|
1
|
|
|
<?php |
|
2
|
|
|
/** |
|
3
|
|
|
* Services: Batch Analysis Service. |
|
4
|
|
|
* |
|
5
|
|
|
* The Batch Analysis service allows to queue analysis operations given a list |
|
6
|
|
|
* of URLs. |
|
7
|
|
|
* |
|
8
|
|
|
* The Batch Analysis service should also allow to queue all the posts/pages |
|
9
|
|
|
* that do not contain any annotation (example annotation: |
|
10
|
|
|
* <span id="urn:enhancement-{uuid}" |
|
11
|
|
|
* class="textannotation disambiguated wl-{class} [wl-[no-]link]" |
|
12
|
|
|
* itemid="{itemid-url}">{label}</span> |
|
13
|
|
|
* |
|
14
|
|
|
* We must identify the batch analysis status according to 3 stages: |
|
15
|
|
|
* 1. BATCH_ANALYSIS_SUBMIT, i.e. a post/page has been submitted for batch |
|
16
|
|
|
* analysis. |
|
17
|
|
|
* 2. BATCH_ANALYSIS_REQUEST, i.e. a post/page batch analysis has been |
|
18
|
|
|
* requested to the remote service. |
|
19
|
|
|
* 3. BATCH_ANALYSIS_SUCCESS / BATCH_ANALYSIS_ERROR: the outcome of the batch |
|
20
|
|
|
* analysis. |
|
21
|
|
|
* |
|
22
|
|
|
* For each state we record the date time, this is especially useful since the |
|
23
|
|
|
* remote service doesn't provide a state management, therefore we need to |
|
24
|
|
|
* define a timeout on the client side. |
|
25
|
|
|
* |
|
26
|
|
|
* Upon reception of the results we need to check whether there are some |
|
27
|
|
|
* potential warning due to interpolation issues, i.e. |
|
28
|
|
|
* |
|
29
|
|
|
* `\w<span id="urn:enhancement-` or `\s</span>` or `</span>\w` |
|
30
|
|
|
* |
|
31
|
|
|
* and in such a case, set a warning BATCH_ANALYSIS_WARNING in order to provide |
|
32
|
|
|
* a list of posts/pages that need manual review and allow the editor to clear |
|
33
|
|
|
* the warning flag. |
|
34
|
|
|
* |
|
35
|
|
|
* All the time-consuming operations must be executed asynchronously. |
|
36
|
|
|
* |
|
37
|
|
|
* Since setting the post meta for a large number of posts may be time consuming |
|
38
|
|
|
* in PHP, we can use prepared queries. |
|
39
|
|
|
* |
|
40
|
|
|
* @since 3.14.0 |
|
41
|
|
|
* @package Wordlift |
|
42
|
|
|
* @subpackage Wordlift/includes |
|
43
|
|
|
*/ |
|
44
|
|
|
|
|
45
|
|
|
/** |
|
46
|
|
|
* Define the {@link Wordlift_Batch_Analysis_Service} class. |
|
47
|
|
|
* |
|
48
|
|
|
* @since 3.14.0 |
|
49
|
|
|
* @package Wordlift |
|
50
|
|
|
* @subpackage Wordlift/includes |
|
51
|
|
|
*/ |
|
52
|
|
|
class Wordlift_Batch_Analysis_Service { |
|
53
|
|
|
|
|
54
|
|
|
/** |
|
55
|
|
|
* The list of states for the Batch Analysis: |
|
56
|
|
|
* - STATE_META_KEY: the batch analysis state meta key, |
|
57
|
|
|
* - STATE_SUBMIT: a post/page has been submitted for analysis, |
|
58
|
|
|
* - STATE_REQUEST: the plugin requested an analysis for the submitted |
|
59
|
|
|
* post/page, |
|
60
|
|
|
* - STATE_SUCCESS: the analysis has completed successfully, |
|
61
|
|
|
* - STATE_ERROR: the analysis returned an error. |
|
62
|
|
|
* |
|
63
|
|
|
* @since 3.14.2 |
|
64
|
|
|
*/ |
|
65
|
|
|
const STATE_META_KEY = '_wl_batch_analysis_state'; |
|
66
|
|
|
const STATE_SUBMIT = 0; |
|
67
|
|
|
const STATE_REQUEST = 1; |
|
68
|
|
|
// ### COMPLETE states. |
|
69
|
|
|
const STATE_SUCCESS = 2; |
|
70
|
|
|
const STATE_ERROR = 2; |
|
71
|
|
|
|
|
72
|
|
|
/** |
|
73
|
|
|
* The submit timestamp meta key. A post may have more than one timestamp. |
|
74
|
|
|
* |
|
75
|
|
|
* @since 3.14.2 |
|
76
|
|
|
*/ |
|
77
|
|
|
const SUBMIT_TIMESTAMP_META_KEY = '_wl_batch_analysis_submit_timestamp'; |
|
78
|
|
|
|
|
79
|
|
|
/** |
|
80
|
|
|
* The request timestamp meta key. A post may have more than one timestamp. |
|
81
|
|
|
* |
|
82
|
|
|
* @since 3.14.2 |
|
83
|
|
|
*/ |
|
84
|
|
|
const REQUEST_TIMESTAMP_META_KEY = '_wl_batch_analysis_request_timestamp'; |
|
85
|
|
|
|
|
86
|
|
|
/** |
|
87
|
|
|
* The complete (success or error) timestamp meta key. A post may have more |
|
88
|
|
|
* than one timestamp. |
|
89
|
|
|
* |
|
90
|
|
|
* @since 3.14.2 |
|
91
|
|
|
*/ |
|
92
|
|
|
const COMPLETE_TIMESTAMP_META_KEY = '_wl_batch_analysis_complete_timestamp'; |
|
93
|
|
|
|
|
94
|
|
|
/** |
|
95
|
|
|
* The options setting meta key. A post may have more than one setting. |
|
96
|
|
|
* |
|
97
|
|
|
* @since 3.14.2 |
|
98
|
|
|
*/ |
|
99
|
|
|
const BATCH_ANALYSIS_OPTIONS_META_KEY = '_wl_batch_analysis_options'; |
|
100
|
|
|
|
|
101
|
|
|
/** |
|
102
|
|
|
* The warning timestamp meta key. A post has only zero/one value. |
|
103
|
|
|
* |
|
104
|
|
|
* @since 3.14.2 |
|
105
|
|
|
*/ |
|
106
|
|
|
const WARNING_META_KEY = '_wl_batch_analysis_warning'; |
|
107
|
|
|
|
|
108
|
|
|
/** |
|
109
|
|
|
* Option name. |
|
110
|
|
|
* |
|
111
|
|
|
* @since 3.14.0 |
|
112
|
|
|
*/ |
|
113
|
|
|
const OPTION_NAME = 'wl_analyze_batch'; |
|
114
|
|
|
|
|
115
|
|
|
/** |
|
116
|
|
|
* Name of waiting to be processed queue array inside the option. |
|
117
|
|
|
* |
|
118
|
|
|
* @since 3.14.0 |
|
119
|
|
|
*/ |
|
120
|
|
|
const ANALYZE_QUEUE = 'queue'; |
|
121
|
|
|
|
|
122
|
|
|
/** |
|
123
|
|
|
* Name of waiting for response queue array inside the option. |
|
124
|
|
|
* |
|
125
|
|
|
* @since 3.14.0 |
|
126
|
|
|
*/ |
|
127
|
|
|
const RESPONSE_QUEUE = 'processing'; |
|
128
|
|
|
|
|
129
|
|
|
/** |
|
130
|
|
|
* Regular expressions that match interpolation errors. |
|
131
|
|
|
* |
|
132
|
|
|
* @since 3.17.0 |
|
133
|
|
|
*/ |
|
134
|
|
|
private static $interpolation_patterns = array( |
|
135
|
|
|
// Matches word before the annotation. |
|
136
|
|
|
'~(\w)<[a-z]+ id="urn:[^"]+" class="[^"]+" itemid="[^"]+">(.*?)<\/[a-z]+>~', |
|
137
|
|
|
// Matches word after the annotation. |
|
138
|
|
|
'~<[a-z]+ id="urn:[^"]+" class="[^"]+" itemid="[^"]+">(.*?)<\/[a-z]+>(\w)~', |
|
139
|
|
|
// Matches space in the beginning of annotation name. |
|
140
|
|
|
'~<[a-z]+ id="urn:[^"]+" class="[^"]+" itemid="[^"]+">(\s)(.*?)<\/[a-z]+>~', |
|
141
|
|
|
); |
|
142
|
|
|
|
|
143
|
|
|
/** |
|
144
|
|
|
* The {@link Wordlift} plugin instance. |
|
145
|
|
|
* |
|
146
|
|
|
* @since 3.14.0 |
|
147
|
|
|
* @access private |
|
148
|
|
|
* @var \Wordlift $plugin The {@link Wordlift} plugin instance. |
|
149
|
|
|
*/ |
|
150
|
|
|
private $plugin; |
|
151
|
|
|
|
|
152
|
|
|
/** |
|
153
|
|
|
* The {@link Wordlift_Configuration_Service} instance. |
|
154
|
|
|
* |
|
155
|
|
|
* @since 3.14.0 |
|
156
|
|
|
* @access private |
|
157
|
|
|
* @var \Wordlift_Configuration_Service $configuration_service The {@link Wordlift_Configuration_Service} instance. |
|
158
|
|
|
*/ |
|
159
|
|
|
private $configuration_service; |
|
160
|
|
|
|
|
161
|
|
|
/** |
|
162
|
|
|
* The {@link Wordlift_Cache_Service} instance. |
|
163
|
|
|
* |
|
164
|
|
|
* @since 3.17.0 |
|
165
|
|
|
* @access protected |
|
166
|
|
|
* @var \Wordlift_Cache_Service $cache_service The {@link Wordlift_Cache_Service} instance. |
|
167
|
|
|
*/ |
|
168
|
|
|
private $cache_service; |
|
169
|
|
|
|
|
170
|
|
|
/** |
|
171
|
|
|
* A {@link Wordlift_Log_Service} instance. |
|
172
|
|
|
* |
|
173
|
|
|
* @since 3.14.2 |
|
174
|
|
|
* @access private |
|
175
|
|
|
* @var \Wordlift_Log_Service $log A {@link Wordlift_Log_Service} instance. |
|
176
|
|
|
*/ |
|
177
|
|
|
private $log; |
|
178
|
|
|
|
|
179
|
|
|
/** |
|
180
|
|
|
* The {@link Class_Wordlift_Batch_Analys_Service} instance. |
|
181
|
|
|
* |
|
182
|
|
|
* @since 3.14.0 |
|
183
|
|
|
* |
|
184
|
|
|
* @param \Wordlift $plugin The {@link Wordlift} plugin instance. |
|
185
|
|
|
* @param \Wordlift_Configuration_Service $configuration_service The {@link Wordlift_Configuration_Service} instance. |
|
186
|
|
|
* @param \Wordlift_Cache_Service $cache_service The {@link Wordlift_Cache_Service} instance. |
|
187
|
|
|
*/ |
|
188
|
|
|
public function __construct( $plugin, $configuration_service, $cache_service ) { |
|
189
|
|
|
|
|
190
|
|
|
$this->log = Wordlift_Log_Service::get_logger( 'Wordlift_Batch_Analysis_Service' ); |
|
191
|
|
|
|
|
192
|
|
|
$this->plugin = $plugin; |
|
193
|
|
|
$this->configuration_service = $configuration_service; |
|
194
|
|
|
$this->cache_service = $cache_service; |
|
195
|
|
|
|
|
196
|
|
|
add_action( 'wl_async_wl_batch_analysis_request', array( |
|
197
|
|
|
$this, |
|
198
|
|
|
'request', |
|
199
|
|
|
) ); |
|
200
|
|
|
add_action( 'wl_async_wl_batch_analysis_complete', array( |
|
201
|
|
|
$this, |
|
202
|
|
|
'complete', |
|
203
|
|
|
) ); |
|
204
|
|
|
|
|
205
|
|
|
} |
|
206
|
|
|
|
|
207
|
|
|
/** |
|
208
|
|
|
* Submit posts for Batch Analysis. |
|
209
|
|
|
* |
|
210
|
|
|
* @since 3.14.2 |
|
211
|
|
|
* |
|
212
|
|
|
* @param array $args { |
|
213
|
|
|
* A list of options for the Batch Analysis. |
|
214
|
|
|
* |
|
215
|
|
|
* @type string $link Either `default`, `no` or `yes` (`default` is used if not specified): |
|
216
|
|
|
* * `default` doesn't set the link option - entities |
|
217
|
|
|
* will be linked if configured so in WordLift settings. |
|
218
|
|
|
* * `yes` links the entities. |
|
219
|
|
|
* * `no` doesn't link the entities. |
|
220
|
|
|
* This value is forwarded to WLS' Batch Analysis end-point. |
|
221
|
|
|
* @type int $min_occurrences The minimum number of occurrences to select |
|
222
|
|
|
* an entity. Default `1`. |
|
223
|
|
|
* @type bool $include_annotated Whether to include annotated posts in selection. |
|
224
|
|
|
* Default `false`. |
|
225
|
|
|
* @type array|int $include Explicitly include the specified {@link WP_Post}s. |
|
226
|
|
|
* @type array|int $exclude Explicitly exclude the specified {@link WP_Post}s. |
|
227
|
|
|
* @type string|null $from An optional date from filter (used in `post_date_gmt`). |
|
228
|
|
|
* @type string|null $to An optional date from filter (used in `post_date_gmt`). |
|
229
|
|
|
* @type array|string $post_type Specify the post type(s), by default only `post`. |
|
230
|
|
|
* } |
|
231
|
|
|
* |
|
232
|
|
|
* @return int The number of submitted {@link WP_Post}s or false on error. |
|
233
|
|
|
*/ |
|
234
|
|
|
public function submit( $args ) { |
|
235
|
|
|
// Parse the parameters. |
|
236
|
|
|
$params = wp_parse_args( $args, array( |
|
237
|
|
|
'links' => 'default', |
|
238
|
|
|
'min_occurrences' => 1, |
|
239
|
|
|
'include_annotated' => false, |
|
240
|
|
|
'exclude' => array(), |
|
241
|
|
|
'from' => null, |
|
242
|
|
|
'to' => null, |
|
243
|
|
|
'post_type' => 'post', |
|
244
|
|
|
) ); |
|
245
|
|
|
|
|
246
|
|
|
// Validation. |
|
247
|
|
|
if ( ! in_array( $params['links'], array( 'default', 'yes', 'no' ) ) ) { |
|
248
|
|
|
wp_die( '`link` must be one of the following: `default`, `yes` or `no`.' ); |
|
249
|
|
|
} |
|
250
|
|
|
|
|
251
|
|
|
if ( ! is_numeric( $params['min_occurrences'] ) || 1 > $params['min_occurrences'] ) { |
|
252
|
|
|
wp_die( '`min_occurrences` must greater or equal 1.' ); |
|
253
|
|
|
} |
|
254
|
|
|
|
|
255
|
|
|
// Get the sql query. |
|
256
|
|
|
$query = Wordlift_Batch_Analysis_Sql_Helper::get_sql( $params ); |
|
257
|
|
|
|
|
258
|
|
|
// Set the post metas and get the value of the posts. |
|
259
|
|
|
$submitted_posts = $this->update_posts_meta( $params, $query ); |
|
260
|
|
|
|
|
261
|
|
|
// Request Batch Analysis (the operation is handled asynchronously). |
|
262
|
|
|
do_action( 'wl_batch_analysis_request' ); |
|
263
|
|
|
|
|
264
|
|
|
// Return the count of the posts. |
|
265
|
|
|
return $submitted_posts; |
|
266
|
|
|
} |
|
267
|
|
|
|
|
268
|
|
|
/** |
|
269
|
|
|
* Submit one or more {@link WP_Posts} for Batch Analysis. |
|
270
|
|
|
* |
|
271
|
|
|
* @param array $args { |
|
272
|
|
|
* An array of arguments. |
|
273
|
|
|
* |
|
274
|
|
|
* @type string $link The link option: `default`, `yes` or |
|
275
|
|
|
* `no`. If not set `default`. |
|
276
|
|
|
* @type int $min_occurrences The minimum number of occurrences. If |
|
277
|
|
|
* not set `1`. |
|
278
|
|
|
* @type array|int $ids An array of {@link WP_Post}s' ids or one |
|
279
|
|
|
* single numeric {@link WP_Post} id. |
|
280
|
|
|
* } |
|
281
|
|
|
* |
|
282
|
|
|
* @return float|int |
|
283
|
|
|
*/ |
|
284
|
|
|
public function submit_posts( $args ) { |
|
285
|
|
|
// Parse the parameters. |
|
286
|
|
|
$params = wp_parse_args( $args, array( |
|
287
|
|
|
'links' => 'default', |
|
288
|
|
|
'min_occurrences' => 1, |
|
289
|
|
|
'ids' => array(), |
|
290
|
|
|
) ); |
|
291
|
|
|
|
|
292
|
|
|
// Validation. |
|
293
|
|
|
if ( empty( $params['ids'] ) ) { |
|
294
|
|
|
wp_die( '`ids` cannot be empty.' ); |
|
295
|
|
|
} |
|
296
|
|
|
|
|
297
|
|
|
// Get the query, |
|
298
|
|
|
$query = Wordlift_Batch_Analysis_Sql_Helper::get_sql_for_ids( $params ); |
|
299
|
|
|
|
|
300
|
|
|
// Set the post metas and get the value of the posts. |
|
301
|
|
|
$submitted_posts = $this->update_posts_meta( $params, $query ); |
|
302
|
|
|
|
|
303
|
|
|
// Request Batch Analysis (the operation is handled asynchronously). |
|
304
|
|
|
do_action( 'wl_batch_analysis_request' ); |
|
305
|
|
|
|
|
306
|
|
|
// Return the count of the posts. |
|
307
|
|
|
return $submitted_posts; |
|
308
|
|
|
} |
|
309
|
|
|
|
|
310
|
|
|
/** |
|
311
|
|
|
* Add metas to the posts that should be analysed. |
|
312
|
|
|
* |
|
313
|
|
|
* @param array $params The request params. |
|
314
|
|
|
* @param string $query The mysql query. |
|
315
|
|
|
* |
|
316
|
|
|
* @since 3.18.0 |
|
317
|
|
|
* |
|
318
|
|
|
* @return int The number of posts found/submitted. |
|
319
|
|
|
*/ |
|
320
|
|
|
public function update_posts_meta( $params, $query ) { |
|
321
|
|
|
global $wpdb; |
|
|
|
|
|
|
322
|
|
|
|
|
323
|
|
|
// Get the link options. |
|
324
|
|
|
$link_options = array( |
|
325
|
|
|
'links' => $params['links'], |
|
326
|
|
|
'min_occurrences' => $params['min_occurrences'], |
|
327
|
|
|
); |
|
328
|
|
|
|
|
329
|
|
|
// Get the posts that should be submitted for analysis. |
|
330
|
|
|
$posts = $wpdb->get_results( $query ); // WPCS: cache ok, db call ok. |
|
331
|
|
|
|
|
332
|
|
|
// Bail if there are no posts found. |
|
333
|
|
|
if ( empty( $posts ) ) { |
|
334
|
|
|
return 0; |
|
335
|
|
|
} |
|
336
|
|
|
|
|
337
|
|
|
// Add the post metas. |
|
338
|
|
|
foreach ( $posts as $p ) { |
|
339
|
|
|
add_post_meta( $p->ID, self::STATE_META_KEY, 0 ); |
|
340
|
|
|
add_post_meta( $p->ID, self::SUBMIT_TIMESTAMP_META_KEY, gmdate( 'Y-m-d H:i:s' ) ); |
|
341
|
|
|
add_post_meta( $p->ID, self::BATCH_ANALYSIS_OPTIONS_META_KEY, $link_options ); |
|
342
|
|
|
} |
|
343
|
|
|
|
|
344
|
|
|
// Finally return the posts count. |
|
345
|
|
|
return count( $posts ); |
|
346
|
|
|
} |
|
347
|
|
|
|
|
348
|
|
|
/** |
|
349
|
|
|
* Cancel the Batch Analysis request for the specified {@link WP_Post}s. |
|
350
|
|
|
* |
|
351
|
|
|
* @since 3.14.2 |
|
352
|
|
|
* |
|
353
|
|
|
* @param int|array $post_ids A single {@link WP_Post}'s id or an array of |
|
354
|
|
|
* {@link WP_Post}s' ids. |
|
355
|
|
|
* |
|
356
|
|
|
* @return false|int The number of cancelled {@link WP_Post}s or false on |
|
357
|
|
|
* error. |
|
358
|
|
|
*/ |
|
359
|
|
|
public function cancel( $post_ids ) { |
|
360
|
|
|
global $wpdb; |
|
|
|
|
|
|
361
|
|
|
|
|
362
|
|
|
return $wpdb->query( $wpdb->prepare( |
|
363
|
|
|
" |
|
364
|
|
|
DELETE FROM $wpdb->postmeta |
|
365
|
|
|
WHERE meta_key = %s |
|
366
|
|
|
AND meta_value IN ( %d, %d ) |
|
367
|
|
|
AND post_id IN( " . implode( ',', wp_parse_id_list( $post_ids ) ) . " ) |
|
368
|
|
|
", |
|
369
|
|
|
self::STATE_META_KEY, |
|
370
|
|
|
self::STATE_SUBMIT, |
|
371
|
|
|
self::STATE_REQUEST |
|
372
|
|
|
) ); // WPCS: cache ok, db call ok. |
|
373
|
|
|
|
|
374
|
|
|
} |
|
375
|
|
|
|
|
376
|
|
|
/** |
|
377
|
|
|
* Request the batch analysis for submitted posts. |
|
378
|
|
|
* |
|
379
|
|
|
* @since 3.14.2 |
|
380
|
|
|
*/ |
|
381
|
|
|
public function request() { |
|
382
|
|
|
|
|
383
|
|
|
$this->log->debug( 'Requesting analysis...' ); |
|
384
|
|
|
|
|
385
|
|
|
// By default 5 posts of any post type are returned. |
|
386
|
|
|
$posts = get_posts( array( |
|
387
|
|
|
'fields' => 'ids', |
|
388
|
|
|
'meta_key' => self::STATE_META_KEY, |
|
389
|
|
|
'meta_value' => self::STATE_SUBMIT, |
|
390
|
|
|
'orderby' => 'ID', |
|
391
|
|
|
'post_type' => 'any', |
|
392
|
|
|
) ); |
|
393
|
|
|
|
|
394
|
|
|
// Bail out if there are no submitted posts. |
|
395
|
|
|
if ( empty( $posts ) ) { |
|
396
|
|
|
$this->log->debug( 'No posts to submit found, checking for completed requests...' ); |
|
397
|
|
|
|
|
398
|
|
|
do_action( 'wl_batch_analysis_complete' ); |
|
399
|
|
|
|
|
400
|
|
|
return; |
|
401
|
|
|
} |
|
402
|
|
|
|
|
403
|
|
|
// Send a request for each post. |
|
404
|
|
|
foreach ( $posts as $id ) { |
|
405
|
|
|
|
|
406
|
|
|
$this->log->debug( "Requesting analysis for post $id..." ); |
|
407
|
|
|
|
|
408
|
|
|
// Send the actual request to the remote service. |
|
409
|
|
|
$result = $this->do_request( $id ); |
|
410
|
|
|
|
|
411
|
|
|
// Set an error if we received an error. |
|
412
|
|
|
if ( is_wp_error( $result ) ) { |
|
413
|
|
|
$this->log->error( "An error occurred while requesting a batch analysis for post $id: " . $result->get_error_message() ); |
|
414
|
|
|
|
|
415
|
|
|
$this->set_state( $id, self::STATE_ERROR ); |
|
416
|
|
|
} |
|
417
|
|
|
|
|
418
|
|
|
} |
|
419
|
|
|
|
|
420
|
|
|
// Call the `wl_batch_analysis_request` action again. This is going |
|
421
|
|
|
// to be handled by the async task. |
|
422
|
|
|
do_action( 'wl_batch_analysis_request' ); |
|
423
|
|
|
|
|
424
|
|
|
} |
|
425
|
|
|
|
|
426
|
|
|
/** |
|
427
|
|
|
* Get the results for the Batch Analysis. |
|
428
|
|
|
* |
|
429
|
|
|
* @since 3.14.2 |
|
430
|
|
|
*/ |
|
431
|
|
|
public function complete() { |
|
432
|
|
|
|
|
433
|
|
|
$this->log->debug( 'Requesting results...' ); |
|
434
|
|
|
|
|
435
|
|
|
// By default 5 posts of any post type are returned. |
|
436
|
|
|
$posts = get_posts( array( |
|
437
|
|
|
'fields' => 'ids', |
|
438
|
|
|
'meta_key' => self::STATE_META_KEY, |
|
439
|
|
|
'meta_value' => self::STATE_REQUEST, |
|
440
|
|
|
'orderby' => 'ID', |
|
441
|
|
|
'post_type' => 'any', |
|
442
|
|
|
) ); |
|
443
|
|
|
|
|
444
|
|
|
// Bail out if there are no submitted posts. |
|
445
|
|
|
if ( empty( $posts ) ) { |
|
446
|
|
|
$this->log->debug( 'No posts in request state found.' ); |
|
447
|
|
|
|
|
448
|
|
|
return; |
|
449
|
|
|
} |
|
450
|
|
|
|
|
451
|
|
|
// Send a request for each post. |
|
452
|
|
|
foreach ( $posts as $id ) { |
|
453
|
|
|
$this->log->debug( "Requesting results for post $id..." ); |
|
454
|
|
|
|
|
455
|
|
|
// Send the actual request to the remote service. |
|
456
|
|
|
$response = $this->do_complete( $id ); |
|
457
|
|
|
|
|
458
|
|
|
// Move to the next item if we don't have a reply for this one. |
|
459
|
|
|
if ( is_wp_error( $response ) || 200 !== (int) $response['response']['code'] || ! isset( $response['body'] ) ) { |
|
460
|
|
|
continue; |
|
461
|
|
|
} |
|
462
|
|
|
|
|
463
|
|
|
$this->log->debug( "Results received for post $id." ); |
|
464
|
|
|
|
|
465
|
|
|
// Save the returned content as new revision. |
|
466
|
|
|
$json = json_decode( $response['body'] ); |
|
467
|
|
|
|
|
468
|
|
|
// Continue if the content isn't set. |
|
469
|
|
|
if ( empty( $json->content ) ) { |
|
470
|
|
|
// The post content is empty, so is should be marked as completed. |
|
471
|
|
|
$this->log->error( "An error occurred while decoding the batch analysis response for post $id: {$response['body']}" ); |
|
472
|
|
|
|
|
473
|
|
|
$this->set_state( $id, self::STATE_ERROR ); |
|
474
|
|
|
continue; |
|
475
|
|
|
} |
|
476
|
|
|
|
|
477
|
|
|
// Set the warning flag if needed. |
|
478
|
|
|
$this->set_warning_based_on_content( $json->content, $id ); |
|
479
|
|
|
|
|
480
|
|
|
// Get the content, cleaned up if there are interpolation errors. |
|
481
|
|
|
$pre_content = $this->fix_interpolation_errors( $json->content, $id ); |
|
482
|
|
|
|
|
483
|
|
|
/** |
|
484
|
|
|
* Filter: 'wl_batch_analysis_update_post_content' - Allow third |
|
485
|
|
|
* parties to perform additional actions when the post content is |
|
486
|
|
|
* updated. |
|
487
|
|
|
* |
|
488
|
|
|
* @since 3.17.0 |
|
489
|
|
|
* @api string $data The {@link WP_Post}'s content. |
|
490
|
|
|
* @api int $id The {@link WP_Post}'s id. |
|
491
|
|
|
*/ |
|
492
|
|
|
$content = apply_filters( 'wl_batch_analysis_update_post_content', $pre_content, $id ); |
|
493
|
|
|
|
|
494
|
|
|
// Update the post content. |
|
495
|
|
|
wp_update_post( array( |
|
496
|
|
|
'ID' => $id, |
|
497
|
|
|
'post_content' => wp_slash( $content ), |
|
498
|
|
|
) ); |
|
499
|
|
|
|
|
500
|
|
|
// Update the status. |
|
501
|
|
|
$this->set_state( $id, self::STATE_SUCCESS ); |
|
502
|
|
|
|
|
503
|
|
|
// Invalidating the cache for the current post. |
|
504
|
|
|
$this->cache_service->delete_cache( $id ); |
|
505
|
|
|
|
|
506
|
|
|
$this->log->debug( "Post $id updated with batch analysis results." ); |
|
507
|
|
|
|
|
508
|
|
|
// Set default entity type term for posts that didn't have any. |
|
509
|
|
|
$this->maybe_set_default_term( $id ); |
|
510
|
|
|
|
|
511
|
|
|
// @todo: implement a kind of timeout that sets an error if the |
|
512
|
|
|
// results haven't been received after a long time. |
|
513
|
|
|
} |
|
514
|
|
|
|
|
515
|
|
|
// Call the `wl_batch_analysis_request` action again. This is going |
|
516
|
|
|
// to be handled by the async task. |
|
517
|
|
|
do_action( 'wl_batch_analysis_complete' ); |
|
518
|
|
|
|
|
519
|
|
|
} |
|
520
|
|
|
|
|
521
|
|
|
/** |
|
522
|
|
|
* Set a warning flag on the {@link WP_Post} if its content has suspicious |
|
523
|
|
|
* interpolations. |
|
524
|
|
|
* |
|
525
|
|
|
* @since 3.14.2 |
|
526
|
|
|
* |
|
527
|
|
|
* @param string $content The {@link WP_Post}'s content. |
|
528
|
|
|
* @param int $post_id The {@link WP_Post}'s id. |
|
529
|
|
|
*/ |
|
530
|
|
|
protected function set_warning_based_on_content( $content, $post_id ) { |
|
531
|
|
|
|
|
532
|
|
|
// Check for suspicious interpolations. |
|
533
|
|
|
$is_warning = $this->has_interpolation_errors( $content ); |
|
534
|
|
|
|
|
535
|
|
|
// Set the warning flag accordingly. |
|
536
|
|
|
$this->set_warning( $post_id, $is_warning ); |
|
537
|
|
|
|
|
538
|
|
|
} |
|
539
|
|
|
|
|
540
|
|
|
private function has_interpolation_errors( $content ) { |
|
541
|
|
|
$matches = array(); |
|
542
|
|
|
|
|
543
|
|
|
// eg: |
|
544
|
|
|
// r-pro<span id="urn:local-text-annotation-oxbgy6139gnjgk1n0oxnq9zg62py29pf" class="textannotation disambiguated wl-thing" itemid="http://data.wordlift.it/be2/entity/developing_country">ne region, has shoul</span>dere |
|
545
|
|
|
|
|
546
|
|
|
return 0 < preg_match_all( '/\w<[a-z]+ id="urn:[^"]+" class="[^"]+" itemid="[^"]+">/', $content, $matches ) |
|
547
|
|
|
|| 0 < preg_match_all( ' /<[a-z]+ id="urn:[^"]+ " class="[^"]+" itemid="[^"]+">\s/', $content, $matches ); |
|
548
|
|
|
} |
|
549
|
|
|
|
|
550
|
|
|
/** |
|
551
|
|
|
* Fix interpolation errors raised by Batch Analysis |
|
552
|
|
|
* |
|
553
|
|
|
* @param string $content The {@link WP_Post}'s content. |
|
554
|
|
|
* @param int $id The {@link WP_Post}'s id. |
|
555
|
|
|
* |
|
556
|
|
|
* @since 3.17.0 |
|
557
|
|
|
* |
|
558
|
|
|
* @return string Post content without interpolations. |
|
559
|
|
|
*/ |
|
560
|
|
|
private function fix_interpolation_errors( $content, $id ) { |
|
561
|
|
|
|
|
562
|
|
|
// Bail out if there are no interpolation errors. |
|
563
|
|
|
if ( ! $this->has_interpolation_errors( $content ) ) { |
|
564
|
|
|
$this->log->trace( "No interpolation errors found for post $id." ); |
|
565
|
|
|
|
|
566
|
|
|
return $content; |
|
567
|
|
|
} |
|
568
|
|
|
|
|
569
|
|
|
$this->log->debug( "Fixing post $id interpolations..." ); |
|
570
|
|
|
|
|
571
|
|
|
// Remove all interpolations from the content. |
|
572
|
|
|
return preg_replace( self::$interpolation_patterns, '$1$2', $content ); |
|
573
|
|
|
} |
|
574
|
|
|
|
|
575
|
|
|
/** |
|
576
|
|
|
* Clear the warning flag for the specified {@link WP_Post}s. |
|
577
|
|
|
* |
|
578
|
|
|
* @since 3.14.2 |
|
579
|
|
|
* |
|
580
|
|
|
* @param int|array $post_ids A single {@link WP_Post}'s id or an array of |
|
581
|
|
|
* {@link WP_Post}s' ids. |
|
582
|
|
|
*/ |
|
583
|
|
|
public function clear_warning( $post_ids ) { |
|
584
|
|
|
|
|
585
|
|
|
foreach ( (array) $post_ids as $post_id ) { |
|
586
|
|
|
delete_post_meta( $post_id, self::WARNING_META_KEY ); |
|
587
|
|
|
} |
|
588
|
|
|
|
|
589
|
|
|
} |
|
590
|
|
|
|
|
591
|
|
|
/** |
|
592
|
|
|
* Set the warning flag for the specified {@link WP_Post}. |
|
593
|
|
|
* |
|
594
|
|
|
* @since 3.14.2 |
|
595
|
|
|
* |
|
596
|
|
|
* @param int $post_id The {@link WP_Post}'s id. |
|
597
|
|
|
* @param bool $value The flag's value. |
|
598
|
|
|
* |
|
599
|
|
|
* @return int|bool Meta ID if the key didn't exist, true on successful update, |
|
600
|
|
|
* false on failure. |
|
601
|
|
|
*/ |
|
602
|
|
|
private function set_warning( $post_id, $value ) { |
|
603
|
|
|
|
|
604
|
|
|
return update_post_meta( $post_id, self::WARNING_META_KEY, ( true === $value ? 'yes' : 'no' ) ); |
|
605
|
|
|
} |
|
606
|
|
|
|
|
607
|
|
|
// /** |
|
608
|
|
|
// * Get the post/page batch analysis state. |
|
609
|
|
|
// * |
|
610
|
|
|
// * @since 3.14.2 |
|
611
|
|
|
// * |
|
612
|
|
|
// * @param int $post_id The {@link WP_Post}'s id. |
|
613
|
|
|
// * |
|
614
|
|
|
// * @return int|string The post state or an empty string if not set. |
|
615
|
|
|
// */ |
|
616
|
|
|
// public function get_state( $post_id ) { |
|
617
|
|
|
// |
|
618
|
|
|
// return get_post_meta( $post_id, self::STATE_META_KEY, true ); |
|
619
|
|
|
// } |
|
620
|
|
|
|
|
621
|
|
|
/** |
|
622
|
|
|
* Set the post/page batch analysis state. |
|
623
|
|
|
* |
|
624
|
|
|
* @since 3.14.2 |
|
625
|
|
|
* |
|
626
|
|
|
* @param int $post_id The {@link WP_Post}'s id. |
|
627
|
|
|
* @param int $value The new state. |
|
628
|
|
|
* |
|
629
|
|
|
* @return int|bool Meta ID if the key didn't exist, true on successful update, |
|
630
|
|
|
* false on failure. |
|
631
|
|
|
*/ |
|
632
|
|
|
private function set_state( $post_id, $value ) { |
|
633
|
|
|
|
|
634
|
|
|
// Update the state. |
|
635
|
|
|
$result = update_post_meta( $post_id, self::STATE_META_KEY, $value ); |
|
636
|
|
|
|
|
637
|
|
|
// Update timestamps as required. |
|
638
|
|
|
switch ( $value ) { |
|
639
|
|
|
|
|
640
|
|
|
// ### REQUEST state. |
|
641
|
|
|
case self::STATE_REQUEST: |
|
642
|
|
|
add_post_meta( $post_id, self::REQUEST_TIMESTAMP_META_KEY, current_time( 'mysql', true ) ); |
|
643
|
|
|
break; |
|
644
|
|
|
|
|
645
|
|
|
// ### SUCCESS/ERROR state. |
|
646
|
|
|
case self::STATE_SUCCESS: |
|
647
|
|
|
case self::STATE_ERROR: |
|
648
|
|
|
add_post_meta( $post_id, self::COMPLETE_TIMESTAMP_META_KEY, current_time( 'mysql', true ) ); |
|
649
|
|
|
break; |
|
650
|
|
|
} |
|
651
|
|
|
|
|
652
|
|
|
// Finally return the result. |
|
653
|
|
|
return $result; |
|
654
|
|
|
} |
|
655
|
|
|
|
|
656
|
|
|
/** |
|
657
|
|
|
* Get the options setting for a {@link WP_Post}. |
|
658
|
|
|
* |
|
659
|
|
|
* If there are multiple link settings, only the last one is returned. |
|
660
|
|
|
* |
|
661
|
|
|
* @since 3.14.2 |
|
662
|
|
|
* |
|
663
|
|
|
* @param int $post_id The {@link WP_Post}'s id. |
|
664
|
|
|
* |
|
665
|
|
|
* @return array The link settings. |
|
666
|
|
|
*/ |
|
667
|
|
|
private function get_options( $post_id ) { |
|
668
|
|
|
|
|
669
|
|
|
$values = get_post_meta( $post_id, self::BATCH_ANALYSIS_OPTIONS_META_KEY ); |
|
670
|
|
|
|
|
671
|
|
|
return end( $values ) ?: array( |
|
672
|
|
|
'links' => 'default', |
|
673
|
|
|
'min_occurrences' => 1, |
|
674
|
|
|
); |
|
675
|
|
|
} |
|
676
|
|
|
|
|
677
|
|
|
/** |
|
678
|
|
|
* Get the array of post IDS waiting in the queue to start processing. |
|
679
|
|
|
* |
|
680
|
|
|
* @since 3.14.0 |
|
681
|
|
|
* |
|
682
|
|
|
* @return array The waiting to be processed post ids queue. |
|
683
|
|
|
*/ |
|
684
|
|
View Code Duplication |
public function waiting_for_analysis() { |
|
|
|
|
|
|
685
|
|
|
|
|
686
|
|
|
return get_posts( array( |
|
687
|
|
|
'posts_per_page' => - 1, |
|
688
|
|
|
'fields' => 'ids', |
|
689
|
|
|
'post_status' => 'any', |
|
690
|
|
|
'meta_key' => self::STATE_META_KEY, |
|
691
|
|
|
'meta_value' => self::STATE_SUBMIT, |
|
692
|
|
|
'orderby' => 'ID', |
|
693
|
|
|
'post_type' => 'any', |
|
694
|
|
|
// Add any because posts from multiple posts types may be waiting. |
|
695
|
|
|
) ); |
|
696
|
|
|
} |
|
697
|
|
|
|
|
698
|
|
|
/** |
|
699
|
|
|
* Get the array of post IDS waiting for response. |
|
700
|
|
|
* |
|
701
|
|
|
* @deprecated |
|
702
|
|
|
* @since 3.14.0 |
|
703
|
|
|
* |
|
704
|
|
|
* @return array The waiting for response post ids queue. |
|
705
|
|
|
*/ |
|
706
|
|
View Code Duplication |
public function waiting_for_response() { |
|
|
|
|
|
|
707
|
|
|
|
|
708
|
|
|
return get_posts( array( |
|
709
|
|
|
'posts_per_page' => - 1, |
|
710
|
|
|
'fields' => 'ids', |
|
711
|
|
|
'post_status' => 'any', |
|
712
|
|
|
'meta_key' => self::STATE_META_KEY, |
|
713
|
|
|
'meta_value' => self::STATE_REQUEST, |
|
714
|
|
|
'orderby' => 'ID', |
|
715
|
|
|
'post_type' => 'any', |
|
716
|
|
|
// Add any because posts from multiple posts types may be waiting. |
|
717
|
|
|
) ); |
|
718
|
|
|
} |
|
719
|
|
|
|
|
720
|
|
|
/** |
|
721
|
|
|
* Request the analysis for the specified {@link WP_Post}. |
|
722
|
|
|
* |
|
723
|
|
|
* @since 3.14.2 |
|
724
|
|
|
* |
|
725
|
|
|
* @param int $post_id The {@link WP_Post}'s id. |
|
726
|
|
|
* |
|
727
|
|
|
* @return WP_Error|array The response or WP_Error on failure. |
|
728
|
|
|
*/ |
|
729
|
|
|
private function do_request( $post_id ) { |
|
730
|
|
|
|
|
731
|
|
|
// Change the state to `REQUEST`. |
|
732
|
|
|
$this->set_state( $post_id, self::STATE_REQUEST ); |
|
733
|
|
|
|
|
734
|
|
|
// Get the post. |
|
735
|
|
|
$post = get_post( $post_id ); |
|
736
|
|
|
|
|
737
|
|
|
// Bail out if the post isn't found. |
|
738
|
|
|
if ( null === $post ) { |
|
739
|
|
|
$this->log->warn( "Post $post_id not found." ); |
|
740
|
|
|
|
|
741
|
|
|
return new WP_Error( 0, "Cannot find post $post_id." ); |
|
742
|
|
|
} |
|
743
|
|
|
|
|
744
|
|
|
// Get the link setting. |
|
745
|
|
|
$options = $this->get_options( $post_id ); |
|
746
|
|
|
|
|
747
|
|
|
$this->log->debug( 'Sending analysis request for post $post_id [ links :: ' . $options['links'] . ', min_occurrences :: ' . $options['min_occurrences'] . ' ] ...' ); |
|
748
|
|
|
|
|
749
|
|
|
// Get the batch analysis URL. |
|
750
|
|
|
$url = $this->configuration_service->get_batch_analysis_url(); |
|
751
|
|
|
|
|
752
|
|
|
// Prepare the POST parameters. |
|
753
|
|
|
$params = array( |
|
754
|
|
|
'id' => $post->ID, |
|
755
|
|
|
'key' => $this->configuration_service->get_key(), |
|
756
|
|
|
'content' => $post->post_content, |
|
757
|
|
|
'contentLanguage' => $this->configuration_service->get_language_code(), |
|
758
|
|
|
'version' => $this->plugin->get_version(), |
|
759
|
|
|
'scope' => 'local', |
|
760
|
|
|
'links' => $options['links'], |
|
761
|
|
|
'minOccurrences' => $options['min_occurrences'], |
|
762
|
|
|
); |
|
763
|
|
|
|
|
764
|
|
|
// Get the HTTP options. |
|
765
|
|
|
$args = array_merge_recursive( unserialize( WL_REDLINK_API_HTTP_OPTIONS ), array( |
|
766
|
|
|
'method' => 'POST', |
|
767
|
|
|
'headers' => array( |
|
768
|
|
|
'Accept' => 'application/json', |
|
769
|
|
|
'Content-type' => 'application/json; charset=UTF-8', |
|
770
|
|
|
), |
|
771
|
|
|
// we need to downgrade the HTTP version in this case since chunked encoding is dumping numbers in the response. |
|
772
|
|
|
'httpversion' => '1.0', |
|
773
|
|
|
'body' => wp_json_encode( $params ), |
|
774
|
|
|
) ); |
|
775
|
|
|
|
|
776
|
|
|
$this->log->debug( "Posting analysis request for post $post_id to $url..." ); |
|
777
|
|
|
|
|
778
|
|
|
// Post the parameter. |
|
779
|
|
|
return wp_remote_post( $url, $args ); |
|
780
|
|
|
} |
|
781
|
|
|
|
|
782
|
|
|
/** |
|
783
|
|
|
* Get the Batch Analysis results for the specified {@link WP_Post}. |
|
784
|
|
|
* |
|
785
|
|
|
* @since 3.14.2 |
|
786
|
|
|
* |
|
787
|
|
|
* @param int $post_id The {@link WP_Post}'s id. |
|
788
|
|
|
* |
|
789
|
|
|
* @return WP_Error|array The response or WP_Error on failure. |
|
790
|
|
|
*/ |
|
791
|
|
|
private function do_complete( $post_id ) { |
|
792
|
|
|
|
|
793
|
|
|
$post = get_post( $post_id ); |
|
794
|
|
|
|
|
795
|
|
|
if ( null === $post ) { |
|
796
|
|
|
// Post was possibly deleted, just bailout. |
|
797
|
|
|
return new WP_Error( 0, "Post $post_id not found . " ); |
|
798
|
|
|
} |
|
799
|
|
|
|
|
800
|
|
|
$url = $this->configuration_service->get_batch_analysis_url(); |
|
801
|
|
|
$key = $this->configuration_service->get_key(); |
|
802
|
|
|
$url = $url . '/' . $post->ID . '?key=' . $key; |
|
803
|
|
|
|
|
804
|
|
|
return wp_remote_get( $url, unserialize( WL_REDLINK_API_HTTP_OPTIONS ) ); |
|
805
|
|
|
} |
|
806
|
|
|
|
|
807
|
|
|
/** |
|
808
|
|
|
* Get the {@link WP_Post}s' ids flagged with warnings. |
|
809
|
|
|
* |
|
810
|
|
|
* @since 3.14.2 |
|
811
|
|
|
* |
|
812
|
|
|
* @return array An array of {@link WP_Post}s' ids. |
|
813
|
|
|
*/ |
|
814
|
|
View Code Duplication |
public function get_warnings() { |
|
|
|
|
|
|
815
|
|
|
|
|
816
|
|
|
return get_posts( array( |
|
817
|
|
|
'fields' => 'ids', |
|
818
|
|
|
'numberposts' => - 1, |
|
819
|
|
|
'post_status' => 'any', |
|
820
|
|
|
'meta_key' => self::WARNING_META_KEY, |
|
821
|
|
|
'meta_value' => 'yes', |
|
822
|
|
|
'post_type' => 'any', |
|
823
|
|
|
// Add any because posts from multiple posts types may be waiting. |
|
824
|
|
|
) ); |
|
825
|
|
|
} |
|
826
|
|
|
|
|
827
|
|
|
/** |
|
828
|
|
|
* Check whether the term has entity type associated and set default term if it hasn't. |
|
829
|
|
|
* |
|
830
|
|
|
* @since 3.17.0 |
|
831
|
|
|
* |
|
832
|
|
|
* @param int $id The {@link WP_Post}'s id. |
|
833
|
|
|
*/ |
|
834
|
|
|
private function maybe_set_default_term( $id ) { |
|
835
|
|
|
// Check whether the post has any of the WordLift entity types. |
|
836
|
|
|
$has_term = has_term( '', Wordlift_Entity_Type_Taxonomy_Service::TAXONOMY_NAME, $id ); |
|
837
|
|
|
|
|
838
|
|
|
// Bail if the term is associated with entity types already. |
|
839
|
|
|
if ( ! empty( $has_term ) ) { |
|
840
|
|
|
return; |
|
841
|
|
|
} |
|
842
|
|
|
|
|
843
|
|
|
// Set the default `article` term. |
|
844
|
|
|
wp_set_object_terms( $id, 'article', Wordlift_Entity_Type_Taxonomy_Service::TAXONOMY_NAME ); |
|
845
|
|
|
|
|
846
|
|
|
} |
|
847
|
|
|
|
|
848
|
|
|
} |
|
849
|
|
|
|
Instead of relying on
globalstate, we recommend one of these alternatives:1. Pass all data via parameters
2. Create a class that maintains your state