|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
declare(strict_types=1); |
|
4
|
|
|
|
|
5
|
|
|
namespace AOE\Crawler\Command; |
|
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\Configuration\ExtensionConfigurationProvider; |
|
23
|
|
|
use AOE\Crawler\Controller\CrawlerController; |
|
24
|
|
|
use AOE\Crawler\Converter\JsonCompatibilityConverter; |
|
25
|
|
|
use AOE\Crawler\Domain\Model\Reason; |
|
26
|
|
|
use AOE\Crawler\Domain\Repository\QueueRepository; |
|
27
|
|
|
use AOE\Crawler\Utility\MessageUtility; |
|
28
|
|
|
use AOE\Crawler\Utility\SignalSlotUtility; |
|
29
|
|
|
use Symfony\Component\Console\Command\Command; |
|
30
|
|
|
use Symfony\Component\Console\Helper\ProgressBar; |
|
31
|
|
|
use Symfony\Component\Console\Input\InputArgument; |
|
32
|
|
|
use Symfony\Component\Console\Input\InputInterface; |
|
33
|
|
|
use Symfony\Component\Console\Input\InputOption; |
|
34
|
|
|
use Symfony\Component\Console\Output\OutputInterface; |
|
35
|
|
|
use TYPO3\CMS\Core\Utility\GeneralUtility; |
|
36
|
|
|
use TYPO3\CMS\Core\Utility\MathUtility; |
|
37
|
|
|
use TYPO3\CMS\Extbase\Object\ObjectManager; |
|
38
|
|
|
|
|
39
|
|
|
class BuildQueueCommand extends Command |
|
40
|
|
|
{ |
|
41
|
|
|
protected function configure(): void |
|
42
|
|
|
{ |
|
43
|
|
|
$this->setDescription('Create entries in the queue that can be processed at once'); |
|
44
|
|
|
|
|
45
|
|
|
$this->setHelp( |
|
46
|
|
|
'Try "typo3 help crawler:buildQueue" to see your options' . chr(10) . chr(10) . |
|
47
|
|
|
'Works as a CLI interface to some functionality from the Web > Info > Site Crawler module; |
|
48
|
|
|
It can put entries in the queue from command line options, return the list of URLs and even execute |
|
49
|
|
|
all entries right away without having to queue them up - this can be useful for immediate re-cache, |
|
50
|
|
|
re-indexing or static publishing from command line.' . chr(10) . chr(10) . |
|
51
|
|
|
' |
|
52
|
|
|
Examples: |
|
53
|
|
|
--- Re-cache pages from page 7 and two levels down, executed immediately |
|
54
|
|
|
$ typo3 crawler:buildQueue 7 defaultConfiguration --depth 2 --mode exec |
|
55
|
|
|
|
|
56
|
|
|
--- Put entries for re-caching pages from page 7 into queue, 4 every minute. |
|
57
|
|
|
$ typo3 crawler:buildQueue 7 defaultConfiguration --depth 0 --mode queue --number 4 |
|
58
|
|
|
' |
|
59
|
|
|
); |
|
60
|
|
|
|
|
61
|
|
|
$this->addArgument( |
|
62
|
|
|
'page', |
|
63
|
|
|
InputArgument::REQUIRED, |
|
64
|
|
|
'The page from where the queue building should start' |
|
65
|
|
|
); |
|
66
|
|
|
|
|
67
|
|
|
$this->addArgument( |
|
68
|
|
|
'conf', |
|
69
|
|
|
InputArgument::REQUIRED, |
|
70
|
|
|
'A comma separated list of crawler configurations' |
|
71
|
|
|
); |
|
72
|
|
|
|
|
73
|
|
|
$this->addOption( |
|
74
|
|
|
'depth', |
|
75
|
|
|
'd', |
|
76
|
|
|
InputOption::VALUE_OPTIONAL, |
|
77
|
|
|
'Tree depth, 0-99\', "How many levels under the \'page_id\' to include.', |
|
78
|
|
|
'0' |
|
79
|
|
|
); |
|
80
|
|
|
|
|
81
|
|
|
$this->addOption( |
|
82
|
|
|
'mode', |
|
83
|
|
|
'm', |
|
84
|
|
|
InputOption::VALUE_OPTIONAL, |
|
85
|
|
|
'Specifies output modes url : Will list URLs which wget could use as input. queue: Will put entries in queue table. exec: Will execute all entries right away!' |
|
86
|
|
|
); |
|
87
|
|
|
|
|
88
|
|
|
$this->addOption( |
|
89
|
|
|
'number', |
|
90
|
|
|
'', |
|
91
|
|
|
InputOption::VALUE_OPTIONAL, |
|
92
|
|
|
'Specifies how many items are put in the queue per minute. Only valid for output mode "queue"', |
|
93
|
|
|
'0' |
|
94
|
|
|
); |
|
95
|
|
|
} |
|
96
|
|
|
|
|
97
|
|
|
/** |
|
98
|
|
|
* Crawler Command - Submitting URLs to be crawled. |
|
99
|
|
|
* |
|
100
|
|
|
* Works as a CLI interface to some functionality from the Web > Info > Site Crawler module; |
|
101
|
|
|
* It can put entries in the queue from command line options, return the list of URLs and even execute |
|
102
|
|
|
* all entries right away without having to queue them up - this can be useful for immediate re-cache, |
|
103
|
|
|
* re-indexing or static publishing from command line. |
|
104
|
|
|
* |
|
105
|
|
|
* Examples: |
|
106
|
|
|
* |
|
107
|
|
|
* --- Re-cache pages from page 7 and two levels down, executed immediately |
|
108
|
|
|
* $ typo3 crawler:buildQueue 7 defaultConfiguration --depth 2 --mode exec |
|
109
|
|
|
* |
|
110
|
|
|
* |
|
111
|
|
|
* --- Put entries for re-caching pages from page 7 into queue, 4 every minute. |
|
112
|
|
|
* $ typo3 crawler:buildQueue 7 defaultConfiguration --depth 0 --mode queue --number 4 |
|
113
|
|
|
*/ |
|
114
|
|
|
protected function execute(InputInterface $input, OutputInterface $output): int |
|
115
|
|
|
{ |
|
116
|
|
|
/** @var JsonCompatibilityConverter $jsonCompatibilityConverter */ |
|
117
|
|
|
$jsonCompatibilityConverter = GeneralUtility::makeInstance(JsonCompatibilityConverter::class); |
|
118
|
|
|
$mode = $input->getOption('mode') ?? 'queue'; |
|
119
|
|
|
|
|
120
|
|
|
$extensionSettings = GeneralUtility::makeInstance(ExtensionConfigurationProvider::class)->getExtensionConfiguration(); |
|
121
|
|
|
|
|
122
|
|
|
$objectManager = GeneralUtility::makeInstance(ObjectManager::class); |
|
123
|
|
|
|
|
124
|
|
|
/** @var CrawlerController $crawlerController */ |
|
125
|
|
|
$crawlerController = $objectManager->get(CrawlerController::class); |
|
126
|
|
|
/** @var QueueRepository $queueRepository */ |
|
127
|
|
|
$queueRepository = $objectManager->get(QueueRepository::class); |
|
128
|
|
|
|
|
129
|
|
|
if ($mode === 'exec') { |
|
130
|
|
|
$crawlerController->registerQueueEntriesInternallyOnly = true; |
|
131
|
|
|
} |
|
132
|
|
|
|
|
133
|
|
|
$pageId = MathUtility::forceIntegerInRange((int) $input->getArgument('page'), 0); |
|
134
|
|
|
if ($pageId === 0) { |
|
135
|
|
|
$message = "Page ${pageId} is not a valid page, please check you root page id and try again."; |
|
136
|
|
|
MessageUtility::addErrorMessage($message); |
|
137
|
|
|
$output->writeln("<info>${message}</info>"); |
|
138
|
|
|
return 1; |
|
139
|
|
|
} |
|
140
|
|
|
|
|
141
|
|
|
$configurationKeys = $this->getConfigurationKeys((string) $input->getArgument('conf')); |
|
142
|
|
|
|
|
143
|
|
|
if ($mode === 'queue' || $mode === 'exec') { |
|
144
|
|
|
$reason = new Reason(); |
|
145
|
|
|
$reason->setReason(Reason::REASON_CLI_SUBMIT); |
|
146
|
|
|
$reason->setDetailText('The cli script of the crawler added to the queue'); |
|
147
|
|
|
|
|
148
|
|
|
$signalPayload = ['reason' => $reason]; |
|
149
|
|
|
SignalSlotUtility::emitSignal( |
|
|
|
|
|
|
150
|
|
|
self::class, |
|
151
|
|
|
SignalSlotUtility::SIGNAL_INVOKE_QUEUE_CHANGE, |
|
152
|
|
|
$signalPayload |
|
153
|
|
|
); |
|
154
|
|
|
} |
|
155
|
|
|
|
|
156
|
|
|
if ($extensionSettings['cleanUpOldQueueEntries']) { |
|
157
|
|
|
$queueRepository->cleanUpOldQueueEntries(); |
|
158
|
|
|
} |
|
159
|
|
|
|
|
160
|
|
|
$crawlerController->setID = GeneralUtility::md5int(microtime()); |
|
161
|
|
|
$crawlerController->getPageTreeAndUrls( |
|
162
|
|
|
$pageId, |
|
163
|
|
|
MathUtility::forceIntegerInRange((int) $input->getOption('depth'), 0, 99), |
|
164
|
|
|
$crawlerController->getCurrentTime(), |
|
165
|
|
|
MathUtility::forceIntegerInRange((int) $input->getOption('number') ?: 30, 1, 1000), |
|
166
|
|
|
$mode === 'queue' || $mode === 'exec', |
|
167
|
|
|
$mode === 'url', |
|
168
|
|
|
[], |
|
169
|
|
|
$configurationKeys |
|
170
|
|
|
); |
|
171
|
|
|
|
|
172
|
|
|
if ($mode === 'url') { |
|
173
|
|
|
$output->writeln('<info>' . implode(PHP_EOL, $crawlerController->downloadUrls) . PHP_EOL . '</info>'); |
|
174
|
|
|
} elseif ($mode === 'exec') { |
|
175
|
|
|
$progressBar = new ProgressBar($output); |
|
176
|
|
|
$output->writeln('<info>Executing ' . count($crawlerController->urlList) . ' requests right away:</info>'); |
|
177
|
|
|
$output->writeln('<info>' . implode(PHP_EOL, $crawlerController->urlList) . '</info>' . PHP_EOL); |
|
178
|
|
|
$output->writeln('<info>Processing</info>' . PHP_EOL); |
|
179
|
|
|
|
|
180
|
|
|
foreach ($progressBar->iterate($crawlerController->queueEntries) as $queueRec) { |
|
181
|
|
|
$p = $jsonCompatibilityConverter->convert($queueRec['parameters']); |
|
182
|
|
|
|
|
183
|
|
|
$progressBar->clear(); |
|
184
|
|
|
$output->writeln('<info>' . $p['url'] . ' (' . implode(',', $p['procInstructions']) . ') => ' . '</info>' . PHP_EOL); |
|
185
|
|
|
$progressBar->display(); |
|
186
|
|
|
|
|
187
|
|
|
$result = $crawlerController->readUrlFromArray($queueRec); |
|
188
|
|
|
|
|
189
|
|
|
$resultContent = $result['content'] ?? ''; |
|
190
|
|
|
$requestResult = $jsonCompatibilityConverter->convert($resultContent); |
|
191
|
|
|
|
|
192
|
|
|
$progressBar->clear(); |
|
193
|
|
|
if (is_array($requestResult)) { |
|
194
|
|
|
$resLog = is_array($requestResult['log']) ? PHP_EOL . chr(9) . chr(9) . implode(PHP_EOL . chr(9) . chr(9), $requestResult['log']) : ''; |
|
195
|
|
|
$output->writeln('<info>OK: ' . $resLog . '</info>' . PHP_EOL); |
|
196
|
|
|
} else { |
|
197
|
|
|
$output->writeln('<error>Error checking Crawler Result: ' . substr(preg_replace('/\s+/', ' ', strip_tags($resultContent)), 0, 30000) . '...' . PHP_EOL . '</error>' . PHP_EOL); |
|
198
|
|
|
} |
|
199
|
|
|
$progressBar->display(); |
|
200
|
|
|
} |
|
201
|
|
|
$output->writeln(''); |
|
202
|
|
|
} elseif ($mode === 'queue') { |
|
203
|
|
|
$output->writeln('<info>Putting ' . count($crawlerController->urlList) . ' entries in queue:</info>' . PHP_EOL); |
|
204
|
|
|
$output->writeln('<info>' . implode(PHP_EOL, $crawlerController->urlList) . '</info>' . PHP_EOL); |
|
205
|
|
|
} else { |
|
206
|
|
|
$output->writeln('<info>' . count($crawlerController->urlList) . ' entries found for processing. (Use "mode" to decide action):</info>' . PHP_EOL); |
|
207
|
|
|
$output->writeln('<info>' . implode(PHP_EOL, $crawlerController->urlList) . '</info>' . PHP_EOL); |
|
208
|
|
|
} |
|
209
|
|
|
|
|
210
|
|
|
return 0; |
|
211
|
|
|
} |
|
212
|
|
|
|
|
213
|
|
|
/** |
|
214
|
|
|
* Obtains configuration keys from the CLI arguments |
|
215
|
|
|
*/ |
|
216
|
|
|
private function getConfigurationKeys(string $conf): array |
|
217
|
|
|
{ |
|
218
|
|
|
$parameter = trim($conf); |
|
219
|
|
|
return ($parameter !== '' ? GeneralUtility::trimExplode(',', $parameter) : []); |
|
220
|
|
|
} |
|
221
|
|
|
} |
|
222
|
|
|
|
AOEpeople /
crawler
Loading history...