|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
namespace Spatie\Crawler\Handlers; |
|
4
|
|
|
|
|
5
|
|
|
use Psr\Http\Message\ResponseInterface; |
|
6
|
|
|
use Psr\Http\Message\StreamInterface; |
|
7
|
|
|
use Psr\Http\Message\UriInterface; |
|
8
|
|
|
use Spatie\Crawler\Crawler; |
|
9
|
|
|
use Spatie\Crawler\CrawlProfile; |
|
10
|
|
|
use Spatie\Crawler\CrawlQueue\CrawlQueue; |
|
11
|
|
|
use Spatie\Crawler\CrawlSubdomains; |
|
12
|
|
|
use Spatie\Crawler\CrawlUrl; |
|
13
|
|
|
use Spatie\Robots\RobotsHeaders; |
|
14
|
|
|
use Spatie\Robots\RobotsMeta; |
|
15
|
|
|
|
|
16
|
|
|
class CrawlRequestFulfilled |
|
17
|
|
|
{ |
|
18
|
|
|
/** @var \Spatie\Crawler\Crawler */ |
|
19
|
|
|
protected $crawler; |
|
20
|
|
|
|
|
21
|
|
|
/** @var \Spatie\Crawler\CrawlQueue\CrawlQueue */ |
|
22
|
|
|
protected $crawlQueue; |
|
23
|
|
|
|
|
24
|
|
|
/** @var \Spatie\Crawler\CrawlProfile */ |
|
25
|
|
|
protected $crawlProfile; |
|
26
|
|
|
|
|
27
|
|
|
/** @var array[\Spatie\Crawler\CrawlObserver] */ |
|
28
|
|
|
protected $crawlObservers; |
|
29
|
|
|
|
|
30
|
|
|
/** @var \Psr\Http\Message\UriInterface */ |
|
31
|
|
|
protected $baseUrl; |
|
32
|
|
|
|
|
33
|
|
|
/** @var int */ |
|
34
|
|
|
protected $maximumResponseSize; |
|
35
|
|
|
|
|
36
|
|
|
/** @var bool */ |
|
37
|
|
|
protected $respectRobots; |
|
38
|
|
|
|
|
39
|
|
|
public function __construct( |
|
40
|
|
|
Crawler $crawler, |
|
41
|
|
|
UriInterface $baseUrl, |
|
42
|
|
|
CrawlQueue $crawlQueue, |
|
43
|
|
|
CrawlProfile $crawlProfile, |
|
44
|
|
|
array $crawlObservers, |
|
45
|
|
|
int $maximumResponseSize, |
|
46
|
|
|
bool $respectRobots |
|
47
|
|
|
) { |
|
48
|
|
|
$this->crawler = $crawler; |
|
49
|
|
|
$this->baseUrl = $baseUrl; |
|
50
|
|
|
$this->crawlQueue = $crawlQueue; |
|
51
|
|
|
$this->crawlProfile = $crawlProfile; |
|
52
|
|
|
$this->crawlObservers = $crawlObservers; |
|
53
|
|
|
$this->maximumResponseSize = $maximumResponseSize; |
|
54
|
|
|
$this->respectRobots = $respectRobots; |
|
55
|
|
|
} |
|
56
|
|
|
|
|
57
|
|
|
public function __invoke(ResponseInterface $response, $index) |
|
58
|
|
|
{ |
|
59
|
|
|
$crawlUrl = $this->crawlQueue->getUrlById($index); |
|
60
|
|
|
|
|
61
|
|
|
$body = $this->convertBodyToString($response->getBody(), $this->maximumResponseSize); |
|
62
|
|
|
|
|
63
|
|
|
$robotsHeaders = RobotsHeaders::create($response->getHeaders()); |
|
64
|
|
|
|
|
65
|
|
|
$robotsMeta = RobotsMeta::create($body); |
|
66
|
|
|
|
|
67
|
|
|
if (! $this->mayIndex($robotsHeaders, $robotsMeta)) { |
|
68
|
|
|
return; |
|
69
|
|
|
} |
|
70
|
|
|
|
|
71
|
|
|
$this->handleCrawled($response, $crawlUrl); |
|
72
|
|
|
|
|
73
|
|
|
if (! $this->crawlProfile instanceof CrawlSubdomains) { |
|
74
|
|
|
if ($crawlUrl->url->getHost() !== $this->baseUrl->getHost()) { |
|
75
|
|
|
return; |
|
76
|
|
|
} |
|
77
|
|
|
} |
|
78
|
|
|
|
|
79
|
|
|
if (! $this->mayFollow($robotsHeaders, $robotsMeta)) { |
|
80
|
|
|
return; |
|
81
|
|
|
} |
|
82
|
|
|
|
|
83
|
|
|
$this->crawler->addAllLinksToCrawlQueue( |
|
84
|
|
|
$body, |
|
85
|
|
|
$crawlUrl->url |
|
86
|
|
|
); |
|
87
|
|
|
} |
|
88
|
|
|
|
|
89
|
|
|
protected function convertBodyToString(StreamInterface $bodyStream, $readMaximumBytes = 1024 * 1024 * 2): string |
|
90
|
|
|
{ |
|
91
|
|
|
$bodyStream->rewind(); |
|
92
|
|
|
|
|
93
|
|
|
$body = $bodyStream->read($readMaximumBytes); |
|
94
|
|
|
|
|
95
|
|
|
return $body; |
|
96
|
|
|
} |
|
97
|
|
|
|
|
98
|
|
|
protected function handleCrawled(ResponseInterface $response, CrawlUrl $crawlUrl) |
|
99
|
|
|
{ |
|
100
|
|
|
foreach ($this->crawlObservers as $crawlObserver) { |
|
101
|
|
|
$crawlObserver->crawled( |
|
102
|
|
|
$crawlUrl->url, |
|
103
|
|
|
$response, |
|
104
|
|
|
$crawlUrl->foundOnUrl |
|
105
|
|
|
); |
|
106
|
|
|
} |
|
107
|
|
|
} |
|
108
|
|
|
|
|
109
|
|
View Code Duplication |
protected function mayIndex(RobotsHeaders $robotsHeaders, RobotsMeta $robotsMeta): bool |
|
|
|
|
|
|
110
|
|
|
{ |
|
111
|
|
|
if (! $this->respectRobots) { |
|
112
|
|
|
return true; |
|
113
|
|
|
} |
|
114
|
|
|
|
|
115
|
|
|
if (! $robotsHeaders->mayIndex()) { |
|
116
|
|
|
return false; |
|
117
|
|
|
} |
|
118
|
|
|
|
|
119
|
|
|
if (! $robotsMeta->mayIndex()) { |
|
120
|
|
|
return false; |
|
121
|
|
|
} |
|
122
|
|
|
|
|
123
|
|
|
return true; |
|
124
|
|
|
} |
|
125
|
|
|
|
|
126
|
|
View Code Duplication |
protected function mayFollow(RobotsHeaders $robotsHeaders, RobotsMeta $robotsMeta): bool |
|
|
|
|
|
|
127
|
|
|
{ |
|
128
|
|
|
if (! $this->respectRobots) { |
|
129
|
|
|
return true; |
|
130
|
|
|
} |
|
131
|
|
|
|
|
132
|
|
|
if (! $robotsHeaders->mayFollow()) { |
|
133
|
|
|
return false; |
|
134
|
|
|
} |
|
135
|
|
|
|
|
136
|
|
|
if (! $robotsMeta->mayFollow()) { |
|
137
|
|
|
return false; |
|
138
|
|
|
} |
|
139
|
|
|
|
|
140
|
|
|
return true; |
|
141
|
|
|
} |
|
142
|
|
|
} |
|
143
|
|
|
|
Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation.
You can also find more detailed suggestions in the “Code” section of your repository.