includes/api/ApiStashEdit.php

   1 <?php
   2 /**
   3  * This program is free software; you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation; either version 2 of the License, or
   6  * (at your option) any later version.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License along
  14  * with this program; if not, write to the Free Software Foundation, Inc.,
  15  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  16  * http://www.gnu.org/copyleft/gpl.html
  17  *
  18  * @file
  19  * @author Aaron Schulz
  20  */
  21
  22 use MediaWiki\Logger\LoggerFactory;
  23 use MediaWiki\MediaWikiServices;
  24
  25 /**
  26  * Prepare an edit in shared cache so that it can be reused on edit
  27  *
  28  * This endpoint can be called via AJAX as the user focuses on the edit
  29  * summary box. By the time of submission, the parse may have already
  30  * finished, and can be immediately used on page save. Certain parser
  31  * functions like {{REVISIONID}} or {{CURRENTTIME}} may cause the cache
  32  * to not be used on edit. Template and files used are check for changes
  33  * since the output was generated. The cache TTL is also kept low for sanity.
  34  *
  35  * @ingroup API
  36  * @since 1.25
  37  */
  38 class ApiStashEdit extends ApiBase {
  39         const ERROR_NONE = 'stashed';
  40         const ERROR_PARSE = 'error_parse';
  41         const ERROR_CACHE = 'error_cache';
  42         const ERROR_UNCACHEABLE = 'uncacheable';
  43         const ERROR_BUSY = 'busy';
  44
  45         const PRESUME_FRESH_TTL_SEC = 30;
  46         const MAX_CACHE_TTL = 300; // 5 minutes
  47
  48         public function execute() {
  49                 $user = $this->getUser();
  50                 $params = $this->extractRequestParams();
  51
  52                 if ( $user->isBot() ) { // sanity
  53                         $this->dieUsage( 'This interface is not supported for bots', 'botsnotsupported' );
  54                 }
  55
  56                 $cache = ObjectCache::getLocalClusterInstance();
  57                 $page = $this->getTitleOrPageId( $params );
  58                 $title = $page->getTitle();
  59
  60                 if ( !ContentHandler::getForModelID( $params['contentmodel'] )
  61                         ->isSupportedFormat( $params['contentformat'] )
  62                 ) {
  63                         $this->dieUsage( 'Unsupported content model/format', 'badmodelformat' );
  64                 }
  65
  66                 $text = null;
  67                 $textHash = null;
  68                 if ( strlen( $params['stashedtexthash'] ) ) {
  69                         // Load from cache since the client indicates the text is the same as last stash
  70                         $textHash = $params['stashedtexthash'];
  71                         $textKey = $cache->makeKey( 'stashedit', 'text', $textHash );
  72                         $text = $cache->get( $textKey );
  73                         if ( !is_string( $text ) ) {
  74                                 $this->dieUsage( 'No stashed text found with the given hash', 'missingtext' );
  75                         }
  76                 } elseif ( $params['text'] !== null ) {
  77                         // Trim and fix newlines so the key SHA1's match (see WebRequest::getText())
  78                         $text = rtrim( str_replace( "\r\n", "\n", $params['text'] ) );
  79                         $textHash = sha1( $text );
  80                 } else {
  81                         $this->dieUsage(
  82                                 'The text or stashedtexthash parameter must be given', 'missingtextparam' );
  83                 }
  84
  85                 $textContent = ContentHandler::makeContent(
  86                         $text, $title, $params['contentmodel'], $params['contentformat'] );
  87
  88                 $page = WikiPage::factory( $title );
  89                 if ( $page->exists() ) {
  90                         // Page exists: get the merged content with the proposed change
  91                         $baseRev = Revision::newFromPageId( $page->getId(), $params['baserevid'] );
  92                         if ( !$baseRev ) {
  93                                 $this->dieUsage( "No revision ID {$params['baserevid']}", 'missingrev' );
  94                         }
  95                         $currentRev = $page->getRevision();
  96                         if ( !$currentRev ) {
  97                                 $this->dieUsage( "No current revision of page ID {$page->getId()}", 'missingrev' );
  98                         }
  99                         // Merge in the new version of the section to get the proposed version
 100                         $editContent = $page->replaceSectionAtRev(
 101                                 $params['section'],
 102                                 $textContent,
 103                                 $params['sectiontitle'],
 104                                 $baseRev->getId()
 105                         );
 106                         if ( !$editContent ) {
 107                                 $this->dieUsage( 'Could not merge updated section.', 'replacefailed' );
 108                         }
 109                         if ( $currentRev->getId() == $baseRev->getId() ) {
 110                                 // Base revision was still the latest; nothing to merge
 111                                 $content = $editContent;
 112                         } else {
 113                                 // Merge the edit into the current version
 114                                 $baseContent = $baseRev->getContent();
 115                                 $currentContent = $currentRev->getContent();
 116                                 if ( !$baseContent || !$currentContent ) {
 117                                         $this->dieUsage( "Missing content for page ID {$page->getId()}", 'missingrev' );
 118                                 }
 119                                 $handler = ContentHandler::getForModelID( $baseContent->getModel() );
 120                                 $content = $handler->merge3( $baseContent, $editContent, $currentContent );
 121                         }
 122                 } else {
 123                         // New pages: use the user-provided content model
 124                         $content = $textContent;
 125                 }
 126
 127                 if ( !$content ) { // merge3() failed
 128                         $this->getResult()->addValue( null,
 129                                 $this->getModuleName(), [ 'status' => 'editconflict' ] );
 130                         return;
 131                 }
 132
 133                 // The user will abort the AJAX request by pressing "save", so ignore that
 134                 ignore_user_abort( true );
 135
 136                 if ( $user->pingLimiter( 'stashedit' ) ) {
 137                         $status = 'ratelimited';
 138                 } else {
 139                         $status = self::parseAndStash( $page, $content, $user, $params['summary'] );
 140                         $textKey = $cache->makeKey( 'stashedit', 'text', $textHash );
 141                         $cache->set( $textKey, $text, self::MAX_CACHE_TTL );
 142                 }
 143
 144                 $stats = MediaWikiServices::getInstance()->getStatsdDataFactory();
 145                 $stats->increment( "editstash.cache_stores.$status" );
 146
 147                 $this->getResult()->addValue(
 148                         null,
 149                         $this->getModuleName(),
 150                         [
 151                                 'status' => $status,
 152                                 'texthash' => $textHash
 153                         ]
 154                 );
 155         }
 156
 157         /**
 158          * @param WikiPage $page
 159          * @param Content $content Edit content
 160          * @param User $user
 161          * @param string $summary Edit summary
 162          * @return integer ApiStashEdit::ERROR_* constant
 163          * @since 1.25
 164          */
 165         public static function parseAndStash( WikiPage $page, Content $content, User $user, $summary ) {
 166                 $cache = ObjectCache::getLocalClusterInstance();
 167                 $logger = LoggerFactory::getInstance( 'StashEdit' );
 168
 169                 $title = $page->getTitle();
 170                 $key = self::getStashKey( $title, self::getContentHash( $content ), $user );
 171
 172                 // Use the master DB for fast blocking locks
 173                 $dbw = wfGetDB( DB_MASTER );
 174                 if ( !$dbw->lock( $key, __METHOD__, 1 ) ) {
 175                         // De-duplicate requests on the same key
 176                         return self::ERROR_BUSY;
 177                 }
 178                 /** @noinspection PhpUnusedLocalVariableInspection */
 179                 $unlocker = new ScopedCallback( function () use ( $dbw, $key ) {
 180                         $dbw->unlock( $key, __METHOD__ );
 181                 } );
 182
 183                 $cutoffTime = time() - self::PRESUME_FRESH_TTL_SEC;
 184
 185                 // Reuse any freshly build matching edit stash cache
 186                 $editInfo = $cache->get( $key );
 187                 if ( $editInfo && wfTimestamp( TS_UNIX, $editInfo->timestamp ) >= $cutoffTime ) {
 188                         $alreadyCached = true;
 189                 } else {
 190                         $format = $content->getDefaultFormat();
 191                         $editInfo = $page->prepareContentForEdit( $content, null, $user, $format, false );
 192                         $alreadyCached = false;
 193                 }
 194
 195                 if ( $editInfo && $editInfo->output ) {
 196                         // Let extensions add ParserOutput metadata or warm other caches
 197                         Hooks::run( 'ParserOutputStashForEdit',
 198                                 [ $page, $content, $editInfo->output, $summary, $user ] );
 199
 200                         if ( $alreadyCached ) {
 201                                 $logger->debug( "Already cached parser output for key '$key' ('$title')." );
 202                                 return self::ERROR_NONE;
 203                         }
 204
 205                         list( $stashInfo, $ttl, $code ) = self::buildStashValue(
 206                                 $editInfo->pstContent,
 207                                 $editInfo->output,
 208                                 $editInfo->timestamp,
 209                                 $user
 210                         );
 211
 212                         if ( $stashInfo ) {
 213                                 $ok = $cache->set( $key, $stashInfo, $ttl );
 214                                 if ( $ok ) {
 215                                         $logger->debug( "Cached parser output for key '$key' ('$title')." );
 216                                         return self::ERROR_NONE;
 217                                 } else {
 218                                         $logger->error( "Failed to cache parser output for key '$key' ('$title')." );
 219                                         return self::ERROR_CACHE;
 220                                 }
 221                         } else {
 222                                 $logger->info( "Uncacheable parser output for key '$key' ('$title') [$code]." );
 223                                 return self::ERROR_UNCACHEABLE;
 224                         }
 225                 }
 226
 227                 return self::ERROR_PARSE;
 228         }
 229
 230         /**
 231          * Check that a prepared edit is in cache and still up-to-date
 232          *
 233          * This method blocks if the prepared edit is already being rendered,
 234          * waiting until rendering finishes before doing final validity checks.
 235          *
 236          * The cache is rejected if template or file changes are detected.
 237          * Note that foreign template or file transclusions are not checked.
 238          *
 239          * The result is a map (pstContent,output,timestamp) with fields
 240          * extracted directly from WikiPage::prepareContentForEdit().
 241          *
 242          * @param Title $title
 243          * @param Content $content
 244          * @param User $user User to get parser options from
 245          * @return stdClass|bool Returns false on cache miss
 246          */
 247         public static function checkCache( Title $title, Content $content, User $user ) {
 248                 if ( $user->isBot() ) {
 249                         return false; // bots never stash - don't pollute stats
 250                 }
 251
 252                 $cache = ObjectCache::getLocalClusterInstance();
 253                 $logger = LoggerFactory::getInstance( 'StashEdit' );
 254                 $stats = MediaWikiServices::getInstance()->getStatsdDataFactory();
 255
 256                 $key = self::getStashKey( $title, self::getContentHash( $content ), $user );
 257                 $editInfo = $cache->get( $key );
 258                 if ( !is_object( $editInfo ) ) {
 259                         $start = microtime( true );
 260                         // We ignore user aborts and keep parsing. Block on any prior parsing
 261                         // so as to use its results and make use of the time spent parsing.
 262                         // Skip this logic if there no master connection in case this method
 263                         // is called on an HTTP GET request for some reason.
 264                         $lb = MediaWikiServices::getInstance()->getDBLoadBalancer();
 265                         $dbw = $lb->getAnyOpenConnection( $lb->getWriterIndex() );
 266                         if ( $dbw && $dbw->lock( $key, __METHOD__, 30 ) ) {
 267                                 $editInfo = $cache->get( $key );
 268                                 $dbw->unlock( $key, __METHOD__ );
 269                         }
 270
 271                         $timeMs = 1000 * max( 0, microtime( true ) - $start );
 272                         $stats->timing( 'editstash.lock_wait_time', $timeMs );
 273                 }
 274
 275                 if ( !is_object( $editInfo ) || !$editInfo->output ) {
 276                         $stats->increment( 'editstash.cache_misses.no_stash' );
 277                         $logger->debug( "Empty cache for key '$key' ('$title'); user '{$user->getName()}'." );
 278                         return false;
 279                 }
 280
 281                 $age = time() - wfTimestamp( TS_UNIX, $editInfo->output->getCacheTime() );
 282                 if ( $age <= self::PRESUME_FRESH_TTL_SEC ) {
 283                         // Assume nothing changed in this time
 284                         $stats->increment( 'editstash.cache_hits.presumed_fresh' );
 285                         $logger->debug( "Timestamp-based cache hit for key '$key' (age: $age sec)." );
 286                 } elseif ( isset( $editInfo->edits ) && $editInfo->edits === $user->getEditCount() ) {
 287                         // Logged-in user made no local upload/template edits in the meantime
 288                         $stats->increment( 'editstash.cache_hits.presumed_fresh' );
 289                         $logger->debug( "Edit count based cache hit for key '$key' (age: $age sec)." );
 290                 } elseif ( $user->isAnon()
 291                         && self::lastEditTime( $user ) < $editInfo->output->getCacheTime()
 292                 ) {
 293                         // Logged-out user made no local upload/template edits in the meantime
 294                         $stats->increment( 'editstash.cache_hits.presumed_fresh' );
 295                         $logger->debug( "Edit check based cache hit for key '$key' (age: $age sec)." );
 296                 } else {
 297                         // User may have changed included content
 298                         $editInfo = false;
 299                 }
 300
 301                 if ( !$editInfo ) {
 302                         $stats->increment( 'editstash.cache_misses.proven_stale' );
 303                         $logger->info( "Stale cache for key '$key'; old key with outside edits. (age: $age sec)" );
 304                 } elseif ( $editInfo->output->getFlag( 'vary-revision' ) ) {
 305                         // This can be used for the initial parse, e.g. for filters or doEditContent(),
 306                         // but a second parse will be triggered in doEditUpdates(). This is not optimal.
 307                         $logger->info( "Cache for key '$key' ('$title') has vary_revision." );
 308                 } elseif ( $editInfo->output->getFlag( 'vary-revision-id' ) ) {
 309                         // Similar to the above if we didn't guess the ID correctly.
 310                         $logger->info( "Cache for key '$key' ('$title') has vary_revision_id." );
 311                 }
 312
 313                 return $editInfo;
 314         }
 315
 316         /**
 317          * @param User $user
 318          * @return string|null TS_MW timestamp or null
 319          */
 320         private static function lastEditTime( User $user ) {
 321                 $time = wfGetDB( DB_REPLICA )->selectField(
 322                         'recentchanges',
 323                         'MAX(rc_timestamp)',
 324                         [ 'rc_user_text' => $user->getName() ],
 325                         __METHOD__
 326                 );
 327
 328                 return wfTimestampOrNull( TS_MW, $time );
 329         }
 330
 331         /**
 332          * Get hash of the content, factoring in model/format
 333          *
 334          * @param Content $content
 335          * @return string
 336          */
 337         private static function getContentHash( Content $content ) {
 338                 return sha1( implode( "\n", [
 339                         $content->getModel(),
 340                         $content->getDefaultFormat(),
 341                         $content->serialize( $content->getDefaultFormat() )
 342                 ] ) );
 343         }
 344
 345         /**
 346          * Get the temporary prepared edit stash key for a user
 347          *
 348          * This key can be used for caching prepared edits provided:
 349          *   - a) The $user was used for PST options
 350          *   - b) The parser output was made from the PST using cannonical matching options
 351          *
 352          * @param Title $title
 353          * @param string $contentHash Result of getContentHash()
 354          * @param User $user User to get parser options from
 355          * @return string
 356          */
 357         private static function getStashKey( Title $title, $contentHash, User $user ) {
 358                 return ObjectCache::getLocalClusterInstance()->makeKey(
 359                         'prepared-edit',
 360                         md5( $title->getPrefixedDBkey() ),
 361                         // Account for the edit model/text
 362                         $contentHash,
 363                         // Account for user name related variables like signatures
 364                         md5( $user->getId() . "\n" . $user->getName() )
 365                 );
 366         }
 367
 368         /**
 369          * Build a value to store in memcached based on the PST content and parser output
 370          *
 371          * This makes a simple version of WikiPage::prepareContentForEdit() as stash info
 372          *
 373          * @param Content $pstContent Pre-Save transformed content
 374          * @param ParserOutput $parserOutput
 375          * @param string $timestamp TS_MW
 376          * @param User $user
 377          * @return array (stash info array, TTL in seconds, info code) or (null, 0, info code)
 378          */
 379         private static function buildStashValue(
 380                 Content $pstContent, ParserOutput $parserOutput, $timestamp, User $user
 381         ) {
 382                 // If an item is renewed, mind the cache TTL determined by config and parser functions.
 383                 // Put an upper limit on the TTL for sanity to avoid extreme template/file staleness.
 384                 $since = time() - wfTimestamp( TS_UNIX, $parserOutput->getTimestamp() );
 385                 $ttl = min( $parserOutput->getCacheExpiry() - $since, self::MAX_CACHE_TTL );
 386                 if ( $ttl <= 0 ) {
 387                         return [ null, 0, 'no_ttl' ];
 388                 }
 389
 390                 // Only store what is actually needed
 391                 $stashInfo = (object)[
 392                         'pstContent' => $pstContent,
 393                         'output'     => $parserOutput,
 394                         'timestamp'  => $timestamp,
 395                         'edits'      => $user->getEditCount()
 396                 ];
 397
 398                 return [ $stashInfo, $ttl, 'ok' ];
 399         }
 400
 401         public function getAllowedParams() {
 402                 return [
 403                         'title' => [
 404                                 ApiBase::PARAM_TYPE => 'string',
 405                                 ApiBase::PARAM_REQUIRED => true
 406                         ],
 407                         'section' => [
 408                                 ApiBase::PARAM_TYPE => 'string',
 409                         ],
 410                         'sectiontitle' => [
 411                                 ApiBase::PARAM_TYPE => 'string'
 412                         ],
 413                         'text' => [
 414                                 ApiBase::PARAM_TYPE => 'text',
 415                                 ApiBase::PARAM_DFLT => null
 416                         ],
 417                         'stashedtexthash' => [
 418                                 ApiBase::PARAM_TYPE => 'string',
 419                                 ApiBase::PARAM_DFLT => null
 420                         ],
 421                         'summary' => [
 422                                 ApiBase::PARAM_TYPE => 'string',
 423                         ],
 424                         'contentmodel' => [
 425                                 ApiBase::PARAM_TYPE => ContentHandler::getContentModels(),
 426                                 ApiBase::PARAM_REQUIRED => true
 427                         ],
 428                         'contentformat' => [
 429                                 ApiBase::PARAM_TYPE => ContentHandler::getAllContentFormats(),
 430                                 ApiBase::PARAM_REQUIRED => true
 431                         ],
 432                         'baserevid' => [
 433                                 ApiBase::PARAM_TYPE => 'integer',
 434                                 ApiBase::PARAM_REQUIRED => true
 435                         ]
 436                 ];
 437         }
 438
 439         public function needsToken() {
 440                 return 'csrf';
 441         }
 442
 443         public function mustBePosted() {
 444                 return true;
 445         }
 446
 447         public function isWriteMode() {
 448                 return true;
 449         }
 450
 451         public function isInternal() {
 452                 return true;
 453         }
 454 }