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
|
|
|
|
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
|
|
|
|
57
|
|
|
/** |
58
|
|
|
* SearchMappingService constructor. |
59
|
|
|
* |
60
|
|
|
* @param ConfigService $configService |
61
|
|
|
* @param MiscService $miscService |
62
|
|
|
*/ |
63
|
|
|
public function __construct(ConfigService $configService, MiscService $miscService) { |
64
|
|
|
$this->configService = $configService; |
65
|
|
|
$this->miscService = $miscService; |
66
|
|
|
} |
67
|
|
|
|
68
|
|
|
|
69
|
|
|
/** |
70
|
|
|
* @param ISearchRequest $request |
71
|
|
|
* @param IDocumentAccess $access |
72
|
|
|
* @param string $providerId |
73
|
|
|
* |
74
|
|
|
* @return array |
75
|
|
|
* @throws ConfigurationException |
76
|
|
|
* @throws SearchQueryGenerationException |
77
|
|
|
*/ |
78
|
|
|
public function generateSearchQuery( |
79
|
|
|
ISearchRequest $request, IDocumentAccess $access, string $providerId |
80
|
|
|
): array { |
81
|
|
|
$query['params'] = $this->generateSearchQueryParams($request, $access, $providerId); |
|
|
|
|
82
|
|
|
|
83
|
|
|
return $query; |
84
|
|
|
} |
85
|
|
|
|
86
|
|
|
|
87
|
|
|
/** |
88
|
|
|
* @param ISearchRequest $request |
89
|
|
|
* @param IDocumentAccess $access |
90
|
|
|
* @param string $providerId |
91
|
|
|
* |
92
|
|
|
* @return array |
93
|
|
|
* @throws ConfigurationException |
94
|
|
|
* @throws SearchQueryGenerationException |
95
|
|
|
*/ |
96
|
|
|
public function generateSearchQueryParams( |
97
|
|
|
ISearchRequest $request, IDocumentAccess $access, string $providerId |
98
|
|
|
): array { |
99
|
|
|
$params = [ |
100
|
|
|
'index' => $this->configService->getElasticIndex(), |
101
|
|
|
'type' => 'standard', |
102
|
|
|
'size' => $request->getSize(), |
103
|
|
|
'from' => (($request->getPage() - 1) * $request->getSize()) |
104
|
|
|
]; |
105
|
|
|
|
106
|
|
|
$bool = []; |
107
|
|
|
$bool['must']['bool']['should'] = $this->generateSearchQueryContent($request); |
108
|
|
|
|
109
|
|
|
$bool['filter'][]['bool']['must'] = ['term' => ['provider' => $providerId]]; |
110
|
|
|
$bool['filter'][]['bool']['should'] = $this->generateSearchQueryAccess($access); |
111
|
|
|
$bool['filter'][]['bool']['should'] = |
112
|
|
|
$this->generateSearchQueryTags('metatags', $request->getMetaTags()); |
113
|
|
|
|
114
|
|
|
$bool['filter'][]['bool']['must'] = |
115
|
|
|
$this->generateSearchQueryTags('subtags', $request->getSubTags(true)); |
116
|
|
|
|
117
|
|
|
$bool['filter'][]['bool']['must'] = |
118
|
|
|
$this->generateSearchSimpleQuery($request->getSimpleQueries()); |
119
|
|
|
|
120
|
|
|
// $bool['filter'][]['bool']['should'] = $this->generateSearchQueryTags($request->getTags()); |
121
|
|
|
|
122
|
|
|
$params['body']['query']['bool'] = $bool; |
123
|
|
|
$params['body']['highlight'] = $this->generateSearchHighlighting($request); |
124
|
|
|
|
125
|
|
|
$this->improveSearchQuerying($request, $params['body']['query']); |
126
|
|
|
|
127
|
|
|
return $params; |
128
|
|
|
} |
129
|
|
|
|
130
|
|
|
|
131
|
|
|
/** |
132
|
|
|
* @param ISearchRequest $request |
133
|
|
|
* @param array $arr |
134
|
|
|
*/ |
135
|
|
|
private function improveSearchQuerying(ISearchRequest $request, array &$arr) { |
136
|
|
|
// $this->improveSearchWildcardQueries($request, $arr); |
137
|
|
|
$this->improveSearchWildcardFilters($request, $arr); |
138
|
|
|
$this->improveSearchRegexFilters($request, $arr); |
139
|
|
|
} |
140
|
|
|
|
141
|
|
|
|
142
|
|
|
// /** |
143
|
|
|
// * @param SearchRequest $request |
144
|
|
|
// * @param array $arr |
145
|
|
|
// */ |
146
|
|
|
// private function improveSearchWildcardQueries(SearchRequest $request, &$arr) { |
147
|
|
|
// |
148
|
|
|
// $queries = $request->getWildcardQueries(); |
149
|
|
|
// foreach ($queries as $query) { |
150
|
|
|
// $wildcards = []; |
151
|
|
|
// foreach ($query as $entry) { |
152
|
|
|
// $wildcards[] = ['wildcard' => $entry]; |
153
|
|
|
// } |
154
|
|
|
// |
155
|
|
|
// array_push($arr['bool']['must']['bool']['should'], $wildcards); |
156
|
|
|
// } |
157
|
|
|
// |
158
|
|
|
// } |
159
|
|
|
|
160
|
|
|
|
161
|
|
|
/** |
162
|
|
|
* @param ISearchRequest $request |
163
|
|
|
* @param array $arr |
164
|
|
|
*/ |
165
|
|
View Code Duplication |
private function improveSearchWildcardFilters(ISearchRequest $request, array &$arr) { |
|
|
|
|
166
|
|
|
|
167
|
|
|
$filters = $request->getWildcardFilters(); |
168
|
|
|
foreach ($filters as $filter) { |
169
|
|
|
$wildcards = []; |
170
|
|
|
foreach ($filter as $entry) { |
171
|
|
|
$wildcards[] = ['wildcard' => $entry]; |
172
|
|
|
} |
173
|
|
|
|
174
|
|
|
$arr['bool']['filter'][]['bool']['should'] = $wildcards; |
175
|
|
|
} |
176
|
|
|
|
177
|
|
|
} |
178
|
|
|
|
179
|
|
|
|
180
|
|
|
/** |
181
|
|
|
* @param ISearchRequest $request |
182
|
|
|
* @param array $arr |
183
|
|
|
*/ |
184
|
|
View Code Duplication |
private function improveSearchRegexFilters(ISearchRequest $request, array &$arr) { |
|
|
|
|
185
|
|
|
|
186
|
|
|
$filters = $request->getRegexFilters(); |
187
|
|
|
foreach ($filters as $filter) { |
188
|
|
|
$regex = []; |
189
|
|
|
foreach ($filter as $entry) { |
190
|
|
|
$regex[] = ['regexp' => $entry]; |
191
|
|
|
} |
192
|
|
|
|
193
|
|
|
$arr['bool']['filter'][]['bool']['should'] = $regex; |
194
|
|
|
} |
195
|
|
|
|
196
|
|
|
} |
197
|
|
|
|
198
|
|
|
|
199
|
|
|
/** |
200
|
|
|
* @param ISearchRequest $request |
201
|
|
|
* |
202
|
|
|
* @return array |
203
|
|
|
* @throws SearchQueryGenerationException |
204
|
|
|
*/ |
205
|
|
|
private function generateSearchQueryContent(ISearchRequest $request): array { |
206
|
|
|
$str = strtolower($request->getSearch()); |
207
|
|
|
|
208
|
|
|
preg_match_all('/[^?]"(?:\\\\.|[^\\\\"])*"|\S+/', " $str ", $words); |
209
|
|
|
$queryContent = []; |
210
|
|
|
foreach ($words[0] as $word) { |
211
|
|
|
try { |
212
|
|
|
$queryContent[] = $this->generateQueryContent(trim($word)); |
213
|
|
|
} catch (QueryContentGenerationException $e) { |
214
|
|
|
continue; |
215
|
|
|
} |
216
|
|
|
} |
217
|
|
|
|
218
|
|
|
if (sizeof($queryContent) === 0) { |
219
|
|
|
throw new SearchQueryGenerationException(); |
220
|
|
|
} |
221
|
|
|
|
222
|
|
|
return $this->generateSearchQueryFromQueryContent($request, $queryContent); |
223
|
|
|
} |
224
|
|
|
|
225
|
|
|
|
226
|
|
|
/** |
227
|
|
|
* @param string $word |
228
|
|
|
* |
229
|
|
|
* @return QueryContent |
230
|
|
|
* @throws QueryContentGenerationException |
231
|
|
|
*/ |
232
|
|
|
private function generateQueryContent(string $word): QueryContent { |
233
|
|
|
|
234
|
|
|
$searchQueryContent = new QueryContent($word); |
235
|
|
|
if (strlen($searchQueryContent->getWord()) === 0) { |
236
|
|
|
throw new QueryContentGenerationException(); |
237
|
|
|
} |
238
|
|
|
|
239
|
|
|
return $searchQueryContent; |
240
|
|
|
} |
241
|
|
|
|
242
|
|
|
|
243
|
|
|
/** |
244
|
|
|
* @param ISearchRequest $request |
245
|
|
|
* @param QueryContent[] $queryContents |
246
|
|
|
* |
247
|
|
|
* @return array |
248
|
|
|
*/ |
249
|
|
|
private function generateSearchQueryFromQueryContent( |
250
|
|
|
ISearchRequest $request, array $queryContents |
251
|
|
|
): array { |
252
|
|
|
|
253
|
|
|
$query = $queryWords = []; |
254
|
|
|
foreach ($queryContents as $queryContent) { |
255
|
|
|
$queryWords[$queryContent->getShould()][] = |
256
|
|
|
$this->generateQueryContentFields($request, $queryContent); |
257
|
|
|
} |
258
|
|
|
|
259
|
|
|
$listShould = array_keys($queryWords); |
260
|
|
|
foreach ($listShould as $itemShould) { |
261
|
|
|
$query[$itemShould][] = $queryWords[$itemShould]; |
262
|
|
|
} |
263
|
|
|
|
264
|
|
|
return ['bool' => $query]; |
265
|
|
|
} |
266
|
|
|
|
267
|
|
|
|
268
|
|
|
/** |
269
|
|
|
* @param ISearchRequest $request |
270
|
|
|
* @param QueryContent $content |
271
|
|
|
* |
272
|
|
|
* @return array |
273
|
|
|
*/ |
274
|
|
|
private function generateQueryContentFields(ISearchRequest $request, QueryContent $content |
275
|
|
|
): array { |
276
|
|
|
$queryFields = []; |
277
|
|
|
|
278
|
|
|
$fields = array_merge(['content', 'title'], $request->getFields()); |
279
|
|
|
foreach ($fields as $field) { |
280
|
|
|
if (!$this->fieldIsOutLimit($request, $field)) { |
281
|
|
|
$queryFields[] = [$content->getMatch() => [$field => $content->getWord()]]; |
282
|
|
|
} |
283
|
|
|
} |
284
|
|
|
|
285
|
|
|
foreach ($request->getWildcardFields() as $field) { |
286
|
|
|
if (!$this->fieldIsOutLimit($request, $field)) { |
287
|
|
|
$queryFields[] = ['wildcard' => [$field => '*' . $content->getWord() . '*']]; |
288
|
|
|
} |
289
|
|
|
} |
290
|
|
|
|
291
|
|
|
$parts = []; |
292
|
|
|
foreach ($this->getPartsFields($request) as $field) { |
293
|
|
|
if (!$this->fieldIsOutLimit($request, $field)) { |
294
|
|
|
$parts[] = $field; |
295
|
|
|
} |
296
|
|
|
} |
297
|
|
|
|
298
|
|
|
if (sizeof($parts) > 0) { |
299
|
|
|
$queryFields[] = [ |
300
|
|
|
'query_string' => [ |
301
|
|
|
'fields' => $parts, |
302
|
|
|
'query' => $content->getWord() |
303
|
|
|
] |
304
|
|
|
]; |
305
|
|
|
} |
306
|
|
|
|
307
|
|
|
return ['bool' => ['should' => $queryFields]]; |
308
|
|
|
} |
309
|
|
|
|
310
|
|
|
|
311
|
|
|
/** |
312
|
|
|
* @param IDocumentAccess $access |
313
|
|
|
* |
314
|
|
|
* @return array |
315
|
|
|
*/ |
316
|
|
|
private function generateSearchQueryAccess(IDocumentAccess $access): array { |
317
|
|
|
|
318
|
|
|
$query = []; |
319
|
|
|
$query[] = ['term' => ['owner' => $access->getViewerId()]]; |
320
|
|
|
$query[] = ['term' => ['users' => $access->getViewerId()]]; |
321
|
|
|
$query[] = ['term' => ['users' => '__all']]; |
322
|
|
|
|
323
|
|
|
foreach ($access->getGroups() as $group) { |
324
|
|
|
$query[] = ['term' => ['groups' => $group]]; |
325
|
|
|
} |
326
|
|
|
|
327
|
|
|
foreach ($access->getCircles() as $circle) { |
328
|
|
|
$query[] = ['term' => ['circles' => $circle]]; |
329
|
|
|
} |
330
|
|
|
|
331
|
|
|
return $query; |
332
|
|
|
} |
333
|
|
|
|
334
|
|
|
|
335
|
|
|
/** |
336
|
|
|
* @param ISearchRequest $request |
337
|
|
|
* @param string $field |
338
|
|
|
* |
339
|
|
|
* @return bool |
340
|
|
|
*/ |
341
|
|
|
private function fieldIsOutLimit(ISearchRequest $request, string $field): bool { |
342
|
|
|
$limit = $request->getLimitFields(); |
343
|
|
|
if (sizeof($limit) === 0) { |
344
|
|
|
return false; |
345
|
|
|
} |
346
|
|
|
|
347
|
|
|
if (in_array($field, $limit)) { |
|
|
|
|
348
|
|
|
return false; |
349
|
|
|
} |
350
|
|
|
|
351
|
|
|
return true; |
352
|
|
|
} |
353
|
|
|
|
354
|
|
|
|
355
|
|
|
/** |
356
|
|
|
* @param string $k |
357
|
|
|
* @param array $tags |
358
|
|
|
* |
359
|
|
|
* @return array |
360
|
|
|
*/ |
361
|
|
|
private function generateSearchQueryTags(string $k, array $tags): array { |
362
|
|
|
|
363
|
|
|
$query = []; |
364
|
|
|
foreach ($tags as $t) { |
365
|
|
|
$query[] = ['term' => [$k => $t]]; |
366
|
|
|
} |
367
|
|
|
|
368
|
|
|
return $query; |
369
|
|
|
} |
370
|
|
|
|
371
|
|
|
|
372
|
|
|
/** |
373
|
|
|
* @param ISearchRequestSimpleQuery[] $queries |
374
|
|
|
* |
375
|
|
|
* @return array |
376
|
|
|
*/ |
377
|
|
|
private function generateSearchSimpleQuery(array $queries): array { |
378
|
|
|
$simpleQuery = []; |
379
|
|
|
foreach ($queries as $query) { |
380
|
|
|
// TODO: manage multiple entries array |
381
|
|
|
|
382
|
|
View Code Duplication |
if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_GTE) { |
|
|
|
|
383
|
|
|
$value = $query->getValues()[0]; |
384
|
|
|
|
385
|
|
|
$simpleQuery[] = ['range' => [$query->getField() => ['gte' => $value]]]; |
386
|
|
|
} |
387
|
|
|
|
388
|
|
View Code Duplication |
if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_LTE) { |
|
|
|
|
389
|
|
|
$value = $query->getValues()[0]; |
390
|
|
|
|
391
|
|
|
$simpleQuery[] = ['range' => [$query->getField() => ['lte' => $value]]]; |
392
|
|
|
} |
393
|
|
|
|
394
|
|
View Code Duplication |
if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_GT) { |
|
|
|
|
395
|
|
|
$value = $query->getValues()[0]; |
396
|
|
|
|
397
|
|
|
$simpleQuery[] = ['range' => [$query->getField() => ['gt' => $value]]]; |
398
|
|
|
} |
399
|
|
|
|
400
|
|
View Code Duplication |
if ($query->getType() === ISearchRequestSimpleQuery::COMPARE_TYPE_INT_LT) { |
|
|
|
|
401
|
|
|
$value = $query->getValues()[0]; |
402
|
|
|
|
403
|
|
|
$simpleQuery[] = ['range' => [$query->getField() => ['lt' => $value]]]; |
404
|
|
|
} |
405
|
|
|
|
406
|
|
|
} |
407
|
|
|
|
408
|
|
|
return $simpleQuery; |
409
|
|
|
} |
410
|
|
|
|
411
|
|
|
|
412
|
|
|
/** |
413
|
|
|
* @param ISearchRequest $request |
414
|
|
|
* |
415
|
|
|
* @return array |
416
|
|
|
*/ |
417
|
|
|
private function generateSearchHighlighting(ISearchRequest $request): array { |
418
|
|
|
|
419
|
|
|
$parts = $this->getPartsFields($request); |
420
|
|
|
$fields = ['content' => new \stdClass()]; |
421
|
|
|
foreach ($parts as $part) { |
422
|
|
|
$fields[$part] = new \stdClass(); |
423
|
|
|
} |
424
|
|
|
|
425
|
|
|
return [ |
426
|
|
|
'fields' => $fields, |
427
|
|
|
'pre_tags' => [''], |
428
|
|
|
'post_tags' => [''] |
429
|
|
|
]; |
430
|
|
|
} |
431
|
|
|
|
432
|
|
|
|
433
|
|
|
/** |
434
|
|
|
* @param string $providerId |
435
|
|
|
* @param string $documentId |
436
|
|
|
* |
437
|
|
|
* @return array |
438
|
|
|
* @throws ConfigurationException |
439
|
|
|
*/ |
440
|
|
|
public function getDocumentQuery(string $providerId, string $documentId): array { |
441
|
|
|
return [ |
442
|
|
|
'index' => $this->configService->getElasticIndex(), |
443
|
|
|
'type' => 'standard', |
444
|
|
|
'id' => $providerId . ':' . $documentId |
445
|
|
|
]; |
446
|
|
|
} |
447
|
|
|
|
448
|
|
|
|
449
|
|
|
/** |
450
|
|
|
* @param ISearchRequest $request |
451
|
|
|
* |
452
|
|
|
* @return array |
453
|
|
|
*/ |
454
|
|
|
private function getPartsFields(ISearchRequest $request) { |
455
|
|
|
return array_map( |
456
|
|
|
function($value) { |
457
|
|
|
return 'parts.' . $value; |
458
|
|
|
}, $request->getParts() |
459
|
|
|
); |
460
|
|
|
} |
461
|
|
|
|
462
|
|
|
} |
463
|
|
|
|
464
|
|
|
|
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:
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 thebar
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.