1
|
|
|
<?php |
2
|
|
|
namespace VDB\Spider; |
3
|
|
|
|
4
|
|
|
use Symfony\Component\EventDispatcher\Event; |
5
|
|
|
use Symfony\Component\EventDispatcher\EventDispatcher; |
6
|
|
|
use Symfony\Component\EventDispatcher\EventDispatcherInterface; |
7
|
|
|
use Symfony\Component\EventDispatcher\GenericEvent; |
8
|
|
|
use VDB\Spider\Discoverer\DiscovererSet; |
9
|
|
|
use VDB\Spider\Event\SpiderEvents; |
10
|
|
|
use VDB\Spider\Exception\QueueException; |
11
|
|
|
use VDB\Spider\QueueManager\QueueManagerInterface; |
12
|
|
|
use VDB\Spider\QueueManager\InMemoryQueueManager; |
13
|
|
|
use VDB\Spider\Uri\DiscoveredUri; |
14
|
|
|
use VDB\Spider\Downloader\DownloaderInterface; |
15
|
|
|
use VDB\Spider\Downloader\Downloader; |
16
|
|
|
use VDB\Uri\UriInterface; |
17
|
|
|
use VDB\Uri\Uri; |
18
|
|
|
|
19
|
|
|
/** |
20
|
|
|
* |
21
|
|
|
*/ |
22
|
|
|
class Spider |
23
|
|
|
{ |
24
|
|
|
/** @var DownloaderInterface */ |
25
|
|
|
private $downloader; |
26
|
|
|
|
27
|
|
|
/** @var QueueManagerInterface */ |
28
|
|
|
private $queueManager; |
29
|
|
|
|
30
|
|
|
/** @var EventDispatcherInterface */ |
31
|
|
|
private $dispatcher; |
32
|
|
|
|
33
|
|
|
/** @var DiscovererSet */ |
34
|
|
|
private $discovererSet; |
35
|
|
|
|
36
|
|
|
/** @var DiscoveredUri The URI of the site to spider */ |
37
|
|
|
private $seed = array(); |
38
|
|
|
|
39
|
|
|
/** @var string the unique id of this spider instance */ |
40
|
|
|
private $spiderId; |
41
|
|
|
|
42
|
|
|
/** |
43
|
|
|
* @param string $seed the URI to start crawling |
44
|
|
|
* @param string|null $spiderId |
45
|
|
|
*/ |
46
|
|
|
public function __construct($seed, $spiderId = null) |
47
|
|
|
{ |
48
|
|
|
$this->setSeed($seed); |
49
|
|
|
if (null !== $spiderId) { |
50
|
|
|
$this->spiderId = $spiderId; |
51
|
|
|
} else { |
52
|
|
|
$this->spiderId = md5($seed . microtime(true)); |
53
|
|
|
} |
54
|
|
|
|
55
|
|
|
// This makes the spider handle signals gracefully and allows us to do cleanup |
56
|
|
|
if (php_sapi_name() == 'cli') { |
57
|
|
|
declare(ticks = 1); |
58
|
|
|
if (function_exists('pcntl_signal')) { |
59
|
|
|
pcntl_signal(SIGTERM, array($this, 'handleSignal')); |
60
|
|
|
pcntl_signal(SIGINT, array($this, 'handleSignal')); |
61
|
|
|
pcntl_signal(SIGHUP, array($this, 'handleSignal')); |
62
|
|
|
pcntl_signal(SIGQUIT, array($this, 'handleSignal')); |
63
|
|
|
} |
64
|
|
|
} |
65
|
|
|
} |
66
|
|
|
|
67
|
|
|
/** |
68
|
|
|
* Starts crawling the URI provided on instantiation |
69
|
|
|
* |
70
|
|
|
* @return void |
71
|
|
|
*/ |
72
|
|
|
public function crawl() |
73
|
|
|
{ |
74
|
|
|
$this->getQueueManager()->addUri($this->seed); |
75
|
|
|
$this->getDownloader()->getPersistenceHandler()->setSpiderId($this->spiderId); |
76
|
|
|
|
77
|
|
|
$this->doCrawl(); |
78
|
|
|
} |
79
|
|
|
|
80
|
|
|
/** |
81
|
|
|
* param DiscovererSet $discovererSet |
82
|
|
|
*/ |
83
|
|
|
public function setDiscovererSet(DiscovererSet $discovererSet) |
84
|
|
|
{ |
85
|
|
|
$this->discovererSet = $discovererSet; |
86
|
|
|
} |
87
|
|
|
|
88
|
|
|
/** |
89
|
|
|
* @return DiscovererSet |
90
|
|
|
*/ |
91
|
|
|
public function getDiscovererSet() |
92
|
|
|
{ |
93
|
|
|
if (!$this->discovererSet) { |
94
|
|
|
$this->discovererSet = new DiscovererSet(); |
95
|
|
|
} |
96
|
|
|
|
97
|
|
|
return $this->discovererSet; |
98
|
|
|
} |
99
|
|
|
|
100
|
|
|
/** |
101
|
|
|
* param QueueManagerInterface $queueManager |
102
|
|
|
*/ |
103
|
|
|
public function setQueueManager(QueueManagerInterface $queueManager) |
104
|
|
|
{ |
105
|
|
|
$this->queueManager = $queueManager; |
106
|
|
|
} |
107
|
|
|
|
108
|
|
|
/** |
109
|
|
|
* @return QueueManagerInterface |
110
|
|
|
*/ |
111
|
|
|
public function getQueueManager() |
112
|
|
|
{ |
113
|
|
|
if (!$this->queueManager) { |
114
|
|
|
$this->queueManager = new InMemoryQueueManager(); |
115
|
|
|
} |
116
|
|
|
|
117
|
|
|
return $this->queueManager; |
118
|
|
|
} |
119
|
|
|
|
120
|
|
|
/** |
121
|
|
|
* @param DownloaderInterface $downloader |
122
|
|
|
* @return $this |
123
|
|
|
*/ |
124
|
|
|
public function setDownloader(DownloaderInterface $downloader) |
125
|
|
|
{ |
126
|
|
|
$this->downloader = $downloader; |
127
|
|
|
|
128
|
|
|
return $this; |
129
|
|
|
} |
130
|
|
|
|
131
|
|
|
/** |
132
|
|
|
* @return DownloaderInterface |
133
|
|
|
*/ |
134
|
|
|
public function getDownloader() |
135
|
|
|
{ |
136
|
|
|
if (!$this->downloader) { |
137
|
|
|
$this->downloader = new Downloader(); |
138
|
|
|
} |
139
|
|
|
return $this->downloader; |
140
|
|
|
} |
141
|
|
|
|
142
|
|
|
/** |
143
|
|
|
* @param EventDispatcherInterface $eventDispatcher |
144
|
|
|
* @return $this |
145
|
|
|
*/ |
146
|
|
|
public function setDispatcher(EventDispatcherInterface $eventDispatcher) |
147
|
|
|
{ |
148
|
|
|
$this->dispatcher = $eventDispatcher; |
149
|
|
|
|
150
|
|
|
return $this; |
151
|
|
|
} |
152
|
|
|
|
153
|
|
|
/** |
154
|
|
|
* @return EventDispatcherInterface |
155
|
|
|
*/ |
156
|
|
|
public function getDispatcher() |
157
|
|
|
{ |
158
|
|
|
if (!$this->dispatcher) { |
159
|
|
|
$this->dispatcher = new EventDispatcher(); |
160
|
|
|
} |
161
|
|
|
return $this->dispatcher; |
162
|
|
|
} |
163
|
|
|
|
164
|
|
|
public function handleSignal($signal) |
165
|
|
|
{ |
166
|
|
|
switch ($signal) { |
167
|
|
|
case SIGTERM: |
168
|
|
|
case SIGKILL: |
169
|
|
|
case SIGINT: |
170
|
|
|
case SIGQUIT: |
171
|
|
|
$this->dispatch(SpiderEvents::SPIDER_CRAWL_USER_STOPPED); |
172
|
|
|
} |
173
|
|
|
} |
174
|
|
|
|
175
|
|
|
/** |
176
|
|
|
* Function that crawls each provided URI |
177
|
|
|
* It applies all processors and listeners set on the Spider |
178
|
|
|
* |
179
|
|
|
* This is a either depth first algorithm as explained here: |
180
|
|
|
* https://en.wikipedia.org/wiki/Depth-first_search#Example |
181
|
|
|
* Note that because we don't do it recursive, but iteratively, |
182
|
|
|
* results will be in a different order from the example, because |
183
|
|
|
* we always take the right-most child first, whereas a recursive |
184
|
|
|
* variant would always take the left-most child first |
185
|
|
|
* |
186
|
|
|
* or |
187
|
|
|
* |
188
|
|
|
* a breadth first algorithm |
189
|
|
|
* |
190
|
|
|
* @return void |
191
|
|
|
*/ |
192
|
|
|
private function doCrawl() |
193
|
|
|
{ |
194
|
|
|
while ($currentUri = $this->getQueueManager()->next()) { |
195
|
|
|
if ($this->getDownloader()->isDownLoadLimitExceeded()) { |
196
|
|
|
break; |
197
|
|
|
} |
198
|
|
|
|
199
|
|
|
if (!$resource = $this->getDownloader()->download($currentUri)) { |
200
|
|
|
continue; |
201
|
|
|
} |
202
|
|
|
|
203
|
|
|
$this->dispatch( |
204
|
|
|
SpiderEvents::SPIDER_CRAWL_RESOURCE_PERSISTED, |
205
|
|
|
new GenericEvent($this, array('uri' => $currentUri)) |
206
|
|
|
); |
207
|
|
|
|
208
|
|
|
// Once the document is enqueued, apply the discoverers to look for more links to follow |
209
|
|
|
$discoveredUris = $this->getDiscovererSet()->discover($resource); |
210
|
|
|
|
211
|
|
|
foreach ($discoveredUris as $uri) { |
212
|
|
|
try { |
213
|
|
|
$this->getQueueManager()->addUri($uri); |
214
|
|
|
} catch (QueueException $e) { |
215
|
|
|
// when the queue size is exceeded, we stop discovering |
216
|
|
|
break; |
217
|
|
|
} |
218
|
|
|
} |
219
|
|
|
} |
220
|
|
|
} |
221
|
|
|
|
222
|
|
|
/** |
223
|
|
|
* A shortcut for EventDispatcher::dispatch() |
224
|
|
|
* |
225
|
|
|
* @param string $eventName |
226
|
|
|
* @param null|Event $event |
227
|
|
|
*/ |
228
|
|
|
private function dispatch($eventName, Event $event = null) |
229
|
|
|
{ |
230
|
|
|
$this->getDispatcher()->dispatch($eventName, $event); |
|
|
|
|
231
|
|
|
} |
232
|
|
|
|
233
|
|
|
/** |
234
|
|
|
* @param string $uri |
235
|
|
|
*/ |
236
|
|
|
private function setSeed($uri) |
237
|
|
|
{ |
238
|
|
|
$this->seed = new DiscoveredUri(new Uri($uri)); |
239
|
|
|
$this->seed->setDepthFound(0); |
240
|
|
|
} |
241
|
|
|
} |
242
|
|
|
|
This check compares calls to functions or methods with their respective definitions. If the call has more arguments than are defined, it raises an issue.
If a function is defined several times with a different number of parameters, the check may pick up the wrong definition and report false positives. One codebase where this has been known to happen is Wordpress. Please note the @ignore annotation hint above.