1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
/** |
4
|
|
|
* File holding the Lingo\LingoParser class. |
5
|
|
|
* |
6
|
|
|
* This file is part of the MediaWiki extension Lingo. |
7
|
|
|
* |
8
|
|
|
* @copyright 2011 - 2018, Stephan Gambke |
9
|
|
|
* @license GPL-2.0-or-later |
10
|
|
|
* |
11
|
|
|
* The Lingo extension is free software: you can redistribute it and/or modify |
12
|
|
|
* it under the terms of the GNU General Public License as published by the Free |
13
|
|
|
* Software Foundation; either version 2 of the License, or (at your option) any |
14
|
|
|
* later version. |
15
|
|
|
* |
16
|
|
|
* The Lingo extension is distributed in the hope that it will be useful, but |
17
|
|
|
* WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
18
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more |
19
|
|
|
* details. |
20
|
|
|
* |
21
|
|
|
* You should have received a copy of the GNU General Public License along |
22
|
|
|
* with this program. If not, see <http://www.gnu.org/licenses/>. |
23
|
|
|
* |
24
|
|
|
* @author Stephan Gambke |
25
|
|
|
* |
26
|
|
|
* @file |
27
|
|
|
* @ingroup Lingo |
28
|
|
|
*/ |
29
|
|
|
namespace Lingo; |
30
|
|
|
|
31
|
|
|
use DOMDocument; |
32
|
|
|
use DOMXPath; |
33
|
|
|
use ObjectCache; |
|
|
|
|
34
|
|
|
use Parser; |
|
|
|
|
35
|
|
|
use StubObject; |
|
|
|
|
36
|
|
|
use Title; |
|
|
|
|
37
|
|
|
use Wikimedia\AtEase\AtEase; |
|
|
|
|
38
|
|
|
|
39
|
|
|
/** |
40
|
|
|
* This class parses the given text and enriches it with definitions for defined |
41
|
|
|
* terms. |
42
|
|
|
* |
43
|
|
|
* Contains a static function to initiate the parsing. |
44
|
|
|
* |
45
|
|
|
* @ingroup Lingo |
46
|
|
|
*/ |
47
|
|
|
class LingoParser { |
48
|
|
|
|
49
|
|
|
const WORD_VALUE = 0; |
50
|
|
|
const WORD_OFFSET = 1; |
51
|
|
|
|
52
|
|
|
private $mLingoTree = null; |
53
|
|
|
|
54
|
|
|
/** |
55
|
|
|
* @var Backend |
56
|
|
|
*/ |
57
|
|
|
private $mLingoBackend = null; |
58
|
|
|
private static $parserSingleton = null; |
59
|
|
|
|
60
|
|
|
// Api params passed in from ApiMakeParserOptions Hook |
61
|
|
|
private $mApiParams = null; |
62
|
|
|
|
63
|
|
|
// The RegEx to split a chunk of text into words |
64
|
|
|
public $regex = null; |
65
|
8 |
|
|
66
|
|
|
/** |
67
|
|
|
* Lingo\LingoParser constructor. |
68
|
8 |
|
* @param MessageLog|null &$messages |
69
|
8 |
|
*/ |
70
|
|
|
public function __construct( MessageLog &$messages = null ) { |
|
|
|
|
71
|
|
|
// The RegEx to split a chunk of text into words |
72
|
|
|
// Words are: placeholders for stripped items, sequences of letters and numbers, single characters that are neither letter nor number |
73
|
|
|
$this->regex = '/' . preg_quote( Parser::MARKER_PREFIX, '/' ) . '.*?' . preg_quote( Parser::MARKER_SUFFIX, '/' ) . '|[\p{L}\p{N}]+|[^\p{L}\p{N}]/u'; |
74
|
|
|
} |
75
|
|
|
|
76
|
7 |
|
/** |
77
|
7 |
|
* @param Parser $mwParser |
78
|
3 |
|
* |
79
|
|
|
* @return bool |
80
|
|
|
*/ |
81
|
7 |
|
public function parse( $mwParser ) { |
82
|
|
|
if ( $this->shouldParse( $mwParser ) ) { |
83
|
|
|
$this->realParse( $mwParser ); |
84
|
|
|
} |
85
|
|
|
|
86
|
|
|
return true; |
87
|
|
|
} |
88
|
|
|
|
89
|
|
|
/** |
90
|
|
|
* @return LingoParser |
91
|
|
|
* @since 2.0.1 |
92
|
|
|
*/ |
93
|
|
|
public static function getInstance() { |
94
|
|
|
if ( !self::$parserSingleton ) { |
95
|
|
|
self::$parserSingleton = new LingoParser(); |
96
|
|
|
|
97
|
|
|
} |
98
|
|
|
|
99
|
|
|
return self::$parserSingleton; |
100
|
|
|
} |
101
|
|
|
|
102
|
|
|
/** |
103
|
|
|
* @return string |
104
|
|
|
*/ |
105
|
|
|
private function getCacheKey() { |
106
|
|
|
// FIXME: If Lingo ever stores the glossary tree per user, then the cache key also needs to include the user id (see T163608) |
|
|
|
|
107
|
|
|
return ObjectCache::getLocalClusterInstance()->makeKey( 'ext', 'lingo', 'lingotree', Tree::TREE_VERSION, get_class( $this->getBackend() ) ); |
108
|
|
|
} |
109
|
|
|
|
110
|
|
|
/** |
111
|
|
|
* @return Backend the backend used by the parser |
112
|
|
|
* @throws \MWException |
113
|
|
|
*/ |
114
|
|
|
public function getBackend() { |
115
|
|
|
if ( $this->mLingoBackend === null ) { |
116
|
|
|
throw new \MWException( 'No Lingo backend available!' ); |
|
|
|
|
117
|
|
|
} |
118
|
|
|
|
119
|
|
|
return $this->mLingoBackend; |
120
|
|
|
} |
121
|
|
|
|
122
|
|
|
/** |
123
|
|
|
* Returns the list of terms in the glossary |
124
|
|
|
* |
125
|
|
|
* @return array an array mapping terms (keys) to descriptions (values) |
126
|
|
|
*/ |
127
|
|
|
public function getLingoArray() { |
128
|
|
|
return $this->getLingoTree()->getTermList(); |
129
|
|
|
} |
130
|
|
|
|
131
|
|
|
/** |
132
|
|
|
* Returns the list of terms in the glossary as a Lingo\Tree |
133
|
|
|
* |
134
|
|
|
* @return Tree a Lingo\Tree mapping terms (keys) to descriptions (values) |
135
|
|
|
*/ |
136
|
|
|
public function getLingoTree() { |
137
|
|
|
// build glossary array only once per request |
138
|
|
|
if ( !$this->mLingoTree ) { |
139
|
|
|
|
140
|
|
|
// use cache if enabled |
141
|
|
|
if ( $this->getBackend()->useCache() ) { |
142
|
|
|
|
143
|
|
|
// Try cache first |
144
|
|
|
global $wgexLingoCacheType; |
145
|
|
|
$cache = ( $wgexLingoCacheType !== null ) ? wfGetCache( $wgexLingoCacheType ) : wfGetMainCache(); |
|
|
|
|
146
|
|
|
$cachekey = $this->getCacheKey(); |
147
|
|
|
$cachedLingoTree = $cache->get( $cachekey ); |
148
|
|
|
|
149
|
|
|
// cache hit? |
150
|
|
|
if ( $cachedLingoTree !== false && $cachedLingoTree !== null ) { |
151
|
|
|
|
152
|
|
|
wfDebug( "Cache hit: Got lingo tree from cache.\n" ); |
|
|
|
|
153
|
|
|
$this->mLingoTree = &$cachedLingoTree; |
154
|
|
|
|
155
|
|
|
wfDebug( "Re-cached lingo tree.\n" ); |
156
|
|
|
} else { |
157
|
|
|
|
158
|
|
|
wfDebug( "Cache miss: Lingo tree not found in cache.\n" ); |
159
|
|
|
$this->mLingoTree =& $this->buildLingo(); |
160
|
|
|
wfDebug( "Cached lingo tree.\n" ); |
161
|
|
|
} |
162
|
|
|
|
163
|
|
|
// Keep for one month |
164
|
|
|
// Limiting the cache validity will allow to purge stale cache |
165
|
|
|
// entries inserted by older versions after one month |
166
|
|
|
$cache->set( $cachekey, $this->mLingoTree, 60 * 60 * 24 * 30 ); |
167
|
|
|
|
168
|
|
|
} else { |
169
|
|
|
wfDebug( "Caching of lingo tree disabled.\n" ); |
170
|
|
|
$this->mLingoTree =& $this->buildLingo(); |
171
|
|
|
} |
172
|
|
|
|
173
|
|
|
} |
174
|
|
|
|
175
|
|
|
return $this->mLingoTree; |
176
|
|
|
} |
177
|
|
|
|
178
|
|
|
/** |
179
|
|
|
* @return Tree |
180
|
|
|
*/ |
181
|
|
|
protected function &buildLingo() { |
182
|
|
|
$lingoTree = new Tree(); |
183
|
|
|
$backend = &$this->mLingoBackend; |
184
|
|
|
|
185
|
|
|
// assemble the result array |
186
|
|
|
while ( $elementData = $backend->next() ) { |
187
|
|
|
$lingoTree->addTerm( $elementData[ Element::ELEMENT_TERM ], $elementData ); |
188
|
|
|
} |
189
|
|
|
|
190
|
|
|
return $lingoTree; |
191
|
|
|
} |
192
|
|
|
|
193
|
|
|
/** |
194
|
|
|
* Parses the given text and enriches applicable terms |
195
|
|
|
* |
196
|
|
|
* This method currently only recognizes terms consisting of max one word |
197
|
3 |
|
* |
198
|
3 |
|
* @param Parser &$parser |
199
|
|
|
* |
200
|
3 |
|
* @return bool |
201
|
3 |
|
*/ |
202
|
|
|
protected function realParse( &$parser ) { |
203
|
|
|
// Parse text identical to options used in includes/api/ApiParse.php |
204
|
|
|
$params = $this->mApiParams; |
205
|
|
|
$text = is_null( $params ) ? $parser->getOutput()->getText() : $parser->getOutput()->getText( [ |
206
|
|
|
'allowTOC' => !$params['disabletoc'], |
207
|
|
|
'enableSectionEditLinks' => !$params['disableeditsection'], |
208
|
|
|
'wrapperDivClass' => $params['wrapoutputclass'], |
209
|
|
|
'deduplicateStyles' => !$params['disablestylededuplication'], |
210
|
|
|
] ); |
211
|
|
|
|
212
|
|
|
if ( $text === null || $text === '' ) { |
213
|
|
|
return true; |
214
|
|
|
} |
215
|
|
|
|
216
|
|
|
// Get array of terms |
217
|
|
|
$glossary = $this->getLingoTree(); |
218
|
|
|
|
219
|
|
|
if ( $glossary == null ) { |
|
|
|
|
220
|
|
|
return true; |
221
|
|
|
} |
222
|
|
|
|
223
|
|
|
// Parse HTML from page |
224
|
|
|
|
225
|
|
|
// TODO: Remove call to \MediaWiki\suppressWarnings() for MW 1.34+. |
|
|
|
|
226
|
|
|
// \Wikimedia\AtEase\AtEase::suppressWarnings() is available from MW 1.34. |
227
|
|
|
if ( method_exists( AtEase::class, 'suppressWarnings' ) ) { |
228
|
|
|
\Wikimedia\AtEase\AtEase::suppressWarnings(); |
229
|
|
|
} else { |
230
|
|
|
\MediaWiki\suppressWarnings(); |
|
|
|
|
231
|
|
|
} |
232
|
|
|
|
233
|
|
|
$doc = new DOMDocument( '1.0', 'utf-8' ); |
234
|
|
|
$doc->loadHTML( '<html><head><meta http-equiv="content-type" content="charset=utf-8"/></head><body>' . $text . '</body></html>' ); |
235
|
|
|
|
236
|
|
|
// TODO: Remove call to \MediaWiki\restoreWarnings() for MW 1.34+. |
|
|
|
|
237
|
|
|
// \Wikimedia\AtEase\AtEase::restoreWarnings() is available from MW 1.34. |
238
|
|
|
if ( method_exists( AtEase::class, 'suppressWarnings' ) ) { |
239
|
|
|
\Wikimedia\AtEase\AtEase::restoreWarnings(); |
240
|
|
|
} else { |
241
|
|
|
\MediaWiki\restoreWarnings(); |
|
|
|
|
242
|
|
|
} |
243
|
|
|
|
244
|
|
|
// Find all text in HTML. |
245
|
|
|
$xpath = new DOMXPath( $doc ); |
246
|
|
|
$textElements = $xpath->query( |
247
|
|
|
"//*[not(ancestor-or-self::*[@class='noglossary'] or ancestor-or-self::a)][text()!=' ']/text()" |
248
|
|
|
); |
249
|
|
|
|
250
|
|
|
// Iterate all HTML text matches |
251
|
|
|
$numberOfTextElements = $textElements->length; |
252
|
|
|
|
253
|
|
|
$definitions = []; |
254
|
|
|
|
255
|
|
|
for ( $textElementIndex = 0; $textElementIndex < $numberOfTextElements; $textElementIndex++ ) { |
256
|
|
|
$textElement = $textElements->item( $textElementIndex ); |
257
|
|
|
|
258
|
|
|
if ( strlen( $textElement->nodeValue ) < $glossary->getMinTermLength() ) { |
259
|
|
|
continue; |
260
|
|
|
} |
261
|
|
|
|
262
|
|
|
$matches = []; |
263
|
|
|
preg_match_all( |
264
|
|
|
$this->regex, |
265
|
|
|
$textElement->nodeValue, |
266
|
|
|
$matches, |
267
|
|
|
PREG_OFFSET_CAPTURE | PREG_PATTERN_ORDER |
268
|
|
|
); |
269
|
|
|
|
270
|
|
|
if ( count( $matches ) === 0 || count( $matches[ 0 ] ) === 0 ) { |
271
|
|
|
continue; |
272
|
|
|
} |
273
|
|
|
|
274
|
|
|
$wordDescriptors = &$matches[ 0 ]; // See __construct() for definition of "word" |
275
|
|
|
$numberOfWordDescriptors = count( $wordDescriptors ); |
276
|
|
|
|
277
|
|
|
$parentNode = &$textElement->parentNode; |
278
|
|
|
|
279
|
|
|
$wordDescriptorIndex = 0; |
280
|
|
|
$changedElem = false; |
281
|
|
|
|
282
|
|
|
while ( $wordDescriptorIndex < $numberOfWordDescriptors ) { |
283
|
|
|
|
284
|
|
|
/** @var \Lingo\Element $definition */ |
285
|
|
|
list( $skippedWords, $usedWords, $definition ) = |
286
|
|
|
$glossary->findNextTerm( $wordDescriptors, $wordDescriptorIndex, $numberOfWordDescriptors ); |
287
|
|
|
|
288
|
|
|
if ( $usedWords > 0 ) { // found a term |
289
|
|
|
|
290
|
|
|
if ( $skippedWords > 0 ) { // skipped some text, insert it as is |
291
|
|
|
|
292
|
|
|
$start = $wordDescriptors[ $wordDescriptorIndex ][ self::WORD_OFFSET ]; |
293
|
|
|
$length = $wordDescriptors[ $wordDescriptorIndex + $skippedWords ][ self::WORD_OFFSET ] - $start; |
294
|
|
|
|
295
|
|
|
$parentNode->insertBefore( |
296
|
|
|
$doc->createTextNode( |
297
|
|
|
substr( $textElement->nodeValue, $start, $length ) |
298
|
|
|
), |
299
|
|
|
$textElement |
300
|
|
|
); |
301
|
|
|
} |
302
|
|
|
|
303
|
|
|
$parentNode->insertBefore( $definition->getFormattedTerm( $doc ), $textElement ); |
304
|
|
|
|
305
|
|
|
$definitions[ $definition->getId() ] = $definition->getFormattedDefinitions(); |
306
|
|
|
|
307
|
|
|
$changedElem = true; |
308
|
|
|
|
309
|
|
|
} else { // did not find any term, just use the rest of the text |
310
|
|
|
|
311
|
|
|
// If we found no term now and no term before, there was no |
312
|
|
|
// term in the whole element. Might as well not change the |
313
|
|
|
// element at all. |
314
|
|
|
|
315
|
|
|
// Only change element if found term before |
316
|
|
|
if ( $changedElem === true ) { |
317
|
|
|
|
318
|
|
|
$start = $wordDescriptors[ $wordDescriptorIndex ][ self::WORD_OFFSET ]; |
319
|
|
|
|
320
|
|
|
$parentNode->insertBefore( |
321
|
|
|
$doc->createTextNode( |
322
|
|
|
substr( $textElement->nodeValue, $start ) |
323
|
|
|
), |
324
|
|
|
$textElement |
325
|
|
|
); |
326
|
|
|
|
327
|
|
|
} |
328
|
|
|
|
329
|
|
|
// In principle superfluous, the loop would run out anyway. Might save a bit of time. |
330
|
|
|
break; |
331
|
|
|
} |
332
|
|
|
|
333
|
|
|
$wordDescriptorIndex += $usedWords + $skippedWords; |
334
|
|
|
} |
335
|
|
|
|
336
|
|
|
if ( $changedElem ) { |
337
|
|
|
$parentNode->removeChild( $textElement ); |
338
|
|
|
} |
339
|
|
|
} |
340
|
|
|
|
341
|
|
|
if ( count( $definitions ) > 0 ) { |
342
|
|
|
|
343
|
|
|
$this->loadModules( $parser ); |
344
|
|
|
|
345
|
|
|
// U - Ungreedy, D - dollar matches only end of string, s - dot matches newlines |
346
|
|
|
$text = preg_replace( '%(^.*<body>)|(</body>.*$)%UDs', '', $doc->saveHTML() ); |
347
|
|
|
$text .= $parser->recursiveTagParseFully( implode( $definitions ) ); |
348
|
|
|
|
349
|
|
|
$parser->getOutput()->setText( $text ); |
350
|
|
|
} |
351
|
|
|
|
352
|
|
|
return true; |
353
|
|
|
} |
354
|
|
|
|
355
|
|
|
/** |
356
|
|
|
* @param Parser &$parser |
357
|
|
|
*/ |
358
|
|
|
protected function loadModules( &$parser ) { |
359
|
|
|
global $wgOut; |
360
|
|
|
|
361
|
|
|
$parserOutput = $parser->getOutput(); |
362
|
|
|
|
363
|
|
|
// load scripts |
364
|
|
|
$parserOutput->addModules( 'ext.Lingo' ); |
365
|
|
|
|
366
|
|
|
if ( !$wgOut->isArticle() ) { |
367
|
|
|
$wgOut->addModules( 'ext.Lingo' ); |
368
|
|
|
} |
369
|
|
|
} |
370
|
|
|
|
371
|
|
|
/** |
372
|
|
|
* Purges the lingo tree from the cache. |
373
|
|
|
* |
374
|
|
|
* @deprecated 2.0.2 |
375
|
|
|
*/ |
376
|
|
|
public static function purgeCache() { |
377
|
|
|
self::getInstance()->purgeGlossaryFromCache(); |
378
|
|
|
} |
379
|
|
|
|
380
|
|
|
/** |
381
|
|
|
* Purges the lingo tree from the cache. |
382
|
|
|
* |
383
|
|
|
* @since 2.0.2 |
384
|
|
|
*/ |
385
|
|
|
public function purgeGlossaryFromCache() { |
386
|
7 |
|
global $wgexLingoCacheType; |
387
|
7 |
|
$cache = ( $wgexLingoCacheType !== null ) ? wfGetCache( $wgexLingoCacheType ) : wfGetMainCache(); |
|
|
|
|
388
|
|
|
$cache->delete( $this->getCacheKey() ); |
389
|
7 |
|
} |
390
|
1 |
|
|
391
|
|
|
/** |
392
|
|
|
* @since 2.0.1 |
393
|
6 |
|
* @param Backend $backend |
394
|
1 |
|
*/ |
395
|
|
|
public function setBackend( Backend $backend ) { |
396
|
|
|
$this->mLingoBackend = $backend; |
397
|
5 |
|
$backend->setLingoParser( $this ); |
398
|
|
|
} |
399
|
5 |
|
|
400
|
1 |
|
/** |
401
|
|
|
* Set parser options from API |
402
|
|
|
* |
403
|
4 |
|
* @param array $params |
404
|
|
|
*/ |
405
|
4 |
|
public function setApiParams( array $params ) { |
406
|
1 |
|
$this->mApiParams = $params; |
407
|
|
|
} |
408
|
|
|
|
409
|
3 |
|
/** |
410
|
|
|
* @param Parser &$parser |
411
|
|
|
* @return bool |
412
|
|
|
*/ |
413
|
|
|
protected function shouldParse( &$parser ) { |
414
|
|
|
global $wgexLingoUseNamespaces; |
415
|
|
|
|
416
|
|
|
if ( !( $parser instanceof Parser || $parser instanceof StubObject ) ) { |
417
|
|
|
return false; |
418
|
|
|
} |
419
|
|
|
|
420
|
|
|
if ( isset( $parser->mDoubleUnderscores[ 'noglossary' ] ) ) { // __NOGLOSSARY__ found in wikitext |
421
|
|
|
return false; |
422
|
|
|
} |
423
|
|
|
|
424
|
|
|
$title = $parser->getTitle(); |
425
|
|
|
|
426
|
|
|
if ( !( $title instanceof Title ) ) { |
427
|
|
|
return false; |
428
|
|
|
} |
429
|
|
|
|
430
|
|
|
$namespace = $title->getNamespace(); |
431
|
|
|
|
432
|
|
|
if ( isset( $wgexLingoUseNamespaces[ $namespace ] ) && $wgexLingoUseNamespaces[ $namespace ] === false ) { |
433
|
|
|
return false; |
434
|
|
|
}; |
435
|
|
|
|
436
|
|
|
return true; |
437
|
|
|
} |
438
|
|
|
} |
439
|
|
|
|
The issue could also be caused by a filter entry in the build configuration. If the path has been excluded in your configuration, e.g.
excluded_paths: ["lib/*"]
, you can move it to the dependency path list as follows:For further information see https://scrutinizer-ci.com/docs/tools/php/php-scrutinizer/#list-dependency-paths