Mass convert NULL -> null. Left strings and comments alone, obviously.
[lhc/web/wiklou.git] / includes / api / ApiQueryRevisions.php
1 <?php
2
3 /*
4 * Created on Sep 7, 2006
5 *
6 * API for MediaWiki 1.8+
7 *
8 * Copyright (C) 2006 Yuri Astrakhan <Firstname><Lastname>@gmail.com
9 *
10 * This program is free software; you can redistribute it and/or modify
11 * it under the terms of the GNU General Public License as published by
12 * the Free Software Foundation; either version 2 of the License, or
13 * (at your option) any later version.
14 *
15 * This program is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 * GNU General Public License for more details.
19 *
20 * You should have received a copy of the GNU General Public License along
21 * with this program; if not, write to the Free Software Foundation, Inc.,
22 * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
23 * http://www.gnu.org/copyleft/gpl.html
24 */
25
26 if (!defined('MEDIAWIKI')) {
27 // Eclipse helper - will be ignored in production
28 require_once ('ApiQueryBase.php');
29 }
30
31 /**
32 * A query action to enumerate revisions of a given page, or show top revisions of multiple pages.
33 * Various pieces of information may be shown - flags, comments, and the actual wiki markup of the rev.
34 * In the enumeration mode, ranges of revisions may be requested and filtered.
35 *
36 * @ingroup API
37 */
38 class ApiQueryRevisions extends ApiQueryBase {
39
40 public function __construct($query, $moduleName) {
41 parent :: __construct($query, $moduleName, 'rv');
42 }
43
44 private $fld_ids = false, $fld_flags = false, $fld_timestamp = false, $fld_size = false,
45 $fld_comment = false, $fld_user = false, $fld_content = false, $fld_tags = false;
46
47 protected function getTokenFunctions() {
48 // tokenname => function
49 // function prototype is func($pageid, $title, $rev)
50 // should return token or false
51
52 // Don't call the hooks twice
53 if(isset($this->tokenFunctions))
54 return $this->tokenFunctions;
55
56 // If we're in JSON callback mode, no tokens can be obtained
57 if(!is_null($this->getMain()->getRequest()->getVal('callback')))
58 return array();
59
60 $this->tokenFunctions = array(
61 'rollback' => array( 'ApiQueryRevisions', 'getRollbackToken' )
62 );
63 wfRunHooks('APIQueryRevisionsTokens', array(&$this->tokenFunctions));
64 return $this->tokenFunctions;
65 }
66
67 public static function getRollbackToken($pageid, $title, $rev)
68 {
69 global $wgUser;
70 if(!$wgUser->isAllowed('rollback'))
71 return false;
72 return $wgUser->editToken(array($title->getPrefixedText(),
73 $rev->getUserText()));
74 }
75
76 public function execute() {
77 $params = $this->extractRequestParams(false);
78
79 // If any of those parameters are used, work in 'enumeration' mode.
80 // Enum mode can only be used when exactly one page is provided.
81 // Enumerating revisions on multiple pages make it extremely
82 // difficult to manage continuations and require additional SQL indexes
83 $enumRevMode = (!is_null($params['user']) || !is_null($params['excludeuser']) ||
84 !is_null($params['limit']) || !is_null($params['startid']) ||
85 !is_null($params['endid']) || $params['dir'] === 'newer' ||
86 !is_null($params['start']) || !is_null($params['end']));
87
88
89 $pageSet = $this->getPageSet();
90 $pageCount = $pageSet->getGoodTitleCount();
91 $revCount = $pageSet->getRevisionCount();
92
93 // Optimization -- nothing to do
94 if ($revCount === 0 && $pageCount === 0)
95 return;
96
97 if ($revCount > 0 && $enumRevMode)
98 $this->dieUsage('The revids= parameter may not be used with the list options (limit, startid, endid, dirNewer, start, end).', 'revids');
99
100 if ($pageCount > 1 && $enumRevMode)
101 $this->dieUsage('titles, pageids or a generator was used to supply multiple pages, but the limit, startid, endid, dirNewer, user, excludeuser, start and end parameters may only be used on a single page.', 'multpages');
102
103 $this->diffto = $this->difftotext = null;
104 if (!is_null($params['difftotext'])) {
105 $this->difftotext = $params['difftotext'];
106 } else if (!is_null($params['diffto'])) {
107 if ($params['diffto'] == 'cur')
108 $params['diffto'] = 0;
109 if ((!ctype_digit($params['diffto']) || $params['diffto'] < 0)
110 && $params['diffto'] != 'prev' && $params['diffto'] != 'next')
111 $this->dieUsage('rvdiffto must be set to a non-negative number, "prev", "next" or "cur"', 'diffto');
112 // Check whether the revision exists and is readable,
113 // DifferenceEngine returns a rather ambiguous empty
114 // string if that's not the case
115 if ($params['diffto'] != 0) {
116 $difftoRev = Revision::newFromID($params['diffto']);
117 if (!$difftoRev)
118 $this->dieUsageMsg(array('nosuchrevid', $params['diffto']));
119 if (!$difftoRev->userCan(Revision::DELETED_TEXT)) {
120 $this->setWarning("Couldn't diff to r{$difftoRev->getID()}: content is hidden");
121 $params['diffto'] = null;
122 }
123 }
124 $this->diffto = $params['diffto'];
125 }
126
127 $db = $this->getDB();
128 $this->addTables(array('page', 'revision'));
129 $this->addFields(Revision::selectFields());
130 $this->addWhere('page_id = rev_page');
131
132 $prop = array_flip($params['prop']);
133
134 // Optional fields
135 $this->fld_ids = isset ($prop['ids']);
136 // $this->addFieldsIf('rev_text_id', $this->fld_ids); // should this be exposed?
137 $this->fld_flags = isset ($prop['flags']);
138 $this->fld_timestamp = isset ($prop['timestamp']);
139 $this->fld_comment = isset ($prop['comment']);
140 $this->fld_size = isset ($prop['size']);
141 $this->fld_user = isset ($prop['user']);
142 $this->token = $params['token'];
143
144 if ( !is_null($this->token) || $pageCount > 0) {
145 $this->addFields( Revision::selectPageFields() );
146 }
147
148 if (isset ($prop['tags'])) {
149 $this->fld_tags = true;
150 $this->addTables('tag_summary');
151 $this->addJoinConds(array('tag_summary' => array('LEFT JOIN', array('rev_id=ts_rev_id'))));
152 $this->addFields('ts_tags');
153 }
154
155 if( !is_null($params['tag']) ) {
156 $this->addTables('change_tag');
157 $this->addJoinConds(array('change_tag' => array('INNER JOIN', array('rev_id=ct_rev_id'))));
158 $this->addWhereFld('ct_tag' , $params['tag']);
159 }
160
161 if (isset($prop['content']) || !is_null($this->difftotext)) {
162
163 // For each page we will request, the user must have read rights for that page
164 foreach ($pageSet->getGoodTitles() as $title) {
165 if( !$title->userCanRead() )
166 $this->dieUsage(
167 'The current user is not allowed to read ' . $title->getPrefixedText(),
168 'accessdenied');
169 }
170
171 $this->addTables('text');
172 $this->addWhere('rev_text_id=old_id');
173 $this->addFields('old_id');
174 $this->addFields(Revision::selectTextFields());
175
176 $this->fld_content = isset($prop['content']);
177
178 $this->expandTemplates = $params['expandtemplates'];
179 $this->generateXML = $params['generatexml'];
180 if(isset($params['section']))
181 $this->section = $params['section'];
182 else
183 $this->section = false;
184 }
185
186 $userMax = ( $this->fld_content ? ApiBase::LIMIT_SML1 : ApiBase::LIMIT_BIG1 );
187 $botMax = ( $this->fld_content ? ApiBase::LIMIT_SML2 : ApiBase::LIMIT_BIG2 );
188 $limit = $params['limit'];
189 if( $limit == 'max' ) {
190 $limit = $this->getMain()->canApiHighLimits() ? $botMax : $userMax;
191 $this->getResult()->addValue( 'limits', $this->getModuleName(), $limit );
192 }
193
194 if ($enumRevMode) {
195
196 // This is mostly to prevent parameter errors (and optimize SQL?)
197 if (!is_null($params['startid']) && !is_null($params['start']))
198 $this->dieUsage('start and startid cannot be used together', 'badparams');
199
200 if (!is_null($params['endid']) && !is_null($params['end']))
201 $this->dieUsage('end and endid cannot be used together', 'badparams');
202
203 if(!is_null($params['user']) && !is_null($params['excludeuser']))
204 $this->dieUsage('user and excludeuser cannot be used together', 'badparams');
205
206 // This code makes an assumption that sorting by rev_id and rev_timestamp produces
207 // the same result. This way users may request revisions starting at a given time,
208 // but to page through results use the rev_id returned after each page.
209 // Switching to rev_id removes the potential problem of having more than
210 // one row with the same timestamp for the same page.
211 // The order needs to be the same as start parameter to avoid SQL filesort.
212
213 if (is_null($params['startid']) && is_null($params['endid']))
214 $this->addWhereRange('rev_timestamp', $params['dir'],
215 $params['start'], $params['end']);
216 else {
217 $this->addWhereRange('rev_id', $params['dir'],
218 $params['startid'], $params['endid']);
219 // One of start and end can be set
220 // If neither is set, this does nothing
221 $this->addWhereRange('rev_timestamp', $params['dir'],
222 $params['start'], $params['end'], false);
223 }
224
225 // must manually initialize unset limit
226 if (is_null($limit))
227 $limit = 10;
228 $this->validateLimit('limit', $limit, 1, $userMax, $botMax);
229
230 // There is only one ID, use it
231 $ids = array_keys($pageSet->getGoodTitles());
232 $this->addWhereFld('rev_page', reset($ids));
233
234 if(!is_null($params['user'])) {
235 $this->addWhereFld('rev_user_text', $params['user']);
236 } elseif (!is_null($params['excludeuser'])) {
237 $this->addWhere('rev_user_text != ' .
238 $db->addQuotes($params['excludeuser']));
239 }
240 if(!is_null($params['user']) || !is_null($params['excludeuser'])) {
241 // Paranoia: avoid brute force searches (bug 17342)
242 $this->addWhere($db->bitAnd('rev_deleted',Revision::DELETED_USER) . ' = 0');
243 }
244 }
245 elseif ($revCount > 0) {
246 $max = $this->getMain()->canApiHighLimits() ? $botMax : $userMax;
247 $revs = $pageSet->getRevisionIDs();
248 if(self::truncateArray($revs, $max))
249 $this->setWarning("Too many values supplied for parameter 'revids': the limit is $max");
250
251 // Get all revision IDs
252 $this->addWhereFld('rev_id', array_keys($revs));
253
254 if(!is_null($params['continue']))
255 $this->addWhere("rev_id >= '" . intval($params['continue']) . "'");
256 $this->addOption('ORDER BY', 'rev_id');
257
258 // assumption testing -- we should never get more then $revCount rows.
259 $limit = $revCount;
260 }
261 elseif ($pageCount > 0) {
262 $max = $this->getMain()->canApiHighLimits() ? $botMax : $userMax;
263 $titles = $pageSet->getGoodTitles();
264 if(self::truncateArray($titles, $max))
265 $this->setWarning("Too many values supplied for parameter 'titles': the limit is $max");
266
267 // When working in multi-page non-enumeration mode,
268 // limit to the latest revision only
269 $this->addWhere('page_id=rev_page');
270 $this->addWhere('page_latest=rev_id');
271
272 // Get all page IDs
273 $this->addWhereFld('page_id', array_keys($titles));
274 // Every time someone relies on equality propagation, god kills a kitten :)
275 $this->addWhereFld('rev_page', array_keys($titles));
276
277 if(!is_null($params['continue']))
278 {
279 $cont = explode('|', $params['continue']);
280 if(count($cont) != 2)
281 $this->dieUsage("Invalid continue param. You should pass the original " .
282 "value returned by the previous query", "_badcontinue");
283 $pageid = intval($cont[0]);
284 $revid = intval($cont[1]);
285 $this->addWhere("rev_page > '$pageid' OR " .
286 "(rev_page = '$pageid' AND " .
287 "rev_id >= '$revid')");
288 }
289 $this->addOption('ORDER BY', 'rev_page, rev_id');
290
291 // assumption testing -- we should never get more then $pageCount rows.
292 $limit = $pageCount;
293 } else
294 ApiBase :: dieDebug(__METHOD__, 'param validation?');
295
296 $this->addOption('LIMIT', $limit +1);
297
298 $data = array ();
299 $count = 0;
300 $res = $this->select(__METHOD__);
301
302 while ($row = $db->fetchObject($res)) {
303
304 if (++ $count > $limit) {
305 // We've reached the one extra which shows that there are additional pages to be had. Stop here...
306 if (!$enumRevMode)
307 ApiBase :: dieDebug(__METHOD__, 'Got more rows then expected'); // bug report
308 $this->setContinueEnumParameter('startid', intval($row->rev_id));
309 break;
310 }
311
312 //
313 $fit = $this->addPageSubItem($row->rev_page, $this->extractRowInfo($row), 'rev');
314 if(!$fit)
315 {
316 if($enumRevMode)
317 $this->setContinueEnumParameter('startid', intval($row->rev_id));
318 else if($revCount > 0)
319 $this->setContinueEnumParameter('continue', intval($row->rev_id));
320 else
321 $this->setContinueEnumParameter('continue', intval($row->rev_page) .
322 '|' . intval($row->rev_id));
323 break;
324 }
325 }
326 $db->freeResult($res);
327 }
328
329 private function extractRowInfo( $row ) {
330 $revision = new Revision( $row );
331 $title = $revision->getTitle();
332 $vals = array ();
333
334 if ($this->fld_ids) {
335 $vals['revid'] = intval($revision->getId());
336 // $vals['oldid'] = intval($row->rev_text_id); // todo: should this be exposed?
337 if (!is_null($revision->getParentId()))
338 $vals['parentid'] = intval($revision->getParentId());
339 }
340
341 if ($this->fld_flags && $revision->isMinor())
342 $vals['minor'] = '';
343
344 if ($this->fld_user) {
345 if ($revision->isDeleted(Revision::DELETED_USER)) {
346 $vals['userhidden'] = '';
347 } else {
348 $vals['user'] = $revision->getUserText();
349 if (!$revision->getUser())
350 $vals['anon'] = '';
351 }
352 }
353
354 if ($this->fld_timestamp) {
355 $vals['timestamp'] = wfTimestamp(TS_ISO_8601, $revision->getTimestamp());
356 }
357
358 if ($this->fld_size && !is_null($revision->getSize())) {
359 $vals['size'] = intval($revision->getSize());
360 }
361
362 if ($this->fld_comment) {
363 if ($revision->isDeleted(Revision::DELETED_COMMENT)) {
364 $vals['commenthidden'] = '';
365 } else {
366 $comment = $revision->getComment();
367 if (strval($comment) !== '')
368 $vals['comment'] = $comment;
369 }
370 }
371
372 if ($this->fld_tags) {
373 if ($row->ts_tags) {
374 $tags = explode(',', $row->ts_tags);
375 $this->getResult()->setIndexedTagName($tags, 'tag');
376 $vals['tags'] = $tags;
377 } else {
378 $vals['tags'] = array();
379 }
380 }
381
382 if(!is_null($this->token))
383 {
384 $tokenFunctions = $this->getTokenFunctions();
385 foreach($this->token as $t)
386 {
387 $val = call_user_func($tokenFunctions[$t], $title->getArticleID(), $title, $revision);
388 if($val === false)
389 $this->setWarning("Action '$t' is not allowed for the current user");
390 else
391 $vals[$t . 'token'] = $val;
392 }
393 }
394
395 $text = null;
396 if ($this->fld_content || !is_null($this->difftotext)) {
397 global $wgParser;
398 $text = $revision->getText();
399 # Expand templates after getting section content because
400 # template-added sections don't count and Parser::preprocess()
401 # will have less input
402 if ($this->section !== false) {
403 $text = $wgParser->getSection( $text, $this->section, false);
404 if($text === false)
405 $this->dieUsage("There is no section {$this->section} in r".$revision->getId(), 'nosuchsection');
406 }
407 }
408 if ($this->fld_content && !$revision->isDeleted(Revision::DELETED_TEXT)) {
409 if ($this->generateXML) {
410 $wgParser->startExternalParse( $title, new ParserOptions(), OT_PREPROCESS );
411 $dom = $wgParser->preprocessToDom( $text );
412 if ( is_callable( array( $dom, 'saveXML' ) ) ) {
413 $xml = $dom->saveXML();
414 } else {
415 $xml = $dom->__toString();
416 }
417 $vals['parsetree'] = $xml;
418
419 }
420 if ($this->expandTemplates) {
421 $text = $wgParser->preprocess( $text, $title, new ParserOptions() );
422 }
423 ApiResult :: setContent($vals, $text);
424 } else if ($this->fld_content) {
425 $vals['texthidden'] = '';
426 }
427
428 if (!is_null($this->diffto) || !is_null($this->difftotext)) {
429 global $wgAPIMaxUncachedDiffs;
430 static $n = 0; // Number of uncached diffs we've had
431 if($n < $wgAPIMaxUncachedDiffs) {
432 $vals['diff'] = array();
433 if(!is_null($this->difftotext)) {
434 $engine = new DifferenceEngine($title);
435 $engine->setText($text, $this->difftotext);
436 } else {
437 $engine = new DifferenceEngine($title, $revision->getID(), $this->diffto);
438 $vals['diff']['from'] = $engine->getOldid();
439 $vals['diff']['to'] = $engine->getNewid();
440 }
441 $difftext = $engine->getDiffBody();
442 ApiResult::setContent($vals['diff'], $difftext);
443 if(!$engine->wasCacheHit())
444 $n++;
445 } else {
446 $vals['diff']['notcached'] = '';
447 }
448 }
449 return $vals;
450 }
451
452 public function getAllowedParams() {
453 return array (
454 'prop' => array (
455 ApiBase :: PARAM_ISMULTI => true,
456 ApiBase :: PARAM_DFLT => 'ids|timestamp|flags|comment|user',
457 ApiBase :: PARAM_TYPE => array (
458 'ids',
459 'flags',
460 'timestamp',
461 'user',
462 'size',
463 'comment',
464 'content',
465 'tags'
466 )
467 ),
468 'limit' => array (
469 ApiBase :: PARAM_TYPE => 'limit',
470 ApiBase :: PARAM_MIN => 1,
471 ApiBase :: PARAM_MAX => ApiBase :: LIMIT_BIG1,
472 ApiBase :: PARAM_MAX2 => ApiBase :: LIMIT_BIG2
473 ),
474 'startid' => array (
475 ApiBase :: PARAM_TYPE => 'integer'
476 ),
477 'endid' => array (
478 ApiBase :: PARAM_TYPE => 'integer'
479 ),
480 'start' => array (
481 ApiBase :: PARAM_TYPE => 'timestamp'
482 ),
483 'end' => array (
484 ApiBase :: PARAM_TYPE => 'timestamp'
485 ),
486 'dir' => array (
487 ApiBase :: PARAM_DFLT => 'older',
488 ApiBase :: PARAM_TYPE => array (
489 'newer',
490 'older'
491 )
492 ),
493 'user' => array(
494 ApiBase :: PARAM_TYPE => 'user'
495 ),
496 'excludeuser' => array(
497 ApiBase :: PARAM_TYPE => 'user'
498 ),
499 'tag' => null,
500 'expandtemplates' => false,
501 'generatexml' => false,
502 'section' => null,
503 'token' => array(
504 ApiBase :: PARAM_TYPE => array_keys($this->getTokenFunctions()),
505 ApiBase :: PARAM_ISMULTI => true
506 ),
507 'continue' => null,
508 'diffto' => null,
509 'difftotext' => null,
510 );
511 }
512
513 public function getParamDescription() {
514 return array (
515 'prop' => 'Which properties to get for each revision.',
516 'limit' => 'limit how many revisions will be returned (enum)',
517 'startid' => 'from which revision id to start enumeration (enum)',
518 'endid' => 'stop revision enumeration on this revid (enum)',
519 'start' => 'from which revision timestamp to start enumeration (enum)',
520 'end' => 'enumerate up to this timestamp (enum)',
521 'dir' => 'direction of enumeration - towards "newer" or "older" revisions (enum)',
522 'user' => 'only include revisions made by user',
523 'excludeuser' => 'exclude revisions made by user',
524 'expandtemplates' => 'expand templates in revision content',
525 'generatexml' => 'generate XML parse tree for revision content',
526 'section' => 'only retrieve the content of this section',
527 'token' => 'Which tokens to obtain for each revision',
528 'continue' => 'When more results are available, use this to continue',
529 'diffto' => array('Revision ID to diff each revision to.',
530 'Use "prev", "next" and "cur" for the previous, next and current revision respectively.'),
531 'difftotext' => array('Text to diff each revision to. Only diffs a limited number of revisions.',
532 'Overrides diffto. If rvsection is set, only that section will be diffed against this text.'),
533 'tag' => 'Only list revisions tagged with this tag',
534 );
535 }
536
537 public function getDescription() {
538 return array (
539 'Get revision information.',
540 'This module may be used in several ways:',
541 ' 1) Get data about a set of pages (last revision), by setting titles or pageids parameter.',
542 ' 2) Get revisions for one given page, by using titles/pageids with start/end/limit params.',
543 ' 3) Get data about a set of revisions by setting their IDs with revids parameter.',
544 'All parameters marked as (enum) may only be used with a single page (#2).'
545 );
546 }
547
548 protected function getExamples() {
549 return array (
550 'Get data with content for the last revision of titles "API" and "Main Page":',
551 ' api.php?action=query&prop=revisions&titles=API|Main%20Page&rvprop=timestamp|user|comment|content',
552 'Get last 5 revisions of the "Main Page":',
553 ' api.php?action=query&prop=revisions&titles=Main%20Page&rvlimit=5&rvprop=timestamp|user|comment',
554 'Get first 5 revisions of the "Main Page":',
555 ' api.php?action=query&prop=revisions&titles=Main%20Page&rvlimit=5&rvprop=timestamp|user|comment&rvdir=newer',
556 'Get first 5 revisions of the "Main Page" made after 2006-05-01:',
557 ' api.php?action=query&prop=revisions&titles=Main%20Page&rvlimit=5&rvprop=timestamp|user|comment&rvdir=newer&rvstart=20060501000000',
558 'Get first 5 revisions of the "Main Page" that were not made made by anonymous user "127.0.0.1"',
559 ' api.php?action=query&prop=revisions&titles=Main%20Page&rvlimit=5&rvprop=timestamp|user|comment&rvexcludeuser=127.0.0.1',
560 'Get first 5 revisions of the "Main Page" that were made by the user "MediaWiki default"',
561 ' api.php?action=query&prop=revisions&titles=Main%20Page&rvlimit=5&rvprop=timestamp|user|comment&rvuser=MediaWiki%20default',
562 );
563 }
564
565 public function getVersion() {
566 return __CLASS__ . ': $Id$';
567 }
568 }