|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/* |
|
4
|
|
|
* This file is part of the ONGR package. |
|
5
|
|
|
* |
|
6
|
|
|
* (c) NFQ Technologies UAB <[email protected]> |
|
7
|
|
|
* |
|
8
|
|
|
* For the full copyright and license information, please view the LICENSE |
|
9
|
|
|
* file that was distributed with this source code. |
|
10
|
|
|
*/ |
|
11
|
|
|
|
|
12
|
|
|
namespace ONGR\ElasticsearchBundle\Mapping; |
|
13
|
|
|
|
|
14
|
|
|
use Doctrine\Common\Cache\CacheProvider; |
|
15
|
|
|
use ONGR\ElasticsearchBundle\Exception\DocumentParserException; |
|
16
|
|
|
use ONGR\ElasticsearchBundle\Exception\MissingDocumentAnnotationException; |
|
17
|
|
|
|
|
18
|
|
|
/** |
|
19
|
|
|
* DocumentParser wrapper for getting bundle documents mapping. |
|
20
|
|
|
*/ |
|
21
|
|
|
class MetadataCollector |
|
22
|
|
|
{ |
|
23
|
|
|
/** |
|
24
|
|
|
* @var DocumentFinder |
|
25
|
|
|
*/ |
|
26
|
|
|
private $finder; |
|
27
|
|
|
|
|
28
|
|
|
/** |
|
29
|
|
|
* @var DocumentParser |
|
30
|
|
|
*/ |
|
31
|
|
|
private $parser; |
|
32
|
|
|
|
|
33
|
|
|
/** |
|
34
|
|
|
* @var CacheProvider |
|
35
|
|
|
*/ |
|
36
|
|
|
private $cache = null; |
|
37
|
|
|
|
|
38
|
|
|
/** |
|
39
|
|
|
* @var bool |
|
40
|
|
|
*/ |
|
41
|
|
|
private $enableCache = false; |
|
42
|
|
|
|
|
43
|
|
|
/** |
|
44
|
|
|
* @param DocumentFinder $finder For finding documents. |
|
45
|
|
|
* @param DocumentParser $parser For reading document annotations. |
|
46
|
|
|
* @param CacheProvider $cache Cache provider to store the meta data for later use. |
|
47
|
|
|
*/ |
|
48
|
|
|
public function __construct($finder, $parser, $cache = null) |
|
49
|
|
|
{ |
|
50
|
|
|
$this->finder = $finder; |
|
51
|
|
|
$this->parser = $parser; |
|
52
|
|
|
$this->cache = $cache; |
|
53
|
|
|
} |
|
54
|
|
|
|
|
55
|
|
|
/** |
|
56
|
|
|
* Enables metadata caching. |
|
57
|
|
|
* |
|
58
|
|
|
* @param bool $enableCache |
|
59
|
|
|
*/ |
|
60
|
|
|
public function setEnableCache($enableCache) |
|
61
|
|
|
{ |
|
62
|
|
|
$this->enableCache = $enableCache; |
|
63
|
|
|
} |
|
64
|
|
|
|
|
65
|
|
|
/** |
|
66
|
|
|
* Fetches bundles mapping from documents. |
|
67
|
|
|
* |
|
68
|
|
|
* @param string[] $bundles Elasticsearch manager config. You can get bundles list from 'mappings' node. |
|
69
|
|
|
* @return array |
|
70
|
|
|
*/ |
|
71
|
|
|
public function getMappings(array $bundles) |
|
72
|
|
|
{ |
|
73
|
|
|
$output = []; |
|
74
|
|
|
foreach ($bundles as $name => $bundleConfig) { |
|
75
|
|
|
// Backward compatibility hack for support. |
|
76
|
|
|
if (!is_array($bundleConfig)) { |
|
77
|
|
|
$name = $bundleConfig; |
|
78
|
|
|
$bundleConfig = []; |
|
79
|
|
|
} |
|
80
|
|
|
$mappings = $this->getBundleMapping($name, $bundleConfig); |
|
81
|
|
|
|
|
82
|
|
|
$alreadyDefinedTypes = array_intersect_key($mappings, $output); |
|
83
|
|
|
if (count($alreadyDefinedTypes)) { |
|
84
|
|
|
throw new \LogicException( |
|
85
|
|
|
implode(',', array_keys($alreadyDefinedTypes)) . |
|
86
|
|
|
' type(s) already defined in other document, you can use the same ' . |
|
87
|
|
|
'type only once in a manager definition.' |
|
88
|
|
|
); |
|
89
|
|
|
} |
|
90
|
|
|
|
|
91
|
|
|
$output = array_merge($output, $mappings); |
|
92
|
|
|
} |
|
93
|
|
|
|
|
94
|
|
|
return $output; |
|
95
|
|
|
} |
|
96
|
|
|
|
|
97
|
|
|
/** |
|
98
|
|
|
* Searches for documents in the bundle and tries to read them. |
|
99
|
|
|
* |
|
100
|
|
|
* @param string $name |
|
101
|
|
|
* @param array $config Bundle configuration |
|
102
|
|
|
* |
|
103
|
|
|
* @return array Empty array on containing zero documents. |
|
104
|
|
|
*/ |
|
105
|
|
|
public function getBundleMapping($name, $config = []) |
|
106
|
|
|
{ |
|
107
|
|
|
if (!is_string($name)) { |
|
108
|
|
|
throw new \LogicException('getBundleMapping() in the Metadata collector expects a string argument only!'); |
|
109
|
|
|
} |
|
110
|
|
|
|
|
111
|
|
|
$cacheName = 'ongr.metadata.mapping.' . md5($name.serialize($config)); |
|
112
|
|
|
|
|
113
|
|
|
$this->enableCache && $mappings = $this->cache->contains($cacheName); |
|
114
|
|
|
|
|
115
|
|
|
if (isset($mappings)) { |
|
116
|
|
|
return $mappings; |
|
117
|
|
|
} |
|
118
|
|
|
|
|
119
|
|
|
$mappings = []; |
|
120
|
|
|
$documentDir = isset($config['document_dir']) ? $config['document_dir'] : DocumentFinder::DOCUMENT_DIR; |
|
121
|
|
|
|
|
122
|
|
|
// Handle the case when single document mapping requested |
|
123
|
|
|
// Usage od ":" in name is deprecated. This if is only for BC. |
|
124
|
|
|
if (strpos($name, ':') !== false) { |
|
125
|
|
|
list($bundle, $documentClass) = explode(':', $name); |
|
126
|
|
|
$documents = $this->finder->getBundleDocumentClasses($bundle); |
|
127
|
|
|
$documents = in_array($documentClass, $documents) ? [$documentClass] : []; |
|
128
|
|
|
} else { |
|
129
|
|
|
$documents = $this->finder->getBundleDocumentClasses($name, $documentDir); |
|
130
|
|
|
$bundle = $name; |
|
131
|
|
|
} |
|
132
|
|
|
|
|
133
|
|
|
$bundleNamespace = $this->finder->getBundleClass($bundle); |
|
134
|
|
|
$bundleNamespace = substr($bundleNamespace, 0, strrpos($bundleNamespace, '\\')); |
|
135
|
|
|
|
|
136
|
|
|
if (!count($documents)) { |
|
137
|
|
|
return []; |
|
138
|
|
|
} |
|
139
|
|
|
|
|
140
|
|
|
// Loop through documents found in bundle. |
|
141
|
|
|
foreach ($documents as $document) { |
|
142
|
|
|
$documentReflection = new \ReflectionClass( |
|
143
|
|
|
$bundleNamespace . |
|
144
|
|
|
'\\' . $documentDir . |
|
145
|
|
|
'\\' . $document |
|
146
|
|
|
); |
|
147
|
|
|
|
|
148
|
|
|
try { |
|
149
|
|
|
$documentMapping = $this->getDocumentReflectionMapping($documentReflection); |
|
150
|
|
|
} catch (MissingDocumentAnnotationException $exception) { |
|
151
|
|
|
// Not a document, just ignore |
|
152
|
|
|
continue; |
|
153
|
|
|
} |
|
154
|
|
|
|
|
155
|
|
|
if (!array_key_exists($documentMapping['type'], $mappings)) { |
|
156
|
|
|
$documentMapping['bundle'] = $bundle; |
|
157
|
|
|
$mappings = array_merge($mappings, [$documentMapping['type'] => $documentMapping]); |
|
158
|
|
|
} else { |
|
159
|
|
|
throw new \LogicException( |
|
160
|
|
|
$bundle . ' has 2 same type names defined in the documents. ' . |
|
161
|
|
|
'Type names must be unique!' |
|
162
|
|
|
); |
|
163
|
|
|
} |
|
164
|
|
|
} |
|
165
|
|
|
|
|
166
|
|
|
$this->enableCache && $this->cache->save($cacheName, $mappings); |
|
167
|
|
|
|
|
168
|
|
|
return $mappings; |
|
169
|
|
|
} |
|
170
|
|
|
|
|
171
|
|
|
/** |
|
172
|
|
|
* @param array $manager |
|
173
|
|
|
* |
|
174
|
|
|
* @return array |
|
175
|
|
|
*/ |
|
176
|
|
|
public function getManagerTypes($manager) |
|
177
|
|
|
{ |
|
178
|
|
|
$mapping = $this->getMappings($manager['mappings']); |
|
179
|
|
|
|
|
180
|
|
|
return array_keys($mapping); |
|
181
|
|
|
} |
|
182
|
|
|
|
|
183
|
|
|
/** |
|
184
|
|
|
* Resolves Elasticsearch type by document class. |
|
185
|
|
|
* |
|
186
|
|
|
* @param string $className FQCN or string in AppBundle:Document format |
|
187
|
|
|
* |
|
188
|
|
|
* @return string |
|
189
|
|
|
*/ |
|
190
|
|
|
public function getDocumentType($className) |
|
191
|
|
|
{ |
|
192
|
|
|
$mapping = $this->getMapping($className); |
|
193
|
|
|
|
|
194
|
|
|
return $mapping['type']; |
|
195
|
|
|
} |
|
196
|
|
|
|
|
197
|
|
|
/** |
|
198
|
|
|
* Retrieves prepared mapping to sent to the elasticsearch client. |
|
199
|
|
|
* |
|
200
|
|
|
* @param array $bundles Manager config. |
|
201
|
|
|
* |
|
202
|
|
|
* @return array|null |
|
203
|
|
|
*/ |
|
204
|
|
|
public function getClientMapping(array $bundles) |
|
205
|
|
|
{ |
|
206
|
|
|
/** @var array $typesMapping Array of filtered mappings for the elasticsearch client*/ |
|
207
|
|
|
$typesMapping = null; |
|
208
|
|
|
|
|
209
|
|
|
/** @var array $mappings All mapping info */ |
|
210
|
|
|
$mappings = $this->getMappings($bundles); |
|
211
|
|
|
|
|
212
|
|
|
foreach ($mappings as $type => $mapping) { |
|
213
|
|
|
if (!empty($mapping['properties'])) { |
|
214
|
|
|
$typesMapping[$type] = array_filter( |
|
215
|
|
|
array_merge( |
|
216
|
|
|
['properties' => $mapping['properties']], |
|
217
|
|
|
$mapping['fields'] |
|
218
|
|
|
), |
|
219
|
|
|
function ($value) { |
|
220
|
|
|
return (bool)$value || is_bool($value); |
|
221
|
|
|
} |
|
222
|
|
|
); |
|
223
|
|
|
} |
|
224
|
|
|
} |
|
225
|
|
|
|
|
226
|
|
|
return $typesMapping; |
|
227
|
|
|
} |
|
228
|
|
|
|
|
229
|
|
|
/** |
|
230
|
|
|
* Prepares analysis node for Elasticsearch client. |
|
231
|
|
|
* |
|
232
|
|
|
* @param array $bundles |
|
233
|
|
|
* @param array $analysisConfig |
|
234
|
|
|
* |
|
235
|
|
|
* @return array |
|
236
|
|
|
*/ |
|
237
|
|
|
public function getClientAnalysis(array $bundles, $analysisConfig = []) |
|
238
|
|
|
{ |
|
239
|
|
|
$cacheName = 'ongr.metadata.analysis.'.md5(serialize($bundles)); |
|
240
|
|
|
$this->enableCache && $typesAnalysis = $this->cache->fetch($cacheName); |
|
241
|
|
|
|
|
242
|
|
|
if (isset($typesAnalysis)) { |
|
243
|
|
|
return $typesAnalysis; |
|
244
|
|
|
} |
|
245
|
|
|
|
|
246
|
|
|
$typesAnalysis = [ |
|
247
|
|
|
'analyzer' => [], |
|
248
|
|
|
'filter' => [], |
|
249
|
|
|
'tokenizer' => [], |
|
250
|
|
|
'char_filter' => [], |
|
251
|
|
|
]; |
|
252
|
|
|
|
|
253
|
|
|
/** @var array $mappings All mapping info */ |
|
254
|
|
|
$mappings = $this->getMappings($bundles); |
|
255
|
|
|
|
|
256
|
|
|
foreach ($mappings as $type => $metadata) { |
|
257
|
|
|
foreach ($metadata['analyzers'] as $analyzerName) { |
|
258
|
|
|
if (isset($analysisConfig['analyzer'][$analyzerName])) { |
|
259
|
|
|
$analyzer = $analysisConfig['analyzer'][$analyzerName]; |
|
260
|
|
|
$typesAnalysis['analyzer'][$analyzerName] = $analyzer; |
|
261
|
|
|
$typesAnalysis['filter'] = $this->getAnalysisNodeConfiguration( |
|
262
|
|
|
'filter', |
|
263
|
|
|
$analyzer, |
|
264
|
|
|
$analysisConfig, |
|
265
|
|
|
$typesAnalysis['filter'] |
|
266
|
|
|
); |
|
267
|
|
|
$typesAnalysis['tokenizer'] = $this->getAnalysisNodeConfiguration( |
|
268
|
|
|
'tokenizer', |
|
269
|
|
|
$analyzer, |
|
270
|
|
|
$analysisConfig, |
|
271
|
|
|
$typesAnalysis['tokenizer'] |
|
272
|
|
|
); |
|
273
|
|
|
$typesAnalysis['char_filter'] = $this->getAnalysisNodeConfiguration( |
|
274
|
|
|
'char_filter', |
|
275
|
|
|
$analyzer, |
|
276
|
|
|
$analysisConfig, |
|
277
|
|
|
$typesAnalysis['char_filter'] |
|
278
|
|
|
); |
|
279
|
|
|
} |
|
280
|
|
|
} |
|
281
|
|
|
} |
|
282
|
|
|
|
|
283
|
|
|
$this->enableCache && $this->cache->save($cacheName, $mappings); |
|
284
|
|
|
|
|
285
|
|
|
return $typesAnalysis; |
|
286
|
|
|
} |
|
287
|
|
|
|
|
288
|
|
|
/** |
|
289
|
|
|
* Prepares analysis node content for Elasticsearch client. |
|
290
|
|
|
* |
|
291
|
|
|
* @param string $type Node type: filter, tokenizer or char_filter |
|
292
|
|
|
* @param array $analyzer Analyzer from which used helpers will be extracted. |
|
293
|
|
|
* @param array $analysisConfig Pre configured analyzers container |
|
294
|
|
|
* @param array $container Current analysis container where prepared helpers will be appended. |
|
295
|
|
|
* |
|
296
|
|
|
* @return array |
|
297
|
|
|
*/ |
|
298
|
|
|
private function getAnalysisNodeConfiguration($type, $analyzer, $analysisConfig, $container = []) |
|
299
|
|
|
{ |
|
300
|
|
|
if (isset($analyzer[$type])) { |
|
301
|
|
|
if (is_array($analyzer[$type])) { |
|
302
|
|
|
foreach ($analyzer[$type] as $filter) { |
|
303
|
|
|
if (isset($analysisConfig[$type][$filter])) { |
|
304
|
|
|
$container[$filter] = $analysisConfig[$type][$filter]; |
|
305
|
|
|
} |
|
306
|
|
|
} |
|
307
|
|
|
} else { |
|
308
|
|
|
if (isset($analysisConfig[$type][$analyzer[$type]])) { |
|
309
|
|
|
$container[$analyzer[$type]] = $analysisConfig[$type][$analyzer[$type]]; |
|
310
|
|
|
} |
|
311
|
|
|
} |
|
312
|
|
|
} |
|
313
|
|
|
return $container; |
|
314
|
|
|
} |
|
315
|
|
|
|
|
316
|
|
|
/** |
|
317
|
|
|
* Gathers annotation data from class. |
|
318
|
|
|
* |
|
319
|
|
|
* @param \ReflectionClass $reflectionClass Document reflection class to read mapping from. |
|
320
|
|
|
* |
|
321
|
|
|
* @return array |
|
322
|
|
|
* @throws DocumentParserException |
|
323
|
|
|
*/ |
|
324
|
|
|
private function getDocumentReflectionMapping(\ReflectionClass $reflectionClass) |
|
325
|
|
|
{ |
|
326
|
|
|
return $this->parser->parse($reflectionClass); |
|
327
|
|
|
} |
|
328
|
|
|
|
|
329
|
|
|
/** |
|
330
|
|
|
* Returns single document mapping metadata. |
|
331
|
|
|
* |
|
332
|
|
|
* @param string $namespace Document namespace |
|
333
|
|
|
* |
|
334
|
|
|
* @return array |
|
335
|
|
|
* @throws DocumentParserException |
|
336
|
|
|
*/ |
|
337
|
|
|
public function getMapping($namespace) |
|
338
|
|
|
{ |
|
339
|
|
|
$cacheName = 'ongr.metadata.document.'.md5($namespace); |
|
340
|
|
|
|
|
341
|
|
|
$namespace = $this->getClassName($namespace); |
|
342
|
|
|
$this->enableCache && $mapping = $this->cache->fetch($cacheName); |
|
343
|
|
|
|
|
344
|
|
|
if (isset($mapping)) { |
|
345
|
|
|
return $mapping; |
|
346
|
|
|
} |
|
347
|
|
|
|
|
348
|
|
|
$mapping = $this->getDocumentReflectionMapping(new \ReflectionClass($namespace)); |
|
349
|
|
|
|
|
350
|
|
|
$this->enableCache && $this->cache->save($cacheName, $mapping); |
|
351
|
|
|
|
|
352
|
|
|
return $mapping; |
|
353
|
|
|
} |
|
354
|
|
|
|
|
355
|
|
|
/** |
|
356
|
|
|
* Returns fully qualified class name. |
|
357
|
|
|
* |
|
358
|
|
|
* @param string $className |
|
359
|
|
|
* |
|
360
|
|
|
* @return string |
|
361
|
|
|
*/ |
|
362
|
|
|
public function getClassName($className) |
|
363
|
|
|
{ |
|
364
|
|
|
return $this->finder->getNamespace($className); |
|
365
|
|
|
} |
|
366
|
|
|
} |
|
367
|
|
|
|