1
|
|
|
<?php |
2
|
|
|
/** |
3
|
|
|
* This file is part of the ZBateson\MailMimeParser project. |
4
|
|
|
* |
5
|
|
|
* @license http://opensource.org/licenses/bsd-license.php BSD |
6
|
|
|
*/ |
7
|
|
|
namespace ZBateson\MailMimeParser; |
8
|
|
|
|
9
|
|
|
/** |
10
|
|
|
* Parses a mail mime message into its component parts. To invoke, call |
11
|
|
|
* MailMimeParser::parse. |
12
|
|
|
* |
13
|
|
|
* @author Zaahid Bateson |
14
|
|
|
*/ |
15
|
|
|
class MessageParser |
16
|
|
|
{ |
17
|
|
|
/** |
18
|
|
|
* @var \ZBateson\MailMimeParser\Message the Message object that the read |
19
|
|
|
* mail mime message will be parsed into |
20
|
|
|
*/ |
21
|
|
|
protected $message; |
22
|
|
|
|
23
|
|
|
/** |
24
|
|
|
* @var \ZBateson\MailMimeParser\MimePartFactory the MimePartFactory object |
25
|
|
|
* used to create parts. |
26
|
|
|
*/ |
27
|
|
|
protected $partFactory; |
28
|
|
|
|
29
|
|
|
/** |
30
|
|
|
* @var \ZBateson\MailMimeParser\PartStreamRegistry the PartStreamRegistry |
31
|
|
|
* object used to register stream parts. |
32
|
|
|
*/ |
33
|
|
|
protected $partStreamRegistry; |
34
|
|
|
|
35
|
|
|
/** |
36
|
|
|
* Sets up the parser with its dependencies. |
37
|
|
|
* |
38
|
|
|
* @param \ZBateson\MailMimeParser\Message $m |
39
|
|
|
* @param \ZBateson\MailMimeParser\MimePartFactory $pf |
40
|
|
|
* @param \ZBateson\MailMimeParser\PartStreamRegistry $psr |
41
|
|
|
*/ |
42
|
|
|
public function __construct(Message $m, MimePartFactory $pf, PartStreamRegistry $psr) |
43
|
|
|
{ |
44
|
|
|
$this->message = $m; |
45
|
|
|
$this->partFactory = $pf; |
46
|
|
|
$this->partStreamRegistry = $psr; |
47
|
|
|
} |
48
|
|
|
|
49
|
|
|
/** |
50
|
|
|
* Parses the passed stream handle into the ZBateson\MailMimeParser\Message |
51
|
|
|
* object and returns it. |
52
|
|
|
* |
53
|
|
|
* @param resource $fhandle the resource handle to the input stream of the |
54
|
|
|
* mime message |
55
|
|
|
* @return \ZBateson\MailMimeParser\Message |
56
|
|
|
*/ |
57
|
|
|
public function parse($fhandle) |
58
|
|
|
{ |
59
|
|
|
$this->partStreamRegistry->register($this->message->getObjectId(), $fhandle); |
60
|
|
|
$this->read($fhandle, $this->message); |
61
|
|
|
return $this->message; |
62
|
|
|
} |
63
|
|
|
|
64
|
|
|
/** |
65
|
|
|
* Ensures the header isn't empty, and contains a colon character, then |
66
|
|
|
* splits it and assigns it to $part |
67
|
|
|
* |
68
|
|
|
* @param string $header |
69
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $part |
70
|
|
|
*/ |
71
|
|
|
private function addRawHeaderToPart($header, MimePart $part) |
72
|
|
|
{ |
73
|
|
|
if (!empty($header) && strpos($header, ':') !== false) { |
74
|
|
|
$a = explode(':', $header, 2); |
75
|
|
|
$part->setRawHeader($a[0], trim($a[1])); |
76
|
|
|
} |
77
|
|
|
} |
78
|
|
|
|
79
|
|
|
/** |
80
|
|
|
* Reads header lines up to an empty line, adding them to the passed $part. |
81
|
|
|
* |
82
|
|
|
* @param resource $handle the resource handle to read from |
83
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $part the current part to add |
84
|
|
|
* headers to |
85
|
|
|
*/ |
86
|
|
|
protected function readHeaders($handle, MimePart $part) |
87
|
|
|
{ |
88
|
|
|
$header = ''; |
89
|
|
|
do { |
90
|
|
|
$line = fgets($handle, 1000); |
91
|
|
|
if ($line[0] !== "\t" && $line[0] !== ' ') { |
92
|
|
|
$this->addRawHeaderToPart($header, $part); |
93
|
|
|
$header = ''; |
94
|
|
|
} else { |
95
|
|
|
$line = ' ' . ltrim($line); |
96
|
|
|
} |
97
|
|
|
$header .= rtrim($line, "\r\n"); |
98
|
|
|
} while (!empty($header)); |
99
|
|
|
} |
100
|
|
|
|
101
|
|
|
/** |
102
|
|
|
* Finds the end of the Mime part at the current read position in $handle |
103
|
|
|
* and sets $boundaryLength to the number of bytes in the part, and |
104
|
|
|
* $endBoundaryFound to true if it's an 'end' boundary, meaning there are no |
105
|
|
|
* further parts for the current mime part (ends with --). |
106
|
|
|
* |
107
|
|
|
* @param resource $handle |
108
|
|
|
* @param string $boundary |
109
|
|
|
* @param int $boundaryLength |
110
|
|
|
* @param boolean $endBoundaryFound |
111
|
|
|
*/ |
112
|
|
|
private function findPartBoundaries($handle, $boundary, &$boundaryLength, &$endBoundaryFound) |
113
|
|
|
{ |
114
|
|
|
do { |
115
|
|
|
$line = fgets($handle); |
116
|
|
|
$boundaryLength = strlen($line); |
117
|
|
|
$test = rtrim($line); |
118
|
|
|
if ($test === "--$boundary") { |
119
|
|
|
break; |
120
|
|
|
} elseif ($test === "--$boundary--") { |
121
|
|
|
$endBoundaryFound = true; |
122
|
|
|
break; |
123
|
|
|
} |
124
|
|
|
} while (!feof($handle)); |
125
|
|
|
} |
126
|
|
|
|
127
|
|
|
/** |
128
|
|
|
* Reads the content of a mime part up to a boundary, or the entire message |
129
|
|
|
* if no boundary is specified. |
130
|
|
|
* |
131
|
|
|
* readPartContent may be called to skip to the first boundary to read its |
132
|
|
|
* headers, in which case $skipPart should be true. |
133
|
|
|
* |
134
|
|
|
* If the end boundary is found, the method returns true. |
135
|
|
|
* |
136
|
|
|
* @param resource $handle the input stream resource |
137
|
|
|
* @param \ZBateson\MailMimeParser\Message $message the current Message |
138
|
|
|
* object |
139
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $part the current MimePart |
140
|
|
|
* object to load the content into. |
141
|
|
|
* @param string $boundary the MIME boundary |
142
|
|
|
* @param boolean $skipPart pass true if the intention is to read up to the |
143
|
|
|
* beginning MIME boundary's headers |
144
|
|
|
* @return boolean if the end boundary is found |
145
|
|
|
*/ |
146
|
|
|
protected function readPartContent($handle, Message $message, MimePart $part, $boundary, $skipPart) |
147
|
|
|
{ |
148
|
|
|
$start = ftell($handle); |
149
|
|
|
$boundaryLength = 0; |
150
|
|
|
$endBoundaryFound = false; |
151
|
|
|
if (!empty($boundary)) { |
152
|
|
|
$this->findPartBoundaries($handle, $boundary, $boundaryLength, $endBoundaryFound); |
153
|
|
|
} else { |
154
|
|
|
fseek($handle, 0, SEEK_END); |
155
|
|
|
} |
156
|
|
|
if (!$skipPart) { |
157
|
|
|
$end = ftell($handle) - $boundaryLength; |
158
|
|
|
$this->partStreamRegistry->attachPartStreamHandle($part, $message, $start, $end); |
159
|
|
|
$message->addPart($part); |
160
|
|
|
} |
161
|
|
|
return $endBoundaryFound; |
162
|
|
|
} |
163
|
|
|
|
164
|
|
|
/** |
165
|
|
|
* Returns the boundary from the parent MimePart, or the current boundary if |
166
|
|
|
* $parent is null |
167
|
|
|
* |
168
|
|
|
* @param string $curBoundary |
169
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $parent |
170
|
|
|
* @return string |
171
|
|
|
*/ |
172
|
|
|
private function getParentBoundary($curBoundary, MimePart $parent = null) |
173
|
|
|
{ |
174
|
|
|
return $parent !== null ? |
175
|
|
|
$parent->getHeaderParameter('Content-Type', 'boundary') : |
176
|
|
|
$curBoundary; |
177
|
|
|
} |
178
|
|
|
|
179
|
|
|
/** |
180
|
|
|
* Instantiates and returns a new MimePart setting the part's parent to |
181
|
|
|
* either the passed $parent, or $message if $parent is null. |
182
|
|
|
* |
183
|
|
|
* @param \ZBateson\MailMimeParser\Message $message |
184
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $parent |
185
|
|
|
* @return \ZBateson\MailMimeParser\MimePart |
186
|
|
|
*/ |
187
|
|
|
private function newMimePartForMessage(Message $message, MimePart $parent = null) |
188
|
|
|
{ |
189
|
|
|
$nextPart = $this->partFactory->newMimePart(); |
190
|
|
|
$nextPart->setParent($parent === null ? $message : $parent); |
191
|
|
|
return $nextPart; |
192
|
|
|
} |
193
|
|
|
|
194
|
|
|
/** |
195
|
|
|
* Keeps reading if an end boundary is found, to find the parent's boundary |
196
|
|
|
* and the part's content. |
197
|
|
|
* |
198
|
|
|
* @param resource $handle |
199
|
|
|
* @param \ZBateson\MailMimeParser\Message $message |
200
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $parent |
201
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $part |
202
|
|
|
* @param string $boundary |
203
|
|
|
* @param bool $skipFirst |
204
|
|
|
* @return \ZBateson\MailMimeParser\MimePart |
205
|
|
|
*/ |
206
|
|
|
private function readMimeMessageBoundaryParts( |
207
|
|
|
$handle, |
208
|
|
|
Message $message, |
209
|
|
|
MimePart $parent, |
210
|
|
|
MimePart $part, |
211
|
|
|
$boundary, |
212
|
|
|
$skipFirst |
213
|
|
|
) { |
214
|
|
|
$skipPart = $skipFirst; |
215
|
|
|
while ($this->readPartContent($handle, $message, $part, $boundary, $skipPart) && $parent !== null) { |
216
|
|
|
$parent = $parent->getParent(); |
217
|
|
|
// $boundary used by next call to readPartContent |
218
|
|
|
$boundary = $this->getParentBoundary($boundary, $parent); |
219
|
|
|
$skipPart = true; |
220
|
|
|
} |
221
|
|
|
return $this->newMimePartForMessage($message, $parent); |
222
|
|
|
} |
223
|
|
|
|
224
|
|
|
/** |
225
|
|
|
* Finds the boundaries for the current MimePart, reads its content and |
226
|
|
|
* creates and returns the next part, setting its parent part accordingly. |
227
|
|
|
* |
228
|
|
|
* @param resource $handle The handle to read from |
229
|
|
|
* @param \ZBateson\MailMimeParser\Message $message The current Message |
230
|
|
|
* @param \ZBateson\MailMimeParser\MimePart $part |
231
|
|
|
* @return MimePart |
232
|
|
|
*/ |
233
|
|
|
protected function readMimeMessagePart($handle, Message $message, MimePart $part) |
234
|
|
|
{ |
235
|
|
|
$boundary = $part->getHeaderParameter('Content-Type', 'boundary'); |
236
|
|
|
$skipFirst = true; |
237
|
|
|
$parent = $part; |
238
|
|
|
|
239
|
|
|
if (empty($boundary) || !preg_match('~multipart/\w+~i', $part->getHeaderValue('Content-Type'))) { |
240
|
|
|
// either there is no boundary (possibly no parent boundary either) and message is read |
241
|
|
|
// till the end, or we're in a boundary already and content should be read till the parent |
242
|
|
|
// boundary is reached |
243
|
|
|
if ($part->getParent() !== null) { |
244
|
|
|
$parent = $part->getParent(); |
245
|
|
|
$boundary = $parent->getHeaderParameter('Content-Type', 'boundary'); |
246
|
|
|
} |
247
|
|
|
$skipFirst = false; |
248
|
|
|
} |
249
|
|
|
return $this->readMimeMessageBoundaryParts($handle, $message, $parent, $part, $boundary, $skipFirst); |
250
|
|
|
} |
251
|
|
|
|
252
|
|
|
/** |
253
|
|
|
* Extracts the filename and end position of a UUEncoded part. |
254
|
|
|
* |
255
|
|
|
* The filename is set to the passed $nextFilename parameter. The end |
256
|
|
|
* position is returned. |
257
|
|
|
* |
258
|
|
|
* @param resource $handle the current file handle |
259
|
|
|
* @param int &$nextMode is assigned the value of the next file mode or null |
260
|
|
|
* if not found |
261
|
|
|
* @param string &$nextFilename is assigned the value of the next filename |
262
|
|
|
* or null if not found |
263
|
|
|
* @param int &$end assigned the offset position within the passed resource |
264
|
|
|
* $handle of the end of the uuencoded part |
265
|
|
|
*/ |
266
|
|
|
private function findNextUUEncodedPartPosition($handle) |
267
|
|
|
{ |
268
|
|
|
$end = ftell($handle); |
269
|
|
|
do { |
270
|
|
|
$line = trim(fgets($handle)); |
271
|
|
|
$matches = null; |
272
|
|
|
if (preg_match('/^begin [0-7]{3} .*$/', $line, $matches)) { |
273
|
|
|
fseek($handle, $end); |
274
|
|
|
break; |
275
|
|
|
} |
276
|
|
|
$end = ftell($handle); |
277
|
|
|
} while (!feof($handle)); |
278
|
|
|
return $end; |
279
|
|
|
} |
280
|
|
|
|
281
|
|
|
/** |
282
|
|
|
* Reads one part of a UUEncoded message and adds it to the passed Message |
283
|
|
|
* as a MimePart. |
284
|
|
|
* |
285
|
|
|
* The method reads up to the first 'begin' part of the message, or to the |
286
|
|
|
* end of the message if no 'begin' exists. |
287
|
|
|
* |
288
|
|
|
* @param resource $handle |
289
|
|
|
* @param \ZBateson\MailMimeParser\Message $message |
290
|
|
|
* @return string |
291
|
|
|
*/ |
292
|
|
|
protected function readUUEncodedOrPlainTextPart($handle, Message $message) |
293
|
|
|
{ |
294
|
|
|
$start = ftell($handle); |
295
|
|
|
$line = trim(fgets($handle)); |
296
|
|
|
$end = $this->findNextUUEncodedPartPosition($handle); |
297
|
|
|
|
298
|
|
|
$part = null; |
299
|
|
|
if (preg_match('/^begin ([0-7]{3}) (.*)$/', $line, $matches)) { |
300
|
|
|
$mode = $matches[1]; |
301
|
|
|
$filename = $matches[2]; |
302
|
|
|
$part = $this->partFactory->newUUEncodedPart($mode, $filename); |
303
|
|
|
$this->partStreamRegistry->attachPartStreamHandle($part, $message, $start, $end); |
304
|
|
|
} else { |
305
|
|
|
$part = $this->partFactory->newNonMimePart(); |
306
|
|
|
$this->partStreamRegistry->attachPartStreamHandle($part, $message, $start, $end); |
307
|
|
|
} |
308
|
|
|
$message->addPart($part); |
309
|
|
|
} |
310
|
|
|
|
311
|
|
|
/** |
312
|
|
|
* Returns true if the passed Message doesn't define either a Content-Type |
313
|
|
|
* or a Mime-Version header. |
314
|
|
|
* |
315
|
|
|
* @param \ZBateson\MailMimeParser\Message $message |
316
|
|
|
* @return bool |
317
|
|
|
*/ |
318
|
|
|
private function isNotMime(Message $message) |
319
|
|
|
{ |
320
|
|
|
$contentType = $message->getHeaderValue('Content-Type'); |
321
|
|
|
$mimeVersion = $message->getHeaderValue('Mime-Version'); |
322
|
|
|
return ($contentType === null && $mimeVersion === null); |
323
|
|
|
} |
324
|
|
|
|
325
|
|
|
/** |
326
|
|
|
* Reads the message from the input stream $handle into $message. |
327
|
|
|
* |
328
|
|
|
* The method will loop to read headers and find and parse multipart-mime |
329
|
|
|
* message parts and uuencoded attachments (as mime-parts), adding them to |
330
|
|
|
* the passed Message object. |
331
|
|
|
* |
332
|
|
|
* @param resource $handle |
333
|
|
|
* @param \ZBateson\MailMimeParser\Message $message |
334
|
|
|
*/ |
335
|
|
|
protected function read($handle, Message $message) |
336
|
|
|
{ |
337
|
|
|
$part = $message; |
338
|
|
|
$this->readHeaders($handle, $message); |
339
|
|
|
$isNotMime = $this->isNotMime($message); |
340
|
|
|
do { |
341
|
|
|
if ($isNotMime) { |
342
|
|
|
$this->readUUEncodedOrPlainTextPart($handle, $message); |
343
|
|
|
} else { |
344
|
|
|
$part = $this->readMimeMessagePart($handle, $message, $part); |
345
|
|
|
$this->readHeaders($handle, $part); |
346
|
|
|
} |
347
|
|
|
} while (!feof($handle)); |
348
|
|
|
} |
349
|
|
|
} |
350
|
|
|
|