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
|
|
|
|
8
|
|
|
namespace ZBateson\MailMimeParser\Header\Consumer; |
9
|
|
|
|
10
|
|
|
use ArrayObject; |
11
|
|
|
use ZBateson\MailMimeParser\Header\IHeaderPart; |
12
|
|
|
use ZBateson\MailMimeParser\Header\Part\CommentPart; |
13
|
|
|
use ZBateson\MailMimeParser\Header\Part\MimeLiteralPart; |
14
|
|
|
use ZBateson\MailMimeParser\Header\Part\SplitParameterToken; |
15
|
|
|
use ZBateson\MailMimeParser\Header\Part\Token; |
16
|
|
|
|
17
|
|
|
/** |
18
|
|
|
* Reads headers separated into parameters consisting of an optional main value, |
19
|
|
|
* and subsequent name/value pairs - for example text/html; charset=utf-8. |
20
|
|
|
* |
21
|
|
|
* A ParameterConsumerService's parts are separated by a semi-colon. Its |
22
|
|
|
* name/value pairs are separated with an '=' character. |
23
|
|
|
* |
24
|
|
|
* Parts may be mime-encoded entities. Additionally, a value can be quoted and |
25
|
|
|
* comments may exist. |
26
|
|
|
* |
27
|
|
|
* @author Zaahid Bateson |
28
|
|
|
*/ |
29
|
|
|
class ParameterConsumerService extends GenericConsumerService |
30
|
|
|
{ |
31
|
|
|
/** |
32
|
|
|
* Returns semi-colon and equals char as token separators. |
33
|
|
|
* |
34
|
|
|
* @return string[] |
35
|
|
|
*/ |
36
|
18 |
|
protected function getTokenSeparators() : array |
37
|
|
|
{ |
38
|
18 |
|
return [';', '=']; |
39
|
|
|
} |
40
|
|
|
|
41
|
|
|
/** |
42
|
|
|
* Overridden to use a specialized regex for finding mime-encoded parts |
43
|
|
|
* (RFC 2047). |
44
|
|
|
* |
45
|
|
|
* Some implementations seem to place mime-encoded parts within quoted |
46
|
|
|
* parameters, and split the mime-encoded parts across multiple split |
47
|
|
|
* parameters. The specialized regex doesn't allow double quotes inside a |
48
|
|
|
* mime encoded part, so it can be "continued" in another parameter. |
49
|
|
|
* |
50
|
|
|
* @return string the regex pattern |
51
|
|
|
*/ |
52
|
18 |
|
protected function getTokenSplitPattern() : string |
53
|
|
|
{ |
54
|
18 |
|
$sChars = \implode('|', $this->getAllTokenSeparators()); |
55
|
18 |
|
$mimePartPattern = MimeLiteralPart::MIME_PART_PATTERN_NO_QUOTES; |
56
|
18 |
|
return '~(' . $mimePartPattern . '|\\\\.|' . $sChars . ')~'; |
57
|
|
|
} |
58
|
|
|
|
59
|
|
|
/** |
60
|
|
|
* Creates and returns a \ZBateson\MailMimeParser\Header\Part\Token out of |
61
|
|
|
* the passed string token and returns it, unless the token is an escaped |
62
|
|
|
* literal, in which case a LiteralPart is returned. |
63
|
|
|
*/ |
64
|
118 |
|
protected function getPartForToken(string $token, bool $isLiteral) : ?IHeaderPart |
65
|
|
|
{ |
66
|
118 |
|
if ($isLiteral) { |
67
|
1 |
|
return $this->partFactory->newLiteralPart($token); |
68
|
|
|
} |
69
|
118 |
|
return $this->partFactory->newToken($token); |
70
|
|
|
} |
71
|
|
|
|
72
|
|
|
/** |
73
|
|
|
* Adds the passed parameter with the given name and value to a |
74
|
|
|
* SplitParameterToken, at the passed index. If one with the given name |
75
|
|
|
* doesn't exist, it is created. |
76
|
|
|
*/ |
77
|
13 |
|
private function addToSplitPart( |
78
|
|
|
ArrayObject $splitParts, |
79
|
|
|
string $name, |
80
|
|
|
string $value, |
81
|
|
|
int $index, |
82
|
|
|
bool $isEncoded |
83
|
|
|
) : ?SplitParameterToken { |
84
|
13 |
|
$ret = null; |
85
|
13 |
|
if (!isset($splitParts[$name])) { |
86
|
13 |
|
$ret = $this->partFactory->newSplitParameterToken($name); |
87
|
13 |
|
$splitParts[$name] = $ret; |
88
|
|
|
} |
89
|
13 |
|
$splitParts[$name]->addPart($value, $isEncoded, $index); |
90
|
13 |
|
return $ret; |
91
|
|
|
} |
92
|
|
|
|
93
|
|
|
/** |
94
|
|
|
* Instantiates and returns either a MimeLiteralPart if $strName is empty, |
95
|
|
|
* a SplitParameterToken if the parameter is a split parameter and is the |
96
|
|
|
* first in a series, null if it's a split parameter but is not the first |
97
|
|
|
* part in its series, or a ParameterPart is returned otherwise. |
98
|
|
|
* |
99
|
|
|
* If the part is a SplitParameterToken, it's added to the passed |
100
|
|
|
* $splitParts as well with its name as a key. |
101
|
|
|
*/ |
102
|
118 |
|
private function getPartFor(string $strName, string $strValue, ArrayObject $splitParts) : ?IHeaderPart |
103
|
|
|
{ |
104
|
118 |
|
if ($strName === '') { |
105
|
118 |
|
return $this->partFactory->newMimeLiteralPart($strValue); |
106
|
117 |
|
} elseif (\preg_match('~^\s*([^\*]+)\*(\d*)(\*)?$~', $strName, $matches)) { |
107
|
13 |
|
return $this->addToSplitPart( |
108
|
13 |
|
$splitParts, |
109
|
13 |
|
$matches[1], |
110
|
13 |
|
$strValue, |
111
|
13 |
|
(int) $matches[2], |
112
|
13 |
|
(($matches[2] === '') || !empty($matches[3])) |
113
|
13 |
|
); |
114
|
|
|
} |
115
|
105 |
|
return $this->partFactory->newParameterPart($strName, $strValue); |
116
|
|
|
} |
117
|
|
|
|
118
|
|
|
/** |
119
|
|
|
* Handles parameter separator tokens during final processing. |
120
|
|
|
* |
121
|
|
|
* If the end token is found, a new IHeaderPart is assigned to the passed |
122
|
|
|
* $combined array. If an '=' character is found, $strCat is assigned to |
123
|
|
|
* $strName and emptied. |
124
|
|
|
* |
125
|
|
|
* Returns true if the token was processed, and false otherwise. |
126
|
|
|
* |
127
|
|
|
*/ |
128
|
118 |
|
private function processTokenPart(string $tokenValue, ArrayObject $combined, ArrayObject $splitParts, string &$strName, string &$strCat) : bool |
129
|
|
|
{ |
130
|
118 |
|
if ($tokenValue === ';') { |
131
|
118 |
|
$combined[] = $this->getPartFor($strName, $strCat, $splitParts); |
132
|
118 |
|
$strName = ''; |
133
|
118 |
|
$strCat = ''; |
134
|
118 |
|
return true; |
135
|
118 |
|
} elseif ($tokenValue === '=' && $strCat !== '') { |
136
|
117 |
|
$strName = $strCat; |
137
|
117 |
|
$strCat = ''; |
138
|
117 |
|
return true; |
139
|
|
|
} |
140
|
118 |
|
return false; |
141
|
|
|
} |
142
|
|
|
|
143
|
|
|
/** |
144
|
|
|
* Loops over parts in the passed array, creating ParameterParts out of any |
145
|
|
|
* parsed SplitParameterTokens, replacing them in the array. |
146
|
|
|
* |
147
|
|
|
* The method then calls filterIgnoreSpaces to filter out empty elements in |
148
|
|
|
* the combined array and returns an array. |
149
|
|
|
* |
150
|
|
|
* @return IHeaderPart[] |
151
|
|
|
*/ |
152
|
118 |
|
private function finalizeParameterParts(ArrayObject $combined) : array |
153
|
|
|
{ |
154
|
118 |
|
foreach ($combined as $key => $part) { |
155
|
118 |
|
if ($part instanceof SplitParameterToken) { |
156
|
13 |
|
$combined[$key] = $this->partFactory->newParameterPart( |
157
|
13 |
|
$part->getName(), |
158
|
13 |
|
$part->getValue(), |
|
|
|
|
159
|
13 |
|
$part->getLanguage() |
160
|
13 |
|
); |
161
|
|
|
} |
162
|
|
|
} |
163
|
118 |
|
return $this->filterIgnoredSpaces($combined->getArrayCopy()); |
164
|
|
|
} |
165
|
|
|
|
166
|
|
|
/** |
167
|
|
|
* Post processing involves creating Part\LiteralPart or Part\ParameterPart |
168
|
|
|
* objects out of created Token and LiteralParts. |
169
|
|
|
* |
170
|
|
|
* @param IHeaderPart[] $parts The parsed parts. |
171
|
|
|
* @return IHeaderPart[] Array of resulting final parts. |
172
|
|
|
*/ |
173
|
118 |
|
protected function processParts(array $parts) : array |
174
|
|
|
{ |
175
|
118 |
|
$combined = new ArrayObject(); |
176
|
118 |
|
$splitParts = new ArrayObject(); |
177
|
118 |
|
$strCat = ''; |
178
|
118 |
|
$strName = ''; |
179
|
118 |
|
$parts[] = $this->partFactory->newToken(';'); |
180
|
118 |
|
foreach ($parts as $part) { |
181
|
118 |
|
$pValue = $part->getValue(); |
182
|
118 |
|
if (($part instanceof Token || $part instanceof CommentPart) && $this->processTokenPart($pValue, $combined, $splitParts, $strName, $strCat)) { |
|
|
|
|
183
|
118 |
|
continue; |
184
|
118 |
|
} elseif ($part instanceof CommentPart) { |
185
|
1 |
|
$combined[] = $part; |
186
|
|
|
} else { |
187
|
118 |
|
$strCat .= $pValue; |
188
|
|
|
} |
189
|
|
|
} |
190
|
118 |
|
return $this->finalizeParameterParts($combined); |
191
|
|
|
} |
192
|
|
|
} |
193
|
|
|
|