|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
declare(strict_types=1); |
|
4
|
|
|
|
|
5
|
|
|
namespace AOE\Crawler\Service; |
|
6
|
|
|
|
|
7
|
|
|
/* |
|
8
|
|
|
* (c) 2020 AOE GmbH <[email protected]> |
|
9
|
|
|
* |
|
10
|
|
|
* This file is part of the TYPO3 Crawler Extension. |
|
11
|
|
|
* |
|
12
|
|
|
* It is free software; you can redistribute it and/or modify it under |
|
13
|
|
|
* the terms of the GNU General Public License, either version 2 |
|
14
|
|
|
* of the License, or any later version. |
|
15
|
|
|
* |
|
16
|
|
|
* For the full copyright and license information, please read the |
|
17
|
|
|
* LICENSE.txt file that was distributed with this source code. |
|
18
|
|
|
* |
|
19
|
|
|
* The TYPO3 project - inspiring people to share! |
|
20
|
|
|
*/ |
|
21
|
|
|
|
|
22
|
|
|
use AOE\Crawler\Controller\CrawlerController; |
|
23
|
|
|
use AOE\Crawler\Exception\CrawlerObjectException; |
|
24
|
|
|
use TYPO3\CMS\Core\Domain\Repository\PageRepository; |
|
25
|
|
|
use TYPO3\CMS\Core\Utility\GeneralUtility; |
|
26
|
|
|
|
|
27
|
|
|
class QueueService |
|
28
|
|
|
{ |
|
29
|
|
|
/** |
|
30
|
|
|
* @var CrawlerController |
|
31
|
|
|
*/ |
|
32
|
|
|
protected $crawlerController; |
|
33
|
|
|
|
|
34
|
|
|
public function addPageToQueue(int $pageUid, int $time = 0): void |
|
35
|
|
|
{ |
|
36
|
|
|
$crawler = $this->findCrawler(); |
|
37
|
|
|
/** |
|
38
|
|
|
* Todo: Switch back to getPage(); when dropping support for TYPO3 9 LTS - TNM |
|
39
|
|
|
* This switch to getPage_noCheck() is needed as TYPO3 9 LTS doesn't return dokType < 200, therefore automatically |
|
40
|
|
|
* adding pages to crawler queue when editing page-titles from the page tree directly was not working. |
|
41
|
|
|
*/ |
|
42
|
|
|
$pageData = GeneralUtility::makeInstance(PageRepository::class)->getPage_noCheck($pageUid, true); |
|
43
|
|
|
$configurations = $crawler->getUrlsForPageRow($pageData); |
|
44
|
|
|
// Currently this is only used from the DataHandlerHook, and we don't know of any allowed/disallowed configurations, |
|
45
|
|
|
// when clearing the cache, therefore we allow all configurations in this case. |
|
46
|
|
|
// This next lines could be skipped as it will return the incomming configurations, but for visibility and |
|
47
|
|
|
// later implementation it's kept as it do no harm. |
|
48
|
|
|
$allowedConfigurations = []; |
|
49
|
|
|
$configurations = ConfigurationService::removeDisallowedConfigurations($allowedConfigurations, $configurations); |
|
50
|
|
|
$downloadUrls = []; |
|
51
|
|
|
$duplicateTrack = []; |
|
52
|
|
|
|
|
53
|
|
|
if (is_array($configurations)) { |
|
|
|
|
|
|
54
|
|
|
foreach ($configurations as $configuration) { |
|
55
|
|
|
//enable inserting of entries |
|
56
|
|
|
$crawler->registerQueueEntriesInternallyOnly = false; |
|
57
|
|
|
$crawler->urlListFromUrlArray( |
|
58
|
|
|
$configuration, |
|
59
|
|
|
$pageData, |
|
60
|
|
|
$time, |
|
61
|
|
|
300, |
|
62
|
|
|
true, |
|
63
|
|
|
false, |
|
64
|
|
|
$duplicateTrack, |
|
65
|
|
|
$downloadUrls, |
|
66
|
|
|
array_keys($this->getCrawlerProcInstructions()) |
|
67
|
|
|
); |
|
68
|
|
|
|
|
69
|
|
|
//reset the queue because the entries have been written to the db |
|
70
|
|
|
unset($crawler->queueEntries); |
|
71
|
|
|
} |
|
72
|
|
|
} |
|
73
|
|
|
} |
|
74
|
|
|
|
|
75
|
|
|
/** |
|
76
|
|
|
* Reads the registered processingInstructions of the crawler |
|
77
|
|
|
*/ |
|
78
|
|
|
private function getCrawlerProcInstructions(): array |
|
79
|
|
|
{ |
|
80
|
|
|
$crawlerProcInstructions = []; |
|
81
|
|
|
if (! empty($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['crawler']['procInstructions'])) { |
|
82
|
|
|
foreach ($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['crawler']['procInstructions'] as $configuration) { |
|
83
|
|
|
$crawlerProcInstructions[$configuration['key']] = $configuration['value']; |
|
84
|
|
|
} |
|
85
|
|
|
} |
|
86
|
|
|
|
|
87
|
|
|
return $crawlerProcInstructions; |
|
88
|
|
|
} |
|
89
|
|
|
|
|
90
|
|
|
/** |
|
91
|
|
|
* Method to get an instance of the internal crawler singleton |
|
92
|
|
|
* |
|
93
|
|
|
* @return CrawlerController Instance of the crawler lib |
|
94
|
|
|
* |
|
95
|
|
|
* @throws CrawlerObjectException |
|
96
|
|
|
*/ |
|
97
|
|
|
private function findCrawler() |
|
98
|
|
|
{ |
|
99
|
|
|
if (! is_object($this->crawlerController)) { |
|
100
|
|
|
$this->crawlerController = GeneralUtility::makeInstance(CrawlerController::class); |
|
101
|
|
|
$this->crawlerController->setID = GeneralUtility::md5int(microtime()); |
|
102
|
|
|
} |
|
103
|
|
|
|
|
104
|
|
|
if (is_object($this->crawlerController)) { |
|
105
|
|
|
return $this->crawlerController; |
|
106
|
|
|
} |
|
107
|
|
|
throw new CrawlerObjectException('no crawler object', 1608465082); |
|
108
|
|
|
} |
|
109
|
|
|
} |
|
110
|
|
|
|
AOEpeople /
crawler
Loading history...