|
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\Header\Consumer; |
|
8
|
|
|
|
|
9
|
|
|
use ZBateson\MailMimeParser\Header\Part\HeaderPart; |
|
10
|
|
|
use ZBateson\MailMimeParser\Header\Part\Token; |
|
11
|
|
|
use Iterator; |
|
12
|
|
|
|
|
13
|
|
|
/** |
|
14
|
|
|
* Extends GenericConsumer to remove its sub consumers. |
|
15
|
|
|
* |
|
16
|
|
|
* Prior to this, subject headers were parsed using the GenericConsumer which |
|
17
|
|
|
* meant if the subject contained text within parentheses, it would not be |
|
18
|
|
|
* included as part of the returned value in a getHeaderValue. Mime-encoded |
|
19
|
|
|
* parts within quotes would be ignored, and backslash characters denoted an |
|
20
|
|
|
* escaped character. |
|
21
|
|
|
* |
|
22
|
|
|
* From testing in ThunderBird and Outlook web mail it seems quoting parts |
|
23
|
|
|
* doesn't have an effect (e.g. quoting a "mime-literal" encoded part still |
|
24
|
|
|
* comes out decoded), and parts in parentheses (comments) are displayed |
|
25
|
|
|
* normally. |
|
26
|
|
|
* |
|
27
|
|
|
* @author Zaahid Bateson |
|
28
|
|
|
*/ |
|
29
|
|
|
class SubjectConsumer extends GenericConsumer |
|
30
|
|
|
{ |
|
31
|
|
|
/** |
|
32
|
|
|
* Returns an empty array |
|
33
|
|
|
* |
|
34
|
|
|
* @return AbstractConsumer[] the sub-consumers |
|
35
|
|
|
*/ |
|
36
|
2 |
|
protected function getSubConsumers() |
|
37
|
|
|
{ |
|
38
|
2 |
|
return []; |
|
39
|
|
|
} |
|
40
|
|
|
|
|
41
|
|
|
/** |
|
42
|
|
|
* Overridden to preserve whitespace. |
|
43
|
|
|
* |
|
44
|
|
|
* Whitespace between two words is preserved unless the whitespace begins |
|
45
|
|
|
* with a newline (\n or \r\n), in which case the entire string of |
|
46
|
|
|
* whitespace is discarded, and a single space ' ' character is used in its |
|
47
|
|
|
* place. |
|
48
|
|
|
* |
|
49
|
|
|
* @param string $token the token |
|
50
|
|
|
* @param bool $isLiteral set to true if the token represents a literal - |
|
51
|
|
|
* e.g. an escaped token |
|
52
|
|
|
* @return \ZBateson\MailMimeParser\Header\Part\HeaderPart|null the |
|
53
|
|
|
* constructed header part or null if the token should be ignored |
|
54
|
|
|
*/ |
|
55
|
2 |
|
protected function getPartForToken($token, $isLiteral) |
|
56
|
|
|
{ |
|
57
|
2 |
|
if ($isLiteral) { |
|
58
|
|
|
return $this->partFactory->newLiteralPart($token); |
|
59
|
2 |
|
} elseif (preg_match('/^\s+$/', $token)) { |
|
60
|
2 |
|
if (preg_match('/^[\r\n]/', $token)) { |
|
61
|
2 |
|
return $this->partFactory->newToken(' '); |
|
62
|
|
|
} |
|
63
|
2 |
|
return $this->partFactory->newToken($token); |
|
64
|
|
|
} |
|
65
|
2 |
|
return $this->partFactory->newInstance($token); |
|
66
|
|
|
} |
|
67
|
|
|
|
|
68
|
|
|
/** |
|
69
|
|
|
* Returns an array of \ZBateson\MailMimeParser\Header\Part\HeaderPart for |
|
70
|
|
|
* the current token on the iterator. |
|
71
|
|
|
* |
|
72
|
|
|
* Overridden from AbstractConsumer to remove special filtering for |
|
73
|
|
|
* backslash escaping, which also seems to not apply to Subject headers at |
|
74
|
|
|
* least in ThunderBird's implementation. |
|
75
|
|
|
* |
|
76
|
|
|
* @param Iterator $tokens |
|
77
|
|
|
* @return \ZBateson\MailMimeParser\Header\Part\HeaderPart[]|array |
|
78
|
|
|
*/ |
|
79
|
2 |
|
protected function getTokenParts(Iterator $tokens) |
|
80
|
|
|
{ |
|
81
|
2 |
|
return $this->getConsumerTokenParts($tokens); |
|
82
|
|
|
} |
|
83
|
|
|
|
|
84
|
|
|
/** |
|
85
|
|
|
* Overridden to not split out backslash characters and its next character |
|
86
|
|
|
* as a special case defined in AbastractConsumer |
|
87
|
|
|
* |
|
88
|
|
|
* @return string the regex pattern |
|
89
|
|
|
*/ |
|
90
|
2 |
|
protected function getTokenSplitPattern() |
|
91
|
|
|
{ |
|
92
|
2 |
|
$sChars = implode('|', $this->getAllTokenSeparators()); |
|
93
|
2 |
|
return '~(' . $sChars . ')~'; |
|
94
|
|
|
} |
|
95
|
|
|
} |
|
96
|
|
|
|
zbateson /
mail-mime-parser
