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\Message; |
8
|
|
|
|
9
|
|
|
use ZBateson\MailMimeParser\Stream\PartStreamRegistry; |
10
|
|
|
use ZBateson\MailMimeParser\Message\Part\PartBuilder; |
11
|
|
|
use ZBateson\MailMimeParser\Message\Part\PartBuilderFactory; |
12
|
|
|
use ZBateson\MailMimeParser\Message\Part\PartFactoryService; |
13
|
|
|
|
14
|
|
|
/** |
15
|
|
|
* Parses a mail mime message into its component parts. To invoke, call |
16
|
|
|
* MailMimeParser::parse. |
17
|
|
|
* |
18
|
|
|
* @author Zaahid Bateson |
19
|
|
|
*/ |
20
|
|
|
class MessageParser |
21
|
|
|
{ |
22
|
|
|
/** |
23
|
|
|
* @var \ZBateson\MailMimeParser\Message\Part\PartFactoryService service |
24
|
|
|
* instance used to create MimePartFactory objects. |
25
|
|
|
*/ |
26
|
|
|
protected $partFactoryService; |
27
|
|
|
|
28
|
|
|
/** |
29
|
|
|
* @var \ZBateson\MailMimeParser\Message\Part\PartBuilderFactory used to |
30
|
|
|
* create PartBuilders |
31
|
|
|
*/ |
32
|
|
|
protected $partBuilderFactory; |
33
|
|
|
|
34
|
|
|
/** |
35
|
|
|
* @var \ZBateson\MailMimeParser\Stream\PartStreamRegistry used for |
36
|
|
|
* registering message part streams. |
37
|
|
|
*/ |
38
|
|
|
protected $partStreamRegistry; |
39
|
|
|
|
40
|
|
|
/** |
41
|
|
|
* Sets up the parser with its dependencies. |
42
|
|
|
* |
43
|
|
|
* @param \ZBateson\MailMimeParser\Message\Part\PartFactoryService $pfs |
44
|
|
|
* @param \ZBateson\MailMimeParser\Message\Part\PartBuilderFactory $pbf |
45
|
|
|
* @param PartStreamRegistry $psr |
46
|
|
|
*/ |
47
|
|
|
public function __construct( |
48
|
|
|
PartFactoryService $pfs, |
49
|
|
|
PartBuilderFactory $pbf, |
50
|
|
|
PartStreamRegistry $psr |
51
|
|
|
) { |
52
|
|
|
$this->partFactoryService = $pfs; |
53
|
|
|
$this->partBuilderFactory = $pbf; |
54
|
|
|
$this->partStreamRegistry = $psr; |
55
|
|
|
} |
56
|
|
|
|
57
|
|
|
/** |
58
|
|
|
* Parses the passed stream handle into a ZBateson\MailMimeParser\Message |
59
|
|
|
* object and returns it. |
60
|
|
|
* |
61
|
|
|
* @param resource $fhandle the resource handle to the input stream of the |
62
|
|
|
* mime message |
63
|
|
|
* @return \ZBateson\MailMimeParser\Message |
64
|
|
|
*/ |
65
|
|
|
public function parse($fhandle) |
66
|
|
|
{ |
67
|
|
|
$messageObjectId = uniqid(); |
68
|
|
|
$this->partStreamRegistry->register($messageObjectId, $fhandle); |
69
|
|
|
$partBuilder = $this->read($fhandle); |
70
|
|
|
return $partBuilder->createMessagePart($messageObjectId); |
71
|
|
|
} |
72
|
|
|
|
73
|
|
|
/** |
74
|
|
|
* Ensures the header isn't empty and contains a colon separator character, |
75
|
|
|
* then splits it and calls $partBuilder->addHeader. |
76
|
|
|
* |
77
|
|
|
* @param string $header |
78
|
|
|
* @param PartBuilder $partBuilder |
79
|
|
|
*/ |
80
|
|
|
private function addRawHeaderToPart($header, PartBuilder $partBuilder) |
81
|
|
|
{ |
82
|
|
|
if ($header !== '' && strpos($header, ':') !== false) { |
83
|
|
|
$a = explode(':', $header, 2); |
84
|
|
|
$partBuilder->addHeader($a[0], trim($a[1])); |
85
|
|
|
} |
86
|
|
|
} |
87
|
|
|
|
88
|
|
|
/** |
89
|
|
|
* Reads header lines up to an empty line, adding them to the passed |
90
|
|
|
* $partBuilder. |
91
|
|
|
* |
92
|
|
|
* @param resource $handle the resource handle to read from |
93
|
|
|
* @param PartBuilder $partBuilder the current part to add headers to |
94
|
|
|
*/ |
95
|
|
|
protected function readHeaders($handle, PartBuilder $partBuilder) |
96
|
|
|
{ |
97
|
|
|
$header = ''; |
98
|
|
|
do { |
99
|
|
|
$line = fgets($handle, 1000); |
100
|
|
|
if (empty($line) || $line[0] !== "\t" && $line[0] !== ' ') { |
101
|
|
|
$this->addRawHeaderToPart($header, $partBuilder); |
102
|
|
|
$header = ''; |
103
|
|
|
} else { |
104
|
|
|
$line = "\r\n" . $line; |
105
|
|
|
} |
106
|
|
|
$header .= rtrim($line, "\r\n"); |
107
|
|
|
} while ($header !== ''); |
108
|
|
|
} |
109
|
|
|
|
110
|
|
|
/** |
111
|
|
|
* Reads lines from the passed $handle, calling $partBuilder->setEndBoundary |
112
|
|
|
* with the passed line until either setEndBoundary returns true or there |
113
|
|
|
* are no more lines to be read. |
114
|
|
|
* |
115
|
|
|
* setEndBoundary returns true if the passed line matches a boundary for the |
116
|
|
|
* $partBuilder itself or any of its parents. |
117
|
|
|
* |
118
|
|
|
* As lines are read, setStreamPartAndContentEndPos is called with the |
119
|
|
|
* passed $handle's read pos (ftell($handle)) to update the position of |
120
|
|
|
* content in the part. |
121
|
|
|
* |
122
|
|
|
* If the entire stream is read and an end boundary was found, i.e. |
123
|
|
|
* $partBuilder->setEndBoundary returns true, true is returned to indicate |
124
|
|
|
* that a content boundary was found. Otherwise false is returned. |
125
|
|
|
* |
126
|
|
|
* @param resource $handle |
127
|
|
|
* @param PartBuilder $partBuilder |
128
|
|
|
* @return boolean true if a mime content boundary was found for |
129
|
|
|
* $partBuilder |
130
|
|
|
*/ |
131
|
|
|
private function findContentBoundary($handle, PartBuilder $partBuilder) |
132
|
|
|
{ |
133
|
|
|
while (!feof($handle)) { |
134
|
|
|
$partBuilder->setStreamPartAndContentEndPos(ftell($handle)); |
135
|
|
|
$line = fgets($handle); |
136
|
|
|
$test = rtrim($line); |
137
|
|
|
if ($partBuilder->setEndBoundary($test)) { |
138
|
|
|
return true; |
139
|
|
|
} |
140
|
|
|
} |
141
|
|
|
$partBuilder->setStreamPartAndContentEndPos(ftell($handle)); |
142
|
|
|
return false; |
143
|
|
|
} |
144
|
|
|
|
145
|
|
|
/** |
146
|
|
|
* Reads content for a non-mime message. If there are uuencoded attachment |
147
|
|
|
* parts in the message (denoted by 'begin' lines), those parts are read and |
148
|
|
|
* added to the passed $partBuilder as children. |
149
|
|
|
* |
150
|
|
|
* @param resource $handle |
151
|
|
|
* @param PartBuilder $partBuilder |
152
|
|
|
* @return string |
153
|
|
|
*/ |
154
|
|
|
protected function readUUEncodedOrPlainTextMessage($handle, PartBuilder $partBuilder) |
155
|
|
|
{ |
156
|
|
|
$partBuilder->setStreamContentStartPos(ftell($handle)); |
157
|
|
|
$part = $partBuilder; |
158
|
|
|
while (!feof($handle)) { |
159
|
|
|
$start = ftell($handle); |
160
|
|
|
$line = trim(fgets($handle)); |
161
|
|
|
if (preg_match('/^begin ([0-7]{3}) (.*)$/', $line, $matches)) { |
162
|
|
|
$part = $this->partBuilderFactory->newPartBuilder( |
163
|
|
|
$this->partFactoryService->getUUEncodedPartFactory() |
164
|
|
|
); |
165
|
|
|
$part->setStreamPartStartPos($start); |
166
|
|
|
// 'begin' line is part of the content |
167
|
|
|
$part->setStreamContentStartPos($start); |
168
|
|
|
$part->setProperty('mode', $matches[1]); |
169
|
|
|
$part->setProperty('filename', $matches[2]); |
170
|
|
|
$partBuilder->addChild($part); |
171
|
|
|
} |
172
|
|
|
$part->setStreamPartAndContentEndPos(ftell($handle)); |
173
|
|
|
} |
174
|
|
|
$partBuilder->setStreamPartEndPos(ftell($handle)); |
175
|
|
|
} |
176
|
|
|
|
177
|
|
|
/** |
178
|
|
|
* Reads content for a single part of a MIME message. |
179
|
|
|
* |
180
|
|
|
* If the part being read is in turn a multipart part, readPart is called on |
181
|
|
|
* it recursively to read its headers and content. |
182
|
|
|
* |
183
|
|
|
* The method tries to read content until a mime boundary (for this part for |
184
|
|
|
* a multipart, or for the parent boundary) is found or EOF is reached. |
185
|
|
|
* |
186
|
|
|
* The start/end positions of the part's content are set on the passed |
187
|
|
|
* $partBuilder as lines are read, as is the part's end position. |
188
|
|
|
* |
189
|
|
|
* @param resource $handle |
190
|
|
|
* @param PartBuilder $partBuilder |
191
|
|
|
*/ |
192
|
|
|
private function readPartContent($handle, PartBuilder $partBuilder) |
193
|
|
|
{ |
194
|
|
|
$partBuilder->setStreamContentStartPos(ftell($handle)); |
195
|
|
|
if ($this->findContentBoundary($handle, $partBuilder) && $partBuilder->isMultiPart()) { |
196
|
|
|
while (!feof($handle) && !$partBuilder->isEndBoundaryFound()) { |
197
|
|
|
$child = $this->partBuilderFactory->newPartBuilder( |
198
|
|
|
$this->partFactoryService->getMimePartFactory() |
199
|
|
|
); |
200
|
|
|
$partBuilder->addChild($child); |
201
|
|
|
$this->readPart($handle, $child); |
202
|
|
|
if ($child->isEndBoundaryFound()) { |
203
|
|
|
$discard = $this->partBuilderFactory->newPartBuilder( |
204
|
|
|
$this->partFactoryService->getMimePartFactory() |
205
|
|
|
); |
206
|
|
|
$discard->setParent($partBuilder); |
207
|
|
|
$this->findContentBoundary($handle, $discard); |
208
|
|
|
} |
209
|
|
|
} |
210
|
|
|
// for non-multipart parts, setStreamContentAndPartEndPos is called |
211
|
|
|
// in findContentBoundary |
212
|
|
|
$partBuilder->setStreamPartEndPos(ftell($handle)); |
213
|
|
|
} |
214
|
|
|
} |
215
|
|
|
|
216
|
|
|
/** |
217
|
|
|
* Reads a part and any of its children, into the passed $partBuilder. |
218
|
|
|
* |
219
|
|
|
* @param resource $handle |
220
|
|
|
* @param PartBuilder $partBuilder |
221
|
|
|
* @param boolean $isMessage |
222
|
|
|
*/ |
223
|
|
|
protected function readPart($handle, PartBuilder $partBuilder, $isMessage = false) |
224
|
|
|
{ |
225
|
|
|
$partBuilder->setStreamPartStartPos(ftell($handle)); |
226
|
|
|
$this->readHeaders($handle, $partBuilder); |
227
|
|
|
if ($isMessage && !$partBuilder->isMime()) { |
228
|
|
|
$this->readUUEncodedOrPlainTextMessage($handle, $partBuilder); |
229
|
|
|
} else { |
230
|
|
|
$this->readPartContent($handle, $partBuilder); |
231
|
|
|
} |
232
|
|
|
} |
233
|
|
|
|
234
|
|
|
/** |
235
|
|
|
* Reads the message from the input stream $handle and returns a PartBuilder |
236
|
|
|
* representing it. |
237
|
|
|
* |
238
|
|
|
* @param resource $handle |
239
|
|
|
* @return PartBuilder |
240
|
|
|
*/ |
241
|
|
|
protected function read($handle) |
242
|
|
|
{ |
243
|
|
|
$partBuilder = $this->partBuilderFactory->newPartBuilder( |
244
|
|
|
$this->partFactoryService->getMessageFactory() |
245
|
|
|
); |
246
|
|
|
$this->readPart($handle, $partBuilder, true); |
247
|
|
|
return $partBuilder; |
248
|
|
|
} |
249
|
|
|
} |
250
|
|
|
|