|
1
|
|
|
<?php declare(strict_types = 1); |
|
2
|
|
|
namespace ApacheSolrForTypo3\Solr\Domain\Search\Statistics; |
|
3
|
|
|
|
|
4
|
|
|
/*************************************************************** |
|
5
|
|
|
* Copyright notice |
|
6
|
|
|
* |
|
7
|
|
|
* (c) 2016 Thomas Hohn <[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 2 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\System\Records\AbstractRepository; |
|
28
|
|
|
use TYPO3\CMS\Core\Database\Query\QueryBuilder; |
|
29
|
|
|
|
|
30
|
|
|
/** |
|
31
|
|
|
* Calculates the SearchQueryStatistics |
|
32
|
|
|
* |
|
33
|
|
|
* @author Thomas Hohn <[email protected]> |
|
34
|
|
|
*/ |
|
35
|
|
|
class StatisticsRepository extends AbstractRepository |
|
36
|
|
|
{ |
|
37
|
|
|
/** |
|
38
|
|
|
* @var string |
|
39
|
|
|
*/ |
|
40
|
|
|
protected $table = 'tx_solr_statistics'; |
|
41
|
|
|
|
|
42
|
|
|
/** |
|
43
|
1 |
|
* Fetches must popular search keys words from the table tx_solr_statistics |
|
44
|
|
|
* |
|
45
|
1 |
|
* @param int $rootPageId |
|
46
|
1 |
|
* @param int $days number of days of history to query |
|
47
|
1 |
|
* @param int $limit |
|
48
|
1 |
|
* @return mixed |
|
49
|
|
|
*/ |
|
50
|
1 |
|
public function getSearchStatistics(int $rootPageId, int $days = 30, $limit = 10) |
|
51
|
1 |
|
{ |
|
52
|
1 |
|
$now = time(); |
|
53
|
1 |
|
$timeStart = (int)($now - 86400 * $days); // 86400 seconds/day |
|
54
|
1 |
|
$limit = (int)$limit; |
|
55
|
1 |
|
|
|
56
|
|
|
return $this->getPreparedQueryBuilderForSearchStatisticsAndTopKeywords($rootPageId, $timeStart, $limit) |
|
57
|
|
|
->execute()->fetchAll(); |
|
58
|
|
|
} |
|
59
|
1 |
|
|
|
60
|
|
|
/** |
|
61
|
1 |
|
* Returns prepared QueryBuilder for two purposes: |
|
62
|
1 |
|
* for getSearchStatistics() and getTopKeyWordsWithOrWithoutHits() methods |
|
63
|
|
|
* |
|
64
|
|
|
* @param int $rootPageId |
|
65
|
|
|
* @param int $timeStart |
|
66
|
1 |
|
* @param int $limit |
|
67
|
|
|
* @return QueryBuilder |
|
68
|
|
|
*/ |
|
69
|
1 |
|
protected function getPreparedQueryBuilderForSearchStatisticsAndTopKeywords(int $rootPageId, int $timeStart, int $limit) : QueryBuilder |
|
70
|
|
|
{ |
|
71
|
1 |
|
$countRows = $this->countByRootPageId($rootPageId); |
|
72
|
|
|
$queryBuilder = $this->getQueryBuilder(); |
|
73
|
|
|
$statisticsQueryBuilder = $queryBuilder |
|
74
|
|
|
->select('keywords') |
|
75
|
|
|
->add('select', $queryBuilder->expr()->count('keywords', 'count'), true) |
|
76
|
|
|
->add('select', $queryBuilder->expr()->avg('num_found', 'hits'), true) |
|
77
|
|
|
->add('select', '(' . $queryBuilder->expr()->count('keywords') . ' * 100 / ' . $countRows . ') AS percent', true) |
|
78
|
|
|
->from($this->table) |
|
79
|
|
|
->andWhere( |
|
80
|
|
|
$queryBuilder->expr()->gt('tstamp', $timeStart), |
|
81
|
|
|
$queryBuilder->expr()->eq('root_pid', $rootPageId) |
|
82
|
1 |
|
) |
|
83
|
|
|
->groupBy('keywords') |
|
84
|
1 |
|
->orderBy('count', 'DESC') |
|
85
|
|
|
->addOrderBy('hits', 'DESC') |
|
86
|
|
|
->addOrderBy('keywords', 'ASC') |
|
87
|
|
|
->setMaxResults($limit); |
|
88
|
|
|
|
|
89
|
|
|
return $statisticsQueryBuilder; |
|
90
|
|
|
} |
|
91
|
|
|
|
|
92
|
|
|
/** |
|
93
|
|
|
* Find Top search keywords with results |
|
94
|
|
|
* |
|
95
|
2 |
|
* @param int $rootPageId |
|
96
|
|
|
* @param int $days number of days of history to query |
|
97
|
2 |
|
* @param int $limit |
|
98
|
|
|
* @return array |
|
99
|
|
|
*/ |
|
100
|
|
|
public function getTopKeyWordsWithHits(int $rootPageId, int $days = 30, int $limit = 10) : array |
|
101
|
|
|
{ |
|
102
|
|
|
return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, false); |
|
103
|
|
|
} |
|
104
|
|
|
|
|
105
|
|
|
/** |
|
106
|
|
|
* Find Top search keywords without results |
|
107
|
|
|
* |
|
108
|
|
|
* @param int $rootPageId |
|
109
|
3 |
|
* @param int $days number of days of history to query |
|
110
|
|
|
* @param int $limit |
|
111
|
3 |
|
* @return array |
|
112
|
3 |
|
*/ |
|
113
|
3 |
|
public function getTopKeyWordsWithoutHits(int $rootPageId, int $days = 30, int $limit = 10) : array |
|
114
|
|
|
{ |
|
115
|
3 |
|
return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, true); |
|
116
|
3 |
|
} |
|
117
|
|
|
|
|
118
|
|
|
/** |
|
119
|
3 |
|
* Find Top search keywords with or without results |
|
120
|
2 |
|
* |
|
121
|
|
|
* @param int $rootPageId |
|
122
|
1 |
|
* @param int $days number of days of history to query |
|
123
|
|
|
* @param int $limit |
|
124
|
|
|
* @param bool $withoutHits |
|
125
|
3 |
|
* @return array |
|
126
|
3 |
|
*/ |
|
127
|
3 |
|
protected function getTopKeyWordsWithOrWithoutHits(int $rootPageId, int $days = 30, int $limit = 10, bool $withoutHits = false) : array |
|
128
|
3 |
|
{ |
|
129
|
3 |
|
$now = time(); |
|
130
|
3 |
|
$timeStart = $now - 86400 * $days; // 86400 seconds/day |
|
131
|
|
|
|
|
132
|
|
|
$queryBuilder = $this->getPreparedQueryBuilderForSearchStatisticsAndTopKeywords($rootPageId, $timeStart, $limit); |
|
133
|
|
|
// Check if we want without or with hits |
|
134
|
3 |
|
if ($withoutHits === true) { |
|
135
|
|
|
$queryBuilder->andWhere($queryBuilder->expr()->eq('num_found', 0)); |
|
136
|
3 |
|
} else { |
|
137
|
|
|
$queryBuilder->andWhere($queryBuilder->expr()->gt('num_found', 0)); |
|
138
|
|
|
} |
|
139
|
|
|
|
|
140
|
|
|
return $queryBuilder->execute()->fetchAll(); |
|
141
|
|
|
} |
|
142
|
|
|
|
|
143
|
|
|
/** |
|
144
|
|
|
* Get number of queries over time |
|
145
|
|
|
* |
|
146
|
4 |
|
* @param int $rootPageId |
|
147
|
|
|
* @param int $days number of days of history to query |
|
148
|
4 |
|
* @param int $bucketSeconds Seconds per bucket |
|
149
|
4 |
|
* @return array [labels, data] |
|
150
|
2 |
|
*/ |
|
151
|
|
|
public function getQueriesOverTime(int $rootPageId, int $days = 30, int $bucketSeconds = 3600) : array |
|
152
|
2 |
|
{ |
|
153
|
2 |
|
$now = time(); |
|
154
|
|
|
$timeStart = $now - 86400 * intval($days); // 86400 seconds/day |
|
155
|
|
|
|
|
156
|
2 |
|
$queryBuilder = $this->getQueryBuilder(); |
|
157
|
2 |
|
$result = $queryBuilder |
|
158
|
|
|
->addSelectLiteral( |
|
159
|
|
|
'FLOOR(`tstamp`/' . $bucketSeconds . ') AS `bucket`', |
|
160
|
2 |
|
'(`tstamp` - (`tstamp` % 86400)) AS `timestamp`', |
|
161
|
|
|
$queryBuilder->expr()->count('*', 'numQueries') |
|
162
|
2 |
|
) |
|
163
|
2 |
|
->from($this->table) |
|
164
|
|
|
->andWhere( |
|
165
|
|
|
$queryBuilder->expr()->gt('tstamp', $timeStart), |
|
166
|
4 |
|
$queryBuilder->expr()->eq('root_pid', $rootPageId) |
|
167
|
|
|
) |
|
168
|
|
|
->groupBy('bucket', 'timestamp') |
|
169
|
|
|
->orderBy('bucket', 'ASC') |
|
170
|
|
|
->execute()->fetchAll(); |
|
171
|
|
|
|
|
172
|
|
|
return $result; |
|
173
|
|
|
} |
|
174
|
|
|
|
|
175
|
|
|
/** |
|
176
|
|
|
* Regurns a result set by given plugin.tx_solr.search.frequentSearches.select configuration. |
|
177
|
|
|
* |
|
178
|
|
|
* @param array $frequentSearchConfiguration |
|
179
|
|
|
* @return array Array of frequent search terms, keys are the terms, values are hits |
|
180
|
|
|
*/ |
|
181
|
|
|
public function getFrequentSearchTermsFromStatisticsByFrequentSearchConfiguration(array $frequentSearchConfiguration) : array |
|
182
|
|
|
{ |
|
183
|
|
|
$queryBuilder = $this->getQueryBuilder(); |
|
184
|
|
|
$resultSet = $queryBuilder |
|
185
|
|
|
->addSelectLiteral( |
|
186
|
|
|
$frequentSearchConfiguration['select.']['SELECT'] |
|
187
|
|
|
) |
|
188
|
|
|
->from($frequentSearchConfiguration['select.']['FROM']) |
|
189
|
|
|
->add('where', $frequentSearchConfiguration['select.']['ADD_WHERE'], true) |
|
190
|
|
|
->add('groupBy', $frequentSearchConfiguration['select.']['GROUP_BY'], true) |
|
191
|
|
|
->add('orderBy', $frequentSearchConfiguration['select.']['ORDER_BY']) |
|
192
|
|
|
->setMaxResults((int)$frequentSearchConfiguration['limit']) |
|
193
|
|
|
->execute()->fetchAll(); |
|
194
|
|
|
|
|
195
|
|
|
return $resultSet; |
|
196
|
|
|
} |
|
197
|
|
|
|
|
198
|
|
|
/** |
|
199
|
|
|
* Persists statistics record |
|
200
|
|
|
* |
|
201
|
2 |
|
* @param array $statisticsRecord |
|
202
|
|
|
* @return void |
|
203
|
|
|
*/ |
|
204
|
2 |
|
public function saveStatisticsRecord(array $statisticsRecord) |
|
205
|
2 |
|
{ |
|
206
|
2 |
|
$queryBuilder = $this->getQueryBuilder(); |
|
207
|
|
|
$queryBuilder->insert($this->table)->values($statisticsRecord)->execute(); |
|
208
|
|
|
} |
|
209
|
1 |
|
|
|
210
|
1 |
|
/** |
|
211
|
1 |
|
* Counts rows for specified site |
|
212
|
1 |
|
* |
|
213
|
1 |
|
* @param int $rootPageId |
|
214
|
|
|
* @return int |
|
215
|
|
|
*/ |
|
216
|
|
|
public function countByRootPageId(int $rootPageId): int |
|
217
|
|
|
{ |
|
218
|
|
|
$queryBuilder = $this->getQueryBuilder(); |
|
219
|
4 |
|
return (int)$this->getQueryBuilder() |
|
220
|
|
|
->count('*') |
|
221
|
4 |
|
->from($this->table) |
|
222
|
|
|
->andWhere($queryBuilder->expr()->eq('root_pid', $rootPageId)) |
|
223
|
|
|
->execute()->fetchColumn(0); |
|
224
|
|
|
} |
|
225
|
|
|
} |
|
226
|
|
|
|
timohund /
ext-solr
