|
1
|
|
|
<?php |
|
2
|
|
|
namespace AOE\Crawler\Hooks; |
|
3
|
|
|
|
|
4
|
|
|
/*************************************************************** |
|
5
|
|
|
* Copyright notice |
|
6
|
|
|
* |
|
7
|
|
|
* (c) 2016 AOE GmbH <[email protected]> |
|
8
|
|
|
* |
|
9
|
|
|
* All rights reserved |
|
10
|
|
|
* |
|
11
|
|
|
* This script is part of the TYPO3 project. The TYPO3 project is |
|
12
|
|
|
* free software; you can redistribute it and/or modify |
|
13
|
|
|
* it under the terms of the GNU General Public License as published by |
|
14
|
|
|
* the Free Software Foundation; either version 3 of the License, or |
|
15
|
|
|
* (at your option) any later version. |
|
16
|
|
|
* |
|
17
|
|
|
* The GNU General Public License can be found at |
|
18
|
|
|
* http://www.gnu.org/copyleft/gpl.html. |
|
19
|
|
|
* |
|
20
|
|
|
* This script is distributed in the hope that it will be useful, |
|
21
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
22
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
23
|
|
|
* GNU General Public License for more details. |
|
24
|
|
|
* |
|
25
|
|
|
* This copyright notice MUST APPEAR in all copies of the script! |
|
26
|
|
|
***************************************************************/ |
|
27
|
|
|
|
|
28
|
|
|
use AOE\Crawler\Controller\CrawlerController; |
|
29
|
|
|
use TYPO3\CMS\Core\Utility\GeneralUtility; |
|
30
|
|
|
|
|
31
|
|
|
/** |
|
32
|
|
|
* Class ProcessCleanUpHook |
|
33
|
|
|
* @package AOE\Crawler\Hooks |
|
34
|
|
|
*/ |
|
35
|
|
|
class ProcessCleanUpHook |
|
36
|
|
|
{ |
|
37
|
|
|
/** |
|
38
|
|
|
* @var CrawlerController |
|
39
|
|
|
*/ |
|
40
|
|
|
private $crawlerController; |
|
41
|
|
|
|
|
42
|
|
|
/** |
|
43
|
|
|
* @var array |
|
44
|
|
|
*/ |
|
45
|
|
|
private $extensionSettings; |
|
46
|
|
|
|
|
47
|
|
|
/** |
|
48
|
|
|
* Main function of process CleanUp Hook. |
|
49
|
|
|
* |
|
50
|
|
|
* @param CrawlerController $crawlerController Crawler Lib class |
|
51
|
|
|
* |
|
52
|
|
|
* @return void |
|
53
|
|
|
*/ |
|
54
|
|
|
public function crawler_init(CrawlerController $crawlerController) |
|
55
|
|
|
{ |
|
56
|
|
|
$this->crawlerController = $crawlerController; |
|
57
|
|
|
$this->extensionSettings = $this->crawlerController->extensionSettings; |
|
58
|
|
|
|
|
59
|
|
|
// Clean Up |
|
60
|
|
|
$this->removeActiveOrphanProcesses(); |
|
61
|
|
|
$this->removeActiveProcessesOlderThanOneHour(); |
|
62
|
|
|
} |
|
63
|
|
|
|
|
64
|
|
|
/** |
|
65
|
|
|
* Remove active processes older than one hour |
|
66
|
|
|
* |
|
67
|
|
|
* @return void |
|
68
|
|
|
*/ |
|
69
|
|
|
private function removeActiveProcessesOlderThanOneHour() |
|
70
|
|
|
{ |
|
71
|
|
|
$results = $this->getActiveProcessesOlderThanOneOHour(); |
|
72
|
|
|
|
|
73
|
|
|
if (!is_array($results)) { |
|
74
|
|
|
return; |
|
75
|
|
|
} |
|
76
|
|
|
foreach ($results as $result) { |
|
77
|
|
|
$systemProcessId = (int)$result['system_process_id']; |
|
78
|
|
|
$processId = $result['process_id']; |
|
79
|
|
|
if ($systemProcessId > 1) { |
|
80
|
|
|
if ($this->doProcessStillExists($systemProcessId)) { |
|
81
|
|
|
$this->killProcess($systemProcessId); |
|
82
|
|
|
} |
|
83
|
|
|
$this->removeProcessFromProcesslist($processId); |
|
84
|
|
|
} |
|
85
|
|
|
} |
|
86
|
|
|
} |
|
87
|
|
|
|
|
88
|
|
|
/** |
|
89
|
|
|
* Removes active orphan processes from process list |
|
90
|
|
|
* |
|
91
|
|
|
* @return void |
|
92
|
|
|
*/ |
|
93
|
|
|
private function removeActiveOrphanProcesses() |
|
94
|
|
|
{ |
|
95
|
|
|
$results = $this->getActiveOrphanProcesses(); |
|
96
|
|
|
|
|
97
|
|
|
if (!is_array($results)) { |
|
98
|
|
|
return; |
|
99
|
|
|
} |
|
100
|
|
|
foreach ($results as $result) { |
|
101
|
|
|
$processExists = false; |
|
102
|
|
|
$systemProcessId = (int)$result['system_process_id']; |
|
103
|
|
|
$processId = $result['process_id']; |
|
104
|
|
|
if ($systemProcessId > 1) { |
|
105
|
|
|
$dispatcherProcesses = $this->findDispatcherProcesses(); |
|
106
|
|
|
if (!is_array($dispatcherProcesses) || empty($dispatcherProcesses)) { |
|
107
|
|
|
$this->removeProcessFromProcesslist($processId); |
|
108
|
|
|
return; |
|
109
|
|
|
} |
|
110
|
|
|
foreach ($dispatcherProcesses as $process) { |
|
111
|
|
|
$responseArray = $this->createResponseArray($process); |
|
112
|
|
|
if ($systemProcessId === (int)$responseArray[1]) { |
|
113
|
|
|
$processExists = true; |
|
114
|
|
|
}; |
|
115
|
|
|
} |
|
116
|
|
|
if (!$processExists) { |
|
117
|
|
|
$this->removeProcessFromProcesslist($processId); |
|
118
|
|
|
} |
|
119
|
|
|
} |
|
120
|
|
|
} |
|
121
|
|
|
} |
|
122
|
|
|
|
|
123
|
|
|
/** |
|
124
|
|
|
* Remove a process from processlist |
|
125
|
|
|
* |
|
126
|
|
|
* @param string $processId Unique process Id. |
|
127
|
|
|
* |
|
128
|
|
|
* @return void |
|
129
|
|
|
*/ |
|
130
|
3 |
|
private function removeProcessFromProcesslist($processId) |
|
131
|
|
|
{ |
|
132
|
3 |
|
$this->getDatabaseConnection()->exec_DELETEquery( |
|
133
|
3 |
|
'tx_crawler_process', |
|
134
|
3 |
|
'process_id = ' . $this->getDatabaseConnection()->fullQuoteStr($processId, 'tx_crawler_process') |
|
135
|
|
|
); |
|
136
|
|
|
|
|
137
|
3 |
|
$this->getDatabaseConnection()->exec_UPDATEquery( |
|
138
|
3 |
|
'tx_crawler_queue', |
|
139
|
3 |
|
'process_id = ' . $this->getDatabaseConnection()->fullQuoteStr($processId, 'tx_crawler_queue'), |
|
140
|
3 |
|
['process_id' => ''] |
|
141
|
|
|
); |
|
142
|
3 |
|
} |
|
143
|
|
|
|
|
144
|
|
|
/** |
|
145
|
|
|
* Create response array |
|
146
|
|
|
* Convert string to array with space character as delimiter, |
|
147
|
|
|
* removes all empty records to have a cleaner array |
|
148
|
|
|
* |
|
149
|
|
|
* @param string $string String to create array from |
|
150
|
|
|
* |
|
151
|
|
|
* @return array |
|
152
|
|
|
* |
|
153
|
|
|
*/ |
|
154
|
2 |
|
private function createResponseArray($string) |
|
155
|
|
|
{ |
|
156
|
2 |
|
$responseArray = GeneralUtility::trimExplode(' ', $string, true); |
|
157
|
2 |
|
$responseArray = array_values($responseArray); |
|
158
|
2 |
|
return $responseArray; |
|
159
|
|
|
} |
|
160
|
|
|
|
|
161
|
|
|
/** |
|
162
|
|
|
* Check if the process still exists |
|
163
|
|
|
* |
|
164
|
|
|
* @param int $pid Process id to be checked. |
|
165
|
|
|
* |
|
166
|
|
|
* @return bool |
|
167
|
|
|
* @codeCoverageIgnore |
|
168
|
|
|
*/ |
|
169
|
|
|
private function doProcessStillExists($pid) |
|
170
|
|
|
{ |
|
171
|
|
|
$doProcessStillExists = false; |
|
172
|
|
|
if (!$this->isOsWindows()) { |
|
173
|
|
|
// Not windows |
|
174
|
|
|
if (file_exists('/proc/' . $pid)) { |
|
175
|
|
|
$doProcessStillExists = true; |
|
176
|
|
|
} |
|
177
|
|
|
} else { |
|
178
|
|
|
// Windows |
|
179
|
|
|
exec('tasklist | find "' . $pid . '"', $returnArray, $returnValue); |
|
180
|
|
|
if (count($returnArray) > 0 && preg_match('/php/i', $returnValue[0])) { |
|
181
|
|
|
$doProcessStillExists = true; |
|
182
|
|
|
} |
|
183
|
|
|
} |
|
184
|
|
|
return $doProcessStillExists; |
|
185
|
|
|
} |
|
186
|
|
|
|
|
187
|
|
|
/** |
|
188
|
|
|
* Kills a process |
|
189
|
|
|
* |
|
190
|
|
|
* @param int $pid Process id to kill |
|
191
|
|
|
* |
|
192
|
|
|
* @return void |
|
193
|
|
|
* @codeCoverageIgnore |
|
194
|
|
|
*/ |
|
195
|
|
|
private function killProcess($pid) |
|
196
|
|
|
{ |
|
197
|
|
|
if (!$this->isOsWindows()) { |
|
198
|
|
|
// Not windows |
|
199
|
|
|
posix_kill($pid, 9); |
|
200
|
|
|
} else { |
|
201
|
|
|
// Windows |
|
202
|
|
|
exec('taskkill /PID ' . $pid); |
|
203
|
|
|
} |
|
204
|
|
|
} |
|
205
|
|
|
|
|
206
|
|
|
/** |
|
207
|
|
|
* Find dispatcher processes |
|
208
|
|
|
* |
|
209
|
|
|
* @return array |
|
210
|
|
|
* @codeCoverageIgnore |
|
211
|
|
|
*/ |
|
212
|
|
|
private function findDispatcherProcesses() |
|
213
|
|
|
{ |
|
214
|
|
|
$returnArray = []; |
|
215
|
|
|
if (!$this->isOsWindows()) { |
|
216
|
|
|
// Not windows |
|
217
|
|
|
exec('ps aux | grep \'cli_dispatcher\'', $returnArray, $returnValue); |
|
218
|
|
|
} else { |
|
219
|
|
|
// Windows |
|
220
|
|
|
exec('tasklist | find \'cli_dispatcher\'', $returnArray, $returnValue); |
|
221
|
|
|
} |
|
222
|
|
|
return $returnArray; |
|
223
|
|
|
} |
|
224
|
|
|
|
|
225
|
|
|
/** |
|
226
|
|
|
* Check if OS is Windows |
|
227
|
|
|
* |
|
228
|
|
|
* @return bool |
|
229
|
|
|
* @codeCoverageIgnore |
|
230
|
|
|
*/ |
|
231
|
|
|
private function isOsWindows() |
|
232
|
|
|
{ |
|
233
|
|
|
if (TYPO3_OS === 'WIN') { |
|
234
|
|
|
return true; |
|
235
|
|
|
} |
|
236
|
|
|
return false; |
|
237
|
|
|
} |
|
238
|
|
|
|
|
239
|
|
|
/** |
|
240
|
|
|
* @return \TYPO3\CMS\Core\Database\DatabaseConnection |
|
241
|
|
|
* @codeCoverageIgnore |
|
242
|
|
|
*/ |
|
243
|
|
|
private function getDatabaseConnection() |
|
244
|
|
|
{ |
|
245
|
|
|
return $GLOBALS['TYPO3_DB']; |
|
246
|
|
|
} |
|
247
|
|
|
|
|
248
|
|
|
/** |
|
249
|
|
|
* @return array|null |
|
250
|
|
|
*/ |
|
251
|
1 |
|
private function getActiveProcessesOlderThanOneOHour() |
|
252
|
|
|
{ |
|
253
|
1 |
|
$results = $this->getDatabaseConnection()->exec_SELECTgetRows( |
|
254
|
1 |
|
'process_id, system_process_id', |
|
255
|
1 |
|
'tx_crawler_process', |
|
256
|
1 |
|
'ttl <= ' . intval(time() - $this->extensionSettings['processMaxRunTime'] - 3600) . ' AND active = 1' |
|
257
|
|
|
); |
|
258
|
|
|
|
|
259
|
1 |
|
return $results; |
|
260
|
|
|
} |
|
261
|
|
|
|
|
262
|
|
|
/** |
|
263
|
|
|
* @return array|null |
|
264
|
|
|
*/ |
|
265
|
1 |
|
private function getActiveOrphanProcesses() |
|
266
|
|
|
{ |
|
267
|
1 |
|
$results = $this->getDatabaseConnection()->exec_SELECTgetRows( |
|
268
|
1 |
|
'process_id, system_process_id', |
|
269
|
1 |
|
'tx_crawler_process', |
|
270
|
1 |
|
'ttl <= ' . intval(time() - $this->extensionSettings['processMaxRunTime']) . ' AND active = 1' |
|
271
|
|
|
); |
|
272
|
|
|
|
|
273
|
1 |
|
return $results; |
|
274
|
|
|
} |
|
275
|
|
|
} |
|
276
|
|
|
|