|
1
|
|
|
<?php |
|
2
|
|
|
/** |
|
3
|
|
|
* @author Matthijs van den Bos <[email protected]> |
|
4
|
|
|
* @copyright 2013 Matthijs van den Bos |
|
5
|
|
|
*/ |
|
6
|
|
|
|
|
7
|
|
|
namespace VDB\Spider\PersistenceHandler; |
|
8
|
|
|
|
|
9
|
|
|
use Symfony\Component\Finder\Finder; |
|
10
|
|
|
use VDB\Spider\Resource; |
|
11
|
|
|
|
|
12
|
|
|
class FileRawResponsePersistenceHandler implements PersistenceHandler |
|
13
|
|
|
{ |
|
14
|
|
|
/** |
|
15
|
|
|
* @var string the path where all spider results should be persisted. |
|
16
|
|
|
* The results will be grouped in a directory by spider ID. |
|
17
|
|
|
*/ |
|
18
|
|
|
private $path = ''; |
|
19
|
|
|
|
|
20
|
|
|
private $spiderId = ''; |
|
21
|
|
|
|
|
22
|
|
|
private $totalSizePersisted = 0; |
|
23
|
|
|
|
|
24
|
|
|
/** @var \Iterator */ |
|
25
|
|
|
private $iterator; |
|
26
|
|
|
|
|
27
|
|
|
/** |
|
28
|
|
|
* @param string $path the path where all spider results should be persisted. |
|
29
|
|
|
* The results will be grouped in a directory by spider ID. |
|
30
|
|
|
*/ |
|
31
|
|
|
public function __construct($path) |
|
32
|
|
|
{ |
|
33
|
|
|
$this->path = $path; |
|
34
|
|
|
} |
|
35
|
|
|
|
|
36
|
|
|
public function setSpiderId($spiderId) |
|
37
|
|
|
{ |
|
38
|
|
|
$this->spiderId = $spiderId; |
|
39
|
|
|
|
|
40
|
|
|
// create the path |
|
41
|
|
|
if (!file_exists($this->getResultPath())) { |
|
42
|
|
|
mkdir($this->getResultPath(), 0700, true); |
|
43
|
|
|
} |
|
44
|
|
|
} |
|
45
|
|
|
|
|
46
|
|
|
public function count() |
|
47
|
|
|
{ |
|
48
|
|
|
return $this->getFinder()->count(); |
|
49
|
|
|
} |
|
50
|
|
|
|
|
51
|
|
|
private function getResultPath() |
|
52
|
|
|
{ |
|
53
|
|
|
return $this->path . DIRECTORY_SEPARATOR . $this->spiderId . DIRECTORY_SEPARATOR; |
|
54
|
|
|
} |
|
55
|
|
|
|
|
56
|
|
|
public function persist(Resource $resource) |
|
57
|
|
|
{ |
|
58
|
|
|
$fileName = urlencode($resource->getUri()->toString()); |
|
59
|
|
|
$file = new \SplFileObject($this->getResultPath() . $fileName, 'w'); |
|
60
|
|
|
$rawResponse = $resource->getResponse()->__toString(); |
|
61
|
|
|
$this->totalSizePersisted += $file->fwrite($rawResponse); |
|
62
|
|
|
} |
|
63
|
|
|
|
|
64
|
|
|
private function getFinder() |
|
65
|
|
|
{ |
|
66
|
|
|
if (!$this->finder instanceof Finder) { |
|
67
|
|
|
$this->finder = Finder::create()->files()->in($this->getResultPath()); |
|
|
|
|
|
|
68
|
|
|
} |
|
69
|
|
|
return $this->finder; |
|
70
|
|
|
} |
|
71
|
|
|
|
|
72
|
|
|
private function getIterator() |
|
73
|
|
|
{ |
|
74
|
|
|
if (!$this->iterator instanceof \Iterator) { |
|
75
|
|
|
$this->iterator = $this->getFinder()->getIterator(); |
|
76
|
|
|
} |
|
77
|
|
|
return $this->iterator; |
|
78
|
|
|
} |
|
79
|
|
|
|
|
80
|
|
|
/** |
|
81
|
|
|
* @return Resource |
|
82
|
|
|
*/ |
|
83
|
|
|
public function current() |
|
84
|
|
|
{ |
|
85
|
|
|
return $this->getIterator()->current()->getContents(); |
|
86
|
|
|
} |
|
87
|
|
|
|
|
88
|
|
|
/** |
|
89
|
|
|
* @return void |
|
90
|
|
|
*/ |
|
91
|
|
|
public function next() |
|
92
|
|
|
{ |
|
93
|
|
|
$this->getIterator()->next(); |
|
94
|
|
|
} |
|
95
|
|
|
|
|
96
|
|
|
/** |
|
97
|
|
|
* @return int |
|
98
|
|
|
*/ |
|
99
|
|
|
public function key() |
|
100
|
|
|
{ |
|
101
|
|
|
return $this->getIterator()->key(); |
|
102
|
|
|
} |
|
103
|
|
|
|
|
104
|
|
|
/** |
|
105
|
|
|
* @return boolean |
|
106
|
|
|
*/ |
|
107
|
|
|
public function valid() |
|
108
|
|
|
{ |
|
109
|
|
|
return $this->getIterator()->valid(); |
|
110
|
|
|
} |
|
111
|
|
|
|
|
112
|
|
|
/** |
|
113
|
|
|
* @return void |
|
114
|
|
|
*/ |
|
115
|
|
|
public function rewind() |
|
116
|
|
|
{ |
|
117
|
|
|
$this->getIterator()->rewind(); |
|
118
|
|
|
} |
|
119
|
|
|
} |
|
120
|
|
|
|
mvdbos /
php-spider
In PHP it is possible to write to properties without declaring them. For example, the following is perfectly valid PHP code:
Generally, it is a good practice to explictly declare properties to avoid accidental typos and provide IDE auto-completion: