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  */
  20
  21 use MediaWiki\Logger\LoggerFactory;
  22 use MediaWiki\MediaWikiServices;
  23 use Wikimedia\ScopedCallback;
  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         const MAX_SIGNATURE_TTL = 60;
  48
  49         public function execute() {
  50                 $user = $this->getUser();
  51                 $params = $this->extractRequestParams();
  52
  53                 if ( $user->isBot() ) { // sanity
  54                         $this->dieWithError( 'apierror-botsnotsupported' );
  55                 }
  56
  57                 $cache = ObjectCache::getLocalClusterInstance();
  58                 $page = $this->getTitleOrPageId( $params );
  59                 $title = $page->getTitle();
  60
  61                 if ( !ContentHandler::getForModelID( $params['contentmodel'] )
  62                         ->isSupportedFormat( $params['contentformat'] )
  63                 ) {
  64                         $this->dieWithError(
  65                                 [ 'apierror-badformat-generic', $params['contentformat'], $params['contentmodel'] ],
  66                                 'badmodelformat'
  67                         );
  68                 }
  69
  70                 $this->requireAtLeastOneParameter( $params, 'stashedtexthash', 'text' );
  71
  72                 $text = null;
  73                 $textHash = null;
  74                 if ( strlen( $params['stashedtexthash'] ) ) {
  75                         // Load from cache since the client indicates the text is the same as last stash
  76                         $textHash = $params['stashedtexthash'];
  77                         if ( !preg_match( '/^[0-9a-f]{40}$/', $textHash ) ) {
  78                                 $this->dieWithError( 'apierror-stashedit-missingtext', 'missingtext' );
  79                         }
  80                         $textKey = $cache->makeKey( 'stashedit', 'text', $textHash );
  81                         $text = $cache->get( $textKey );
  82                         if ( !is_string( $text ) ) {
  83                                 $this->dieWithError( 'apierror-stashedit-missingtext', 'missingtext' );
  84                         }
  85                 } elseif ( $params['text'] !== null ) {
  86                         // Trim and fix newlines so the key SHA1's match (see WebRequest::getText())
  87                         $text = rtrim( str_replace( "\r\n", "\n", $params['text'] ) );
  88                         $textHash = sha1( $text );
  89                 } else {
  90                         $this->dieWithError( [
  91                                 'apierror-missingparam-at-least-one-of',
  92                                 Message::listParam( [ '<var>stashedtexthash</var>', '<var>text</var>' ] ),
  93                                 2,
  94                         ], 'missingparam' );
  95                 }
  96
  97                 $textContent = ContentHandler::makeContent(
  98                         $text, $title, $params['contentmodel'], $params['contentformat'] );
  99
 100                 $page = WikiPage::factory( $title );
 101                 if ( $page->exists() ) {
 102                         // Page exists: get the merged content with the proposed change
 103                         $baseRev = Revision::newFromPageId( $page->getId(), $params['baserevid'] );
 104                         if ( !$baseRev ) {
 105                                 $this->dieWithError( [ 'apierror-nosuchrevid', $params['baserevid'] ] );
 106                         }
 107                         $currentRev = $page->getRevision();
 108                         if ( !$currentRev ) {
 109                                 $this->dieWithError( [ 'apierror-missingrev-pageid', $page->getId() ], 'missingrev' );
 110                         }
 111                         // Merge in the new version of the section to get the proposed version
 112                         $editContent = $page->replaceSectionAtRev(
 113                                 $params['section'],
 114                                 $textContent,
 115                                 $params['sectiontitle'],
 116                                 $baseRev->getId()
 117                         );
 118                         if ( !$editContent ) {
 119                                 $this->dieWithError( 'apierror-sectionreplacefailed', 'replacefailed' );
 120                         }
 121                         if ( $currentRev->getId() == $baseRev->getId() ) {
 122                                 // Base revision was still the latest; nothing to merge
 123                                 $content = $editContent;
 124                         } else {
 125                                 // Merge the edit into the current version
 126                                 $baseContent = $baseRev->getContent();
 127                                 $currentContent = $currentRev->getContent();
 128                                 if ( !$baseContent || !$currentContent ) {
 129                                         $this->dieWithError( [ 'apierror-missingcontent-pageid', $page->getId() ], 'missingrev' );
 130                                 }
 131                                 $handler = ContentHandler::getForModelID( $baseContent->getModel() );
 132                                 $content = $handler->merge3( $baseContent, $editContent, $currentContent );
 133                         }
 134                 } else {
 135                         // New pages: use the user-provided content model
 136                         $content = $textContent;
 137                 }
 138
 139                 if ( !$content ) { // merge3() failed
 140                         $this->getResult()->addValue( null,
 141                                 $this->getModuleName(), [ 'status' => 'editconflict' ] );
 142                         return;
 143                 }
 144
 145                 // The user will abort the AJAX request by pressing "save", so ignore that
 146                 ignore_user_abort( true );
 147
 148                 if ( $user->pingLimiter( 'stashedit' ) ) {
 149                         $status = 'ratelimited';
 150                 } else {
 151                         $status = self::parseAndStash( $page, $content, $user, $params['summary'] );
 152                         $textKey = $cache->makeKey( 'stashedit', 'text', $textHash );
 153                         $cache->set( $textKey, $text, self::MAX_CACHE_TTL );
 154                 }
 155
 156                 $stats = MediaWikiServices::getInstance()->getStatsdDataFactory();
 157                 $stats->increment( "editstash.cache_stores.$status" );
 158
 159                 $this->getResult()->addValue(
 160                         null,
 161                         $this->getModuleName(),
 162                         [
 163                                 'status' => $status,
 164                                 'texthash' => $textHash
 165                         ]
 166                 );
 167         }
 168
 169         /**
 170          * @param WikiPage $page
 171          * @param Content $content Edit content
 172          * @param User $user
 173          * @param string $summary Edit summary
 174          * @return string ApiStashEdit::ERROR_* constant
 175          * @since 1.25
 176          */
 177         public static function parseAndStash( WikiPage $page, Content $content, User $user, $summary ) {
 178                 $cache = ObjectCache::getLocalClusterInstance();
 179                 $logger = LoggerFactory::getInstance( 'StashEdit' );
 180
 181                 $title = $page->getTitle();
 182                 $key = self::getStashKey( $title, self::getContentHash( $content ), $user );
 183
 184                 // Use the master DB for fast blocking locks
 185                 $dbw = wfGetDB( DB_MASTER );
 186                 if ( !$dbw->lock( $key, __METHOD__, 1 ) ) {
 187                         // De-duplicate requests on the same key
 188                         return self::ERROR_BUSY;
 189                 }
 190                 /** @noinspection PhpUnusedLocalVariableInspection */
 191                 $unlocker = new ScopedCallback( function () use ( $dbw, $key ) {
 192                         $dbw->unlock( $key, __METHOD__ );
 193                 } );
 194
 195                 $cutoffTime = time() - self::PRESUME_FRESH_TTL_SEC;
 196
 197                 // Reuse any freshly build matching edit stash cache
 198                 $editInfo = $cache->get( $key );
 199                 if ( $editInfo && wfTimestamp( TS_UNIX, $editInfo->timestamp ) >= $cutoffTime ) {
 200                         $alreadyCached = true;
 201                 } else {
 202                         $format = $content->getDefaultFormat();
 203                         $editInfo = $page->prepareContentForEdit( $content, null, $user, $format, false );
 204                         $alreadyCached = false;
 205                 }
 206
 207                 if ( $editInfo && $editInfo->output ) {
 208                         // Let extensions add ParserOutput metadata or warm other caches
 209                         Hooks::run( 'ParserOutputStashForEdit',
 210                                 [ $page, $content, $editInfo->output, $summary, $user ] );
 211
 212                         if ( $alreadyCached ) {
 213                                 $logger->debug( "Already cached parser output for key '$key' ('$title')." );
 214                                 return self::ERROR_NONE;
 215                         }
 216
 217                         list( $stashInfo, $ttl, $code ) = self::buildStashValue(
 218                                 $editInfo->pstContent,
 219                                 $editInfo->output,
 220                                 $editInfo->timestamp,
 221                                 $user
 222                         );
 223
 224                         if ( $stashInfo ) {
 225                                 $ok = $cache->set( $key, $stashInfo, $ttl );
 226                                 if ( $ok ) {
 227                                         $logger->debug( "Cached parser output for key '$key' ('$title')." );
 228                                         return self::ERROR_NONE;
 229                                 } else {
 230                                         $logger->error( "Failed to cache parser output for key '$key' ('$title')." );
 231                                         return self::ERROR_CACHE;
 232                                 }
 233                         } else {
 234                                 $logger->info( "Uncacheable parser output for key '$key' ('$title') [$code]." );
 235                                 return self::ERROR_UNCACHEABLE;
 236                         }
 237                 }
 238
 239                 return self::ERROR_PARSE;
 240         }
 241
 242         /**
 243          * Check that a prepared edit is in cache and still up-to-date
 244          *
 245          * This method blocks if the prepared edit is already being rendered,
 246          * waiting until rendering finishes before doing final validity checks.
 247          *
 248          * The cache is rejected if template or file changes are detected.
 249          * Note that foreign template or file transclusions are not checked.
 250          *
 251          * The result is a map (pstContent,output,timestamp) with fields
 252          * extracted directly from WikiPage::prepareContentForEdit().
 253          *
 254          * @param Title $title
 255          * @param Content $content
 256          * @param User $user User to get parser options from
 257          * @return stdClass|bool Returns false on cache miss
 258          */
 259         public static function checkCache( Title $title, Content $content, User $user ) {
 260                 if ( $user->isBot() ) {
 261                         return false; // bots never stash - don't pollute stats
 262                 }
 263
 264                 $cache = ObjectCache::getLocalClusterInstance();
 265                 $logger = LoggerFactory::getInstance( 'StashEdit' );
 266                 $stats = MediaWikiServices::getInstance()->getStatsdDataFactory();
 267
 268                 $key = self::getStashKey( $title, self::getContentHash( $content ), $user );
 269                 $editInfo = $cache->get( $key );
 270                 if ( !is_object( $editInfo ) ) {
 271                         $start = microtime( true );
 272                         // We ignore user aborts and keep parsing. Block on any prior parsing
 273                         // so as to use its results and make use of the time spent parsing.
 274                         // Skip this logic if there no master connection in case this method
 275                         // is called on an HTTP GET request for some reason.
 276                         $lb = MediaWikiServices::getInstance()->getDBLoadBalancer();
 277                         $dbw = $lb->getAnyOpenConnection( $lb->getWriterIndex() );
 278                         if ( $dbw && $dbw->lock( $key, __METHOD__, 30 ) ) {
 279                                 $editInfo = $cache->get( $key );
 280                                 $dbw->unlock( $key, __METHOD__ );
 281                         }
 282
 283                         $timeMs = 1000 * max( 0, microtime( true ) - $start );
 284                         $stats->timing( 'editstash.lock_wait_time', $timeMs );
 285                 }
 286
 287                 if ( !is_object( $editInfo ) || !$editInfo->output ) {
 288                         $stats->increment( 'editstash.cache_misses.no_stash' );
 289                         $logger->debug( "Empty cache for key '$key' ('$title'); user '{$user->getName()}'." );
 290                         return false;
 291                 }
 292
 293                 $age = time() - wfTimestamp( TS_UNIX, $editInfo->output->getCacheTime() );
 294                 if ( $age <= self::PRESUME_FRESH_TTL_SEC ) {
 295                         // Assume nothing changed in this time
 296                         $stats->increment( 'editstash.cache_hits.presumed_fresh' );
 297                         $logger->debug( "Timestamp-based cache hit for key '$key' (age: $age sec)." );
 298                 } elseif ( isset( $editInfo->edits ) && $editInfo->edits === $user->getEditCount() ) {
 299                         // Logged-in user made no local upload/template edits in the meantime
 300                         $stats->increment( 'editstash.cache_hits.presumed_fresh' );
 301                         $logger->debug( "Edit count based cache hit for key '$key' (age: $age sec)." );
 302                 } elseif ( $user->isAnon()
 303                         && self::lastEditTime( $user ) < $editInfo->output->getCacheTime()
 304                 ) {
 305                         // Logged-out user made no local upload/template edits in the meantime
 306                         $stats->increment( 'editstash.cache_hits.presumed_fresh' );
 307                         $logger->debug( "Edit check based cache hit for key '$key' (age: $age sec)." );
 308                 } else {
 309                         // User may have changed included content
 310                         $editInfo = false;
 311                 }
 312
 313                 if ( !$editInfo ) {
 314                         $stats->increment( 'editstash.cache_misses.proven_stale' );
 315                         $logger->info( "Stale cache for key '$key'; old key with outside edits. (age: $age sec)" );
 316                 } elseif ( $editInfo->output->getFlag( 'vary-revision' ) ) {
 317                         // This can be used for the initial parse, e.g. for filters or doEditContent(),
 318                         // but a second parse will be triggered in doEditUpdates(). This is not optimal.
 319                         $logger->info( "Cache for key '$key' ('$title') has vary_revision." );
 320                 } elseif ( $editInfo->output->getFlag( 'vary-revision-id' ) ) {
 321                         // Similar to the above if we didn't guess the ID correctly.
 322                         $logger->info( "Cache for key '$key' ('$title') has vary_revision_id." );
 323                 }
 324
 325                 return $editInfo;
 326         }
 327
 328         /**
 329          * @param User $user
 330          * @return string|null TS_MW timestamp or null
 331          */
 332         private static function lastEditTime( User $user ) {
 333                 $time = wfGetDB( DB_REPLICA )->selectField(
 334                         'recentchanges',
 335                         'MAX(rc_timestamp)',
 336                         [ 'rc_user_text' => $user->getName() ],
 337                         __METHOD__
 338                 );
 339
 340                 return wfTimestampOrNull( TS_MW, $time );
 341         }
 342
 343         /**
 344          * Get hash of the content, factoring in model/format
 345          *
 346          * @param Content $content
 347          * @return string
 348          */
 349         private static function getContentHash( Content $content ) {
 350                 return sha1( implode( "\n", [
 351                         $content->getModel(),
 352                         $content->getDefaultFormat(),
 353                         $content->serialize( $content->getDefaultFormat() )
 354                 ] ) );
 355         }
 356
 357         /**
 358          * Get the temporary prepared edit stash key for a user
 359          *
 360          * This key can be used for caching prepared edits provided:
 361          *   - a) The $user was used for PST options
 362          *   - b) The parser output was made from the PST using cannonical matching options
 363          *
 364          * @param Title $title
 365          * @param string $contentHash Result of getContentHash()
 366          * @param User $user User to get parser options from
 367          * @return string
 368          */
 369         private static function getStashKey( Title $title, $contentHash, User $user ) {
 370                 return ObjectCache::getLocalClusterInstance()->makeKey(
 371                         'prepared-edit',
 372                         md5( $title->getPrefixedDBkey() ),
 373                         // Account for the edit model/text
 374                         $contentHash,
 375                         // Account for user name related variables like signatures
 376                         md5( $user->getId() . "\n" . $user->getName() )
 377                 );
 378         }
 379
 380         /**
 381          * Build a value to store in memcached based on the PST content and parser output
 382          *
 383          * This makes a simple version of WikiPage::prepareContentForEdit() as stash info
 384          *
 385          * @param Content $pstContent Pre-Save transformed content
 386          * @param ParserOutput $parserOutput
 387          * @param string $timestamp TS_MW
 388          * @param User $user
 389          * @return array (stash info array, TTL in seconds, info code) or (null, 0, info code)
 390          */
 391         private static function buildStashValue(
 392                 Content $pstContent, ParserOutput $parserOutput, $timestamp, User $user
 393         ) {
 394                 // If an item is renewed, mind the cache TTL determined by config and parser functions.
 395                 // Put an upper limit on the TTL for sanity to avoid extreme template/file staleness.
 396                 $since = time() - wfTimestamp( TS_UNIX, $parserOutput->getTimestamp() );
 397                 $ttl = min( $parserOutput->getCacheExpiry() - $since, self::MAX_CACHE_TTL );
 398
 399                 // Avoid extremely stale user signature timestamps (T84843)
 400                 if ( $parserOutput->getFlag( 'user-signature' ) ) {
 401                         $ttl = min( $ttl, self::MAX_SIGNATURE_TTL );
 402                 }
 403
 404                 if ( $ttl <= 0 ) {
 405                         return [ null, 0, 'no_ttl' ];
 406                 }
 407
 408                 // Only store what is actually needed
 409                 $stashInfo = (object)[
 410                         'pstContent' => $pstContent,
 411                         'output'     => $parserOutput,
 412                         'timestamp'  => $timestamp,
 413                         'edits'      => $user->getEditCount()
 414                 ];
 415
 416                 return [ $stashInfo, $ttl, 'ok' ];
 417         }
 418
 419         public function getAllowedParams() {
 420                 return [
 421                         'title' => [
 422                                 ApiBase::PARAM_TYPE => 'string',
 423                                 ApiBase::PARAM_REQUIRED => true
 424                         ],
 425                         'section' => [
 426                                 ApiBase::PARAM_TYPE => 'string',
 427                         ],
 428                         'sectiontitle' => [
 429                                 ApiBase::PARAM_TYPE => 'string'
 430                         ],
 431                         'text' => [
 432                                 ApiBase::PARAM_TYPE => 'text',
 433                                 ApiBase::PARAM_DFLT => null
 434                         ],
 435                         'stashedtexthash' => [
 436                                 ApiBase::PARAM_TYPE => 'string',
 437                                 ApiBase::PARAM_DFLT => null
 438                         ],
 439                         'summary' => [
 440                                 ApiBase::PARAM_TYPE => 'string',
 441                         ],
 442                         'contentmodel' => [
 443                                 ApiBase::PARAM_TYPE => ContentHandler::getContentModels(),
 444                                 ApiBase::PARAM_REQUIRED => true
 445                         ],
 446                         'contentformat' => [
 447                                 ApiBase::PARAM_TYPE => ContentHandler::getAllContentFormats(),
 448                                 ApiBase::PARAM_REQUIRED => true
 449                         ],
 450                         'baserevid' => [
 451                                 ApiBase::PARAM_TYPE => 'integer',
 452                                 ApiBase::PARAM_REQUIRED => true
 453                         ]
 454                 ];
 455         }
 456
 457         public function needsToken() {
 458                 return 'csrf';
 459         }
 460
 461         public function mustBePosted() {
 462                 return true;
 463         }
 464
 465         public function isWriteMode() {
 466                 return true;
 467         }
 468
 469         public function isInternal() {
 470                 return true;
 471         }
 472 }