1
|
|
|
<?php |
2
|
|
|
namespace ApacheSolrForTypo3\Solr\Domain\Search\Statistics; |
3
|
|
|
|
4
|
|
|
/*************************************************************** |
5
|
|
|
* Copyright notice |
6
|
|
|
* |
7
|
|
|
* (c) 2009-2015 Ingo Renner <[email protected]> |
8
|
|
|
* All rights reserved |
9
|
|
|
* |
10
|
|
|
* This script is part of the TYPO3 project. The TYPO3 project is |
11
|
|
|
* free software; you can redistribute it and/or modify |
12
|
|
|
* it under the terms of the GNU General Public License as published by |
13
|
|
|
* the Free Software Foundation; either version 3 of the License, or |
14
|
|
|
* (at your option) any later version. |
15
|
|
|
* |
16
|
|
|
* The GNU General Public License can be found at |
17
|
|
|
* http://www.gnu.org/copyleft/gpl.html. |
18
|
|
|
* |
19
|
|
|
* This script is distributed in the hope that it will be useful, |
20
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
21
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
22
|
|
|
* GNU General Public License for more details. |
23
|
|
|
* |
24
|
|
|
* This copyright notice MUST APPEAR in all copies of the script! |
25
|
|
|
***************************************************************/ |
26
|
|
|
|
27
|
|
|
use ApacheSolrForTypo3\Solr\Domain\Search\Query\Query; |
28
|
|
|
use ApacheSolrForTypo3\Solr\Domain\Search\ResultSet\SearchResultSet; |
29
|
|
|
use ApacheSolrForTypo3\Solr\Domain\Search\ResultSet\SearchResultSetProcessor; |
30
|
|
|
use ApacheSolrForTypo3\Solr\HtmlContentExtractor; |
31
|
|
|
use TYPO3\CMS\Core\Utility\GeneralUtility; |
32
|
|
|
use TYPO3\CMS\Frontend\Controller\TypoScriptFrontendController; |
33
|
|
|
|
34
|
|
|
/** |
35
|
|
|
* Writes statistics after searches have been conducted. |
36
|
|
|
* |
37
|
|
|
* @author Ingo Renner <[email protected]> |
38
|
|
|
* @author Dimitri Ebert <[email protected]> |
39
|
|
|
* @author Timo Hund <[email protected]> |
40
|
|
|
*/ |
41
|
|
|
class StatisticsWriterProcessor implements SearchResultSetProcessor |
42
|
|
|
{ |
43
|
|
|
/** |
44
|
|
|
* @var StatisticsRepository |
45
|
|
|
*/ |
46
|
|
|
protected $statisticsRepository; |
47
|
|
|
|
48
|
|
|
/** |
49
|
|
|
* @param StatisticsRepository $statisticsRepository |
50
|
|
|
*/ |
51
|
32 |
|
public function __construct(StatisticsRepository $statisticsRepository = null) |
52
|
|
|
{ |
53
|
32 |
|
$this->statisticsRepository = isset($statisticsRepository) ? $statisticsRepository : GeneralUtility::makeInstance(StatisticsRepository::class); |
54
|
32 |
|
} |
55
|
|
|
|
56
|
|
|
/** |
57
|
|
|
* @param SearchResultSet $resultSet |
58
|
|
|
* @return SearchResultSet |
59
|
|
|
*/ |
60
|
32 |
|
public function process(SearchResultSet $resultSet) { |
61
|
32 |
|
$searchRequest = $resultSet->getUsedSearchRequest(); |
62
|
32 |
|
$response = $resultSet->getResponse(); |
63
|
32 |
|
$configuration = $searchRequest->getContextTypoScriptConfiguration(); |
64
|
32 |
|
$keywords = $this->getProcessedKeywords($resultSet->getUsedQuery(), $configuration->getSearchFrequentSearchesUseLowercaseKeywords()); |
65
|
|
|
|
66
|
32 |
|
if (empty($keywords)) { |
67
|
|
|
// do not track empty queries |
68
|
4 |
|
return $resultSet; |
69
|
|
|
} |
70
|
|
|
|
71
|
28 |
|
$filters = $searchRequest->getActiveFacets(); |
72
|
28 |
|
$sorting = $this->sanitizeString($searchRequest->getSorting()); |
73
|
28 |
|
$page = (int)$searchRequest->getPage(); |
74
|
28 |
|
$ipMaskLength = (int)$configuration->getStatisticsAnonymizeIP(); |
75
|
|
|
|
76
|
28 |
|
$TSFE = $this->getTSFE(); |
77
|
|
|
$statisticData = [ |
78
|
28 |
|
'pid' => $TSFE->id, |
79
|
28 |
|
'root_pid' => $TSFE->tmpl->rootLine[0]['uid'], |
80
|
28 |
|
'tstamp' => $this->getTime(), |
81
|
28 |
|
'language' => $TSFE->sys_language_uid, |
82
|
28 |
|
'num_found' => isset($response->response->numFound) ? (int)$response->response->numFound : 0, |
83
|
28 |
|
'suggestions_shown' => is_object($response->spellcheck->suggestions) ? (int)get_object_vars($response->spellcheck->suggestions) : 0, |
84
|
28 |
|
'time_total' => isset($response->debug->timing->time) ? $response->debug->timing->time : 0, |
85
|
28 |
|
'time_preparation' => isset($response->debug->timing->prepare->time) ? $response->debug->timing->prepare->time : 0, |
86
|
28 |
|
'time_processing' => isset($response->debug->timing->process->time) ? $response->debug->timing->process->time : 0, |
87
|
28 |
|
'feuser_id' => (int)$TSFE->fe_user->user['uid'], |
88
|
28 |
|
'cookie' => $TSFE->fe_user->id, |
89
|
28 |
|
'ip' => $this->applyIpMask((string)$this->getUserIp(), $ipMaskLength), |
90
|
28 |
|
'page' => (int)$page, |
91
|
28 |
|
'keywords' => $keywords, |
92
|
28 |
|
'filters' => serialize($filters), |
93
|
28 |
|
'sorting' => $sorting, |
94
|
28 |
|
'parameters' => serialize($response->responseHeader->params) |
95
|
|
|
]; |
96
|
|
|
|
97
|
28 |
|
$this->statisticsRepository->saveStatisticsRecord($statisticData); |
98
|
|
|
|
99
|
28 |
|
return $resultSet; |
100
|
|
|
} |
101
|
|
|
|
102
|
|
|
/** |
103
|
|
|
* @param Query $query |
104
|
|
|
* @param boolean $lowerCaseQuery |
105
|
|
|
* @return string |
106
|
|
|
*/ |
107
|
32 |
|
protected function getProcessedKeywords(Query $query, $lowerCaseQuery = false) |
108
|
|
|
{ |
109
|
32 |
|
$keywords = $query->getQueryStringContainer()->getKeywords(); |
110
|
32 |
|
$keywords = $this->sanitizeString($keywords); |
111
|
32 |
|
if ($lowerCaseQuery) { |
112
|
|
|
$keywords = mb_strtolower($keywords); |
113
|
|
|
} |
114
|
|
|
|
115
|
32 |
|
return $keywords; |
116
|
|
|
} |
117
|
|
|
|
118
|
|
|
/** |
119
|
|
|
* Sanitizes a string |
120
|
|
|
* |
121
|
|
|
* @param $string String to sanitize |
122
|
|
|
* @return string Sanitized string |
123
|
|
|
*/ |
124
|
32 |
|
protected function sanitizeString($string) |
125
|
|
|
{ |
126
|
|
|
// clean content |
127
|
32 |
|
$string = HtmlContentExtractor::cleanContent($string); |
128
|
32 |
|
$string = html_entity_decode($string, ENT_QUOTES, 'UTF-8'); |
129
|
32 |
|
$string = filter_var(strip_tags($string), FILTER_SANITIZE_STRING); // after entity decoding we might have tags again |
130
|
32 |
|
$string = trim($string); |
131
|
|
|
|
132
|
32 |
|
return $string; |
133
|
|
|
} |
134
|
|
|
|
135
|
|
|
/** |
136
|
|
|
* Internal function to mask portions of the visitor IP address |
137
|
|
|
* |
138
|
|
|
* @param string $ip IP address in network address format |
139
|
|
|
* @param int $maskLength Number of octets to reset |
140
|
|
|
* @return string |
141
|
|
|
*/ |
142
|
28 |
|
protected function applyIpMask(string $ip, int $maskLength): string |
143
|
|
|
{ |
144
|
28 |
|
if (empty($ip) || $maskLength === 0) { |
145
|
28 |
|
return $ip; |
146
|
|
|
} |
147
|
|
|
|
148
|
|
|
// IPv4 or mapped IPv4 in IPv6 |
149
|
|
|
if (filter_var($ip, FILTER_VALIDATE_IP, FILTER_FLAG_IPV4)) { |
150
|
|
|
return $this->applyIpV4Mask($ip, $maskLength); |
151
|
|
|
} |
152
|
|
|
|
153
|
|
|
return $this->applyIpV6Mask($ip, $maskLength); |
154
|
|
|
} |
155
|
|
|
|
156
|
|
|
/** |
157
|
|
|
* Apply a mask filter on the ip v4 address. |
158
|
|
|
* |
159
|
|
|
* @param string $ip |
160
|
|
|
* @param int $maskLength |
161
|
|
|
* @return string |
162
|
|
|
*/ |
163
|
|
|
protected function applyIpV4Mask($ip, $maskLength) |
164
|
|
|
{ |
165
|
|
|
$i = strlen($ip); |
166
|
|
|
if ($maskLength > $i) { |
167
|
|
|
$maskLength = $i; |
168
|
|
|
} |
169
|
|
|
|
170
|
|
|
while ($maskLength-- > 0) { |
171
|
|
|
$ip[--$i] = chr(0); |
172
|
|
|
} |
173
|
|
|
return (string)$ip; |
174
|
|
|
} |
175
|
|
|
|
176
|
|
|
/** |
177
|
|
|
* Apply a mask filter on the ip v6 address. |
178
|
|
|
* |
179
|
|
|
* @param string $ip |
180
|
|
|
* @param int $maskLength |
181
|
|
|
* @return int |
182
|
|
|
*/ |
183
|
|
|
protected function applyIpV6Mask($ip, $maskLength):int |
184
|
|
|
{ |
185
|
|
|
$masks = ['ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff', 'ffff:ffff:ffff:ffff::', 'ffff:ffff:ffff:0000::', 'ffff:ff00:0000:0000::']; |
186
|
|
|
$packedAddress = inet_pton($masks[$maskLength]); |
187
|
|
|
$binaryString = pack('a16', $packedAddress); |
188
|
|
|
return (string)($ip & $binaryString); |
|
|
|
|
189
|
|
|
} |
190
|
|
|
|
191
|
|
|
/** |
192
|
|
|
* @return TypoScriptFrontendController |
193
|
|
|
*/ |
194
|
27 |
|
protected function getTSFE() |
195
|
|
|
{ |
196
|
27 |
|
return $GLOBALS['TSFE']; |
197
|
|
|
} |
198
|
|
|
|
199
|
|
|
/** |
200
|
|
|
* @return string |
201
|
|
|
*/ |
202
|
27 |
|
protected function getUserIp() |
203
|
|
|
{ |
204
|
27 |
|
return GeneralUtility::getIndpEnv('REMOTE_ADDR'); |
205
|
|
|
} |
206
|
|
|
|
207
|
|
|
/** |
208
|
|
|
* @return mixed |
209
|
|
|
*/ |
210
|
27 |
|
protected function getTime() |
211
|
|
|
{ |
212
|
27 |
|
return $GLOBALS['EXEC_TIME']; |
213
|
|
|
} |
214
|
|
|
} |
215
|
|
|
|