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