|
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
|
|
|
/** |
|
29
|
|
|
* Class TsfeHook |
|
30
|
|
|
* @package AOE\Crawler\Hooks |
|
31
|
|
|
*/ |
|
32
|
|
|
class TsfeHook |
|
33
|
|
|
{ |
|
34
|
|
|
/** |
|
35
|
|
|
* Initialization hook (called after database connection) |
|
36
|
|
|
* Takes the "HTTP_X_T3CRAWLER" header and looks up queue record and verifies if the session comes from the system (by comparing hashes) |
|
37
|
|
|
* |
|
38
|
|
|
* @param array Parameters from frontend |
|
39
|
|
|
* @param object TSFE object (reference under PHP5) |
|
40
|
|
|
* @return void |
|
41
|
|
|
* |
|
42
|
|
|
* TODO: Write Unit test |
|
43
|
|
|
*/ |
|
44
|
|
|
public function fe_init(&$params, $ref) |
|
|
|
|
|
|
45
|
|
|
{ |
|
46
|
|
|
|
|
47
|
|
|
// Authenticate crawler request: |
|
48
|
|
|
if (isset($_SERVER['HTTP_X_T3CRAWLER'])) { |
|
49
|
|
|
//@todo: ask service to exclude current call for special reasons: for example no relevance because the language version is not affected |
|
50
|
|
|
|
|
51
|
|
|
list($queueId, $hash) = explode(':', $_SERVER['HTTP_X_T3CRAWLER']); |
|
52
|
|
|
list($queueRec) = $GLOBALS['TYPO3_DB']->exec_SELECTgetRows('*', 'tx_crawler_queue', 'qid=' . intval($queueId)); |
|
53
|
|
|
|
|
54
|
|
|
// If a crawler record was found and hash was matching, set it up: |
|
55
|
|
|
if (is_array($queueRec) && $hash === md5($queueRec['qid'] . '|' . $queueRec['set_id'] . '|' . $GLOBALS['TYPO3_CONF_VARS']['SYS']['encryptionKey'])) { |
|
56
|
|
|
$params['pObj']->applicationData['tx_crawler']['running'] = true; |
|
57
|
|
|
$params['pObj']->applicationData['tx_crawler']['parameters'] = unserialize($queueRec['parameters']); |
|
58
|
|
|
$params['pObj']->applicationData['tx_crawler']['log'] = []; |
|
59
|
|
|
} else { |
|
60
|
|
|
die('No crawler entry found!'); |
|
61
|
|
|
} |
|
62
|
|
|
} |
|
63
|
|
|
} |
|
64
|
|
|
|
|
65
|
|
|
/** |
|
66
|
|
|
* Initialization of FE-user, setting the user-group list if applicable. |
|
67
|
|
|
* |
|
68
|
|
|
* @param array Parameters from frontend |
|
69
|
|
|
* @param object TSFE object |
|
70
|
|
|
* @return void |
|
71
|
|
|
* |
|
72
|
|
|
* TODO: Write Unit test |
|
73
|
|
|
*/ |
|
74
|
|
|
public function fe_feuserInit(&$params, $ref) |
|
|
|
|
|
|
75
|
|
|
{ |
|
76
|
|
|
if ($params['pObj']->applicationData['tx_crawler']['running']) { |
|
77
|
|
|
$grList = $params['pObj']->applicationData['tx_crawler']['parameters']['feUserGroupList']; |
|
78
|
|
|
if ($grList) { |
|
79
|
|
|
if (!is_array($params['pObj']->fe_user->user)) { |
|
80
|
|
|
$params['pObj']->fe_user->user = []; |
|
81
|
|
|
} |
|
82
|
|
|
$params['pObj']->fe_user->user['usergroup'] = $grList; |
|
83
|
|
|
$params['pObj']->applicationData['tx_crawler']['log'][] = 'User Groups: ' . $grList; |
|
84
|
|
|
} |
|
85
|
|
|
} |
|
86
|
|
|
} |
|
87
|
|
|
|
|
88
|
|
|
/** |
|
89
|
|
|
* Whether to output rendered content or not. If the crawler is running, the rendered output is never outputted! |
|
90
|
|
|
* |
|
91
|
|
|
* @param array Parameters from frontend |
|
92
|
|
|
* @param object TSFE object |
|
93
|
|
|
* @return void |
|
94
|
|
|
* |
|
95
|
|
|
* TODO: Write Unit test |
|
96
|
|
|
*/ |
|
97
|
|
|
public function fe_isOutputting(&$params, $ref) |
|
|
|
|
|
|
98
|
|
|
{ |
|
99
|
|
|
if ($params['pObj']->applicationData['tx_crawler']['running']) { |
|
100
|
|
|
$params['enableOutput'] = false; |
|
101
|
|
|
} |
|
102
|
|
|
} |
|
103
|
|
|
|
|
104
|
|
|
/** |
|
105
|
|
|
* Concluding: Outputting serialized information instead of letting rendered content out. |
|
106
|
|
|
* |
|
107
|
|
|
* @param array Parameters from frontend |
|
108
|
|
|
* @param object TSFE object |
|
109
|
|
|
* @return void |
|
110
|
|
|
*/ |
|
111
|
|
|
public function fe_eofe(&$params, $ref) |
|
|
|
|
|
|
112
|
|
|
{ |
|
113
|
|
|
if ($params['pObj']->applicationData['tx_crawler']['running']) { |
|
114
|
|
|
$params['pObj']->applicationData['tx_crawler']['vars'] = [ |
|
115
|
|
|
'id' => $params['pObj']->id, |
|
116
|
|
|
'gr_list' => $params['pObj']->gr_list, |
|
117
|
|
|
'no_cache' => $params['pObj']->no_cache, |
|
118
|
|
|
]; |
|
119
|
|
|
|
|
120
|
|
|
/** |
|
121
|
|
|
* Required because some extensions (staticpub) might never be requested to run due to some Core side effects |
|
122
|
|
|
* and since this is considered as error the crawler should handle it properly |
|
123
|
|
|
*/ |
|
124
|
|
|
if (is_array($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['crawler']['pollSuccess'])) { |
|
125
|
|
|
foreach ($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['crawler']['pollSuccess'] as $pollable) { |
|
126
|
|
|
if (is_array($params['pObj']->applicationData['tx_crawler']['content']['parameters']['procInstructions']) && in_array($pollable, $params['pObj']->applicationData['tx_crawler']['content']['parameters']['procInstructions'])) { |
|
127
|
|
|
if (empty($params['pObj']->applicationData['tx_crawler']['success'][$pollable])) { |
|
128
|
|
|
$params['pObj']->applicationData['tx_crawler']['errorlog'][] = 'Error: Pollable extension (' . $pollable . ') did not complete successfully.'; |
|
129
|
|
|
} |
|
130
|
|
|
} |
|
131
|
|
|
} |
|
132
|
|
|
} |
|
133
|
|
|
|
|
134
|
|
|
// Output log data for crawler (serialized content): |
|
135
|
|
|
$str = serialize($params['pObj']->applicationData['tx_crawler']); |
|
136
|
|
|
//just make sure that no other output distracts this |
|
137
|
|
|
ob_clean(); |
|
138
|
|
|
header('Content-Length: ' . strlen($str)); |
|
139
|
|
|
echo $str; |
|
140
|
|
|
// Exit since we don't want anymore output! |
|
141
|
|
|
exit; |
|
142
|
|
|
} |
|
143
|
|
|
} |
|
144
|
|
|
} |
|
145
|
|
|
|
AOEpeople /
crawler
Loading history...
This check looks from parameters that have been defined for a function or method, but which are not used in the method body.