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