|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
|
|
4
|
|
|
namespace Wordlift\Analysis\Response; |
|
5
|
|
|
|
|
6
|
|
|
use stdClass; |
|
7
|
|
|
|
|
8
|
|
|
class Analysis_Response_Ops { |
|
9
|
|
|
|
|
10
|
|
|
/** |
|
11
|
|
|
* The analysis response json. |
|
12
|
|
|
* |
|
13
|
|
|
* @since 3.21.5 |
|
14
|
|
|
* @access private |
|
15
|
|
|
* @var mixed $json Holds the analysis response json. |
|
16
|
|
|
*/ |
|
17
|
|
|
private $json; |
|
18
|
|
|
|
|
19
|
|
|
/** |
|
20
|
|
|
* Holds the {@link Wordlift_Entity_Uri_Service}. |
|
21
|
|
|
* |
|
22
|
|
|
* @since 3.21.5 |
|
23
|
|
|
* @access private |
|
24
|
|
|
* @var \Wordlift_Entity_Uri_Service $entity_uri_service The {@link Wordlift_Entity_Uri_Service} instance. |
|
25
|
|
|
*/ |
|
26
|
|
|
private $entity_uri_service; |
|
27
|
|
|
|
|
28
|
|
|
private $entity_service; |
|
29
|
|
|
|
|
30
|
|
|
/** |
|
31
|
|
|
* @var \Wordlift_Entity_Type_Service |
|
32
|
|
|
*/ |
|
33
|
|
|
private $entity_type_service; |
|
34
|
|
|
/** |
|
35
|
|
|
* @var \Wordlift_Post_Image_Storage |
|
36
|
|
|
*/ |
|
37
|
|
|
private $post_image_storage; |
|
38
|
|
|
|
|
39
|
|
|
/** |
|
40
|
|
|
* Analysis_Response_Ops constructor. |
|
41
|
|
|
* |
|
42
|
|
|
* @param \Wordlift_Entity_Uri_Service $entity_uri_service The {@link Wordlift_Entity_Uri_Service}. |
|
43
|
|
|
* @param \Wordlift_Entity_Service $entity_service The {@link Wordlift_Entity_Service}. |
|
44
|
|
|
* @param \Wordlift_Entity_Type_Service $entity_type_service The {@link Wordlift_Entity_Type_Service}. |
|
45
|
|
|
* @param \Wordlift_Post_Image_Storage $post_image_storage A {@link Wordlift_Post_Image_Storage} instance. |
|
46
|
|
|
* @param mixed $json The analysis response json. |
|
47
|
|
|
* |
|
48
|
|
|
* @since 3.21.5 |
|
49
|
|
|
*/ |
|
50
|
|
|
public function __construct( $entity_uri_service, $entity_service, $entity_type_service, $post_image_storage, $json ) { |
|
51
|
|
|
|
|
52
|
|
|
$this->json = $json; |
|
53
|
|
|
$this->entity_uri_service = $entity_uri_service; |
|
54
|
|
|
$this->entity_service = $entity_service; |
|
55
|
|
|
$this->entity_type_service = $entity_type_service; |
|
56
|
|
|
$this->post_image_storage = $post_image_storage; |
|
57
|
|
|
|
|
58
|
|
|
} |
|
59
|
|
|
|
|
60
|
|
|
/** |
|
61
|
|
|
* Switches remote entities, i.e. entities with id outside the local dataset, to local entities. |
|
62
|
|
|
* |
|
63
|
|
|
* The function takes all the entities that have an id which is not local. For each remote entity, a list of URIs |
|
64
|
|
|
* is built comprising the entity id and the sameAs. Then a query is issued in the local database to find potential |
|
65
|
|
|
* matches from the local vocabulary. |
|
66
|
|
|
* |
|
67
|
|
|
* If found, the entity id is swapped with the local id and the remote id is added to the sameAs. |
|
68
|
|
|
* |
|
69
|
|
|
* @return Analysis_Response_Ops The current Analysis_Response_Ops instance. |
|
70
|
|
|
*/ |
|
71
|
|
|
public function make_entities_local() { |
|
72
|
|
|
|
|
73
|
|
|
if ( ! isset( $this->json->entities ) ) { |
|
74
|
|
|
return $this; |
|
75
|
|
|
} |
|
76
|
|
|
|
|
77
|
|
|
// Get the URIs. |
|
78
|
|
|
$uris = array_keys( get_object_vars( $this->json->entities ) ); |
|
79
|
|
|
|
|
80
|
|
|
// Filter only the external URIs. |
|
81
|
|
|
$entity_uri_service = $this->entity_uri_service; |
|
82
|
|
|
$external_uris = array_filter( $uris, function ( $item ) use ( $entity_uri_service ) { |
|
83
|
|
|
return ! $entity_uri_service->is_internal( $item ); |
|
84
|
|
|
} ); |
|
85
|
|
|
|
|
86
|
|
|
// Preload the URIs. |
|
87
|
|
|
$entity_uri_service->preload_uris( $external_uris ); |
|
88
|
|
|
|
|
89
|
|
|
$mappings = array(); |
|
90
|
|
|
foreach ( $external_uris as $external_uri ) { |
|
91
|
|
|
$entity = $entity_uri_service->get_entity( $external_uri ); |
|
92
|
|
|
if ( null !== $entity ) { |
|
93
|
|
|
|
|
94
|
|
|
// Get the internal URI. |
|
95
|
|
|
$internal_uri = $this->entity_service->get_uri( $entity->ID ); |
|
96
|
|
|
$mappings[ $external_uri ] = $internal_uri; |
|
97
|
|
|
} |
|
98
|
|
|
} |
|
99
|
|
|
|
|
100
|
|
|
foreach ( $mappings as $external_uri => $internal_uri ) { |
|
101
|
|
|
|
|
102
|
|
|
// Move the data from the external URI to the internal URI. |
|
103
|
|
|
if ( ! isset( $this->json->entities->{$internal_uri} ) ) { |
|
104
|
|
|
$this->json->entities->{$internal_uri} = $this->json->entities->{$external_uri}; |
|
105
|
|
|
} |
|
106
|
|
|
|
|
107
|
|
|
// Ensure sameAs is an array. |
|
108
|
|
|
if ( ! isset( $this->json->entities->{$internal_uri}->sameAs ) |
|
109
|
|
|
|| ! is_array( $this->json->entities->{$internal_uri}->sameAs ) ) { |
|
110
|
|
|
$this->json->entities->{$internal_uri}->sameAs = array(); |
|
111
|
|
|
} |
|
112
|
|
|
|
|
113
|
|
|
// Add the external URI as sameAs. |
|
114
|
|
|
$this->json->entities->{$internal_uri}->sameAs[] = $external_uri; |
|
115
|
|
|
|
|
116
|
|
|
// Finally remove the external URI. |
|
117
|
|
|
unset( $this->json->entities->{$external_uri} ); |
|
118
|
|
|
} |
|
119
|
|
|
|
|
120
|
|
|
if ( isset( $this->json->annotations ) ) { |
|
121
|
|
|
foreach ( $this->json->annotations as $key => $annotation ) { |
|
122
|
|
|
if ( isset( $annotation->entityMatches ) ) { |
|
123
|
|
|
foreach ( $annotation->entityMatches as $match ) { |
|
124
|
|
|
if ( isset( $match->entityId ) && isset( $mappings[ $match->entityId ] ) ) { |
|
125
|
|
|
$match->entityId = $mappings[ $match->entityId ]; |
|
126
|
|
|
} |
|
127
|
|
|
} |
|
128
|
|
|
} |
|
129
|
|
|
} |
|
130
|
|
|
} |
|
131
|
|
|
|
|
132
|
|
|
return $this; |
|
133
|
|
|
} |
|
134
|
|
|
|
|
135
|
|
|
/** |
|
136
|
|
|
* Add occurrences by parsing the provided html content. |
|
137
|
|
|
* |
|
138
|
|
|
* @param string $content The html content with annotations. |
|
139
|
|
|
* |
|
140
|
|
|
* @return Analysis_Response_Ops The {@link Analysis_Response_Ops} instance. |
|
141
|
|
|
* |
|
142
|
|
|
* @since 3.23.7 refactor the regex pattern to take into account that there might be css classes between textannotation |
|
143
|
|
|
* and disambiguated. |
|
144
|
|
|
* |
|
145
|
|
|
* @link https://github.com/insideout10/wordlift-plugin/issues/1001 |
|
146
|
|
|
*/ |
|
147
|
|
|
public function add_occurrences( $content ) { |
|
148
|
|
|
|
|
149
|
|
|
// Try to get all the disambiguated annotations and bail out if an error occurs. |
|
150
|
|
|
if ( false === preg_match_all( |
|
151
|
|
|
'|<span\s+id="([^"]+)"\s+class="textannotation\s+(?:\S+\s+)?disambiguated(?=[\s"])[^"]*"\s+itemid="([^"]*)">(.*?)</span>|', |
|
152
|
|
|
$content, |
|
153
|
|
|
$matches, |
|
154
|
|
|
PREG_OFFSET_CAPTURE |
|
155
|
|
|
) ) { |
|
156
|
|
|
return $this; |
|
157
|
|
|
} |
|
158
|
|
|
|
|
159
|
|
|
if ( empty( $matches ) ) { |
|
160
|
|
|
return $this; |
|
161
|
|
|
} |
|
162
|
|
|
|
|
163
|
|
|
$parse_data = array_reduce( range( 0, count( $matches[1] ) - 1 ), function ( $carry, $i ) use ( $matches ) { |
|
164
|
|
|
if ( empty( $matches[0] ) ) { |
|
165
|
|
|
return $carry; |
|
166
|
|
|
} |
|
167
|
|
|
|
|
168
|
|
|
$start = $matches[0][ $i ][1]; |
|
169
|
|
|
$end = $start + strlen( $matches[0][ $i ][0] ); |
|
170
|
|
|
$annotation_id = $matches[1][ $i ][0]; |
|
171
|
|
|
$item_id = $matches[2][ $i ][0]; |
|
172
|
|
|
$text = $matches[3][ $i ][0]; |
|
173
|
|
|
|
|
174
|
|
|
$annotation = new StdClass; |
|
175
|
|
|
$annotation->annotationId = $annotation_id; |
|
176
|
|
|
$annotation->start = $start; |
|
177
|
|
|
$annotation->end = $end; |
|
178
|
|
|
$annotation->text = $text; |
|
179
|
|
|
|
|
180
|
|
|
$entity_match = new StdClass; |
|
181
|
|
|
$entity_match->confidence = 100; |
|
182
|
|
|
$entity_match->entityId = $item_id; |
|
183
|
|
|
$annotation->entityMatches[] = $entity_match; |
|
184
|
|
|
|
|
185
|
|
|
$carry['annotations'][ $annotation_id ] = $annotation; |
|
186
|
|
|
$carry['occurrences'][ $item_id ][] = $annotation_id; |
|
187
|
|
|
|
|
188
|
|
|
return $carry; |
|
189
|
|
|
}, array( 'annotations' => array(), 'occurrences' => array(), ) ); |
|
190
|
|
|
|
|
191
|
|
|
$annotations = $parse_data['annotations']; |
|
192
|
|
|
$occurrences = $parse_data['occurrences']; |
|
193
|
|
|
|
|
194
|
|
|
foreach ( array_keys( $occurrences ) as $item_id ) { |
|
195
|
|
|
|
|
196
|
|
|
// If the entity isn't there, add it. |
|
197
|
|
|
if ( ! isset( $this->json->entities->{$item_id} ) ) { |
|
198
|
|
|
$entity = $this->get_local_entity( $item_id ); |
|
199
|
|
|
|
|
200
|
|
|
// Entity not found in the local vocabulary, continue to the next one. |
|
201
|
|
|
if ( false === $entity ) { |
|
202
|
|
|
continue; |
|
203
|
|
|
} |
|
204
|
|
|
|
|
205
|
|
|
$this->json->entities->{$item_id} = $entity; |
|
206
|
|
|
} |
|
207
|
|
|
} |
|
208
|
|
|
|
|
209
|
|
|
// Here we're adding back some data structures required by the client-side code. |
|
210
|
|
|
// |
|
211
|
|
|
// We're adding: |
|
212
|
|
|
// 1. the .entities[entity_id].occurrences array with the annotations' ids. |
|
213
|
|
|
// 2. the .entities[entity_id].annotations[annotation_id] = { id: annotation_id } map. |
|
214
|
|
|
// |
|
215
|
|
|
// Before 3.23.0 this was done by the client-side code located in src/coffee/editpost-widget/app.services.AnalysisService.coffee |
|
216
|
|
|
// function `preselect`, which was called by src/coffee/editpost-widget/app.services.EditorService.coffee in |
|
217
|
|
|
// `embedAnalysis`. |
|
218
|
|
|
foreach ( $this->json->entities as $id => $entity ) { |
|
219
|
|
|
$this->json->entities->{$id}->occurrences = isset( $occurrences[ $id ] ) ? $occurrences[ $id ] : array();; |
|
220
|
|
|
|
|
221
|
|
|
foreach ( $this->json->entities->{$id}->occurrences as $annotation_id ) { |
|
222
|
|
|
$this->json->entities->{$id}->annotations[ $annotation_id ] = array( |
|
223
|
|
|
'id' => $annotation_id, |
|
224
|
|
|
); |
|
225
|
|
|
} |
|
226
|
|
|
} |
|
227
|
|
|
|
|
228
|
|
|
// Add the missing annotations. This allows the analysis response to work also if we didn't receive results |
|
229
|
|
|
// from the analysis API. |
|
230
|
|
|
foreach ( $annotations as $annotation_id => $annotation ) { |
|
231
|
|
|
|
|
232
|
|
|
if ( ! isset( $this->json->annotations->{$annotation_id} ) ) { |
|
233
|
|
|
$this->json->annotations->{$annotation_id} = $annotation; |
|
234
|
|
|
} |
|
235
|
|
|
|
|
236
|
|
|
} |
|
237
|
|
|
|
|
238
|
|
|
return $this; |
|
239
|
|
|
} |
|
240
|
|
|
|
|
241
|
|
|
private function get_local_entity( $uri ) { |
|
242
|
|
|
|
|
243
|
|
|
$entity = $this->entity_uri_service->get_entity( $uri ); |
|
244
|
|
|
|
|
245
|
|
|
if ( null === $entity ) { |
|
246
|
|
|
return false; |
|
247
|
|
|
} |
|
248
|
|
|
|
|
249
|
|
|
$type = $this->entity_type_service->get( $entity->ID ); |
|
250
|
|
|
$images = $this->post_image_storage->get( $entity->ID ); |
|
251
|
|
|
|
|
252
|
|
|
return (object) array( |
|
253
|
|
|
'id' => $uri, |
|
254
|
|
|
'label' => $entity->post_title, |
|
255
|
|
|
'description' => $entity->post_content, |
|
256
|
|
|
'sameAs' => wl_schema_get_value( $entity->ID, 'sameAs' ), |
|
257
|
|
|
'mainType' => str_replace( 'wl-', '', $type['css_class'] ), |
|
258
|
|
|
'types' => wl_get_entity_rdf_types( $entity->ID ), |
|
259
|
|
|
'images' => $images, |
|
260
|
|
|
); |
|
261
|
|
|
} |
|
262
|
|
|
|
|
263
|
|
|
/** |
|
264
|
|
|
* Return the JSON response. |
|
265
|
|
|
* |
|
266
|
|
|
* @return mixed The JSON response. |
|
267
|
|
|
* @since 3.24.2 |
|
268
|
|
|
*/ |
|
269
|
|
|
public function get_json() { |
|
270
|
|
|
|
|
271
|
|
|
return $this->json; |
|
272
|
|
|
} |
|
273
|
|
|
|
|
274
|
|
|
/** |
|
275
|
|
|
* Get the string representation of the JSON. |
|
276
|
|
|
* |
|
277
|
|
|
* @return false|string The string representation or false in case of error. |
|
278
|
|
|
*/ |
|
279
|
|
|
public function to_string() { |
|
280
|
|
|
|
|
281
|
|
|
// Add the `JSON_UNESCAPED_UNICODE` only for PHP 5.4+. |
|
282
|
|
|
$options = ( version_compare( PHP_VERSION, '5.4', '>=' ) |
|
283
|
|
|
? 256 : 0 ); |
|
284
|
|
|
|
|
285
|
|
|
return wp_json_encode( $this->json, $options ); |
|
286
|
|
|
} |
|
287
|
|
|
|
|
288
|
|
|
/** |
|
289
|
|
|
* Create an Analysis_Response_Ops instance given the provided JSON structure. |
|
290
|
|
|
* |
|
291
|
|
|
* @param mixed $json The JSON structure. |
|
292
|
|
|
* |
|
293
|
|
|
* @return Analysis_Response_Ops A new Analysis_Response_Ops instance. |
|
294
|
|
|
*/ |
|
295
|
|
|
public static function create( $json ) { |
|
296
|
|
|
|
|
297
|
|
|
return new static( |
|
298
|
|
|
\Wordlift_Entity_Uri_Service::get_instance(), |
|
299
|
|
|
\Wordlift_Entity_Service::get_instance(), |
|
300
|
|
|
\Wordlift_Entity_Type_Service::get_instance(), |
|
301
|
|
|
\Wordlift_Storage_Factory::get_instance()->post_images(), |
|
302
|
|
|
$json ); |
|
303
|
|
|
} |
|
304
|
|
|
|
|
305
|
|
|
/** |
|
306
|
|
|
* Create an Analysis_Response_Ops instance given the provided http response. |
|
307
|
|
|
* |
|
308
|
|
|
* @param array $response { |
|
309
|
|
|
* |
|
310
|
|
|
* @type string $body The response body. |
|
311
|
|
|
* } |
|
312
|
|
|
* |
|
313
|
|
|
* @return Analysis_Response_Ops A new Analysis_Response_Ops instance. |
|
314
|
|
|
* @throws \Exception if the provided response doesn't contain a `body` element. |
|
315
|
|
|
*/ |
|
316
|
|
|
public static function create_with_response( $response ) { |
|
317
|
|
|
|
|
318
|
|
|
if ( ! isset( $response['body'] ) ) { |
|
319
|
|
|
throw new \Exception( "`body` is required in response." ); |
|
320
|
|
|
} |
|
321
|
|
|
|
|
322
|
|
|
return static::create( json_decode( $response['body'] ) ); |
|
323
|
|
|
} |
|
324
|
|
|
|
|
325
|
|
|
} |
|
326
|
|
|
|