Added ApiResult::NO_SIZE_CHECK flag for addValue()
[lhc/web/wiklou.git] / includes / api / ApiQuery.php
1 <?php
2 /**
3 *
4 *
5 * Created on Sep 7, 2006
6 *
7 * Copyright © 2006 Yuri Astrakhan "<Firstname><Lastname>@gmail.com"
8 *
9 * This program is free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License as published by
11 * the Free Software Foundation; either version 2 of the License, or
12 * (at your option) any later version.
13 *
14 * This program is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 * GNU General Public License for more details.
18 *
19 * You should have received a copy of the GNU General Public License along
20 * with this program; if not, write to the Free Software Foundation, Inc.,
21 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
22 * http://www.gnu.org/copyleft/gpl.html
23 *
24 * @file
25 */
26
27 /**
28 * This is the main query class. It behaves similar to ApiMain: based on the
29 * parameters given, it will create a list of titles to work on (an ApiPageSet
30 * object), instantiate and execute various property/list/meta modules, and
31 * assemble all resulting data into a single ApiResult object.
32 *
33 * In generator mode, a generator will be executed first to populate a second
34 * ApiPageSet object, and that object will be used for all subsequent modules.
35 *
36 * @ingroup API
37 */
38 class ApiQuery extends ApiBase {
39
40 /**
41 * List of Api Query prop modules
42 * @var array
43 */
44 private static $QueryPropModules = array(
45 'categories' => 'ApiQueryCategories',
46 'categoryinfo' => 'ApiQueryCategoryInfo',
47 'contributors' => 'ApiQueryContributors',
48 'duplicatefiles' => 'ApiQueryDuplicateFiles',
49 'extlinks' => 'ApiQueryExternalLinks',
50 'images' => 'ApiQueryImages',
51 'imageinfo' => 'ApiQueryImageInfo',
52 'info' => 'ApiQueryInfo',
53 'links' => 'ApiQueryLinks',
54 'iwlinks' => 'ApiQueryIWLinks',
55 'langlinks' => 'ApiQueryLangLinks',
56 'pageprops' => 'ApiQueryPageProps',
57 'redirects' => 'ApiQueryRedirects',
58 'revisions' => 'ApiQueryRevisions',
59 'stashimageinfo' => 'ApiQueryStashImageInfo',
60 'templates' => 'ApiQueryLinks',
61 );
62
63 /**
64 * List of Api Query list modules
65 * @var array
66 */
67 private static $QueryListModules = array(
68 'allcategories' => 'ApiQueryAllCategories',
69 'allfileusages' => 'ApiQueryAllLinks',
70 'allimages' => 'ApiQueryAllImages',
71 'alllinks' => 'ApiQueryAllLinks',
72 'allpages' => 'ApiQueryAllPages',
73 'allredirects' => 'ApiQueryAllLinks',
74 'alltransclusions' => 'ApiQueryAllLinks',
75 'allusers' => 'ApiQueryAllUsers',
76 'backlinks' => 'ApiQueryBacklinks',
77 'blocks' => 'ApiQueryBlocks',
78 'categorymembers' => 'ApiQueryCategoryMembers',
79 'deletedrevs' => 'ApiQueryDeletedrevs',
80 'embeddedin' => 'ApiQueryBacklinks',
81 'exturlusage' => 'ApiQueryExtLinksUsage',
82 'filearchive' => 'ApiQueryFilearchive',
83 'imageusage' => 'ApiQueryBacklinks',
84 'iwbacklinks' => 'ApiQueryIWBacklinks',
85 'langbacklinks' => 'ApiQueryLangBacklinks',
86 'logevents' => 'ApiQueryLogEvents',
87 'pageswithprop' => 'ApiQueryPagesWithProp',
88 'pagepropnames' => 'ApiQueryPagePropNames',
89 'prefixsearch' => 'ApiQueryPrefixSearch',
90 'protectedtitles' => 'ApiQueryProtectedTitles',
91 'querypage' => 'ApiQueryQueryPage',
92 'random' => 'ApiQueryRandom',
93 'recentchanges' => 'ApiQueryRecentChanges',
94 'search' => 'ApiQuerySearch',
95 'tags' => 'ApiQueryTags',
96 'usercontribs' => 'ApiQueryContributions',
97 'users' => 'ApiQueryUsers',
98 'watchlist' => 'ApiQueryWatchlist',
99 'watchlistraw' => 'ApiQueryWatchlistRaw',
100 );
101
102 /**
103 * List of Api Query meta modules
104 * @var array
105 */
106 private static $QueryMetaModules = array(
107 'allmessages' => 'ApiQueryAllMessages',
108 'siteinfo' => 'ApiQuerySiteinfo',
109 'userinfo' => 'ApiQueryUserInfo',
110 'filerepoinfo' => 'ApiQueryFileRepoInfo',
111 );
112
113 /**
114 * @var ApiPageSet
115 */
116 private $mPageSet;
117
118 private $mParams;
119 private $mNamedDB = array();
120 private $mModuleMgr;
121
122 /**
123 * @param ApiMain $main
124 * @param string $action
125 */
126 public function __construct( ApiMain $main, $action ) {
127 parent::__construct( $main, $action );
128
129 $this->mModuleMgr = new ApiModuleManager( $this );
130
131 // Allow custom modules to be added in LocalSettings.php
132 $config = $this->getConfig();
133 $this->mModuleMgr->addModules( self::$QueryPropModules, 'prop' );
134 $this->mModuleMgr->addModules( $config->get( 'APIPropModules' ), 'prop' );
135 $this->mModuleMgr->addModules( self::$QueryListModules, 'list' );
136 $this->mModuleMgr->addModules( $config->get( 'APIListModules' ), 'list' );
137 $this->mModuleMgr->addModules( self::$QueryMetaModules, 'meta' );
138 $this->mModuleMgr->addModules( $config->get( 'APIMetaModules' ), 'meta' );
139
140 // Create PageSet that will process titles/pageids/revids/generator
141 $this->mPageSet = new ApiPageSet( $this );
142 }
143
144 /**
145 * Overrides to return this instance's module manager.
146 * @return ApiModuleManager
147 */
148 public function getModuleManager() {
149 return $this->mModuleMgr;
150 }
151
152 /**
153 * Get the query database connection with the given name.
154 * If no such connection has been requested before, it will be created.
155 * Subsequent calls with the same $name will return the same connection
156 * as the first, regardless of the values of $db and $groups
157 * @param string $name Name to assign to the database connection
158 * @param int $db One of the DB_* constants
159 * @param array $groups Query groups
160 * @return DatabaseBase
161 */
162 public function getNamedDB( $name, $db, $groups ) {
163 if ( !array_key_exists( $name, $this->mNamedDB ) ) {
164 $this->profileDBIn();
165 $this->mNamedDB[$name] = wfGetDB( $db, $groups );
166 $this->profileDBOut();
167 }
168
169 return $this->mNamedDB[$name];
170 }
171
172 /**
173 * Gets the set of pages the user has requested (or generated)
174 * @return ApiPageSet
175 */
176 public function getPageSet() {
177 return $this->mPageSet;
178 }
179
180 /**
181 * Get the array mapping module names to class names
182 * @deprecated since 1.21, use getModuleManager()'s methods instead
183 * @return array array(modulename => classname)
184 */
185 public function getModules() {
186 wfDeprecated( __METHOD__, '1.21' );
187
188 return $this->getModuleManager()->getNamesWithClasses();
189 }
190
191 /**
192 * Get the generators array mapping module names to class names
193 * @deprecated since 1.21, list of generators is maintained by ApiPageSet
194 * @return array array(modulename => classname)
195 */
196 public function getGenerators() {
197 wfDeprecated( __METHOD__, '1.21' );
198 $gens = array();
199 foreach ( $this->mModuleMgr->getNamesWithClasses() as $name => $class ) {
200 if ( is_subclass_of( $class, 'ApiQueryGeneratorBase' ) ) {
201 $gens[$name] = $class;
202 }
203 }
204
205 return $gens;
206 }
207
208 /**
209 * Get whether the specified module is a prop, list or a meta query module
210 * @deprecated since 1.21, use getModuleManager()->getModuleGroup()
211 * @param string $moduleName Name of the module to find type for
212 * @return string|null
213 */
214 function getModuleType( $moduleName ) {
215 return $this->getModuleManager()->getModuleGroup( $moduleName );
216 }
217
218 /**
219 * @return ApiFormatRaw|null
220 */
221 public function getCustomPrinter() {
222 // If &exportnowrap is set, use the raw formatter
223 if ( $this->getParameter( 'export' ) &&
224 $this->getParameter( 'exportnowrap' )
225 ) {
226 return new ApiFormatRaw( $this->getMain(),
227 $this->getMain()->createPrinterByName( 'xml' ) );
228 } else {
229 return null;
230 }
231 }
232
233 /**
234 * Query execution happens in the following steps:
235 * #1 Create a PageSet object with any pages requested by the user
236 * #2 If using a generator, execute it to get a new ApiPageSet object
237 * #3 Instantiate all requested modules.
238 * This way the PageSet object will know what shared data is required,
239 * and minimize DB calls.
240 * #4 Output all normalization and redirect resolution information
241 * #5 Execute all requested modules
242 */
243 public function execute() {
244 $this->mParams = $this->extractRequestParams();
245
246 // Instantiate requested modules
247 $allModules = array();
248 $this->instantiateModules( $allModules, 'prop' );
249 $propModules = array_keys( $allModules );
250 $this->instantiateModules( $allModules, 'list' );
251 $this->instantiateModules( $allModules, 'meta' );
252
253 // Filter modules based on continue parameter
254 list( $generatorDone, $modules ) = $this->getResult()->beginContinuation(
255 $this->mParams['continue'], $allModules, $propModules
256 );
257
258 if ( !$generatorDone ) {
259 // Query modules may optimize data requests through the $this->getPageSet()
260 // object by adding extra fields from the page table.
261 foreach ( $modules as $module ) {
262 $module->requestExtraData( $this->mPageSet );
263 }
264 // Populate page/revision information
265 $this->mPageSet->execute();
266 // Record page information (title, namespace, if exists, etc)
267 $this->outputGeneralPageInfo();
268 } else {
269 $this->mPageSet->executeDryRun();
270 }
271
272 $cacheMode = $this->mPageSet->getCacheMode();
273
274 // Execute all unfinished modules
275 /** @var $module ApiQueryBase */
276 foreach ( $modules as $module ) {
277 $params = $module->extractRequestParams();
278 $cacheMode = $this->mergeCacheMode(
279 $cacheMode, $module->getCacheMode( $params ) );
280 $module->profileIn();
281 $module->execute();
282 wfRunHooks( 'APIQueryAfterExecute', array( &$module ) );
283 $module->profileOut();
284 }
285
286 // Set the cache mode
287 $this->getMain()->setCacheMode( $cacheMode );
288
289 // Write the continuation data into the result
290 $this->getResult()->endContinuation(
291 $this->mParams['continue'] === null ? 'raw' : 'standard'
292 );
293 }
294
295 /**
296 * Update a cache mode string, applying the cache mode of a new module to it.
297 * The cache mode may increase in the level of privacy, but public modules
298 * added to private data do not decrease the level of privacy.
299 *
300 * @param string $cacheMode
301 * @param string $modCacheMode
302 * @return string
303 */
304 protected function mergeCacheMode( $cacheMode, $modCacheMode ) {
305 if ( $modCacheMode === 'anon-public-user-private' ) {
306 if ( $cacheMode !== 'private' ) {
307 $cacheMode = 'anon-public-user-private';
308 }
309 } elseif ( $modCacheMode === 'public' ) {
310 // do nothing, if it's public already it will stay public
311 } else { // private
312 $cacheMode = 'private';
313 }
314
315 return $cacheMode;
316 }
317
318 /**
319 * Create instances of all modules requested by the client
320 * @param array $modules To append instantiated modules to
321 * @param string $param Parameter name to read modules from
322 */
323 private function instantiateModules( &$modules, $param ) {
324 $wasPosted = $this->getRequest()->wasPosted();
325 if ( isset( $this->mParams[$param] ) ) {
326 foreach ( $this->mParams[$param] as $moduleName ) {
327 $instance = $this->mModuleMgr->getModule( $moduleName, $param );
328 if ( $instance === null ) {
329 ApiBase::dieDebug( __METHOD__, 'Error instantiating module' );
330 }
331 if ( !$wasPosted && $instance->mustBePosted() ) {
332 $this->dieUsageMsgOrDebug( array( 'mustbeposted', $moduleName ) );
333 }
334 // Ignore duplicates. TODO 2.0: die()?
335 if ( !array_key_exists( $moduleName, $modules ) ) {
336 $modules[$moduleName] = $instance;
337 }
338 }
339 }
340 }
341
342 /**
343 * Appends an element for each page in the current pageSet with the
344 * most general information (id, title), plus any title normalizations
345 * and missing or invalid title/pageids/revids.
346 */
347 private function outputGeneralPageInfo() {
348 $pageSet = $this->getPageSet();
349 $result = $this->getResult();
350
351 // We don't check for a full result set here because we can't be adding
352 // more than 380K. The maximum revision size is in the megabyte range,
353 // and the maximum result size must be even higher than that.
354
355 $values = $pageSet->getNormalizedTitlesAsResult( $result );
356 if ( $values ) {
357 $result->addValue( 'query', 'normalized', $values );
358 }
359 $values = $pageSet->getConvertedTitlesAsResult( $result );
360 if ( $values ) {
361 $result->addValue( 'query', 'converted', $values );
362 }
363 $values = $pageSet->getInterwikiTitlesAsResult( $result, $this->mParams['iwurl'] );
364 if ( $values ) {
365 $result->addValue( 'query', 'interwiki', $values );
366 }
367 $values = $pageSet->getRedirectTitlesAsResult( $result );
368 if ( $values ) {
369 $result->addValue( 'query', 'redirects', $values );
370 }
371 $values = $pageSet->getMissingRevisionIDsAsResult( $result );
372 if ( $values ) {
373 $result->addValue( 'query', 'badrevids', $values );
374 }
375
376 // Page elements
377 $pages = array();
378
379 // Report any missing titles
380 foreach ( $pageSet->getMissingTitles() as $fakeId => $title ) {
381 $vals = array();
382 ApiQueryBase::addTitleInfo( $vals, $title );
383 $vals['missing'] = '';
384 $pages[$fakeId] = $vals;
385 }
386 // Report any invalid titles
387 foreach ( $pageSet->getInvalidTitles() as $fakeId => $title ) {
388 $pages[$fakeId] = array( 'title' => $title, 'invalid' => '' );
389 }
390 // Report any missing page ids
391 foreach ( $pageSet->getMissingPageIDs() as $pageid ) {
392 $pages[$pageid] = array(
393 'pageid' => $pageid,
394 'missing' => ''
395 );
396 }
397 // Report special pages
398 /** @var $title Title */
399 foreach ( $pageSet->getSpecialTitles() as $fakeId => $title ) {
400 $vals = array();
401 ApiQueryBase::addTitleInfo( $vals, $title );
402 $vals['special'] = '';
403 if ( $title->isSpecialPage() &&
404 !SpecialPageFactory::exists( $title->getDBkey() )
405 ) {
406 $vals['missing'] = '';
407 } elseif ( $title->getNamespace() == NS_MEDIA &&
408 !wfFindFile( $title )
409 ) {
410 $vals['missing'] = '';
411 }
412 $pages[$fakeId] = $vals;
413 }
414
415 // Output general page information for found titles
416 foreach ( $pageSet->getGoodTitles() as $pageid => $title ) {
417 $vals = array();
418 $vals['pageid'] = $pageid;
419 ApiQueryBase::addTitleInfo( $vals, $title );
420 $pages[$pageid] = $vals;
421 }
422
423 if ( count( $pages ) ) {
424 if ( $this->mParams['indexpageids'] ) {
425 $pageIDs = array_keys( $pages );
426 // json treats all map keys as strings - converting to match
427 $pageIDs = array_map( 'strval', $pageIDs );
428 $result->setIndexedTagName( $pageIDs, 'id' );
429 $result->addValue( 'query', 'pageids', $pageIDs );
430 }
431
432 $result->setIndexedTagName( $pages, 'page' );
433 $result->addValue( 'query', 'pages', $pages );
434 }
435 if ( $this->mParams['export'] ) {
436 $this->doExport( $pageSet, $result );
437 }
438 }
439
440 /**
441 * This method is called by the generator base when generator in the smart-continue
442 * mode tries to set 'query-continue' value. ApiQuery stores those values separately
443 * until the post-processing when it is known if the generation should continue or repeat.
444 * @deprecated @since 1.24
445 * @param ApiQueryGeneratorBase $module Generator module
446 * @param string $paramName
447 * @param mixed $paramValue
448 * @return bool True if processed, false if this is a legacy continue
449 */
450 public function setGeneratorContinue( $module, $paramName, $paramValue ) {
451 wfDeprecated( __METHOD__, '1.24' );
452 $this->getResult()->setGeneratorContinueParam( $module, $paramName, $paramValue );
453 return $this->getParameter( 'continue' ) !== null;
454 }
455
456 /**
457 * @param ApiPageSet $pageSet Pages to be exported
458 * @param ApiResult $result Result to output to
459 */
460 private function doExport( $pageSet, $result ) {
461 $exportTitles = array();
462 $titles = $pageSet->getGoodTitles();
463 if ( count( $titles ) ) {
464 $user = $this->getUser();
465 /** @var $title Title */
466 foreach ( $titles as $title ) {
467 if ( $title->userCan( 'read', $user ) ) {
468 $exportTitles[] = $title;
469 }
470 }
471 }
472
473 $exporter = new WikiExporter( $this->getDB() );
474 // WikiExporter writes to stdout, so catch its
475 // output with an ob
476 ob_start();
477 $exporter->openStream();
478 foreach ( $exportTitles as $title ) {
479 $exporter->pageByTitle( $title );
480 }
481 $exporter->closeStream();
482 $exportxml = ob_get_contents();
483 ob_end_clean();
484
485 // Don't check the size of exported stuff
486 // It's not continuable, so it would cause more
487 // problems than it'd solve
488 if ( $this->mParams['exportnowrap'] ) {
489 $result->reset();
490 // Raw formatter will handle this
491 $result->addValue( null, 'text', $exportxml, ApiResult::NO_SIZE_CHECK );
492 $result->addValue( null, 'mime', 'text/xml', ApiResult::NO_SIZE_CHECK );
493 } else {
494 $r = array();
495 ApiResult::setContent( $r, $exportxml );
496 $result->addValue( 'query', 'export', $r, ApiResult::NO_SIZE_CHECK );
497 }
498 }
499
500 public function getAllowedParams( $flags = 0 ) {
501 $result = array(
502 'prop' => array(
503 ApiBase::PARAM_ISMULTI => true,
504 ApiBase::PARAM_TYPE => $this->mModuleMgr->getNames( 'prop' )
505 ),
506 'list' => array(
507 ApiBase::PARAM_ISMULTI => true,
508 ApiBase::PARAM_TYPE => $this->mModuleMgr->getNames( 'list' )
509 ),
510 'meta' => array(
511 ApiBase::PARAM_ISMULTI => true,
512 ApiBase::PARAM_TYPE => $this->mModuleMgr->getNames( 'meta' )
513 ),
514 'indexpageids' => false,
515 'export' => false,
516 'exportnowrap' => false,
517 'iwurl' => false,
518 'continue' => null,
519 );
520 if ( $flags ) {
521 $result += $this->getPageSet()->getFinalParams( $flags );
522 }
523
524 return $result;
525 }
526
527 /**
528 * Override the parent to generate help messages for all available query modules.
529 * @return string
530 */
531 public function makeHelpMsg() {
532
533 // Use parent to make default message for the query module
534 $msg = parent::makeHelpMsg();
535
536 $querySeparator = str_repeat( '--- ', 12 );
537 $moduleSeparator = str_repeat( '*** ', 14 );
538 $msg .= "\n$querySeparator Query: Prop $querySeparator\n\n";
539 $msg .= $this->makeHelpMsgHelper( 'prop' );
540 $msg .= "\n$querySeparator Query: List $querySeparator\n\n";
541 $msg .= $this->makeHelpMsgHelper( 'list' );
542 $msg .= "\n$querySeparator Query: Meta $querySeparator\n\n";
543 $msg .= $this->makeHelpMsgHelper( 'meta' );
544 $msg .= "\n\n$moduleSeparator Modules: continuation $moduleSeparator\n\n";
545
546 return $msg;
547 }
548
549 /**
550 * For all modules of a given group, generate help messages and join them together
551 * @param string $group Module group
552 * @return string
553 */
554 private function makeHelpMsgHelper( $group ) {
555 $moduleDescriptions = array();
556
557 $moduleNames = $this->mModuleMgr->getNames( $group );
558 sort( $moduleNames );
559 foreach ( $moduleNames as $name ) {
560 /**
561 * @var $module ApiQueryBase
562 */
563 $module = $this->mModuleMgr->getModule( $name );
564
565 $msg = ApiMain::makeHelpMsgHeader( $module, $group );
566 $msg2 = $module->makeHelpMsg();
567 if ( $msg2 !== false ) {
568 $msg .= $msg2;
569 }
570 if ( $module instanceof ApiQueryGeneratorBase ) {
571 $msg .= "Generator:\n This module may be used as a generator\n";
572 }
573 $moduleDescriptions[] = $msg;
574 }
575
576 return implode( "\n", $moduleDescriptions );
577 }
578
579 public function shouldCheckMaxlag() {
580 return true;
581 }
582
583 public function getParamDescription() {
584 return $this->getPageSet()->getFinalParamDescription() + array(
585 'prop' => 'Which properties to get for the titles/revisions/pageids. ' .
586 'Module help is available below',
587 'list' => 'Which lists to get. Module help is available below',
588 'meta' => 'Which metadata to get about the site. Module help is available below',
589 'indexpageids' => 'Include an additional pageids section listing all returned page IDs',
590 'export' => 'Export the current revisions of all given or generated pages',
591 'exportnowrap' => 'Return the export XML without wrapping it in an ' .
592 'XML result (same format as Special:Export). Can only be used with export',
593 'iwurl' => 'Whether to get the full URL if the title is an interwiki link',
594 'continue' => array(
595 'When present, formats query-continue as key-value pairs that ' .
596 'should simply be merged into the original request.',
597 'This parameter must be set to an empty string in the initial query.',
598 'This parameter is recommended for all new development, and ' .
599 'will be made default in the next API version.'
600 ),
601 );
602 }
603
604 public function getDescription() {
605 return array(
606 'Query API module allows applications to get needed pieces of data ' .
607 'from the MediaWiki databases,',
608 'and is loosely based on the old query.php interface.',
609 'All data modifications will first have to use query to acquire a ' .
610 'token to prevent abuse from malicious sites.'
611 );
612 }
613
614 public function getPossibleErrors() {
615 return array_merge(
616 parent::getPossibleErrors(),
617 $this->getPageSet()->getFinalPossibleErrors()
618 );
619 }
620
621 public function getExamples() {
622 return array(
623 'api.php?action=query&prop=revisions&meta=siteinfo&' .
624 'titles=Main%20Page&rvprop=user|comment&continue=',
625 'api.php?action=query&generator=allpages&gapprefix=API/&prop=revisions&continue=',
626 );
627 }
628
629 public function getHelpUrls() {
630 return array(
631 'https://www.mediawiki.org/wiki/API:Query',
632 'https://www.mediawiki.org/wiki/API:Meta',
633 'https://www.mediawiki.org/wiki/API:Properties',
634 'https://www.mediawiki.org/wiki/API:Lists',
635 );
636 }
637 }