SearchMappingService::generateSearchQueryContent() - Code Metrics - Inspection of "Add support for searching in external files (fixin..." - nextcloud/fulltextsearch_elasticsearch - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Pull Request — master (#100)

by Robin

created 2021-01-11 21:46 UTC

SearchMappingService::generateSearchQueryContent() A

↳ Parent: SearchMappingService

Complexity

Conditions	4
Paths	6

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
dl	0
loc	19
rs	9.6333
c	0
b	0
f	0
cc	4
nc	6
nop	1

<?php
declare(strict_types=1);


/**
 * FullTextSearch_Elasticsearch - Use Elasticsearch to index the content of your nextcloud
 *
 * This file is licensed under the Affero General Public License version 3 or
 * later. See the COPYING file.
 *
 * @author Maxence Lange <[email protected]>
 * @copyright 2018
 * @license GNU AGPL version 3 or any later version
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU Affero General Public License as
 * published by the Free Software Foundation, either version 3 of the
 * License, or (at your option) any later version.
 *
 * This program 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 Affero General Public License for more details.
 *
 * You should have received a copy of the GNU Affero General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 *
 */


namespace OCA\FullTextSearch_Elasticsearch\Service;


use OCA\FullTextSearch_Elasticsearch\Exceptions\ConfigurationException;
use OCA\FullTextSearch_Elasticsearch\Exceptions\QueryContentGenerationException;
use OCA\FullTextSearch_Elasticsearch\Exceptions\SearchQueryGenerationException;
use OCA\FullTextSearch_Elasticsearch\Model\QueryContent;
use OCP\FullTextSearch\Model\IDocumentAccess;
use OCP\FullTextSearch\Model\ISearchRequest;
use OCP\FullTextSearch\Model\ISearchRequestSimpleQuery;
use stdClass;

/**
 * Class SearchMappingService
 *
 * @package OCA\FullTextSearch_Elasticsearch\Service
 */
class SearchMappingService {

	/** @var ConfigService */
	private $configService;

	/** @var MiscService */
	private $miscService;

	/** @var IUserStoragesService */
	private $userStoragesService;

	/**
	 * SearchMappingService constructor.
	 *
	 * @param ConfigService $configService
	 * @param MiscService $miscService
	 * @param null|IUserStoragesService $userStoragesService
	 */
	public function __construct(ConfigService $configService, MiscService $miscService, IUserStoragesService $userStoragesService = null) {
		$this->configService = $configService;
		$this->miscService = $miscService;
		$this->userStoragesService = $userStoragesService;
	}


	/**
	 * @param ISearchRequest $request
	 * @param IDocumentAccess $access
	 * @param string $providerId
	 *
	 * @return array
	 * @throws ConfigurationException
	 * @throws SearchQueryGenerationException
	 */
	public function generateSearchQuery(
		ISearchRequest $request, IDocumentAccess $access, string $providerId
	): array {
		$query['params'] = $this->generateSearchQueryParams($request, $access, $providerId);
foreach ($collection as $item) {
    $myArray['foo'] = $item->getFoo();

    if ($item->hasBar()) {
        $myArray['bar'] = $item->getBar();
    }

    // do something with $myArray
}

		return $query;
	}


	/**
	 * @param ISearchRequest $request
	 * @param IDocumentAccess $access
	 * @param string $providerId
	 *
	 * @return array
	 * @throws ConfigurationException
	 * @throws SearchQueryGenerationException
	 */
	public function generateSearchQueryParams(
		ISearchRequest $request, IDocumentAccess $access, string $providerId
	): array {
		$params = [
			'index' => $this->configService->getElasticIndex(),
			'type'  => 'standard',
			'size'  => $request->getSize(),
			'from'  => (($request->getPage() - 1) * $request->getSize())
		];

		$bool = [];
		if ($request->getSearch() !== '') {
			$bool['must']['bool'] = $this->generateSearchQueryContent($request);
		}

		$bool['filter'][]['bool']['must'] = ['term' => ['provider' => $providerId]];
		$bool['filter'][]['bool']['should'] = $this->generateSearchQueryAccess($access);
		$bool['filter'][]['bool']['should'] =
			$this->generateSearchQueryTags('metatags', $request->getMetaTags());

		$bool['filter'][]['bool']['must'] =
			$this->generateSearchQueryTags('subtags', $request->getSubTags(true));

		$bool['filter'][]['bool']['must'] =
			$this->generateSearchSimpleQuery($request->getSimpleQueries());

//		$bool['filter'][]['bool']['should'] = $this->generateSearchQueryTags($request->getTags());

		$params['body']['query']['bool'] = $bool;
		$params['body']['highlight'] = $this->generateSearchHighlighting($request);

		$this->improveSearchQuerying($request, $params['body']['query']);

		return $params;
	}


	/**
	 * @param ISearchRequest $request
	 * @param array $arr
	 */
	private function improveSearchQuerying(ISearchRequest $request, array &$arr) {
//		$this->improveSearchWildcardQueries($request, $arr);
		$this->improveSearchWildcardFilters($request, $arr);
		$this->improveSearchRegexFilters($request, $arr);
	}


//	/**
//	 * @param SearchRequest $request
//	 * @param array $arr
//	 */
//	private function improveSearchWildcardQueries(SearchRequest $request, &$arr) {
//
//		$queries = $request->getWildcardQueries();
//		foreach ($queries as $query) {
//			$wildcards = [];
//			foreach ($query as $entry) {
//				$wildcards[] = ['wildcard' => $entry];
//			}
//
//			array_push($arr['bool']['must']['bool']['should'], $wildcards);
//		}
//
//	}


	/**
	 * @param ISearchRequest $request
	 * @param array $arr
	 */
	private function improveSearchWildcardFilters(ISearchRequest $request, array &$arr) {


		$filters = $request->getWildcardFilters();
		foreach ($filters as $filter) {
			$wildcards = [];
			foreach ($filter as $entry) {
				$wildcards[] = ['wildcard' => $entry];
			}

			$arr['bool']['filter'][]['bool']['should'] = $wildcards;
		}

	}


	/**
	 * @param ISearchRequest $request
	 * @param array $arr
	 */
	private function improveSearchRegexFilters(ISearchRequest $request, array &$arr) {


		$filters = $request->getRegexFilters();
		foreach ($filters as $filter) {
			$regex = [];
			foreach ($filter as $entry) {
				$regex[] = ['regexp' => $entry];
			}

			$arr['bool']['filter'][]['bool']['should'] = $regex;
		}

	}


	/**
	 * @param ISearchRequest $request
	 *
	 * @return array
	 * @throws SearchQueryGenerationException
	 */
	private function generateSearchQueryContent(ISearchRequest $request): array {
		$str = strtolower($request->getSearch());

		preg_match_all('/[^?]"(?:\\\\.|[^\\\\"])*"|\S+/', " $str ", $words);
		$queryContent = [];
		foreach ($words[0] as $word) {
			try {
				$queryContent[] = $this->generateQueryContent(trim($word));
			} catch (QueryContentGenerationException $e) {
				continue;
			}
		}

		if (sizeof($queryContent) === 0) {
			throw new SearchQueryGenerationException();
		}

		return $this->generateSearchQueryFromQueryContent($request, $queryContent);
	}


	/**
	 * @param string $word
	 *
	 * @return QueryContent
	 * @throws QueryContentGenerationException
	 */
	private function generateQueryContent(string $word): QueryContent {

		$searchQueryContent = new QueryContent($word);
		if (strlen($searchQueryContent->getWord()) === 0) {
			throw new QueryContentGenerationException();
		}

		return $searchQueryContent;
	}


	/**
	 * @param ISearchRequest $request
	 * @param QueryContent[] $contents
	 *
	 * @return array
	 */
	private function generateSearchQueryFromQueryContent(ISearchRequest $request, array $contents): array {
		$query = [];
		foreach ($contents as $content) {
			if (!array_key_exists($content->getShould(), $query)) {
				$query[$content->getShould()] = [];
			}

			if ($content->getShould() === 'must') {
				$query[$content->getShould()][] =
					['bool' => ['should' => $this->generateQueryContentFields($request, $content)]];
			} else {
				$query[$content->getShould()] = array_merge(
					$query[$content->getShould()], $this->generateQueryContentFields($request, $content)
				);
			}
		}

		return $query;
	}


	/**
	 * @param ISearchRequest $request
	 * @param QueryContent $content
	 *
	 * @return array
	 */
	private function generateQueryContentFields(ISearchRequest $request, QueryContent $content): array {
		$queryFields = [];

		$fields = array_merge(['content', 'title'], $request->getFields());
		foreach ($fields as $field) {
			if (!$this->fieldIsOutLimit($request, $field)) {
				$queryFields[] = [$content->getMatch() => [$field => $content->getWord()]];
			}
		}

		foreach ($request->getWildcardFields() as $field) {
			if (!$this->fieldIsOutLimit($request, $field)) {
				$queryFields[] = ['wildcard' => [$field => '*' . $content->getWord() . '*']];
			}
		}

		$parts = [];
		foreach ($this->getPartsFields($request) as $field) {
			if (!$this->fieldIsOutLimit($request, $field)) {
				$parts[] = $field;
			}
		}

		if (sizeof($parts) > 0) {
			$queryFields[] = [
				'query_string' => [
					'fields' => $parts,
					'query'  => $content->getWord()
				]
			];
		}

		return $queryFields;
	}


	/**
	 * @param IDocumentAccess $access
	 *
	 * @return array
	 */
	private function generateSearchQueryAccess(IDocumentAccess $access): array {

		$query = [];
		$query[] = ['term' => ['owner' => $access->getViewerId()]];
		$query[] = ['term' => ['users' => $access->getViewerId()]];
		$query[] = ['term' => ['users' => '__all']];

		foreach ($access->getGroups() as $group) {
			$query[] = ['term' => ['groups' => $group]];
		}

		foreach ($access->getCircles() as $circle) {
			$query[] = ['term' => ['circles' => $circle]];
		}

		$externalFilesConditions = $this->getExternalFilesConditions();
		if (!empty($externalFilesConditions)) {
			$query[] = ['bool' => ['must' => $externalFilesConditions]];
		}
		
		return $query;
	}

	/**
	 * @return array
	 */
	private function getExternalFileShares() : array {
		if (!$this->userStoragesService) {
			return [];
		}
		return $this->userStoragesService->getAllStoragesForUser();
	}

	/**
	 * Generates condition array for external files
	 * @return array
	 */
	private function getExternalFilesConditions(): array {
		// TODO :: normally we should check if user want's to search 
		// external files with "$request->getOption('files_external', '1') === '1'"
		$externalFileShares = $this->getExternalFileShares();
		if (empty($externalFileShares)) {
			return [];
		}
		$allowedExternalShares = [];
		foreach ($externalFileShares as $fileShare) {
			// If any external share is mounted as root, every
			// path is allowed
			if ($fileShare === '/') {
				$allowedExternalShares = [];
				break;
			}
			$allowedExternalShares[] = ['prefix' => ['title' => $fileShare]];
		}
		$externalFilesConditions = [];
		$externalFilesConditions[] = ['term' => ['source' => 'files_external']];
		$externalFilesConditions[] = ['term' => ['owner' => '']];
		if (!empty($allowedExternalShares)) {
			$externalFilesConditions[] = ['bool' => ['should' => $allowedExternalShares]];
		}

		return $externalFilesConditions;
	}

	/**
	 * @param ISearchRequest $request
	 * @param string $field
	 *
	 * @return bool
	 */
	private function fieldIsOutLimit(ISearchRequest $request, string $field): bool {
		$limit = $request->getLimitFields();
		if (sizeof($limit) === 0) {
			return false;
		}

		if (in_array($field, $limit)) {

			return false;
		}

		return true;
	}


	/**
	 * @param string $k
	 * @param array $tags
	 *
	 * @return array
	 */
	private function generateSearchQueryTags(string $k, array $tags): array {

		$query = [];
		foreach ($tags as $t) {
			$query[] = ['term' => [$k => $t]];
		}

		return $query;
	}


	/**
	 * @param ISearchRequestSimpleQuery[] $queries
	 *
	 * @return array
	 */
	private function generateSearchSimpleQuery(array $queries): array {
		$simpleQuery = [];
		foreach ($queries as $query) {
			// TODO: manage multiple entries array

			if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_KEYWORD) {

				$value = $query->getValues()[0];
				$simpleQuery[] = ['term' => [$query->getField() => $value]];
			}

			if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_WILDCARD) {

				$value = $query->getValues()[0];
				$simpleQuery[] = ['wildcard' => [$query->getField() => $value]];
			}

			if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_EQ) {

				$value = $query->getValues()[0];
				$simpleQuery[] = ['term' => [$query->getField() => $value]];
			}

			if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_GTE) {

				$value = $query->getValues()[0];
				$simpleQuery[] = ['range' => [$query->getField() => ['gte' => $value]]];
			}

			if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_LTE) {

				$value = $query->getValues()[0];
				$simpleQuery[] = ['range' => [$query->getField() => ['lte' => $value]]];
			}

			if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_GT) {

				$value = $query->getValues()[0];
				$simpleQuery[] = ['range' => [$query->getField() => ['gt' => $value]]];
			}

			if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_LT) {

				$value = $query->getValues()[0];
				$simpleQuery[] = ['range' => [$query->getField() => ['lt' => $value]]];
			}

		}

		return $simpleQuery;
	}


	/**
	 * @param ISearchRequest $request
	 *
	 * @return array
	 */
	private function generateSearchHighlighting(ISearchRequest $request): array {

		$parts = $this->getPartsFields($request);
		$fields = ['content' => new stdClass()];
		foreach ($parts as $part) {
			$fields[$part] = new stdClass();
		}

		return [
			'fields'    => $fields,
			'pre_tags'  => [''],
			'post_tags' => ['']
		];
	}


	/**
	 * @param string $providerId
	 * @param string $documentId
	 *
	 * @return array
	 * @throws ConfigurationException
	 */
	public function getDocumentQuery(string $providerId, string $documentId): array {
		return [
			'index' => $this->configService->getElasticIndex(),
			'type'  => 'standard',
			'id'    => $providerId . ':' . $documentId
		];
	}


	/**
	 * @param ISearchRequest $request
	 *
	 * @return array
	 */
	private function getPartsFields(ISearchRequest $request) {
		return array_map(
			function($value) {
				return 'parts.' . $value;
			}, $request->getParts()
		);
	}

}



1		<?php
2		declare(strict_types=1);
3
4
5		/**
6		* FullTextSearch_Elasticsearch - Use Elasticsearch to index the content of your nextcloud
7		*
8		* This file is licensed under the Affero General Public License version 3 or
9		* later. See the COPYING file.
10		*
11		* @author Maxence Lange <[email protected]>
12		* @copyright 2018
13		* @license GNU AGPL version 3 or any later version
14		*
15		* This program is free software: you can redistribute it and/or modify
16		* it under the terms of the GNU Affero General Public License as
17		* published by the Free Software Foundation, either version 3 of the
18		* License, or (at your option) any later version.
19		*
20		* This program 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 Affero General Public License for more details.
24		*
25		* You should have received a copy of the GNU Affero General Public License
26		* along with this program. If not, see <http://www.gnu.org/licenses/>.
27		*
28		*/
29
30
31		namespace OCA\FullTextSearch_Elasticsearch\Service;
32
33
34		use OCA\FullTextSearch_Elasticsearch\Exceptions\ConfigurationException;
35		use OCA\FullTextSearch_Elasticsearch\Exceptions\QueryContentGenerationException;
36		use OCA\FullTextSearch_Elasticsearch\Exceptions\SearchQueryGenerationException;
37		use OCA\FullTextSearch_Elasticsearch\Model\QueryContent;
38		use OCP\FullTextSearch\Model\IDocumentAccess;
39		use OCP\FullTextSearch\Model\ISearchRequest;
40		use OCP\FullTextSearch\Model\ISearchRequestSimpleQuery;
41		use stdClass;
42
43		/**
44		* Class SearchMappingService
45		*
46		* @package OCA\FullTextSearch_Elasticsearch\Service
47		*/
48		class SearchMappingService {
49
50		/** @var ConfigService */
51		private $configService;
52
53		/** @var MiscService */
54		private $miscService;
55
56		/** @var IUserStoragesService */
57		private $userStoragesService;
58
59		/**
60		* SearchMappingService constructor.
61		*
62		* @param ConfigService $configService
63		* @param MiscService $miscService
64		* @param null\|IUserStoragesService $userStoragesService
65		*/
66		public function __construct(ConfigService $configService, MiscService $miscService, IUserStoragesService $userStoragesService = null) {
67		$this->configService = $configService;
68		$this->miscService = $miscService;
69		$this->userStoragesService = $userStoragesService;
70		}
71
72
73		/**
74		* @param ISearchRequest $request
75		* @param IDocumentAccess $access
76		* @param string $providerId
77		*
78		* @return array
79		* @throws ConfigurationException
80		* @throws SearchQueryGenerationException
81		*/
82		public function generateSearchQuery(
83		ISearchRequest $request, IDocumentAccess $access, string $providerId
84		): array {
85		$query['params'] = $this->generateSearchQueryParams($request, $access, $providerId);
		0 ignored issues – show Coding Style Comprehensibility introduced 2018-09-04 13:31 UTC by Report Bug Copy Issue Report `$query` was never initialized. Although not strictly required by PHP, it is generally a good practice to add `$query = array();` before regardless. Adding an explicit array definition is generally preferable to implicit array definition as it guarantees a stable state of the code. Let’s take a look at an example: foreach ($collection as $item) { $myArray['foo'] = $item->getFoo(); if ($item->hasBar()) { $myArray['bar'] = $item->getBar(); } // do something with $myArray } As you can see in this example, the array `$myArray` is initialized the first time when the foreach loop is entered. You can also see that the value of the `bar` key is only written conditionally; thus, its value might result from a previous iteration. This might or might not be intended. To make your intention clear, your code more readible and to avoid accidental bugs, we recommend to add an explicit initialization `$myArray = array()` either outside or inside the foreach loop. Loading history...
86
87		return $query;
88		}
89
90
91		/**
92		* @param ISearchRequest $request
93		* @param IDocumentAccess $access
94		* @param string $providerId
95		*
96		* @return array
97		* @throws ConfigurationException
98		* @throws SearchQueryGenerationException
99		*/
100		public function generateSearchQueryParams(
101		ISearchRequest $request, IDocumentAccess $access, string $providerId
102		): array {
103		$params = [
104		'index' => $this->configService->getElasticIndex(),
105		'type' => 'standard',
106		'size' => $request->getSize(),
107		'from' => (($request->getPage() - 1) * $request->getSize())
108		];
109
110		$bool = [];
111		if ($request->getSearch() !== '') {
112		$bool['must']['bool'] = $this->generateSearchQueryContent($request);
113		}
114
115		$bool['filter'][]['bool']['must'] = ['term' => ['provider' => $providerId]];
116		$bool['filter'][]['bool']['should'] = $this->generateSearchQueryAccess($access);
117		$bool['filter'][]['bool']['should'] =
118		$this->generateSearchQueryTags('metatags', $request->getMetaTags());
119
120		$bool['filter'][]['bool']['must'] =
121		$this->generateSearchQueryTags('subtags', $request->getSubTags(true));
122
123		$bool['filter'][]['bool']['must'] =
124		$this->generateSearchSimpleQuery($request->getSimpleQueries());
125
126		// $bool['filter'][]['bool']['should'] = $this->generateSearchQueryTags($request->getTags());
127
128		$params['body']['query']['bool'] = $bool;
129		$params['body']['highlight'] = $this->generateSearchHighlighting($request);
130
131		$this->improveSearchQuerying($request, $params['body']['query']);
132
133		return $params;
134		}
135
136
137		/**
138		* @param ISearchRequest $request
139		* @param array $arr
140		*/
141		private function improveSearchQuerying(ISearchRequest $request, array &$arr) {
142		// $this->improveSearchWildcardQueries($request, $arr);
143		$this->improveSearchWildcardFilters($request, $arr);
144		$this->improveSearchRegexFilters($request, $arr);
145		}
146
147
148		// /**
149		// * @param SearchRequest $request
150		// * @param array $arr
151		// */
152		// private function improveSearchWildcardQueries(SearchRequest $request, &$arr) {
153		//
154		// $queries = $request->getWildcardQueries();
155		// foreach ($queries as $query) {
156		// $wildcards = [];
157		// foreach ($query as $entry) {
158		// $wildcards[] = ['wildcard' => $entry];
159		// }
160		//
161		// array_push($arr['bool']['must']['bool']['should'], $wildcards);
162		// }
163		//
164		// }
165
166
167		/**
168		* @param ISearchRequest $request
169		* @param array $arr
170		*/
171	View Code Duplication	private function improveSearchWildcardFilters(ISearchRequest $request, array &$arr) {
		0 ignored issues – show Duplication introduced 2018-03-15 11:51 UTC by Report Bug Copy Issue Report This method seems to be duplicated in your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
172
173		$filters = $request->getWildcardFilters();
174		foreach ($filters as $filter) {
175		$wildcards = [];
176		foreach ($filter as $entry) {
177		$wildcards[] = ['wildcard' => $entry];
178		}
179
180		$arr['bool']['filter'][]['bool']['should'] = $wildcards;
181		}
182
183		}
184
185
186		/**
187		* @param ISearchRequest $request
188		* @param array $arr
189		*/
190	View Code Duplication	private function improveSearchRegexFilters(ISearchRequest $request, array &$arr) {
		0 ignored issues – show Duplication introduced 2018-05-30 12:57 UTC by Report Bug Copy Issue Report This method seems to be duplicated in your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
191
192		$filters = $request->getRegexFilters();
193		foreach ($filters as $filter) {
194		$regex = [];
195		foreach ($filter as $entry) {
196		$regex[] = ['regexp' => $entry];
197		}
198
199		$arr['bool']['filter'][]['bool']['should'] = $regex;
200		}
201
202		}
203
204
205		/**
206		* @param ISearchRequest $request
207		*
208		* @return array
209		* @throws SearchQueryGenerationException
210		*/
211		private function generateSearchQueryContent(ISearchRequest $request): array {
212		$str = strtolower($request->getSearch());
213
214		preg_match_all('/[^?]"(?:\\\\.\|[^\\\\"])*"\|\S+/', " $str ", $words);
215		$queryContent = [];
216		foreach ($words[0] as $word) {
217		try {
218		$queryContent[] = $this->generateQueryContent(trim($word));
219		} catch (QueryContentGenerationException $e) {
220		continue;
221		}
222		}
223
224		if (sizeof($queryContent) === 0) {
225		throw new SearchQueryGenerationException();
226		}
227
228		return $this->generateSearchQueryFromQueryContent($request, $queryContent);
229		}
230
231
232		/**
233		* @param string $word
234		*
235		* @return QueryContent
236		* @throws QueryContentGenerationException
237		*/
238		private function generateQueryContent(string $word): QueryContent {
239
240		$searchQueryContent = new QueryContent($word);
241		if (strlen($searchQueryContent->getWord()) === 0) {
242		throw new QueryContentGenerationException();
243		}
244
245		return $searchQueryContent;
246		}
247
248
249		/**
250		* @param ISearchRequest $request
251		* @param QueryContent[] $contents
252		*
253		* @return array
254		*/
255		private function generateSearchQueryFromQueryContent(ISearchRequest $request, array $contents): array {
256		$query = [];
257		foreach ($contents as $content) {
258		if (!array_key_exists($content->getShould(), $query)) {
259		$query[$content->getShould()] = [];
260		}
261
262		if ($content->getShould() === 'must') {
263		$query[$content->getShould()][] =
264		['bool' => ['should' => $this->generateQueryContentFields($request, $content)]];
265		} else {
266		$query[$content->getShould()] = array_merge(
267		$query[$content->getShould()], $this->generateQueryContentFields($request, $content)
268		);
269		}
270		}
271
272		return $query;
273		}
274
275
276		/**
277		* @param ISearchRequest $request
278		* @param QueryContent $content
279		*
280		* @return array
281		*/
282		private function generateQueryContentFields(ISearchRequest $request, QueryContent $content): array {
283		$queryFields = [];
284
285		$fields = array_merge(['content', 'title'], $request->getFields());
286		foreach ($fields as $field) {
287		if (!$this->fieldIsOutLimit($request, $field)) {
288		$queryFields[] = [$content->getMatch() => [$field => $content->getWord()]];
289		}
290		}
291
292		foreach ($request->getWildcardFields() as $field) {
293		if (!$this->fieldIsOutLimit($request, $field)) {
294		$queryFields[] = ['wildcard' => [$field => '' . $content->getWord() . '']];
295		}
296		}
297
298		$parts = [];
299		foreach ($this->getPartsFields($request) as $field) {
300		if (!$this->fieldIsOutLimit($request, $field)) {
301		$parts[] = $field;
302		}
303		}
304
305		if (sizeof($parts) > 0) {
306		$queryFields[] = [
307		'query_string' => [
308		'fields' => $parts,
309		'query' => $content->getWord()
310		]
311		];
312		}
313
314		return $queryFields;
315		}
316
317
318		/**
319		* @param IDocumentAccess $access
320		*
321		* @return array
322		*/
323		private function generateSearchQueryAccess(IDocumentAccess $access): array {
324
325		$query = [];
326		$query[] = ['term' => ['owner' => $access->getViewerId()]];
327		$query[] = ['term' => ['users' => $access->getViewerId()]];
328		$query[] = ['term' => ['users' => '__all']];
329
330		foreach ($access->getGroups() as $group) {
331		$query[] = ['term' => ['groups' => $group]];
332		}
333
334		foreach ($access->getCircles() as $circle) {
335		$query[] = ['term' => ['circles' => $circle]];
336		}
337
338		$externalFilesConditions = $this->getExternalFilesConditions();
339		if (!empty($externalFilesConditions)) {
340		$query[] = ['bool' => ['must' => $externalFilesConditions]];
341		}
342
343		return $query;
344		}
345
346		/**
347		* @return array
348		*/
349		private function getExternalFileShares() : array {
350		if (!$this->userStoragesService) {
351		return [];
352		}
353		return $this->userStoragesService->getAllStoragesForUser();
354		}
355
356		/**
357		* Generates condition array for external files
358		* @return array
359		*/
360		private function getExternalFilesConditions(): array {
361		// TODO :: normally we should check if user want's to search
362		// external files with "$request->getOption('files_external', '1') === '1'"
363		$externalFileShares = $this->getExternalFileShares();
364		if (empty($externalFileShares)) {
365		return [];
366		}
367		$allowedExternalShares = [];
368		foreach ($externalFileShares as $fileShare) {
369		// If any external share is mounted as root, every
370		// path is allowed
371		if ($fileShare === '/') {
372		$allowedExternalShares = [];
373		break;
374		}
375		$allowedExternalShares[] = ['prefix' => ['title' => $fileShare]];
376		}
377		$externalFilesConditions = [];
378		$externalFilesConditions[] = ['term' => ['source' => 'files_external']];
379		$externalFilesConditions[] = ['term' => ['owner' => '']];
380		if (!empty($allowedExternalShares)) {
381		$externalFilesConditions[] = ['bool' => ['should' => $allowedExternalShares]];
382		}
383
384		return $externalFilesConditions;
385		}
386
387		/**
388		* @param ISearchRequest $request
389		* @param string $field
390		*
391		* @return bool
392		*/
393		private function fieldIsOutLimit(ISearchRequest $request, string $field): bool {
394		$limit = $request->getLimitFields();
395		if (sizeof($limit) === 0) {
396		return false;
397		}
398
399		if (in_array($field, $limit)) {
		0 ignored issues – show Unused Code introduced 2018-06-13 18:14 UTC by Report Bug Copy Issue Report This `if` statement, and the following `return` statement can be replaced with `return !in_array($field, $limit);`. Loading history...
400		return false;
401		}
402
403		return true;
404		}
405
406
407		/**
408		* @param string $k
409		* @param array $tags
410		*
411		* @return array
412		*/
413		private function generateSearchQueryTags(string $k, array $tags): array {
414
415		$query = [];
416		foreach ($tags as $t) {
417		$query[] = ['term' => [$k => $t]];
418		}
419
420		return $query;
421		}
422
423
424		/**
425		* @param ISearchRequestSimpleQuery[] $queries
426		*
427		* @return array
428		*/
429		private function generateSearchSimpleQuery(array $queries): array {
430		$simpleQuery = [];
431		foreach ($queries as $query) {
432		// TODO: manage multiple entries array
433
434	View Code Duplication	if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_KEYWORD) {
		0 ignored issues – show Duplication introduced 2019-07-15 15:36 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
435		$value = $query->getValues()[0];
436		$simpleQuery[] = ['term' => [$query->getField() => $value]];
437		}
438
439	View Code Duplication	if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_WILDCARD) {
		0 ignored issues – show Duplication introduced 2019-07-15 15:36 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
440		$value = $query->getValues()[0];
441		$simpleQuery[] = ['wildcard' => [$query->getField() => $value]];
442		}
443
444	View Code Duplication	if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_EQ) {
		0 ignored issues – show Duplication introduced 2019-07-15 15:36 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
445		$value = $query->getValues()[0];
446		$simpleQuery[] = ['term' => [$query->getField() => $value]];
447		}
448
449	View Code Duplication	if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_GTE) {
		0 ignored issues – show Duplication introduced 2019-07-15 15:36 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
450		$value = $query->getValues()[0];
451		$simpleQuery[] = ['range' => [$query->getField() => ['gte' => $value]]];
452		}
453
454	View Code Duplication	if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_LTE) {
		0 ignored issues – show Duplication introduced 2019-07-19 15:08 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
455		$value = $query->getValues()[0];
456		$simpleQuery[] = ['range' => [$query->getField() => ['lte' => $value]]];
457		}
458
459	View Code Duplication	if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_GT) {
		0 ignored issues – show Duplication introduced 2019-07-19 18:31 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
460		$value = $query->getValues()[0];
461		$simpleQuery[] = ['range' => [$query->getField() => ['gt' => $value]]];
462		}
463
464	View Code Duplication	if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_LT) {
		0 ignored issues – show Duplication introduced 2020-01-09 18:05 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
465		$value = $query->getValues()[0];
466		$simpleQuery[] = ['range' => [$query->getField() => ['lt' => $value]]];
467		}
468
469		}
470
471		return $simpleQuery;
472		}
473
474
475		/**
476		* @param ISearchRequest $request
477		*
478		* @return array
479		*/
480		private function generateSearchHighlighting(ISearchRequest $request): array {
481
482		$parts = $this->getPartsFields($request);
483		$fields = ['content' => new stdClass()];
484		foreach ($parts as $part) {
485		$fields[$part] = new stdClass();
486		}
487
488		return [
489		'fields' => $fields,
490		'pre_tags' => [''],
491		'post_tags' => ['']
492		];
493		}
494
495
496		/**
497		* @param string $providerId
498		* @param string $documentId
499		*
500		* @return array
501		* @throws ConfigurationException
502		*/
503		public function getDocumentQuery(string $providerId, string $documentId): array {
504		return [
505		'index' => $this->configService->getElasticIndex(),
506		'type' => 'standard',
507		'id' => $providerId . ':' . $documentId
508		];
509		}
510
511
512		/**
513		* @param ISearchRequest $request
514		*
515		* @return array
516		*/
517		private function getPartsFields(ISearchRequest $request) {
518		return array_map(
519		function($value) {
520		return 'parts.' . $value;
521		}, $request->getParts()
522		);
523		}
524
525		}
526
527

nextcloud / fulltextsearch_elasticsearch

Pull Request — master (#100)

SearchMappingService::generateSearchQueryContent() A

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like