59177427 |
1 | <?php |
2ba13803 |
2 | |
35586184 |
3 | /** |
4 | * mime.php |
5 | * |
15e6162e |
6 | * Copyright (c) 1999-2002 The SquirrelMail Project Team |
35586184 |
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 | */ |
b74ba498 |
14 | |
35586184 |
15 | require_once('../functions/imap.php'); |
16 | require_once('../functions/attachment_common.php'); |
8beafbbc |
17 | |
35586184 |
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. **/ |
b74ba498 |
22 | |
88cb1b4d |
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; |
35586184 |
28 | } |
b74ba498 |
29 | |
451f74a2 |
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 | **/ |
77b88425 |
36 | var $header = '', $entities = array(); |
37 | |
451f74a2 |
38 | function addEntity ($msg) { |
39 | $this->entities[] = $msg; |
40 | } |
41 | } |
8beafbbc |
42 | |
451f74a2 |
43 | /* --------------------------------------------------------------------------------- */ |
44 | /* MIME DECODING */ |
45 | /* --------------------------------------------------------------------------------- */ |
b74ba498 |
46 | |
451f74a2 |
47 | /* This function gets the structure of a message and stores it in the "message" class. |
48 | * It will return this object for use with all relevant header information and |
49 | * fully parsed into the standard "message" object format. |
50 | */ |
51 | function mime_structure ($imap_stream, $header) { |
52 | |
451f74a2 |
53 | $ssid = sqimap_session_id(); |
54 | $lsid = strlen( $ssid ); |
55 | $id = $header->id; |
56 | fputs ($imap_stream, "$ssid FETCH $id BODYSTRUCTURE\r\n"); |
57 | // |
58 | // This should use sqimap_read_data instead of reading it itself |
59 | // |
77b88425 |
60 | $read = fgets ($imap_stream, 9216); |
451f74a2 |
61 | $bodystructure = ''; |
62 | while ( substr($read, 0, $lsid) <> $ssid && |
63 | !feof( $imap_stream ) ) { |
64 | $bodystructure .= $read; |
77b88425 |
65 | $read = fgets ($imap_stream, 9216); |
451f74a2 |
66 | } |
67 | $read = $bodystructure; |
77b88425 |
68 | |
451f74a2 |
69 | // isolate the body structure and remove beginning and end parenthesis |
70 | $read = trim(substr ($read, strpos(strtolower($read), 'bodystructure') + 13)); |
71 | $read = trim(substr ($read, 0, -1)); |
72 | $end = mime_match_parenthesis(0, $read); |
73 | while ($end == strlen($read)-1) { |
74 | $read = trim(substr ($read, 0, -1)); |
75 | $read = trim(substr ($read, 1)); |
76 | $end = mime_match_parenthesis(0, $read); |
77 | } |
77b88425 |
78 | |
451f74a2 |
79 | $msg = mime_parse_structure ($read, 0); |
80 | $msg->header = $header; |
77b88425 |
81 | |
451f74a2 |
82 | return( $msg ); |
83 | } |
b74ba498 |
84 | |
451f74a2 |
85 | /* this starts the parsing of a particular structure. It is called recursively, |
86 | * so it can be passed different structures. It returns an object of type |
87 | * $message. |
88 | * First, it checks to see if it is a multipart message. If it is, then it |
89 | * handles that as it sees is necessary. If it is just a regular entity, |
90 | * then it parses it and adds the necessary header information (by calling out |
91 | * to mime_get_elements() |
92 | */ |
93 | function mime_parse_structure ($structure, $ent_id) { |
94 | |
95 | $msg = new message(); |
96 | if ($structure{0} == '(') { |
97 | $ent_id = mime_new_element_level($ent_id); |
98 | $start = $end = -1; |
99 | do { |
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 ($structure{$end+1} == '('); |
108 | } else { |
109 | // parse the elements |
110 | $msg = mime_get_element ($structure, $msg, $ent_id); |
111 | } |
112 | return $msg; |
113 | } |
e4a256af |
114 | |
451f74a2 |
115 | /* Increments the element ID. An element id can look like any of |
116 | * the following: 1, 1.2, 4.3.2.4.1, etc. This function increments |
117 | * the last number of the element id, changing 1.2 to 1.3. |
118 | */ |
119 | function mime_increment_id ($id) { |
120 | |
77b88425 |
121 | if (strpos($id, '.')) { |
122 | $first = substr($id, 0, strrpos($id, '.')); |
123 | $last = substr($id, strrpos($id, '.')+1); |
451f74a2 |
124 | $last++; |
77b88425 |
125 | $new = $first . '.' .$last; |
451f74a2 |
126 | } else { |
127 | $new = $id + 1; |
128 | } |
77b88425 |
129 | |
451f74a2 |
130 | return $new; |
131 | } |
132 | |
133 | /* |
134 | * See comment for mime_increment_id(). |
135 | * This adds another level on to the entity_id changing 1.3 to 1.3.0 |
136 | * NOTE: 1.3.0 is not a valid element ID. It MUST be incremented |
137 | * before it can be used. I left it this way so as not to have |
138 | * to make a special case if it is the first entity_id. It |
139 | * always increments it, and that works fine. |
140 | */ |
141 | function mime_new_element_level ($id) { |
142 | |
77b88425 |
143 | if (!$id) { |
144 | $id = 0; |
145 | } else { |
146 | $id = $id . '.0'; |
147 | } |
451f74a2 |
148 | |
77b88425 |
149 | return( $id ); |
451f74a2 |
150 | } |
151 | |
152 | function mime_get_element (&$structure, $msg, $ent_id) { |
153 | |
154 | $elem_num = 1; |
155 | $msg->header = new msg_header(); |
156 | $msg->header->entity_id = $ent_id; |
157 | $properties = array(); |
158 | |
159 | while (strlen($structure) > 0) { |
160 | $structure = trim($structure); |
161 | $char = $structure{0}; |
162 | |
163 | if (strtolower(substr($structure, 0, 3)) == 'nil') { |
164 | $text = ''; |
165 | $structure = substr($structure, 3); |
166 | } else if ($char == '"') { |
167 | // loop through until we find the matching quote, and return that as a string |
168 | $pos = 1; |
169 | $text = ''; |
170 | while ( ($char = $structure{$pos} ) <> '"' && $pos < strlen($structure)) { |
171 | $text .= $char; |
172 | $pos++; |
173 | } |
174 | $structure = substr($structure, strlen($text) + 2); |
175 | } else if ($char == '(') { |
176 | // comment me |
177 | $end = mime_match_parenthesis (0, $structure); |
178 | $sub = substr($structure, 1, $end-1); |
179 | $properties = mime_get_props($properties, $sub); |
180 | $structure = substr($structure, strlen($sub) + 2); |
181 | } else { |
182 | // loop through until we find a space or an end parenthesis |
183 | $pos = 0; |
184 | $char = $structure{$pos}; |
185 | $text = ''; |
186 | while ($char != ' ' && $char != ')' && $pos < strlen($structure)) { |
187 | $text .= $char; |
188 | $pos++; |
189 | $char = $structure{$pos}; |
190 | } |
191 | $structure = substr($structure, strlen($text)); |
192 | } |
193 | |
194 | // This is where all the text parts get put into the header |
195 | switch ($elem_num) { |
196 | case 1: |
197 | $msg->header->type0 = strtolower($text); |
198 | break; |
199 | case 2: |
200 | $msg->header->type1 = strtolower($text); |
201 | break; |
202 | case 4: // Id |
203 | // Invisimail enclose images with <> |
204 | $msg->header->id = str_replace( '<', '', str_replace( '>', '', $text ) ); |
205 | break; |
206 | case 5: |
207 | $msg->header->description = $text; |
208 | break; |
209 | case 6: |
210 | $msg->header->encoding = strtolower($text); |
211 | break; |
212 | case 7: |
213 | $msg->header->size = $text; |
214 | break; |
215 | default: |
216 | if ($msg->header->type0 == 'text' && $elem_num == 8) { |
217 | // This is a plain text message, so lets get the number of lines |
218 | // that it contains. |
219 | $msg->header->num_lines = $text; |
220 | |
221 | } else if ($msg->header->type0 == 'message' && $msg->header->type1 == 'rfc822' && $elem_num == 8) { |
222 | // This is an encapsulated message, so lets start all over again and |
223 | // parse this message adding it on to the existing one. |
224 | $structure = trim($structure); |
225 | if ( $structure{0} == '(' ) { |
226 | $e = mime_match_parenthesis (0, $structure); |
227 | $structure = substr($structure, 0, $e); |
228 | $structure = substr($structure, 1); |
229 | $m = mime_parse_structure($structure, $msg->header->entity_id); |
230 | |
231 | // the following conditional is there to correct a bug that wasn't |
232 | // incrementing the entity IDs correctly because of the special case |
233 | // that message/rfc822 is. This fixes it fine. |
234 | if (substr($structure, 1, 1) != '(') |
235 | $m->header->entity_id = mime_increment_id(mime_new_element_level($ent_id)); |
236 | |
237 | // Now we'll go through and reformat the results. |
238 | if ($m->entities) { |
239 | for ($i=0; $i < count($m->entities); $i++) { |
240 | $msg->addEntity($m->entities[$i]); |
241 | } |
242 | } else { |
243 | $msg->addEntity($m); |
244 | } |
245 | $structure = ""; |
246 | } |
247 | } |
248 | break; |
249 | } |
250 | $elem_num++; |
251 | $text = ""; |
252 | } |
253 | // loop through the additional properties and put those in the various headers |
f0c4dc12 |
254 | // if ($msg->header->type0 != 'message') { |
451f74a2 |
255 | for ($i=0; $i < count($properties); $i++) { |
256 | $msg->header->{$properties[$i]['name']} = $properties[$i]['value']; |
257 | } |
f0c4dc12 |
258 | // } |
451f74a2 |
259 | |
260 | return $msg; |
261 | } |
262 | |
263 | /* |
264 | * I did most of the MIME stuff yesterday (June 20, 2000), but I couldn't |
265 | * figure out how to do this part, so I decided to go to bed. I woke up |
266 | * in the morning and had a flash of insight. I went to the white-board |
267 | * and scribbled it out, then spent a bit programming it, and this is the |
268 | * result. Nothing complicated, but I think my brain was fried yesterday. |
269 | * Funny how that happens some times. |
270 | * |
271 | * This gets properties in a nested parenthesisized list. For example, |
272 | * this would get passed something like: ("attachment" ("filename" "luke.tar.gz")) |
273 | * This returns an array called $props with all paired up properties. |
274 | * It ignores the "attachment" for now, maybe that should change later |
275 | * down the road. In this case, what is returned is: |
276 | * $props[0]["name"] = "filename"; |
277 | * $props[0]["value"] = "luke.tar.gz"; |
278 | */ |
279 | function mime_get_props ($props, $structure) { |
280 | |
281 | while (strlen($structure) > 0) { |
282 | $structure = trim($structure); |
283 | $char = $structure{0}; |
284 | |
285 | if ($char == '"') { |
286 | $pos = 1; |
287 | $tmp = ''; |
288 | while ( ( $char = $structure{$pos} ) != '"' && |
289 | $pos < strlen($structure)) { |
290 | $tmp .= $char; |
291 | $pos++; |
292 | } |
293 | $structure = trim(substr($structure, strlen($tmp) + 2)); |
294 | $char = $structure{0}; |
295 | |
296 | if ($char == '"') { |
297 | $pos = 1; |
298 | $value = ''; |
299 | while ( ( $char = $structure{$pos} ) != '"' && |
300 | $pos < strlen($structure) ) { |
301 | $value .= $char; |
302 | $pos++; |
303 | } |
304 | $structure = trim(substr($structure, strlen($tmp) + 2)); |
305 | |
306 | $k = count($props); |
307 | $props[$k]['name'] = strtolower($tmp); |
308 | $props[$k]['value'] = $value; |
309 | } else if ($char == '(') { |
310 | $end = mime_match_parenthesis (0, $structure); |
311 | $sub = substr($structure, 1, $end-1); |
312 | if (! isset($props)) |
313 | $props = array(); |
314 | $props = mime_get_props($props, $sub); |
315 | $structure = substr($structure, strlen($sub) + 2); |
316 | } |
317 | return $props; |
318 | } else if ($char == '(') { |
319 | $end = mime_match_parenthesis (0, $structure); |
320 | $sub = substr($structure, 1, $end-1); |
321 | $props = mime_get_props($props, $sub); |
322 | $structure = substr($structure, strlen($sub) + 2); |
323 | return $props; |
324 | } else { |
325 | return $props; |
326 | } |
327 | } |
328 | } |
329 | |
330 | /* |
331 | * Matches parenthesis. It will return the position of the matching |
332 | * parenthesis in $structure. For instance, if $structure was: |
333 | * ("text" "plain" ("val1name", "1") nil ... ) |
334 | * x x |
335 | * then this would return 42 to match up those two. |
336 | */ |
337 | function mime_match_parenthesis ($pos, $structure) { |
338 | |
339 | $j = strlen( $structure ); |
340 | |
341 | // ignore all extra characters |
342 | // If inside of a string, skip string -- Boundary IDs and other |
343 | // things can have ) in them. |
344 | if ( $structure{$pos} != '(' ) { |
345 | return( $j ); |
346 | } |
347 | |
348 | while ( $pos < $j ) { |
349 | $pos++; |
350 | if ($structure{$pos} == ')') { |
8beafbbc |
351 | return $pos; |
451f74a2 |
352 | } elseif ($structure{$pos} == '"') { |
b74ba498 |
353 | $pos++; |
451f74a2 |
354 | while ( $structure{$pos} != '"' && |
355 | $pos < $j ) { |
356 | if (substr($structure, $pos, 2) == '\\"') { |
b74ba498 |
357 | $pos++; |
451f74a2 |
358 | } elseif (substr($structure, $pos, 2) == '\\\\') { |
b74ba498 |
359 | $pos++; |
451f74a2 |
360 | } |
b74ba498 |
361 | $pos++; |
5ffe5a7e |
362 | } |
451f74a2 |
363 | } elseif ( $structure{$pos} == '(' ) { |
8beafbbc |
364 | $pos = mime_match_parenthesis ($pos, $structure); |
451f74a2 |
365 | } |
366 | } |
367 | echo _("Error decoding mime structure. Report this as a bug!") . '<br>'; |
368 | return( $pos ); |
369 | } |
370 | |
346817d4 |
371 | function mime_fetch_body($imap_stream, $id, $ent_id ) { |
09a4bde3 |
372 | |
373 | /* |
374 | * do a bit of error correction. If we couldn't find the entity id, just guess |
375 | * that it is the first one. That is usually the case anyway. |
376 | */ |
377 | if (!$ent_id) { |
451f74a2 |
378 | $ent_id = 1; |
09a4bde3 |
379 | } |
346817d4 |
380 | |
6ab1bd9e |
381 | $cmd = "FETCH $id BODY[$ent_id]"; |
346817d4 |
382 | $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message); |
77b88425 |
383 | |
384 | do { |
e976319c |
385 | $topline = trim(array_shift( $data )); |
386 | } while( $topline && $topline[0] == '*' && !preg_match( '/\* [0-9]+ FETCH.*/i', $topline )) ; |
451f74a2 |
387 | $wholemessage = implode('', $data); |
388 | if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) { |
77b88425 |
389 | |
451f74a2 |
390 | $ret = substr( $wholemessage, 0, $regs[1] ); |
391 | /* |
392 | There is some information in the content info header that could be important |
393 | in order to parse html messages. Let's get them here. |
394 | */ |
395 | if ( $ret{0} == '<' ) { |
1c72b151 |
396 | $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message); |
e5ea9327 |
397 | /* BASE within HTML documents is illegal (see w3 spec) |
398 | * $base = ''; |
399 | * $k = 10; |
400 | * foreach( $data as $d ) { |
401 | * if ( substr( $d, 0, 13 ) == 'Content-Base:' ) { |
402 | * $j = strlen( $d ); |
403 | * $i = 13; |
404 | * $base = ''; |
405 | * while ( $i < $j && |
406 | * ( !isNoSep( $d{$i} ) || $d{$i} == '"' ) ) |
407 | * $i++; |
408 | * while ( $i < $j ) { |
409 | * if ( isNoSep( $d{$i} ) ) |
410 | * $base .= $d{$i}; |
411 | * $i++; |
412 | * } |
413 | * $k = 0; |
414 | * } elseif ( $k == 1 && !isnosep( $d{0} ) ) { |
415 | * $base .= substr( $d, 1 ); |
416 | * } |
417 | * $k++; |
418 | * } |
419 | * if ( $base <> '' ) { |
420 | * $ret = "<base href=\"$base\">" . $ret; |
421 | * } |
7e235a1a |
422 | * */ |
451f74a2 |
423 | } |
424 | } else if (ereg('"([^"]*)"', $topline, $regs)) { |
425 | $ret = $regs[1]; |
426 | } else { |
427 | global $where, $what, $mailbox, $passed_id, $startMessage; |
e5ea9327 |
428 | $par = 'mailbox=' . urlencode($mailbox) . "&passed_id=$passed_id"; |
451f74a2 |
429 | if (isset($where) && isset($what)) { |
e5ea9327 |
430 | $par .= '&where='. urlencode($where) . "&what=" . urlencode($what); |
a3daaaf3 |
431 | } else { |
e5ea9327 |
432 | $par .= "&startMessage=$startMessage&show_more=0"; |
451f74a2 |
433 | } |
e5ea9327 |
434 | $par .= '&response=' . urlencode($response) . |
435 | '&message=' . urlencode($message). |
436 | '&topline=' . urlencode($topline); |
a019eeb8 |
437 | |
346817d4 |
438 | echo '<tt><br>' . |
439 | '<table width="80%"><tr>' . |
440 | '<tr><td colspan=2>' . |
451f74a2 |
441 | _("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!") . |
346817d4 |
442 | " <A HREF=\"../src/retrievalerror.php?$par\"><br>" . |
443 | _("Submit message") . '</A><BR> ' . |
444 | '</td></tr>' . |
445 | '<td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" . |
446 | '<td><b>' . _("Response:") . "</td><td>$response</td></tr>" . |
447 | '<td><b>' . _("Message:") . "</td><td>$message</td></tr>" . |
448 | '<td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" . |
449 | "</table><BR></tt></font><hr>"; |
450 | |
1c72b151 |
451 | $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message); |
451f74a2 |
452 | array_shift($data); |
453 | $wholemessage = implode('', $data); |
a019eeb8 |
454 | |
346817d4 |
455 | $ret = $wholemessage; |
a3daaaf3 |
456 | } |
451f74a2 |
457 | return( $ret ); |
458 | } |
d4467150 |
459 | |
451f74a2 |
460 | function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) { |
461 | // do a bit of error correction. If we couldn't find the entity id, just guess |
462 | // that it is the first one. That is usually the case anyway. |
463 | if (!$ent_id) { |
464 | $ent_id = 1; |
465 | } |
466 | $sid = sqimap_session_id(); |
467 | // Don't kill the connection if the browser is over a dialup |
468 | // and it would take over 30 seconds to download it. |
b7206e1d |
469 | |
470 |