1
|
|
|
<?php namespace Nord\Lumen\Elasticsearch\Search\Query\FullText; |
2
|
|
|
|
3
|
|
|
use Nord\Lumen\Elasticsearch\Exceptions\InvalidArgument; |
4
|
|
|
use Nord\Lumen\Elasticsearch\Search\Traits\HasField; |
5
|
|
|
use Nord\Lumen\Elasticsearch\Search\Query\Traits\HasType; |
6
|
|
|
use Nord\Lumen\Elasticsearch\Search\Query\Traits\HasValue; |
7
|
|
|
|
8
|
|
|
/** |
9
|
|
|
* A family of match queries that accepts text/numerics/dates, analyzes them, and constructs a query. |
10
|
|
|
* |
11
|
|
|
* @see https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-match-query.html |
12
|
|
|
*/ |
13
|
|
|
class MatchQuery extends AbstractQuery |
14
|
|
|
{ |
15
|
|
|
use HasField; |
16
|
|
|
use HasType; |
17
|
|
|
use HasValue; |
18
|
|
|
|
19
|
|
|
const OPERATOR_OR = 'or'; |
20
|
|
|
const OPERATOR_AND = 'and'; |
21
|
|
|
|
22
|
|
|
const ZERO_TERM_QUERY_NONE = 'none'; |
23
|
|
|
const ZERO_TERM_QUERY_ALL = 'all'; |
24
|
|
|
|
25
|
|
|
const TYPE_PHRASE = 'phrase'; |
26
|
|
|
const TYPE_PHRASE_PREFIX = 'phrase_prefix'; |
27
|
|
|
|
28
|
|
|
/** |
29
|
|
|
* @var string The operator flag can be set to "or" or "and" to control the boolean clauses (defaults to "or"). |
30
|
|
|
*/ |
31
|
|
|
private $operator; |
32
|
|
|
|
33
|
|
|
/** |
34
|
|
|
* @var string If the analyzer used removes all tokens in a query like a "stop" filter does, the default behavior |
35
|
|
|
* is to match no documents at all. In order to change that the zero_terms_query option can be used, which accepts |
36
|
|
|
* "none" (default) and "all" which corresponds to a "match_all" query. |
37
|
|
|
*/ |
38
|
|
|
private $zeroTermsQuery; |
39
|
|
|
|
40
|
|
|
/** |
41
|
|
|
* @var float The match query supports a cutoff_frequency that allows specifying an absolute or relative document |
42
|
|
|
* frequency where high frequency terms are moved into an optional subquery and are only scored if one of the low |
43
|
|
|
* frequency (below the cutoff) terms in the case of an or operator or all of the low frequency terms in the case |
44
|
|
|
* of an and operator match. |
45
|
|
|
*/ |
46
|
|
|
private $cutOffFrequency; |
47
|
|
|
|
48
|
|
|
/** |
49
|
|
|
* @var int A phrase query matches terms up to a configurable slop (which defaults to 0) in any order. |
50
|
|
|
* Transposed terms have a slop of 2. |
51
|
|
|
*/ |
52
|
|
|
private $slop; |
53
|
|
|
|
54
|
|
|
/** |
55
|
|
|
* @var int A "phrase_prefix" query option that controls to how many prefixes the last term will be expanded. |
56
|
|
|
* It is highly recommended to set it to an acceptable value to control the execution time of the query. |
57
|
|
|
*/ |
58
|
|
|
private $maxExpansions; |
59
|
|
|
|
60
|
|
|
/** |
61
|
|
|
* @var string The analyzer can be set to control which analyzer will perform the analysis process on the text. It |
62
|
|
|
* defaults to the field explicit mapping definition, or the default search analyzer, |
63
|
|
|
*/ |
64
|
|
|
private $analyzer; |
65
|
|
|
|
66
|
|
|
/** |
67
|
|
|
* @inheritdoc |
68
|
|
|
*/ |
69
|
|
|
public function toArray() |
70
|
|
|
{ |
71
|
|
|
$match = ['query' => $this->getValue()]; |
72
|
|
|
|
73
|
|
|
$match = $this->applyOptions($match); |
74
|
|
|
|
75
|
|
|
if (count($match) === 1 && isset($match['query'])) { |
76
|
|
|
$match = $match['query']; |
77
|
|
|
} |
78
|
|
|
|
79
|
|
|
return ['match' => [$this->getField() => $match]]; |
80
|
|
|
} |
81
|
|
|
|
82
|
|
|
|
83
|
|
|
/** |
84
|
|
|
* @param string $operator |
85
|
|
|
* @return MatchQuery |
86
|
|
|
* @throws InvalidArgument |
87
|
|
|
*/ |
88
|
|
|
public function setOperator($operator) |
89
|
|
|
{ |
90
|
|
|
$this->operator = $operator; |
91
|
|
|
return $this; |
92
|
|
|
} |
93
|
|
|
|
94
|
|
|
|
95
|
|
|
/** |
96
|
|
|
* @return string |
97
|
|
|
*/ |
98
|
|
|
public function getOperator() |
99
|
|
|
{ |
100
|
|
|
return $this->operator; |
101
|
|
|
} |
102
|
|
|
|
103
|
|
|
|
104
|
|
|
/** |
105
|
|
|
* @param string $zeroTermsQuery |
106
|
|
|
* @return MatchQuery |
107
|
|
|
* @throws InvalidArgument |
108
|
|
|
*/ |
109
|
|
|
public function setZeroTermsQuery($zeroTermsQuery) |
110
|
|
|
{ |
111
|
|
|
$this->zeroTermsQuery = $zeroTermsQuery; |
112
|
|
|
return $this; |
113
|
|
|
} |
114
|
|
|
|
115
|
|
|
|
116
|
|
|
/** |
117
|
|
|
* @return string |
118
|
|
|
*/ |
119
|
|
|
public function getZeroTermsQuery() |
120
|
|
|
{ |
121
|
|
|
return $this->zeroTermsQuery; |
122
|
|
|
} |
123
|
|
|
|
124
|
|
|
|
125
|
|
|
/** |
126
|
|
|
* @param float $cutOffFrequency |
127
|
|
|
* @return MatchQuery |
128
|
|
|
* @throws InvalidArgument |
129
|
|
|
*/ |
130
|
|
|
public function setCutOffFrequency($cutOffFrequency) |
131
|
|
|
{ |
132
|
|
|
$this->assertCutOffFrequency($cutOffFrequency); |
133
|
|
|
$this->cutOffFrequency = $cutOffFrequency; |
134
|
|
|
return $this; |
135
|
|
|
} |
136
|
|
|
|
137
|
|
|
|
138
|
|
|
/** |
139
|
|
|
* @return float |
140
|
|
|
*/ |
141
|
|
|
public function getCutOffFrequency() |
142
|
|
|
{ |
143
|
|
|
return $this->cutOffFrequency; |
144
|
|
|
} |
145
|
|
|
|
146
|
|
|
|
147
|
|
|
/** |
148
|
|
|
* @param string $type |
149
|
|
|
* @return MatchQuery |
150
|
|
|
* @throws InvalidArgument |
151
|
|
|
*/ |
152
|
|
|
public function setType($type) |
153
|
|
|
{ |
154
|
|
|
$this->type = $type; |
155
|
|
|
return $this; |
156
|
|
|
} |
157
|
|
|
|
158
|
|
|
|
159
|
|
|
/** |
160
|
|
|
* @param int $slop |
161
|
|
|
* @return MatchQuery |
162
|
|
|
* @throws InvalidArgument |
163
|
|
|
*/ |
164
|
|
|
public function setSlop($slop) |
165
|
|
|
{ |
166
|
|
|
$this->assertSlop($slop); |
167
|
|
|
$this->slop = $slop; |
168
|
|
|
return $this; |
169
|
|
|
} |
170
|
|
|
|
171
|
|
|
|
172
|
|
|
/** |
173
|
|
|
* @return int |
174
|
|
|
*/ |
175
|
|
|
public function getSlop() |
176
|
|
|
{ |
177
|
|
|
return $this->slop; |
178
|
|
|
} |
179
|
|
|
|
180
|
|
|
|
181
|
|
|
/** |
182
|
|
|
* @param int $maxExpansions |
183
|
|
|
* @return MatchQuery |
184
|
|
|
* @throws InvalidArgument |
185
|
|
|
*/ |
186
|
|
|
public function setMaxExpansions($maxExpansions) |
187
|
|
|
{ |
188
|
|
|
$this->assertMaxExpansions($maxExpansions); |
189
|
|
|
$this->maxExpansions = $maxExpansions; |
190
|
|
|
return $this; |
191
|
|
|
} |
192
|
|
|
|
193
|
|
|
|
194
|
|
|
/** |
195
|
|
|
* @return int |
196
|
|
|
*/ |
197
|
|
|
public function getMaxExpansions() |
198
|
|
|
{ |
199
|
|
|
return $this->maxExpansions; |
200
|
|
|
} |
201
|
|
|
|
202
|
|
|
|
203
|
|
|
/** |
204
|
|
|
* @param string $analyzer |
205
|
|
|
* @return MatchQuery |
206
|
|
|
*/ |
207
|
|
|
public function setAnalyzer($analyzer) |
208
|
|
|
{ |
209
|
|
|
$this->analyzer = $analyzer; |
210
|
|
|
return $this; |
211
|
|
|
} |
212
|
|
|
|
213
|
|
|
|
214
|
|
|
/** |
215
|
|
|
* @return string |
216
|
|
|
*/ |
217
|
|
|
public function getAnalyzer() |
218
|
|
|
{ |
219
|
|
|
return $this->analyzer; |
220
|
|
|
} |
221
|
|
|
|
222
|
|
|
|
223
|
|
|
/** |
224
|
|
|
* @param array $match |
225
|
|
|
* @return array |
226
|
|
|
*/ |
227
|
|
|
protected function applyOptions(array $match) |
228
|
|
|
{ |
229
|
|
|
$operator = $this->getOperator(); |
230
|
|
|
if (!is_null($operator)) { |
231
|
|
|
$match['operator'] = $operator; |
232
|
|
|
} |
233
|
|
|
$zeroTermsQuery = $this->getZeroTermsQuery(); |
234
|
|
|
if (!is_null($zeroTermsQuery)) { |
235
|
|
|
$match['zero_terms_query'] = $zeroTermsQuery; |
236
|
|
|
} |
237
|
|
|
$cutOffFreq = $this->getCutOffFrequency(); |
238
|
|
|
if (!is_null($cutOffFreq)) { |
239
|
|
|
$match['cutoff_frequency'] = $cutOffFreq; |
240
|
|
|
} |
241
|
|
|
$type = $this->getType(); |
242
|
|
|
if (!is_null($type)) { |
243
|
|
|
$match['type'] = $type; |
244
|
|
|
$slop = $this->getSlop(); |
245
|
|
|
if (!is_null($slop)) { |
246
|
|
|
$match['slop'] = $slop; |
247
|
|
|
} |
248
|
|
|
if ($match['type'] === self::TYPE_PHRASE_PREFIX) { |
249
|
|
|
$maxExp = $this->getMaxExpansions(); |
250
|
|
|
if (!is_null($maxExp)) { |
251
|
|
|
$match['max_expansions'] = $maxExp; |
252
|
|
|
} |
253
|
|
|
} |
254
|
|
|
} |
255
|
|
|
$analyzer = $this->getAnalyzer(); |
256
|
|
|
if (!is_null($analyzer)) { |
257
|
|
|
$match['analyzer'] = $analyzer; |
258
|
|
|
} |
259
|
|
|
|
260
|
|
|
return $match; |
261
|
|
|
} |
262
|
|
|
|
263
|
|
|
|
264
|
|
|
/** |
265
|
|
|
* @param float $cutOffFrequency |
266
|
|
|
* @throws InvalidArgument |
267
|
|
|
*/ |
268
|
|
|
protected function assertCutOffFrequency($cutOffFrequency) |
269
|
|
|
{ |
270
|
|
|
if (!is_float($cutOffFrequency)) { |
271
|
|
|
throw new InvalidArgument(sprintf( |
272
|
|
|
'Match Query `cutoff_frequency` must be a float value, "%s" given.', |
273
|
|
|
gettype($cutOffFrequency) |
274
|
|
|
)); |
275
|
|
|
} |
276
|
|
|
} |
277
|
|
|
|
278
|
|
|
|
279
|
|
|
/** |
280
|
|
|
* @param string $type |
281
|
|
|
* @throws InvalidArgument |
282
|
|
|
*/ |
283
|
|
|
protected function assertType($type) |
284
|
|
|
{ |
285
|
|
|
$validTypes = [self::TYPE_PHRASE, self::TYPE_PHRASE_PREFIX]; |
286
|
|
|
if (!in_array($type, $validTypes)) { |
287
|
|
|
throw new InvalidArgument(sprintf( |
288
|
|
|
'Match Query `type` must be one of "%s", "%s" given.', |
289
|
|
|
implode(', ', $validTypes), |
290
|
|
|
$type |
291
|
|
|
)); |
292
|
|
|
} |
293
|
|
|
} |
294
|
|
|
|
295
|
|
|
|
296
|
|
|
/** |
297
|
|
|
* @param int $slop |
298
|
|
|
* @throws InvalidArgument |
299
|
|
|
*/ |
300
|
|
|
protected function assertSlop($slop) |
301
|
|
|
{ |
302
|
|
|
if (!is_int($slop)) { |
303
|
|
|
throw new InvalidArgument(sprintf( |
304
|
|
|
'Match Query `slop` must be an integer, "%s" given.', |
305
|
|
|
gettype($slop) |
306
|
|
|
)); |
307
|
|
|
} |
308
|
|
|
} |
309
|
|
|
|
310
|
|
|
|
311
|
|
|
/** |
312
|
|
|
* @param int $maxExpansions |
313
|
|
|
* @throws InvalidArgument |
314
|
|
|
*/ |
315
|
|
|
protected function assertMaxExpansions($maxExpansions) |
316
|
|
|
{ |
317
|
|
|
if (!is_int($maxExpansions)) { |
318
|
|
|
throw new InvalidArgument(sprintf( |
319
|
|
|
'Match Query `max_expansions` must be an integer, "%s" given.', |
320
|
|
|
gettype($maxExpansions) |
321
|
|
|
)); |
322
|
|
|
} |
323
|
|
|
} |
324
|
|
|
} |
325
|
|
|
|