|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
declare(strict_types=1); |
|
4
|
|
|
|
|
5
|
|
|
namespace AOE\Crawler\Command; |
|
6
|
|
|
|
|
7
|
|
|
/*************************************************************** |
|
8
|
|
|
* Copyright notice |
|
9
|
|
|
* |
|
10
|
|
|
* (c) 2019 AOE GmbH <[email protected]> |
|
11
|
|
|
* |
|
12
|
|
|
* All rights reserved |
|
13
|
|
|
* |
|
14
|
|
|
* This script is part of the TYPO3 project. The TYPO3 project is |
|
15
|
|
|
* free software; you can redistribute it and/or modify |
|
16
|
|
|
* it under the terms of the GNU General Public License as published by |
|
17
|
|
|
* the Free Software Foundation; either version 3 of the License, or |
|
18
|
|
|
* (at your option) any later version. |
|
19
|
|
|
* |
|
20
|
|
|
* The GNU General Public License can be found at |
|
21
|
|
|
* http://www.gnu.org/copyleft/gpl.html. |
|
22
|
|
|
* |
|
23
|
|
|
* This script is distributed in the hope that it will be useful, |
|
24
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
25
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
26
|
|
|
* GNU General Public License for more details. |
|
27
|
|
|
* |
|
28
|
|
|
* This copyright notice MUST APPEAR in all copies of the script! |
|
29
|
|
|
***************************************************************/ |
|
30
|
|
|
|
|
31
|
|
|
use AOE\Crawler\Controller\CrawlerController; |
|
32
|
|
|
use AOE\Crawler\Crawler; |
|
33
|
|
|
use AOE\Crawler\Domain\Repository\ProcessRepository; |
|
34
|
|
|
use AOE\Crawler\Domain\Repository\QueueRepository; |
|
35
|
|
|
use Symfony\Component\Console\Command\Command; |
|
36
|
|
|
use Symfony\Component\Console\Input\InputInterface; |
|
37
|
|
|
use Symfony\Component\Console\Input\InputOption; |
|
38
|
|
|
use Symfony\Component\Console\Output\OutputInterface; |
|
39
|
|
|
use TYPO3\CMS\Core\Utility\GeneralUtility; |
|
40
|
|
|
use TYPO3\CMS\Extbase\Object\ObjectManager; |
|
41
|
|
|
|
|
42
|
|
|
class ProcessQueueCommand extends Command |
|
43
|
|
|
{ |
|
44
|
|
|
public const CLI_STATUS_NOTHING_PROCCESSED = 0; |
|
45
|
|
|
|
|
46
|
|
|
//queue not empty |
|
47
|
|
|
public const CLI_STATUS_REMAIN = 1; |
|
48
|
|
|
|
|
49
|
|
|
//(some) queue items where processed |
|
50
|
|
|
public const CLI_STATUS_PROCESSED = 2; |
|
51
|
|
|
|
|
52
|
|
|
//instance didn't finish |
|
53
|
|
|
public const CLI_STATUS_ABORTED = 4; |
|
54
|
|
|
|
|
55
|
|
|
public const CLI_STATUS_POLLABLE_PROCESSED = 8; |
|
56
|
|
|
|
|
57
|
|
|
/** |
|
58
|
|
|
* Crawler Command - Crawling the URLs from the queue |
|
59
|
|
|
* |
|
60
|
|
|
* Examples: |
|
61
|
|
|
* |
|
62
|
|
|
* --- Will trigger the crawler which starts to process the queue entries |
|
63
|
|
|
* $ typo3 crawler:crawlQueue |
|
64
|
|
|
* |
|
65
|
|
|
* @return int |
|
66
|
|
|
*/ |
|
67
|
2 |
|
public function execute(InputInterface $input, OutputInterface $output) |
|
68
|
|
|
{ |
|
69
|
2 |
|
$amount = $input->getOption('amount'); |
|
70
|
2 |
|
$sleeptime = $input->getOption('sleeptime'); |
|
71
|
2 |
|
$sleepafter = $input->getOption('sleepafter'); |
|
72
|
|
|
|
|
73
|
2 |
|
$objectManager = GeneralUtility::makeInstance(ObjectManager::class); |
|
74
|
|
|
|
|
75
|
2 |
|
$result = self::CLI_STATUS_NOTHING_PROCCESSED; |
|
76
|
|
|
|
|
77
|
|
|
/** @var CrawlerController $crawlerController */ |
|
78
|
2 |
|
$crawlerController = $objectManager->get(CrawlerController::class); |
|
79
|
|
|
/** @var QueueRepository $queueRepository */ |
|
80
|
2 |
|
$queueRepository = $objectManager->get(QueueRepository::class); |
|
81
|
|
|
/** @var ProcessRepository $processRepository */ |
|
82
|
2 |
|
$processRepository = $objectManager->get(ProcessRepository::class); |
|
83
|
|
|
|
|
84
|
|
|
/** @var Crawler $crawler */ |
|
85
|
2 |
|
$crawler = GeneralUtility::makeInstance(Crawler::class); |
|
86
|
|
|
|
|
87
|
2 |
|
if (! $crawler->isDisabled() && $crawlerController->CLI_checkAndAcquireNewProcess($crawlerController->CLI_buildProcessId())) { |
|
88
|
2 |
|
$countInARun = $amount ? (int) $amount : (int) $crawlerController->extensionSettings['countInARun']; |
|
89
|
2 |
|
$sleepAfterFinish = $sleepafter ? (int) $sleepafter : (int) $crawlerController->extensionSettings['sleepAfterFinish']; |
|
90
|
2 |
|
$sleepTime = $sleeptime ? (int) $sleeptime : (int) $crawlerController->extensionSettings['sleepTime']; |
|
91
|
|
|
|
|
92
|
|
|
try { |
|
93
|
|
|
// Run process: |
|
94
|
2 |
|
$result = $crawlerController->CLI_run($countInARun, $sleepTime, $sleepAfterFinish); |
|
95
|
|
|
} catch (\Throwable $e) { |
|
96
|
|
|
$output->writeln('<warning>' . get_class($e) . ': ' . $e->getMessage() . '</warning>'); |
|
97
|
|
|
$result = self::CLI_STATUS_ABORTED; |
|
98
|
|
|
} |
|
99
|
|
|
|
|
100
|
|
|
// Cleanup |
|
101
|
2 |
|
$processRepository->deleteProcessesWithoutItemsAssigned(); |
|
102
|
2 |
|
$processRepository->markRequestedProcessesAsNotActive([$crawlerController->CLI_buildProcessId()]); |
|
103
|
2 |
|
$queueRepository->unsetProcessScheduledAndProcessIdForQueueEntries([$crawlerController->CLI_buildProcessId()]); |
|
104
|
|
|
|
|
105
|
2 |
|
$output->writeln('<info>Unprocessed Items remaining:' . count($queueRepository->getUnprocessedItems()) . ' (' . $crawlerController->CLI_buildProcessId() . ')</info>'); |
|
106
|
2 |
|
$result |= (count($queueRepository->getUnprocessedItems()) > 0 ? self::CLI_STATUS_REMAIN : self::CLI_STATUS_NOTHING_PROCCESSED); |
|
107
|
|
|
} else { |
|
108
|
|
|
$result |= self::CLI_STATUS_ABORTED; |
|
109
|
|
|
} |
|
110
|
|
|
|
|
111
|
2 |
|
$output->writeln($result); |
|
112
|
2 |
|
return $result & self::CLI_STATUS_ABORTED; |
|
113
|
|
|
} |
|
114
|
|
|
|
|
115
|
2 |
|
protected function configure(): void |
|
116
|
|
|
{ |
|
117
|
2 |
|
$this->setDescription('Trigger the crawler to process the queue entries'); |
|
118
|
|
|
|
|
119
|
2 |
|
$this->setHelp( |
|
120
|
2 |
|
'Crawler Command - Crawling the URLs from the queue' . chr(10) . chr(10) . |
|
121
|
2 |
|
' |
|
122
|
|
|
Examples: |
|
123
|
|
|
--- Will trigger the crawler which starts to process the queue entries |
|
124
|
|
|
$ typo3 crawler:processqueue --amount 15 --sleepafter 5 --sleeptime 2 |
|
125
|
|
|
' |
|
126
|
|
|
); |
|
127
|
2 |
|
$this->addOption( |
|
128
|
2 |
|
'amount', |
|
129
|
2 |
|
'', |
|
130
|
2 |
|
InputOption::VALUE_OPTIONAL, |
|
131
|
2 |
|
'How many pages should be crawled during that run', |
|
132
|
|
|
0 |
|
133
|
|
|
); |
|
134
|
|
|
|
|
135
|
2 |
|
$this->addOption( |
|
136
|
2 |
|
'sleepafter', |
|
137
|
2 |
|
'', |
|
138
|
2 |
|
InputOption::VALUE_OPTIONAL, |
|
139
|
2 |
|
'Amount of milliseconds which the system should use to relax between crawls', |
|
140
|
|
|
0 |
|
141
|
|
|
); |
|
142
|
|
|
|
|
143
|
2 |
|
$this->addOption( |
|
144
|
2 |
|
'sleeptime', |
|
145
|
2 |
|
'', |
|
146
|
2 |
|
InputOption::VALUE_OPTIONAL, |
|
147
|
2 |
|
'Amount of seconds which the system should use to relax after all crawls are done.' |
|
148
|
|
|
); |
|
149
|
2 |
|
} |
|
150
|
|
|
} |
|
151
|
|
|
|
AOEpeople /
crawler
