StatisticsRepository::getSearchStatistics() - Code Metrics - Inspection of "Merge pull request #1032 from thomashohn/task/mast..." - TYPO3-Solr/ext-solr - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( fb242c...3f99e9 )

by Timo

created 2017-02-19 10:55 UTC

StatisticsRepository::getSearchStatistics() B

↳ Parent: StatisticsRepository

Complexity

Conditions	2
Paths	2

Size

Total Lines	31
Code Lines	21

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	19
CRAP Score	2.0552

Importance

Changes

Metric	Value
c	0
b	0
f	0
dl	0
loc	31
ccs	19
cts	25
cp	0.76
rs	8.8571
cc	2
eloc	21
nc	2
nop	3
crap	2.0552

<?php

namespace ApacheSolrForTypo3\Solr\Domain\Search\Statistics;

/***************************************************************
 *  Copyright notice
 *
 *  (c) 2016 Thomas Hohn <[email protected]>
 *  All rights reserved
 *
 *  This script is part of the TYPO3 project. The TYPO3 project is
 *  free software; you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License as published by
 *  the Free Software Foundation; either version 2 of the License, or
 *  (at your option) any later version.
 *
 *  The GNU General Public License can be found at
 *  http://www.gnu.org/copyleft/gpl.html.
 *
 *  This script is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  This copyright notice MUST APPEAR in all copies of the script!
 ***************************************************************/

/**
 * Calculates the SearchQueryStatistics
 *
 * @author Thomas Hohn <[email protected]>
 */
class StatisticsRepository
{
    /**
     * Fetches must popular search keys words from the table tx_solr_statistics
     *
     * @param int $rootPageId
     * @param int $days number of days of history to query
     * @param int $limit
     * @return mixed
     */
    public function getSearchStatistics($rootPageId, $days = 30, $limit = 10)
    {
        $now = time();
        $timeStart = (int) ($now - 86400 * intval($days)); // 86400 seconds/day
        $rootPageId = (int) $rootPageId;
        $limit = (int) $limit;

        $statisticsRows = (array)$this->getDatabase()->exec_SELECTgetRows(
            'keywords, count(keywords) as count, num_found as hits',
            'tx_solr_statistics',
            'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId,
            'keywords, num_found',
            'count DESC, hits DESC, keywords ASC',
            $limit
        );

        $statisticsRows = $this->mergeRowsWithSameKeyword($statisticsRows);

        $sumCount = $statisticsRows['sumCount'];
        foreach ($statisticsRows as $statisticsRow) {
            $sumCount += $statisticsRow['count'];
        }

        $statisticsRows = array_map(function ($row) use ($sumCount) {
            $row['percent'] = $row['count'] * 100 / $sumCount;

            return $row;
        }, $statisticsRows);

        return $statisticsRows;
    }

    /**
     * Find Top search keywords with results
     *
     * @param int $rootPageId
     * @param int $days number of days of history to query
     * @param int $limit
     * @return array
     */
    public function getTopKeyWordsWithHits($rootPageId, $days = 30, $limit = 10)
    {
        return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, false);
    }

    /**
     * Find Top search keywords without results
     *
     * @param int $rootPageId
     * @param int $days number of days of history to query
     * @param int $limit
     * @return array
     */
    public function getTopKeyWordsWithoutHits($rootPageId, $days = 30, $limit = 10)
    {
        return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, true);
    }

    /**
     * Find Top search keywords with or without results
     *
     * @param int $rootPageId
     * @param int $days number of days of history to query
     * @param int $limit
     * @param bool $withoutHits
     * @return array
     */
    protected function getTopKeyWordsWithOrWithoutHits($rootPageId, $days = 30, $limit, $withoutHits)
    {
        $rootPageId = (int) $rootPageId;
        $limit = (int) $limit;
        $withoutHits = (bool) $withoutHits;

        $now = time();
        $timeStart = $now - 86400 * intval($days); // 86400 seconds/day

        // Check if we want without or with hits
        if ($withoutHits === true) {
            $comparisonOperator = '=';
        } else {
            $comparisonOperator = '>';
        }

        $statisticsRows = (array)$this->getDatabase()->exec_SELECTgetRows(
            'keywords, count(keywords) as count, num_found as hits',
            'tx_solr_statistics',
            'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId . ' AND num_found ' . $comparisonOperator . ' 0',
            'keywords, num_found',
            'count DESC, hits DESC, keywords ASC',
            $limit
        );

        $statisticsRows = $this->mergeRowsWithSameKeyword($statisticsRows);

        return $statisticsRows;
    }

    /**
     * This method groups rows with the same term and different count and hits
     * and calculates the average.
     *
     * @param array $statisticsRows
     * @return array
     */
    protected function mergeRowsWithSameKeyword(array $statisticsRows)
    {
        $result = [];
        foreach ($statisticsRows as $statisticsRow) {
            $term = html_entity_decode($statisticsRow['keywords'], ENT_QUOTES);

            $mergedRow = isset($result[$term]) ? $result[$term] : ['mergedrows' => 0, 'count' => 0];
            $mergedRow['mergedrows']++;

                // for the hits we need to take the average
            $avgHits = $this->getAverageFromField($mergedRow, $statisticsRow, 'hits');
            $mergedRow['hits'] = (int) $avgHits;

                // for the count we need to take the sum, because it's the sum of searches
            $mergedRow['count'] = $mergedRow['count'] + $statisticsRow['count'];

            $mergedRow['keywords'] = $term;
            $result[$term] = $mergedRow;
        }

        return array_values($result);
    }

    /**
     * Get number of queries over time
     *
     * @param int $rootPageId
     * @param int $days number of days of history to query
     * @param int $bucketSeconds Seconds per bucket
     * @return array [labels, data]
     */
    public function getQueriesOverTime($rootPageId, $days = 30, $bucketSeconds = 3600)
    {
        $now = time();
        $timeStart = $now - 86400 * intval($days); // 86400 seconds/day

        $queries = $this->getDatabase()->exec_SELECTgetRows(
            'FLOOR(tstamp/' . $bucketSeconds . ') AS bucket, unix_timestamp(from_unixtime(tstamp, "%y-%m-%d")) as timestamp, COUNT(*) AS numQueries',
            'tx_solr_statistics',
            'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId,
            'bucket, timestamp',
            'bucket ASC'
        );

        return $queries;
    }

    /**
     * This method is used to get an average value from merged statistic rows.
     *
     * @param array $mergedRow
     * @param array $statisticsRow
     * @param string $fieldName
     * @return float|int
     */
    protected function getAverageFromField(array &$mergedRow, array $statisticsRow,  $fieldName)
    {
        // when this is the first row we can take it.
        if ($mergedRow['mergedrows'] === 1) {
            $avgCount = $statisticsRow[$fieldName];
            return $avgCount;
        }

        $oldAverage = $mergedRow[$fieldName];
        $oldMergeRows = $mergedRow['mergedrows'] - 1;
        $oldCount = $oldAverage * $oldMergeRows;
        $avgCount = (($oldCount + $statisticsRow[$fieldName]) / $mergedRow['mergedrows']);
        return $avgCount;
    }

    /**
     * @return \TYPO3\CMS\Core\Database\DatabaseConnection
     */
    protected function getDatabase()
    {
        return $GLOBALS['TYPO3_DB'];
    }
}


1		<?php
2
3		namespace ApacheSolrForTypo3\Solr\Domain\Search\Statistics;
4
5		/***************************************************************
6		* Copyright notice
7		*
8		* (c) 2016 Thomas Hohn <[email protected]>
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 2 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		* Calculates the SearchQueryStatistics
30		*
31		* @author Thomas Hohn <[email protected]>
32		*/
33		class StatisticsRepository
34		{
35		/**
36		* Fetches must popular search keys words from the table tx_solr_statistics
37		*
38		* @param int $rootPageId
39		* @param int $days number of days of history to query
40		* @param int $limit
41		* @return mixed
42		*/
43	1	public function getSearchStatistics($rootPageId, $days = 30, $limit = 10)
44		{
45	1	$now = time();
46	1	$timeStart = (int) ($now - 86400 * intval($days)); // 86400 seconds/day
47	1	$rootPageId = (int) $rootPageId;
48	1	$limit = (int) $limit;
49
50	1	$statisticsRows = (array)$this->getDatabase()->exec_SELECTgetRows(
51	1	'keywords, count(keywords) as count, num_found as hits',
52	1	'tx_solr_statistics',
53	1	'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId,
54	1	'keywords, num_found',
55	1	'count DESC, hits DESC, keywords ASC',
56		$limit
57	1	);
58
59	1	$statisticsRows = $this->mergeRowsWithSameKeyword($statisticsRows);
60
61	1	$sumCount = $statisticsRows['sumCount'];
62	1	foreach ($statisticsRows as $statisticsRow) {
63		$sumCount += $statisticsRow['count'];
64	1	}
65
66	1	$statisticsRows = array_map(function ($row) use ($sumCount) {
67		$row['percent'] = $row['count'] * 100 / $sumCount;
68
69		return $row;
70	1	}, $statisticsRows);
71
72	1	return $statisticsRows;
73		}
74
75		/**
76		* Find Top search keywords with results
77		*
78		* @param int $rootPageId
79		* @param int $days number of days of history to query
80		* @param int $limit
81		* @return array
82		*/
83	1	public function getTopKeyWordsWithHits($rootPageId, $days = 30, $limit = 10)
84		{
85	1	return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, false);
86		}
87
88		/**
89		* Find Top search keywords without results
90		*
91		* @param int $rootPageId
92		* @param int $days number of days of history to query
93		* @param int $limit
94		* @return array
95		*/
96	2	public function getTopKeyWordsWithoutHits($rootPageId, $days = 30, $limit = 10)
97		{
98	2	return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, true);
99		}
100
101		/**
102		* Find Top search keywords with or without results
103		*
104		* @param int $rootPageId
105		* @param int $days number of days of history to query
106		* @param int $limit
107		* @param bool $withoutHits
108		* @return array
109		*/
110	3	protected function getTopKeyWordsWithOrWithoutHits($rootPageId, $days = 30, $limit, $withoutHits)
111		{
112	3	$rootPageId = (int) $rootPageId;
113	3	$limit = (int) $limit;
114	3	$withoutHits = (bool) $withoutHits;
115
116	3	$now = time();
117	3	$timeStart = $now - 86400 * intval($days); // 86400 seconds/day
118
119		// Check if we want without or with hits
120	3	if ($withoutHits === true) {
121	2	$comparisonOperator = '=';
122	2	} else {
123	1	$comparisonOperator = '>';
124		}
125
126	3	$statisticsRows = (array)$this->getDatabase()->exec_SELECTgetRows(
127	3	'keywords, count(keywords) as count, num_found as hits',
128	3	'tx_solr_statistics',
129	3	'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId . ' AND num_found ' . $comparisonOperator . ' 0',
130	3	'keywords, num_found',
131	3	'count DESC, hits DESC, keywords ASC',
132		$limit
133	3	);
134
135	3	$statisticsRows = $this->mergeRowsWithSameKeyword($statisticsRows);
136
137	3	return $statisticsRows;
138		}
139
140		/**
141		* This method groups rows with the same term and different count and hits
142		* and calculates the average.
143		*
144		* @param array $statisticsRows
145		* @return array
146		*/
147	4	protected function mergeRowsWithSameKeyword(array $statisticsRows)
148		{
149	4	$result = [];
150	4	foreach ($statisticsRows as $statisticsRow) {
151	2	$term = html_entity_decode($statisticsRow['keywords'], ENT_QUOTES);
152
153	2	$mergedRow = isset($result[$term]) ? $result[$term] : ['mergedrows' => 0, 'count' => 0];
154	2	$mergedRow['mergedrows']++;
155
156		// for the hits we need to take the average
157	2	$avgHits = $this->getAverageFromField($mergedRow, $statisticsRow, 'hits');
158	2	$mergedRow['hits'] = (int) $avgHits;
159
160		// for the count we need to take the sum, because it's the sum of searches
161	2	$mergedRow['count'] = $mergedRow['count'] + $statisticsRow['count'];
162
163	2	$mergedRow['keywords'] = $term;
164	2	$result[$term] = $mergedRow;
165	4	}
166
167	4	return array_values($result);
168		}
169
170		/**
171		* Get number of queries over time
172		*
173		* @param int $rootPageId
174		* @param int $days number of days of history to query
175		* @param int $bucketSeconds Seconds per bucket
176		* @return array [labels, data]
177		*/
178		public function getQueriesOverTime($rootPageId, $days = 30, $bucketSeconds = 3600)
179		{
180		$now = time();
181		$timeStart = $now - 86400 * intval($days); // 86400 seconds/day
182
183		$queries = $this->getDatabase()->exec_SELECTgetRows(
184		'FLOOR(tstamp/' . $bucketSeconds . ') AS bucket, unix_timestamp(from_unixtime(tstamp, "%y-%m-%d")) as timestamp, COUNT(*) AS numQueries',
185		'tx_solr_statistics',
186		'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId,
187		'bucket, timestamp',
188		'bucket ASC'
189		);
190
191		return $queries;
192		}
193
194		/**
195		* This method is used to get an average value from merged statistic rows.
196		*
197		* @param array $mergedRow
198		* @param array $statisticsRow
199		* @param string $fieldName
200		* @return float\|int
201		*/
202	2	protected function getAverageFromField(array &$mergedRow, array $statisticsRow, $fieldName)
203		{
204		// when this is the first row we can take it.
205	2	if ($mergedRow['mergedrows'] === 1) {
206	2	$avgCount = $statisticsRow[$fieldName];
207	2	return $avgCount;
208		}
209
210	1	$oldAverage = $mergedRow[$fieldName];
211	1	$oldMergeRows = $mergedRow['mergedrows'] - 1;
212	1	$oldCount = $oldAverage * $oldMergeRows;
213	1	$avgCount = (($oldCount + $statisticsRow[$fieldName]) / $mergedRow['mergedrows']);
214	1	return $avgCount;
215		}
216
217		/**
218		* @return \TYPO3\CMS\Core\Database\DatabaseConnection
219		*/
220	4	protected function getDatabase()
221		{
222	4	return $GLOBALS['TYPO3_DB'];
223		}
224		}
225

TYPO3-Solr / ext-solr

Push — master ( fb242c...3f99e9 )

StatisticsRepository::getSearchStatistics() B

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like