7c8940c1b9a36e95bc711779acea2b31648241c7
[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 $endline = fgets($imap_stream, 1024);
63 $read = strtolower($read);
64
65 if ($debug_mime) echo "<tt>$read</tt><br><br>";
66 // isolate the body structure and remove beginning and end parenthesis
67 $read = trim(substr ($read, strpos($read, "bodystructure") + 13));
68 $read = trim(substr ($read, 0, -1));
69 $end = mime_match_parenthesis(0, $read);
70 while ($end == strlen($read)-1) {
71 $read = trim(substr ($read, 0, -1));
72 $read = trim(substr ($read, 1));
73 $end = mime_match_parenthesis(0, $read);
74 }
75
76 if ($debug_mime) echo "<tt>$read</tt><br><br>";
77
78 $msg = mime_parse_structure ($read, 0);
79 $msg->header = $header;
80 return $msg;
81 }
82
83 // this starts the parsing of a particular structure. It is called recursively,
84 // so it can be passed different structures. It returns an object of type
85 // $message.
86 // First, it checks to see if it is a multipart message. If it is, then it
87 // handles that as it sees is necessary. If it is just a regular entity,
88 // then it parses it and adds the necessary header information (by calling out
89 // to mime_get_elements()
90 function mime_parse_structure ($structure, $ent_id) {
91 global $debug_mime;
92 if ($debug_mime) echo "<font color=008800><tt>START: mime_parse_structure()</tt></font><br>";
93 $msg = new message();
94 if (substr($structure, 0, 1) == "(") {
95 $ent_id = mime_new_element_level($ent_id);
96 $start = $end = -1;
97 if ($debug_mime) echo "<br><font color=0000aa><tt>$structure</tt></font><br>";
98 do {
99 if ($debug_mime) echo "<font color=008800><tt>Found entity...</tt></font><br>";
100 $start = $end+1;
101 $end = mime_match_parenthesis ($start, $structure);
102
103 $element = substr($structure, $start+1, ($end - $start)-1);
104 $ent_id = mime_increment_id ($ent_id);
105 $newmsg = mime_parse_structure ($element, $ent_id);
106 $msg->addEntity ($newmsg);
107 } while (substr($structure, $end+1, 1) == "(");
108 } else {
109 // parse the elements
110 if ($debug_mime) echo "<br><font color=0000aa><tt>$structure</tt></font><br>";
111 $msg = mime_get_element (&$structure, $msg, $ent_id);
112 if ($debug_mime) echo "<br>";
113 }
114 return $msg;
115 if ($debug_mime) echo "<font color=008800><tt>&nbsp;&nbsp;END: mime_parse_structure()</tt></font><br>";
116 }
117
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.
121 function mime_increment_id ($id) {
122 global $debug_mime;
123 if (strpos($id, ".")) {
124 $first = substr($id, 0, strrpos($id, "."));
125 $last = substr($id, strrpos($id, ".")+1);
126 $last++;
127 $new = $first . "." .$last;
128 } else {
129 $new = $id + 1;
130 }
131 if ($debug_mime) echo "<b>INCREMENT: $new</b><br>";
132 return $new;
133 }
134
135 // See comment for mime_increment_id().
136 // This adds another level on to the entity_id changing 1.3 to 1.3.0
137 // NOTE: 1.3.0 is not a valid element ID. It MUST be incremented
138 // before it can be used. I left it this way so as not to have
139 // to make a special case if it is the first entity_id. It
140 // always increments it, and that works fine.
141 function mime_new_element_level ($id) {
142 if (!$id) $id = 0;
143 else $id = $id . ".0";
144
145 return $id;
146 }
147
148 function mime_get_element (&$structure, $msg, $ent_id) {
149 global $debug_mime;
150 $elem_num = 1;
151 $msg->header = new msg_header();
152 $msg->header->entity_id = $ent_id;
153
154 while (strlen($structure) > 0) {
155 $structure = trim($structure);
156 $char = substr($structure, 0, 1);
157
158 if (substr($structure, 0, 3) == "nil") {
159 $text = "";
160 $structure = substr($structure, 3);
161 } else if ($char == "\"") {
162 // loop through until we find the matching quote, and return that as a string
163 $pos = 1;
164 $char = substr($structure, $pos, 1);
165 while ($char != "\"" && $pos < strlen($structure)) {
166 $text .= $char;
167 $pos++;
168 $char = substr($structure, $pos, 1);
169 }
170 $structure = substr($structure, strlen($text) + 2);
171 } else if ($char == "(") {
172 // comment me
173 $end = mime_match_parenthesis (0, $structure);
174 $sub = substr($structure, 1, $end-1);
175 $properties = mime_get_props($properties, $sub);
176 $structure = substr($structure, strlen($sub) + 2);
177 } else {
178 // loop through until we find a space or an end parenthesis
179 $pos = 0;
180 $char = substr($structure, $pos, 1);
181 while ($char != " " && $char != ")" && $pos < strlen($structure)) {
182 $text .= $char;
183 $pos++;
184 $char = substr($structure, $pos, 1);
185 }
186 $structure = substr($structure, strlen($text));
187 }
188 if ($debug_mime) echo "<tt>$elem_num : $text</tt><br>";
189
190 // This is where all the text parts get put into the header
191 switch ($elem_num) {
192 case 1:
193 $msg->header->type0 = $text;
194 if ($debug_mime) echo "<tt>type0 = $text</tt><br>";
195 break;
196 case 2:
197 $msg->header->type1 = $text;
198 if ($debug_mime) echo "<tt>type1 = $text</tt><br>";
199 break;
200 case 5:
201 $msg->header->description = $text;
202 if ($debug_mime) echo "<tt>description = $text</tt><br>";
203 break;
204 case 6:
205 $msg->header->encoding = $text;
206 if ($debug_mime) echo "<tt>encoding = $text</tt><br>";
207 break;
208 case 7:
209 $msg->header->size = $text;
210 if ($debug_mime) echo "<tt>size = $text</tt><br>";
211 break;
212 default:
213 if ($msg->header->type0 == "text" && $elem_num == 8) {
214 // This is a plain text message, so lets get the number of lines
215 // that it contains.
216 $msg->header->num_lines = $text;
217 if ($debug_mime) echo "<tt>num_lines = $text</tt><br>";
218
219 } else if ($msg->header->type0 == "message" && $msg->header->type1 == "rfc822" && $elem_num == 8) {
220 // This is an encapsulated message, so lets start all over again and
221 // parse this message adding it on to the existing one.
222 $structure = trim($structure);
223 if (substr($structure, 0, 1) == "(") {
224 $e = mime_match_parenthesis (0, $structure);
225 $structure = substr($structure, 0, $e);
226 $structure = substr($structure, 1);
227 $m = mime_parse_structure($structure, $msg->header->entity_id);
228
229 // the following conditional is there to correct a bug that wasn't
230 // incrementing the entity IDs correctly because of the special case
231 // that message/rfc822 is. This fixes it fine.
232 if (substr($structure, 1, 1) != "(")
233 $m->header->entity_id = mime_increment_id(mime_new_element_level($ent_id));
234
235 // Now we'll go through and reformat the results.
236 if ($m->entities) {
237 for ($i=0; $i < count($m->entities); $i++) {
238 $msg->addEntity($m->entities[$i]);
239 }
240 } else {
241 $msg->addEntity($m);
242 }
243 $structure = "";
244 }
245 }
246 break;
247 }
248 $elem_num++;
249 $text = "";
250 }
251 // loop through the additional properties and put those in the various headers
252 if ($msg->header->type0 != "message") {
253 for ($i=0; $i < count($properties); $i++) {
254 $msg->header->{$properties[$i]["name"]} = $properties[$i]["value"];
255 if ($debug_mime) echo "<tt>".$properties[$i]["name"]." = " . $properties[$i]["value"] . "</tt><br>";
256 }
257 }
258
259 return $msg;
260 }
261
262 // I did most of the MIME stuff yesterday (June 20, 2000), but I couldn't
263 // figure out how to do this part, so I decided to go to bed. I woke up
264 // in the morning and had a flash of insight. I went to the white-board
265 // and scribbled it out, then spent a bit programming it, and this is the
266 // result. Nothing complicated, but I think my brain was fried yesterday.
267 // Funny how that happens some times.
268 //
269 // This gets properties in a nested parenthesisized list. For example,
270 // this would get passed something like: ("attachment" ("filename" "luke.tar.gz"))
271 // This returns an array called $props with all paired up properties.
272 // It ignores the "attachment" for now, maybe that should change later
273 // down the road. In this case, what is returned is:
274 // $props[0]["name"] = "filename";
275 // $props[0]["value"] = "luke.tar.gz";
276 function mime_get_props ($props, $structure) {
277 global $debug_mime;
278 while (strlen($structure) > 0) {
279 $structure = trim($structure);
280 $char = substr($structure, 0, 1);
281
282 if ($char == "\"") {
283 $pos = 1;
284 $char = substr($structure, $pos, 1);
285 while ($char != "\"" && $pos < strlen($structure)) {
286 $tmp .= $char;
287 $pos++;
288 $char = substr($structure, $pos, 1);
289 }
290 $structure = trim(substr($structure, strlen($tmp) + 2));
291 $char = substr($structure, 0, 1);
292
293 if ($char == "\"") {
294 $pos = 1;
295 $char = substr($structure, $pos, 1);
296 while ($char != "\"" && $pos < strlen($structure)) {
297 $value .= $char;
298 $pos++;
299 $char = substr($structure, $pos, 1);
300 }
301 $structure = trim(substr($structure, strlen($tmp) + 2));
302
303 $k = count($props);
304 $props[$k]["name"] = $tmp;
305 $props[$k]["value"] = $value;
306 } else if ($char == "(") {
307 $end = mime_match_parenthesis (0, $structure);
308 $sub = substr($structure, 1, $end-1);
309 $props = mime_get_props($props, $sub);
310 $structure = substr($structure, strlen($sub) + 2);
311 }
312 return $props;
313 } else if ($char == "(") {
314 $end = mime_match_parenthesis (0, $structure);
315 $sub = substr($structure, 1, $end-1);
316 $props = mime_get_props($props, $sub);
317 $structure = substr($structure, strlen($sub) + 2);
318 return $props;
319 } else {
320 return $props;
321 }
322 }
323 }
324
325 // Matches parenthesis. It will return the position of the matching
326 // parenthesis in $structure. For instance, if $structure was:
327 // ("text" "plain" ("val1name", "1") nil ... )
328 // x x
329 // then this would return 42 to match up those two.
330 function mime_match_parenthesis ($pos, $structure) {
331 $char = substr($structure, $pos, 1);
332
333 // ignore all extra characters
334 while ($pos < strlen($structure)) {
335 $pos++;
336 $char = substr($structure, $pos, 1);
337 if ($char == ")") {
338 return $pos;
339 } else if ($char == "(") {
340 $pos = mime_match_parenthesis ($pos, $structure);
341 }
342 }
343 }
344
345 function mime_fetch_body ($imap_stream, $id, $ent_id) {
346 // do a bit of error correction. If we couldn't find the entity id, just guess
347 // that it is the first one. That is usually the case anyway.
348 if (!$ent_id) $ent_id = 1;
349
350 fputs ($imap_stream, "a001 FETCH $id BODY[$ent_id]\r\n");
351 $topline = fgets ($imap_stream, 1024);
352 $size = substr ($topline, strpos($topline, "{")+1);
353 $size = substr ($size, 0, strpos($size, "}"));
354 $read = fread ($imap_stream, $size);
355 return $read;
356 }
357
358 /* -[ END MIME DECODING ]----------------------------------------------------------- */
359
360
361
362 /** This is the first function called. It decides if this is a multipart
363 message or if it should be handled as a single entity
364 **/
365 function decodeMime ($imap_stream, $body, $header) {
366 global $username, $key, $imapServerAddress, $imapPort;
367 return mime_structure ($imap_stream, $header);
368 }
369
370 // This is here for debugging purposese. It will print out a list
371 // of all the entity IDs that are in the $message object.
372 function listEntities ($message) {
373 if ($message) {
374 if ($message->header->entity_id)
375 echo "<tt>" . $message->header->entity_id . " : " . $message->header->type0 . "/" . $message->header->type1 . "<br>";
376 for ($i = 0; $message->entities[$i]; $i++) {
377 $msg = listEntities($message->entities[$i], $ent_id);
378 if ($msg)
379 return $msg;
380 }
381 }
382 }
383
384 // returns a $message object for a particular entity id
385 function getEntity ($message, $ent_id) {
386 if ($message) {
387 if ($message->header->entity_id == $ent_id && strlen($ent_id) == strlen($message->header->entity_id)) {
388 return $message;
389 } else {
390 for ($i = 0; $message->entities[$i]; $i++) {
391 $msg = getEntity ($message->entities[$i], $ent_id);
392 if ($msg)
393 return $msg;
394 }
395 }
396 }
397 }
398
399 // figures out what entity to display and returns the $message object
400 // for that entity.
401 function findDisplayEntity ($message) {
402 if ($message) {
403 if ($message->header->type0 == "text") {
404 if ($message->header->type1 == "plain" ||
405 $message->header->type1 == "html") {
406 return $message->header->entity_id;
407 }
408 } else {
409 for ($i=0; $message->entities[$i]; $i++) {
410 return findDisplayEntity($message->entities[$i]);
411 }
412 }
413 }
414 }
415
416 /** This returns a parsed string called $body. That string can then
417 be displayed as the actual message in the HTML. It contains
418 everything needed, including HTML Tags, Attachments at the
419 bottom, etc.
420 **/
421 function formatBody($imap_stream, $message, $color, $wrap_at) {
422 // this if statement checks for the entity to show as the
423 // primary message. To add more of them, just put them in the
424 // order that is their priority.
425 global $username, $key, $imapServerAddress, $imapPort;
426
427 $id = $message->header->id;
428 $urlmailbox = urlencode($message->header->mailbox);
429
430 // Get the right entity and redefine message to be this entity
431 $ent_num = findDisplayEntity ($message);
432 $body_message = getEntity($message, $ent_num);
433
434 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
435 $body = decodeBody($body, $body_message->header->encoding);
436
437 // If there are other types that shouldn't be formatted, add
438 // them here
439 if ($message->header->type1 != "html") {
440 $body = translateText($body, $wrap_at, $body_message->header->charset);
441 }
442
443 $body .= "<SMALL><CENTER><A HREF=\"../src/download.php?absolute_dl=true&passed_id=$id&passed_ent_id=$ent_num&mailbox=$urlmailbox\">". _("Download this as a file") ."</A></CENTER><BR></SMALL>";
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 $filename = decodeHeader($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 $body .= "&nbsp;(<a href=\"../src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&passed_ent_id=$ent\">"._("download")."</a>)\n";
481 $body .= "</TT><BR>";
482 $num++;
483 }
484 return $body;
485 } else {
486 for ($i = 0; $i < count($message->entities); $i++) {
487 $body .= formatAttachments ($message->entities[$i], $ent_id, $mailbox, $id);
488 }
489 return $body;
490 }
491 }
492 }
493
494
495 /** this function decodes the body depending on the encoding type. **/
496 function decodeBody($body, $encoding) {
497 $body = str_replace("\r\n", "\n", $body);
498 $encoding = strtolower($encoding);
499
500 if ($encoding == "quoted-printable") {
501 $body = quoted_printable_decode($body);
502
503 while (ereg("=\n", $body))
504 $body = ereg_replace ("=\n", "", $body);
505 } else if ($encoding == "base64") {
506 $body = base64_decode($body);
507 }
508
509 // All other encodings are returned raw.
510 return $body;
511 }
512
513
514 // This functions decode strings that is encoded according to
515 // RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
516 function decodeHeader ($string) {
517 if (eregi('=\?([^?]+)\?(q|b)\?([^?]+)\?=',
518 $string, $res)) {
519 if (ucfirst($res[2]) == "B") {
520 $replace = base64_decode($res[3]);
521 } else {
522 $replace = ereg_replace("_", " ", $res[3]);
523 $replace = quoted_printable_decode($replace);
524 }
525
526 $replace = charset_decode ($res[1], $replace);
527
528 $string = eregi_replace
529 ('=\?([^?]+)\?(q|b)\?([^?]+)\?=',
530 $replace, $string);
531 // In case there should be more encoding in the string: recurse
532 return (decodeHeader($string));
533 } else
534 return ($string);
535 }
536
537 // Encode a string according to RFC 1522 for use in headers if it
538 // contains 8-bit characters or anything that looks like it should
539 // be encoded.
540 function encodeHeader ($string) {
541 global $default_charset;
542
543 // Encode only if the string contains 8-bit characters or =?
544 if (ereg("([\200-\377])|=\\?", $string)) {
545 $newstring = "=?$default_charset?Q?";
546
547 // First the special characters
548 $string = str_replace("=", "=3D", $string);
549 $string = str_replace("?", "=3F", $string);
550 $string = str_replace("_", "=5F", $string);
551 $string = str_replace(" ", "_", $string);
552
553
554 while (ereg("([\200-\377])", $string, $regs)) {
555 $replace = $regs[1];
556 $insert = "=" . strtoupper(bin2hex($replace));
557 $string = str_replace($replace, $insert, $string);
558 }
559
560 $newstring = "=?$default_charset?Q?".$string."?=";
561
562 return $newstring;
563 }
564
565 return $string;
566 }
567
568 ?>