single entity
362 **/
363 function decodeMime ($body, $header) {
364 global $username, $key, $imapServerAddress, $imapPort;
365 $imap_stream = sqimap_login($username, $key, $imapServerAddress, $imapPort, 0);
366 sqimap_mailbox_select($imap_stream, $header->mailbox);
367
368 return mime_structure ($imap_stream, $header);
369 }
370
371 // This is here for debugging purposese. It will print out a list
372 // of all the entity IDs that are in the $message object.
373 function listEntities ($message) {
374 if ($message) {
375 if ($message->header->entity_id)
376 echo "<tt>" . $message->header->entity_id . " : " . $message->header->type0 . "/" . $message->header->type1 . "<br>";
377 for ($i = 0; $message->entities[$i]; $i++) {
378 $msg = listEntities($message->entities[$i], $ent_id);
379 if ($msg)
380 return $msg;
381 }
382 }
383 }
384
385 // returns a $message object for a particular entity id
386 function getEntity ($message, $ent_id) {
387 if ($message) {
388 if ($message->header->entity_id == $ent_id && strlen($ent_id) == strlen($message->header->entity_id)) {
389 return $message;
390 } else {
391 for ($i = 0; $message->entities[$i]; $i++) {
392 $msg = getEntity ($message->entities[$i], $ent_id);
393 if ($msg)
394 return $msg;
395 }
396 }
397 }
398 }
399
400 // figures out what entity to display and returns the $message object
401 // for that entity.
402 function findDisplayEntity ($message) {
403 if ($message) {
404 if ($message->header->type0 == "text") {
405 if ($message->header->type1 == "plain" ||
406 $message->header->type1 == "html") {
407 return $message->header->entity_id;
408 }
409 } else {
410 for ($i=0; $message->entities[$i]; $i++) {
411 return findDisplayEntity($message->entities[$i]);
412 }
413 }
414 }
415 }
416
417 /** This returns a parsed string called $body. That string can then
418 be displayed as the actual message in the HTML. It contains
419 everything needed, including HTML Tags, Attachments at the
420 bottom, etc.
421 **/
422 function formatBody($message, $color, $wrap_at) {
423 // this if statement checks for the entity to show as the
424 // primary message. To add more of them, just put them in the
425 // order that is their priority.
426 global $username, $key, $imapServerAddress, $imapPort;
427
428 $id = $message->header->id;
429 $urlmailbox = urlencode($message->header->mailbox);
430
431 $imap_stream = sqimap_login($username, $key, $imapServerAddress, $imapPort, 0);
432 sqimap_mailbox_select($imap_stream, $message->header->mailbox);
433
434 $ent_num = findDisplayEntity ($message);
435 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
436
437 // If there are other types that shouldn't be formatted, add
438 // them here
439 if ($message->header->type1 != "html") {
440 $body = translateText($body, $wrap_at, $charset);
441 }
442
443 $body .= "<BR><SMALL><CENTER><A HREF=\"../src/download.php?absolute_dl=true&passed_id=$id&passed_ent_id=$ent_num&mailbox=$urlmailbox\">". _("Download this as a file") ."</A></CENTER><BR></SMALL>";
444
445 /** Display the ATTACHMENTS: message if there's more than one part **/
446 if ($message->entities) {
447 $body .= "<TABLE WIDTH=100% CELLSPACING=0 CELLPADDING=4 BORDER=0><TR><TD BGCOLOR=\"$color[0]\">";
448 $body .= "<TT><B>ATTACHMENTS:</B></TT>";
449 $body .= "</TD></TR><TR><TD BGCOLOR=\"$color[0]\">";
450 $num = 0;
451
452 /** make this recurisve at some point **/
453 $body .= formatAttachments ($message, $ent_num, $message->header->mailbox, $id);
454 $body .= "</TD></TR></TABLE>";
455 }
456 return $body;
457 }
458
459 // A recursive function that returns a list of attachments with links
460 // to where to download these attachments
461 function formatAttachments ($message, $ent_id, $mailbox, $id) {
462 if ($message) {
463 if (!$message->entities) {
464 $type0 = strtolower($message->header->type0);
465 $type1 = strtolower($message->header->type1);
466
467 if ($message->header->entity_id != $ent_id) {
468 $filename = $message->header->filename;
469 if (trim($filename) == "") {
470 $display_filename = "untitled-".$message->header->entity_id;
471 } else {
472 $display_filename = $filename;
473 }
474
475 $urlMailbox = urlencode($mailbox);
476 $ent = urlencode($message->header->entity_id);
477 $body .= "<TT>&nbsp;&nbsp;&nbsp;<A HREF=\"../src/download.php?passed_id=$id&mailbox=$urlMailbox&passed_ent_id=$ent\">" . $display_filename . "</A>&nbsp;&nbsp;(TYPE: $type0/$type1)";
478 if ($message->header->description)
479 $body .= "&nbsp;&nbsp;<b>" . htmlspecialchars($message->header->description)."</b>";
480 $body .= "</TT><BR>";
481 $num++;
482 }
483 return $body;
484 } else {
485 for ($i = 0; $i < count($message->entities); $i++) {
486 $body .= formatAttachments ($message->entities[$i], $ent_id, $mailbox, $id);
487 }
488 return $body;
489 }
490 }
491 }
492
493
494 /** this function decodes the body depending on the encoding type. **/
495 function decodeBody($body, $encoding) {
496 $encoding = strtolower($encoding);
497
498 if ($encoding == "quoted-printable") {
499 $body = quoted_printable_decode($body);
500
501 while (ereg("=\n", $body))
502 $body = ereg_replace ("=\n", "", $body);
503 } else if ($encoding == "base64") {
504 $body = base64_decode($body);
505 }
506
507 // All other encodings are returned raw.
508 return $body;
509 }
510
511
512 // This functions decode strings that is encoded according to
513 // RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
514 function decodeHeader ($string) {
515 if (eregi('=\?([^?]+)\?(q|b)\?([^?]+)\?=',
516 $string, $res)) {
517 if (ucfirst($res[2]) == "B") {
518 $replace = base64_decode($res[3]);
519 } else {
520 $replace = ereg_replace("_", " ", $res[3]);
521 $replace = quoted_printable_decode($replace);
522 }
523
524 $replace = charset_decode ($res[1], $replace);
525
526 $string = eregi_replace
527 ('=\?([^?]+)\?(q|b)\?([^?]+)\?=',
528 $replace, $string);
529 // In case there should be more encoding in the string: recurse
530 return (decodeHeader($string));
531 } else
532 return ($string);
533 }
534
535 // Encode a string according to RFC 1522 for use in headers if it
536 // contains 8-bit characters
537 function encodeHeader ($string) {
538 global $default_charset;
539
540 // Encode only if the string contains 8-bit characters
541 if (ereg("[\200-\377]", $string)) {
542 $newstring = "=?$default_charset?Q?";
543 $newstring .= str_replace(" ", "_", $string);
544
545 while (ereg("([\200-\377])", $newstring, $regs)) {
546 $replace = $regs[1];
547 $insert = "=" . bin2hex($replace);
548 $newstring = str_replace($replace, $insert, $newstring);
549 }
550
551 $newstring .= "?=";
552
553 return $newstring;
554 }
555
556 return $string;
557 }
558
559 ?>