commit | author | age
|
f7427f
|
1 |
<?php |
AM |
2 |
|
|
3 |
/** |
|
4 |
+-----------------------------------------------------------------------+ |
|
5 |
| This file is part of the Roundcube Webmail client | |
|
6 |
| Copyright (C) 2005-2015, The Roundcube Dev Team | |
|
7 |
| Copyright (C) 2011-2015, Kolab Systems AG | |
|
8 |
| | |
|
9 |
| Licensed under the GNU General Public License version 3 or | |
|
10 |
| any later version with exceptions for skins & plugins. | |
|
11 |
| See the README file for a full license statement. | |
|
12 |
| | |
|
13 |
| PURPOSE: | |
|
14 |
| MIME message parsing utilities derived from Mail_mimeDecode | |
|
15 |
+-----------------------------------------------------------------------+ |
|
16 |
| Author: Thomas Bruederli <roundcube@gmail.com> | |
|
17 |
| Author: Aleksander Machniak <alec@alec.pl> | |
|
18 |
| Author: Richard Heyes <richard@phpguru.org> | |
|
19 |
+-----------------------------------------------------------------------+ |
|
20 |
*/ |
|
21 |
|
|
22 |
/** |
|
23 |
* Class for parsing MIME messages |
|
24 |
* |
|
25 |
* @package Framework |
|
26 |
* @subpackage Storage |
|
27 |
* @author Aleksander Machniak <alec@alec.pl> |
|
28 |
*/ |
|
29 |
class rcube_mime_decode |
|
30 |
{ |
|
31 |
/** |
|
32 |
* Class configuration parameters. |
|
33 |
* |
|
34 |
* @var array |
|
35 |
*/ |
|
36 |
protected $params = array( |
|
37 |
'include_bodies' => true, |
|
38 |
'decode_bodies' => true, |
|
39 |
'decode_headers' => true, |
|
40 |
'crlf' => "\r\n", |
|
41 |
'default_charset' => RCUBE_CHARSET, |
|
42 |
); |
|
43 |
|
|
44 |
|
|
45 |
/** |
|
46 |
* Constructor. |
|
47 |
* |
|
48 |
* Sets up the object, initialise the variables, and splits and |
|
49 |
* stores the header and body of the input. |
|
50 |
* |
|
51 |
* @param array $params An array of various parameters that determine |
|
52 |
* various things: |
|
53 |
* include_bodies - Whether to include the body in the returned |
|
54 |
* object. |
|
55 |
* decode_bodies - Whether to decode the bodies |
|
56 |
* of the parts. (Transfer encoding) |
|
57 |
* decode_headers - Whether to decode headers |
|
58 |
* crlf - CRLF type to use (CRLF/LF/CR) |
|
59 |
*/ |
|
60 |
public function __construct($params = array()) |
|
61 |
{ |
|
62 |
if (!empty($params)) { |
|
63 |
$this->params = array_merge($this->params, (array) $params); |
|
64 |
} |
|
65 |
} |
|
66 |
|
|
67 |
/** |
|
68 |
* Performs the decoding process. |
|
69 |
* |
53fa08
|
70 |
* @param string $input The input to decode |
AM |
71 |
* @param bool $convert Convert result to rcube_message_part structure |
f7427f
|
72 |
* |
AM |
73 |
* @return object|bool Decoded results or False on failure |
|
74 |
*/ |
53fa08
|
75 |
public function decode($input, $convert = true) |
f7427f
|
76 |
{ |
AM |
77 |
list($header, $body) = $this->splitBodyHeader($input); |
|
78 |
|
53fa08
|
79 |
$struct = $this->do_decode($header, $body); |
f7427f
|
80 |
|
53fa08
|
81 |
if ($struct && $convert) { |
f7427f
|
82 |
$struct = $this->structure_part($struct); |
AM |
83 |
} |
|
84 |
|
|
85 |
return $struct; |
|
86 |
} |
|
87 |
|
|
88 |
/** |
|
89 |
* Performs the decoding. Decodes the body string passed to it |
|
90 |
* If it finds certain content-types it will call itself in a |
|
91 |
* recursive fashion |
|
92 |
* |
|
93 |
* @param string $headers Header section |
|
94 |
* @param string $body Body section |
|
95 |
* @param string $default_ctype Default content type |
|
96 |
* |
|
97 |
* @return object|bool Decoded results or False on error |
|
98 |
*/ |
|
99 |
protected function do_decode($headers, $body, $default_ctype = 'text/plain') |
|
100 |
{ |
|
101 |
$return = new stdClass; |
|
102 |
$headers = $this->parseHeaders($headers); |
|
103 |
|
|
104 |
while (list($key, $value) = each($headers)) { |
|
105 |
$header_name = strtolower($value['name']); |
|
106 |
|
|
107 |
if (isset($return->headers[$header_name]) && !is_array($return->headers[$header_name])) { |
|
108 |
$return->headers[$header_name] = array($return->headers[$header_name]); |
|
109 |
$return->headers[$header_name][] = $value['value']; |
|
110 |
} |
|
111 |
else if (isset($return->headers[$header_name])) { |
|
112 |
$return->headers[$header_name][] = $value['value']; |
|
113 |
} |
|
114 |
else { |
|
115 |
$return->headers[$header_name] = $value['value']; |
|
116 |
} |
|
117 |
|
|
118 |
switch ($header_name) { |
|
119 |
case 'content-type': |
|
120 |
$content_type = $this->parseHeaderValue($value['value']); |
|
121 |
|
|
122 |
if (preg_match('/([0-9a-z+.-]+)\/([0-9a-z+.-]+)/i', $content_type['value'], $regs)) { |
|
123 |
$return->ctype_primary = $regs[1]; |
|
124 |
$return->ctype_secondary = $regs[2]; |
|
125 |
} |
|
126 |
|
|
127 |
if (isset($content_type['other'])) { |
|
128 |
while (list($p_name, $p_value) = each($content_type['other'])) { |
|
129 |
$return->ctype_parameters[$p_name] = $p_value; |
|
130 |
} |
|
131 |
} |
|
132 |
|
|
133 |
break; |
|
134 |
|
|
135 |
case 'content-disposition'; |
|
136 |
$content_disposition = $this->parseHeaderValue($value['value']); |
|
137 |
$return->disposition = $content_disposition['value']; |
|
138 |
|
|
139 |
if (isset($content_disposition['other'])) { |
|
140 |
while (list($p_name, $p_value) = each($content_disposition['other'])) { |
|
141 |
$return->d_parameters[$p_name] = $p_value; |
|
142 |
} |
|
143 |
} |
|
144 |
|
|
145 |
break; |
|
146 |
|
|
147 |
case 'content-transfer-encoding': |
|
148 |
$content_transfer_encoding = $this->parseHeaderValue($value['value']); |
|
149 |
break; |
|
150 |
} |
|
151 |
} |
|
152 |
|
|
153 |
if (isset($content_type)) { |
|
154 |
$ctype = strtolower($content_type['value']); |
|
155 |
|
|
156 |
switch ($ctype) { |
|
157 |
case 'text/plain': |
|
158 |
$encoding = isset($content_transfer_encoding) ? $content_transfer_encoding['value'] : '7bit'; |
|
159 |
|
|
160 |
if ($this->params['include_bodies']) { |
|
161 |
$return->body = $this->params['decode_bodies'] ? rcube_mime::decode($body, $encoding) : $body; |
|
162 |
} |
|
163 |
|
|
164 |
break; |
|
165 |
|
|
166 |
case 'text/html': |
|
167 |
$encoding = isset($content_transfer_encoding) ? $content_transfer_encoding['value'] : '7bit'; |
|
168 |
|
|
169 |
if ($this->params['include_bodies']) { |
|
170 |
$return->body = $this->params['decode_bodies'] ? rcube_mime::decode($body, $encoding) : $body; |
|
171 |
} |
|
172 |
|
|
173 |
break; |
|
174 |
|
|
175 |
case 'multipart/digest': |
|
176 |
case 'multipart/alternative': |
|
177 |
case 'multipart/related': |
|
178 |
case 'multipart/mixed': |
c9e2ab
|
179 |
case 'multipart/signed': |
AM |
180 |
case 'multipart/encrypted': |
f7427f
|
181 |
if (!isset($content_type['other']['boundary'])) { |
AM |
182 |
return false; |
|
183 |
} |
|
184 |
|
|
185 |
$default_ctype = $ctype === 'multipart/digest' ? 'message/rfc822' : 'text/plain'; |
|
186 |
$parts = $this->boundarySplit($body, $content_type['other']['boundary']); |
|
187 |
|
|
188 |
for ($i = 0; $i < count($parts); $i++) { |
|
189 |
list($part_header, $part_body) = $this->splitBodyHeader($parts[$i]); |
|
190 |
$return->parts[] = $this->do_decode($part_header, $part_body, $default_ctype); |
|
191 |
} |
|
192 |
|
|
193 |
break; |
|
194 |
|
|
195 |
case 'message/rfc822': |
|
196 |
$obj = new rcube_mime_decode($this->params); |
53fa08
|
197 |
$return->parts[] = $obj->decode($body, false); |
f7427f
|
198 |
unset($obj); |
AM |
199 |
break; |
|
200 |
|
|
201 |
default: |
|
202 |
if ($this->params['include_bodies']) { |
|
203 |
$return->body = $this->params['decode_bodies'] ? rcube_mime::decode($body, $content_transfer_encoding['value']) : $body; |
|
204 |
} |
|
205 |
|
|
206 |
break; |
|
207 |
} |
|
208 |
} |
|
209 |
else { |
|
210 |
$ctype = explode('/', $default_ctype); |
|
211 |
$return->ctype_primary = $ctype[0]; |
|
212 |
$return->ctype_secondary = $ctype[1]; |
|
213 |
|
|
214 |
if ($this->params['include_bodies']) { |
|
215 |
$return->body = $this->params['decode_bodies'] ? rcube_mime::decode($body) : $body; |
|
216 |
} |
|
217 |
} |
|
218 |
|
|
219 |
return $return; |
|
220 |
} |
|
221 |
|
|
222 |
/** |
|
223 |
* Given a string containing a header and body |
|
224 |
* section, this function will split them (at the first |
|
225 |
* blank line) and return them. |
|
226 |
* |
|
227 |
* @param string $input Input to split apart |
|
228 |
* |
|
229 |
* @return array Contains header and body section |
|
230 |
*/ |
|
231 |
protected function splitBodyHeader($input) |
|
232 |
{ |
|
233 |
$pos = strpos($input, $this->params['crlf'] . $this->params['crlf']); |
|
234 |
if ($pos === false) { |
|
235 |
return false; |
|
236 |
} |
|
237 |
|
|
238 |
$crlf_len = strlen($this->params['crlf']); |
|
239 |
$header = substr($input, 0, $pos); |
|
240 |
$body = substr($input, $pos + 2 * $crlf_len); |
|
241 |
|
|
242 |
if (substr_compare($body, $this->params['crlf'], -$crlf_len) === 0) { |
|
243 |
$body = substr($body, 0, -$crlf_len); |
|
244 |
} |
|
245 |
|
|
246 |
return array($header, $body); |
|
247 |
} |
|
248 |
|
|
249 |
/** |
|
250 |
* Parse headers given in $input and return as assoc array. |
|
251 |
* |
|
252 |
* @param string $input Headers to parse |
|
253 |
* |
|
254 |
* @return array Contains parsed headers |
|
255 |
*/ |
|
256 |
protected function parseHeaders($input) |
|
257 |
{ |
|
258 |
if ($input !== '') { |
|
259 |
// Unfold the input |
|
260 |
$input = preg_replace('/' . $this->params['crlf'] . "(\t| )/", ' ', $input); |
|
261 |
$headers = explode($this->params['crlf'], trim($input)); |
|
262 |
|
|
263 |
foreach ($headers as $value) { |
|
264 |
$hdr_name = substr($value, 0, $pos = strpos($value, ':')); |
|
265 |
$hdr_value = substr($value, $pos+1); |
|
266 |
|
|
267 |
if ($hdr_value[0] == ' ') { |
|
268 |
$hdr_value = substr($hdr_value, 1); |
|
269 |
} |
|
270 |
|
|
271 |
$return[] = array( |
|
272 |
'name' => $hdr_name, |
|
273 |
'value' => $this->params['decode_headers'] ? $this->decodeHeader($hdr_value) : $hdr_value, |
|
274 |
); |
|
275 |
} |
|
276 |
} |
|
277 |
else { |
|
278 |
$return = array(); |
|
279 |
} |
|
280 |
|
|
281 |
return $return; |
|
282 |
} |
|
283 |
|
|
284 |
/** |
|
285 |
* Function to parse a header value, extract first part, and any secondary |
|
286 |
* parts (after ;) This function is not as robust as it could be. |
|
287 |
* Eg. header comments in the wrong place will probably break it. |
|
288 |
* |
|
289 |
* @param string $input Header value to parse |
|
290 |
* |
|
291 |
* @return array Contains parsed result |
|
292 |
*/ |
|
293 |
protected function parseHeaderValue($input) |
|
294 |
{ |
|
295 |
$parts = preg_split('/;\s*/', $input); |
|
296 |
|
|
297 |
if (!empty($parts)) { |
|
298 |
$return['value'] = trim($parts[0]); |
|
299 |
|
|
300 |
for ($n = 1; $n < count($parts); $n++) { |
|
301 |
if (preg_match_all('/(([[:alnum:]]+)="?([^"]*)"?\s?;?)+/i', $parts[$n], $matches)) { |
|
302 |
for ($i = 0; $i < count($matches[2]); $i++) { |
|
303 |
$return['other'][strtolower($matches[2][$i])] = $matches[3][$i]; |
|
304 |
} |
|
305 |
} |
|
306 |
} |
|
307 |
} |
|
308 |
else { |
|
309 |
$return['value'] = trim($input); |
|
310 |
} |
|
311 |
|
|
312 |
return $return; |
|
313 |
} |
|
314 |
|
|
315 |
/** |
|
316 |
* This function splits the input based on the given boundary |
|
317 |
* |
|
318 |
* @param string $input Input to parse |
|
319 |
* @param string $boundary Boundary |
|
320 |
* |
|
321 |
* @return array Contains array of resulting mime parts |
|
322 |
*/ |
|
323 |
protected function boundarySplit($input, $boundary) |
|
324 |
{ |
|
325 |
$tmp = explode('--' . $boundary, $input); |
|
326 |
|
|
327 |
for ($i = 1; $i < count($tmp)-1; $i++) { |
|
328 |
$parts[] = $tmp[$i]; |
|
329 |
} |
|
330 |
|
|
331 |
return $parts; |
|
332 |
} |
|
333 |
|
|
334 |
/** |
|
335 |
* Given a header, this function will decode it according to RFC2047. |
|
336 |
* Probably not *exactly* conformant, but it does pass all the given |
|
337 |
* examples (in RFC2047). |
|
338 |
* |
|
339 |
* @param string $input Input header value to decode |
|
340 |
* |
|
341 |
* @return string Decoded header value |
|
342 |
*/ |
|
343 |
protected function decodeHeader($input) |
|
344 |
{ |
|
345 |
return rcube_mime::decode_mime_string($input, $this->params['default_charset']); |
|
346 |
} |
|
347 |
|
|
348 |
/** |
|
349 |
* Recursive method to convert a rcube_mime_decode structure |
|
350 |
* into a rcube_message_part object. |
|
351 |
* |
|
352 |
* @param object $part A message part struct |
|
353 |
* @param int $count Part count |
|
354 |
* @param string $parent Parent MIME ID |
|
355 |
* |
|
356 |
* @return object rcube_message_part |
|
357 |
* @see self::decode() |
|
358 |
*/ |
|
359 |
protected function structure_part($part, $count = 0, $parent = '') |
|
360 |
{ |
|
361 |
$struct = new rcube_message_part; |
|
362 |
$struct->mime_id = $part->mime_id ?: (empty($parent) ? (string)$count : "$parent.$count"); |
|
363 |
$struct->headers = $part->headers; |
|
364 |
$struct->mimetype = $part->ctype_primary . '/' . $part->ctype_secondary; |
|
365 |
$struct->ctype_primary = $part->ctype_primary; |
|
366 |
$struct->ctype_secondary = $part->ctype_secondary; |
|
367 |
$struct->ctype_parameters = $part->ctype_parameters; |
|
368 |
|
|
369 |
if ($part->headers['content-transfer-encoding']) { |
|
370 |
$struct->encoding = $part->headers['content-transfer-encoding']; |
|
371 |
} |
|
372 |
|
|
373 |
if ($part->ctype_parameters['charset']) { |
|
374 |
$struct->charset = $part->ctype_parameters['charset']; |
|
375 |
} |
|
376 |
|
|
377 |
$part_charset = $struct->charset ?: $this->params['default_charset']; |
|
378 |
|
|
379 |
// determine filename |
|
380 |
if (($filename = $part->d_parameters['filename']) || ($filename = $part->ctype_parameters['name'])) { |
|
381 |
if (!$this->params['decode_headers']) { |
|
382 |
$filename = $this->decodeHeader($filename); |
|
383 |
} |
|
384 |
|
|
385 |
$struct->filename = $filename; |
|
386 |
} |
|
387 |
|
|
388 |
$struct->body = $part->body; |
|
389 |
$struct->size = strlen($part->body); |
|
390 |
$struct->disposition = $part->disposition; |
|
391 |
|
|
392 |
$count = 0; |
|
393 |
foreach ((array)$part->parts as $child_part) { |
|
394 |
$struct->parts[] = $this->structure_part($child_part, ++$count, $struct->mime_id); |
|
395 |
} |
|
396 |
|
|
397 |
return $struct; |
|
398 |
} |
|
399 |
} |