StatisticsRepository - Code Metrics - Inspection of "[TASK] Add statistics graph (#623)" - TYPO3-Solr/ext-solr - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 7682df...b1b8c7 )

by Timo

created 2016-08-18 11:16 UTC

StatisticsRepository A

↳ Parent: Project

Complexity

Total Complexity

Size/Duplication

Total Lines	193
Duplicated Lines	0 %

Coupling/Cohesion

Components	1
Dependencies	1

Importance

Changes	2
Bugs	0	Features	2

Metric	Value
c	2
b	0
f	2
dl	0
loc	193
rs	10
wmc	13
lcom	1
cbo	1

8 Methods

Rating	Name	Size	Complexity
A	getTopKeyWordsWithHits()	4	1
B	getSearchStatistics()	31	2
A	getTopKeyWordsWithoutHits()	4	1
B	getTopKeyWordsWithOrWithoutHits()	29	2
A	mergeRowsWithSameKeyword()	22	3
A	getQueriesOverTime()	15	1
A	getAverageFromField()	14	2
A	getDatabase()	4	1

<?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]>
 * @package TYPO3
 * @subpackage solr
 */
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 = $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);
/**
 * @return array|string
 */
function returnsDifferentValues($x) {
    if ($x) {
        return 'foo';
    }

    return array();
}

$x = returnsDifferentValues($y);
if (is_array($x)) {
    // $x is an array.
}

        $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 = $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 array|string
 */
function returnsDifferentValues($x) {
    if ($x) {
        return 'foo';
    }

    return array();
}

$x = returnsDifferentValues($y);
if (is_array($x)) {
    // $x is an array.
}

        return $statisticsRows;
    }


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

            $mergedRow = isset($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'] = $statisticsRow['keywords'];
            $result[$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, tstamp, COUNT(*) AS numQueries',
            'tx_solr_statistics',
            'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId,
            'bucket',
            'bucket ASC'
        );

        return $queries;
    }


    /**
     * This method is used to get an average value from merged statistic rows.
     *
     * @param array $mergedRow
     * @param array $statisticsRow
     * @return float|int
     */
    protected function getAverageFromField(array &$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()
// Bad
class Router
{
    public function generate($path)
    {
        return $_SERVER['HOST'].$path;
    }
}

// Better
class Router
{
    private $host;

    public function __construct($host)
    {
        $this->host = $host;
    }

    public function generate($path)
    {
        return $this->host.$path;
    }
}

class Controller
{
    public function myAction(Request $request)
    {
        // Instead of
        $page = isset($_GET['page']) ? intval($_GET['page']) : 1;

        // Better (assuming you use the Symfony2 request)
        $page = $request->query->get('page', 1);
    }
}
    {
        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			/**
30			* Calculates the SearchQueryStatistics
31			*
32			* @author Thomas Hohn <[email protected]>
33			* @package TYPO3
34			* @subpackage solr
35			*/
36			class StatisticsRepository
37			{
38			/**
39			* Fetches must popular search keys words from the table tx_solr_statistics
40			*
41			* @param int $rootPageId
42			* @param int $days number of days of history to query
43			* @param int $limit
44			* @return mixed
45			*/
46			public function getSearchStatistics($rootPageId, $days = 30, $limit = 10)
47			{
48			$now = time();
49			$timeStart = (int) ($now - 86400 * intval($days)); // 86400 seconds/day
50			$rootPageId = (int) $rootPageId;
51			$limit = (int) $limit;
52
53			$statisticsRows = $this->getDatabase()->exec_SELECTgetRows(
54			'keywords, count(keywords) as count, num_found as hits',
55			'tx_solr_statistics',
56			'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId,
57			'keywords, num_found',
58			'count DESC, hits DESC, keywords ASC',
59			$limit
60			);
61
62			$statisticsRows = $this->mergeRowsWithSameKeyword($statisticsRows);
			1 ignored issue – show Bug introduced 2016-08-18 11:15 UTC by Report Bug Copy Issue Report It seems like `$statisticsRows` can also be of type `null`; however, `ApacheSolrForTypo3\Solr\...geRowsWithSameKeyword()` does only seem to accept `array`, maybe add an additional type check? If a method or function can return multiple different values and unless you are sure that you only can receive a single value in this context, we recommend to add an additional type check: /** * @return array\|string */ function returnsDifferentValues($x) { if ($x) { return 'foo'; } return array(); } $x = returnsDifferentValues($y); if (is_array($x)) { // $x is an array. } If this a common case that PHP Analyzer should handle natively, please let us know by opening an issue. Loading history...
63
64			$sumCount = $statisticsRows['sumCount'];
65			foreach ($statisticsRows as $statisticsRow) {
66			$sumCount += $statisticsRow['count'];
67			}
68
69			$statisticsRows = array_map(function ($row) use ($sumCount) {
70			$row['percent'] = $row['count'] * 100 / $sumCount;
71
72			return $row;
73			}, $statisticsRows);
74
75			return $statisticsRows;
76			}
77
78			/**
79			* Find Top search keywords with results
80			*
81			* @param int $rootPageId
82			* @param int $days number of days of history to query
83			* @param int $limit
84			* @return array
85			*/
86			public function getTopKeyWordsWithHits($rootPageId, $days = 30, $limit = 10)
87			{
88			return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, false);
89			}
90
91			/**
92			* Find Top search keywords without results
93			*
94			* @param int $rootPageId
95			* @param int $days number of days of history to query
96			* @param int $limit
97			* @return array
98			*/
99			public function getTopKeyWordsWithoutHits($rootPageId, $days = 30, $limit = 10)
100			{
101			return $this->getTopKeyWordsWithOrWithoutHits($rootPageId, $days, $limit, true);
102			}
103
104			/**
105			* Find Top search keywords with or without results
106			*
107			* @param int $rootPageId
108			* @param int $days number of days of history to query
109			* @param int $limit
110			* @param bool $withoutHits
111			* @return array
112			*/
113			protected function getTopKeyWordsWithOrWithoutHits($rootPageId, $days = 30, $limit, $withoutHits)
114			{
115			$rootPageId = (int) $rootPageId;
116			$limit = (int) $limit;
117			$withoutHits = (bool) $withoutHits;
118
119			$now = time();
120			$timeStart = $now - 86400 * intval($days); // 86400 seconds/day
121
122			// Check if we want without or with hits
123			if ($withoutHits === true) {
124			$comparisonOperator = '=';
125			} else {
126			$comparisonOperator = '>';
127			}
128
129			$statisticsRows = $this->getDatabase()->exec_SELECTgetRows(
130			'keywords, count(keywords) as count, num_found as hits',
131			'tx_solr_statistics',
132			'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId . ' AND num_found ' . $comparisonOperator . ' 0',
133			'keywords, num_found',
134			'count DESC, hits DESC, keywords ASC',
135			$limit
136			);
137
138			$statisticsRows = $this->mergeRowsWithSameKeyword($statisticsRows);
			0 ignored issues – show Bug introduced 2016-08-18 11:15 UTC by Report Bug Copy Issue Report It seems like `$statisticsRows` can also be of type `null`; however, `ApacheSolrForTypo3\Solr\...geRowsWithSameKeyword()` does only seem to accept `array`, maybe add an additional type check? If a method or function can return multiple different values and unless you are sure that you only can receive a single value in this context, we recommend to add an additional type check: /** * @return array\|string */ function returnsDifferentValues($x) { if ($x) { return 'foo'; } return array(); } $x = returnsDifferentValues($y); if (is_array($x)) { // $x is an array. } If this a common case that PHP Analyzer should handle natively, please let us know by opening an issue. Loading history...
139
140			return $statisticsRows;
141			}
142
143
144			/**
145			* This method groups rows with the same term and different cound and hits
146			* and calculates the average.
147			*
148			* @param array $statisticsRows
149			* @return array
150			*/
151			protected function mergeRowsWithSameKeyword(array $statisticsRows)
152			{
153			$result = [];
154			foreach ($statisticsRows as $statisticsRow) {
155			$term = $statisticsRow['keywords'];
156
157			$mergedRow = isset($result[$term]) ? $result[$term] : ['mergedrows' => 0, 'count' => 0];
158			$mergedRow['mergedrows']++;
159
160			// for the hits we need to take the average
161			$avgHits = $this->getAverageFromField($mergedRow, $statisticsRow, 'hits');
162			$mergedRow['hits'] = (int) $avgHits;
163
164			// for the count we need to take the sum, because it's the sum of searches
165			$mergedRow['count'] = $mergedRow['count'] + $statisticsRow['count'];
166
167			$mergedRow['keywords'] = $statisticsRow['keywords'];
168			$result[$term] = $mergedRow;
169			}
170
171			return array_values($result);
172			}
173
174			/**
175			* Get number of queries over time
176			*
177			* @param int $rootPageId
178			* @param int $days number of days of history to query
179			* @param int $bucketSeconds Seconds per bucket
180			* @return array [labels, data]
181			*/
182			public function getQueriesOverTime($rootPageId, $days = 30, $bucketSeconds = 3600)
183			{
184			$now = time();
185			$timeStart = $now - 86400 * intval($days); // 86400 seconds/day
186
187			$queries = $this->getDatabase()->exec_SELECTgetRows(
188			'FLOOR(tstamp/' . $bucketSeconds . ') AS bucket, tstamp, COUNT(*) AS numQueries',
189			'tx_solr_statistics',
190			'tstamp > ' . $timeStart . ' AND root_pid = ' . $rootPageId,
191			'bucket',
192			'bucket ASC'
193			);
194
195			return $queries;
196			}
197
198
199			/**
200			* This method is used to get an average value from merged statistic rows.
201			*
202			* @param array $mergedRow
203			* @param array $statisticsRow
204			* @return float\|int
205			*/
206			protected function getAverageFromField(array &$mergedRow, array $statisticsRow, $fieldName)
207			{
208			// when this is the first row we can take it.
209			if ($mergedRow['mergedrows'] === 1) {
210			$avgCount = $statisticsRow[$fieldName];
211			return $avgCount;
212			}
213
214			$oldAverage = $mergedRow[$fieldName];
215			$oldMergeRows = $mergedRow['mergedrows'] - 1;
216			$oldCount = $oldAverage * $oldMergeRows;
217			$avgCount = (($oldCount + $statisticsRow[$fieldName]) / $mergedRow['mergedrows']);
218			return $avgCount;
219			}
220
221			/**
222			* @return \TYPO3\CMS\Core\Database\DatabaseConnection
223			*/
224			protected function getDatabase()
			1 ignored issue – show Coding Style introduced 2016-08-17 21:45 UTC by Report Bug Copy Issue Report `getDatabase` uses the super-global variable `$GLOBALS` which is generally not recommended. Instead of super-globals, we recommend to explicitly inject the dependencies of your class. This makes your code less dependent on global state and it becomes generally more testable: // Bad class Router { public function generate($path) { return $_SERVER['HOST'].$path; } } // Better class Router { private $host; public function __construct($host) { $this->host = $host; } public function generate($path) { return $this->host.$path; } } class Controller { public function myAction(Request $request) { // Instead of $page = isset($_GET['page']) ? intval($_GET['page']) : 1; // Better (assuming you use the Symfony2 request) $page = $request->query->get('page', 1); } } Loading history...
225			{
226			return $GLOBALS['TYPO3_DB'];
227			}
228			}
229

TYPO3-Solr / ext-solr

Push — master ( 7682df...b1b8c7 )

StatisticsRepository A

Complexity

Size/Duplication

Coupling/Cohesion

Importance

8 Methods

Duplication Side-by-Side

Filter issues like