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 /**
  23  * Prepare and edit in shared cache so that it can be reused on edit
  24  *
  25  * This endpoint can be called via AJAX as the user focuses on the edit
  26  * summary box. By the time of submission, the parse may have already
  27  * finished, and can be immediately used on page save. Certain parser
  28  * functions like {{REVISIONID}} or {{CURRENTTIME}} may cause the cache
  29  * to not be used on edit. Template and files used are check for changes
  30  * since the output was generated. The cache TTL is also kept low for sanity.
  31  *
  32  * @ingroup API
  33  * @since 1.25
  34  */
  35 class ApiStashEdit extends ApiBase {
  36         const ERROR_NONE = 'stashed';
  37         const ERROR_PARSE = 'error_parse';
  38         const ERROR_CACHE = 'error_cache';
  39         const ERROR_UNCACHEABLE = 'uncacheable';
  40
  41         public function execute() {
  42                 global $wgMemc;
  43
  44                 $user = $this->getUser();
  45                 $params = $this->extractRequestParams();
  46
  47                 $page = $this->getTitleOrPageId( $params );
  48                 $title = $page->getTitle();
  49
  50                 if ( !ContentHandler::getForModelID( $params['contentmodel'] )
  51                         ->isSupportedFormat( $params['contentformat'] )
  52                 ) {
  53                         $this->dieUsage( "Unsupported content model/format", 'badmodelformat' );
  54                 }
  55
  56                 // Trim and fix newlines so the key SHA1's match (see RequestContext::getText())
  57                 $text = rtrim( str_replace( "\r\n", "\n", $params['text'] ) );
  58                 $textContent = ContentHandler::makeContent(
  59                         $text, $title, $params['contentmodel'], $params['contentformat'] );
  60
  61                 $page = WikiPage::factory( $title );
  62                 if ( $page->exists() ) {
  63                         // Page exists: get the merged content with the proposed change
  64                         $baseRev = Revision::newFromPageId( $page->getId(), $params['baserevid'] );
  65                         if ( !$baseRev ) {
  66                                 $this->dieUsage( "No revision ID {$params['baserevid']}", 'missingrev' );
  67                         }
  68                         $currentRev = $page->getRevision();
  69                         if ( !$currentRev ) {
  70                                 $this->dieUsage( "No current revision of page ID {$page->getId()}", 'missingrev' );
  71                         }
  72                         // Merge in the new version of the section to get the proposed version
  73                         $editContent = $page->replaceSectionAtRev(
  74                                 $params['section'],
  75                                 $textContent,
  76                                 $params['sectiontitle'],
  77                                 $baseRev->getId()
  78                         );
  79                         if ( !$editContent ) {
  80                                 $this->dieUsage( "Could not merge updated section.", 'replacefailed' );
  81                         }
  82                         if ( $currentRev->getId() == $baseRev->getId() ) {
  83                                 // Base revision was still the latest; nothing to merge
  84                                 $content = $editContent;
  85                         } else {
  86                                 // Merge the edit into the current version
  87                                 $baseContent = $baseRev->getContent();
  88                                 $currentContent = $currentRev->getContent();
  89                                 if ( !$baseContent || !$currentContent ) {
  90                                         $this->dieUsage( "Missing content for page ID {$page->getId()}", 'missingrev' );
  91                                 }
  92                                 $handler = ContentHandler::getForModelID( $baseContent->getModel() );
  93                                 $content = $handler->merge3( $baseContent, $editContent, $currentContent );
  94                         }
  95                 } else {
  96                         // New pages: use the user-provided content model
  97                         $content = $textContent;
  98                 }
  99
 100                 if ( !$content ) { // merge3() failed
 101                         $this->getResult()->addValue( null,
 102                                 $this->getModuleName(), array( 'status' => 'editconflict' ) );
 103                         return;
 104                 }
 105
 106                 // The user will abort the AJAX request by pressing "save", so ignore that
 107                 ignore_user_abort( true );
 108
 109                 // Get a key based on the source text, format, and user preferences
 110                 $key = self::getStashKey( $title, $content, $user );
 111                 // De-duplicate requests on the same key
 112                 if ( $user->pingLimiter( 'stashedit' ) ) {
 113                         $status = 'ratelimited';
 114                 } elseif ( $wgMemc->lock( $key, 0, 30 ) ) {
 115                         $unlocker = new ScopedCallback( function() use ( $key ) {
 116                                 global $wgMemc;
 117                                 $wgMemc->unlock( $key );
 118                         } );
 119                         $status = self::parseAndStash( $page, $content, $user );
 120                 } else {
 121                         $status = 'busy';
 122                 }
 123
 124                 $this->getResult()->addValue( null, $this->getModuleName(), array( 'status' => $status ) );
 125         }
 126
 127         /**
 128          * @param WikiPage $page
 129          * @param Content $content
 130          * @param User $user
 131          * @return integer ApiStashEdit::ERROR_* constant
 132          * @since 1.25
 133          */
 134         public static function parseAndStash( WikiPage $page, Content $content, User $user ) {
 135                 global $wgMemc;
 136
 137                 $format = $content->getDefaultFormat();
 138                 $editInfo = $page->prepareContentForEdit( $content, null, $user, $format, false );
 139
 140                 if ( $editInfo && $editInfo->output ) {
 141                         $key = self::getStashKey( $page->getTitle(), $content, $user );
 142
 143                         list( $stashInfo, $ttl ) = self::buildStashValue(
 144                                 $editInfo->pstContent, $editInfo->output, $editInfo->timestamp
 145                         );
 146
 147                         if ( $stashInfo ) {
 148                                 $ok = $wgMemc->set( $key, $stashInfo, $ttl );
 149                                 if ( $ok ) {
 150                                         wfDebugLog( 'StashEdit', "Cached parser output for key '$key'." );
 151                                         return self::ERROR_NONE;
 152                                 } else {
 153                                         wfDebugLog( 'StashEdit', "Failed to cache parser output for key '$key'." );
 154                                         return self::ERROR_CACHE;
 155                                 }
 156                         } else {
 157                                 wfDebugLog( 'StashEdit', "Uncacheable parser output for key '$key'." );
 158                                 return self::ERROR_UNCACHEABLE;
 159                         }
 160                 }
 161
 162                 return self::ERROR_PARSE;
 163         }
 164
 165         /**
 166          * Attempt to cache PST content and corresponding parser output in passing
 167          *
 168          * This method can be called when the output was already generated for other
 169          * reasons. Parsing should not be done just to call this method, however.
 170          * $pstOpts must be that of the user doing the edit preview. If $pOpts does
 171          * not match the options of WikiPage::makeParserOptions( 'canonical' ), this
 172          * will do nothing. Provided the values are cacheable, they will be stored
 173          * in memcached so that final edit submission might make use of them.
 174          *
 175          * @param Article|WikiPage $page Page title
 176          * @param Content $content Proposed page content
 177          * @param Content $pstContent The result of preSaveTransform() on $content
 178          * @param ParserOutput $pOut The result of getParserOutput() on $pstContent
 179          * @param ParserOptions $pstOpts Options for $pstContent (MUST be for prospective author)
 180          * @param ParserOptions $pOpts Options for $pOut
 181          * @param string $timestamp TS_MW timestamp of parser output generation
 182          * @return bool Success
 183          */
 184         public static function stashEditFromPreview(
 185                 Page $page, Content $content, Content $pstContent, ParserOutput $pOut,
 186                 ParserOptions $pstOpts, ParserOptions $pOpts, $timestamp
 187         ) {
 188                 global $wgMemc;
 189
 190                 // getIsPreview() controls parser function behavior that references things
 191                 // like user/revision that don't exists yet. The user/text should already
 192                 // be set correctly by callers, just double check the preview flag.
 193                 if ( !$pOpts->getIsPreview() ) {
 194                         return false; // sanity
 195                 } elseif ( $pOpts->getIsSectionPreview() ) {
 196                         return false; // short-circuit (need the full content)
 197                 }
 198
 199                 // PST parser options are for the user (handles signatures, etc...)
 200                 $user = $pstOpts->getUser();
 201                 // Get a key based on the source text, format, and user preferences
 202                 $key = self::getStashKey( $page->getTitle(), $content, $user );
 203
 204                 // Parser output options must match cannonical options.
 205                 // Treat some options as matching that are different but don't matter.
 206                 $canonicalPOpts = $page->makeParserOptions( 'canonical' );
 207                 $canonicalPOpts->setIsPreview( true ); // force match
 208                 $canonicalPOpts->setTimestamp( $pOpts->getTimestamp() ); // force match
 209                 if ( !$pOpts->matches( $canonicalPOpts ) ) {
 210                         wfDebugLog( 'StashEdit', "Uncacheable preview output for key '$key' (options)." );
 211                         return false;
 212                 }
 213
 214                 // Build a value to cache with a proper TTL
 215                 list( $stashInfo, $ttl ) = self::buildStashValue( $pstContent, $pOut, $timestamp );
 216                 if ( !$stashInfo ) {
 217                         wfDebugLog( 'StashEdit', "Uncacheable parser output for key '$key' (rev/TTL)." );
 218                         return false;
 219                 }
 220
 221                 $ok = $wgMemc->set( $key, $stashInfo, $ttl );
 222                 if ( !$ok ) {
 223                         wfDebugLog( 'StashEdit', "Failed to cache preview parser output for key '$key'." );
 224                 } else {
 225                         wfDebugLog( 'StashEdit', "Cached preview output for key '$key'." );
 226                 }
 227
 228                 return $ok;
 229         }
 230
 231         /**
 232          * Check that a prepared edit is in cache and still up-to-date
 233          *
 234          * This method blocks if the prepared edit is already being rendered,
 235          * waiting until rendering finishes before doing final validity checks.
 236          *
 237          * The cache is rejected if template or file changes are detected.
 238          * Note that foreign template or file transclusions are not checked.
 239          *
 240          * The result is a map (pstContent,output,timestamp) with fields
 241          * extracted directly from WikiPage::prepareContentForEdit().
 242          *
 243          * @param Title $title
 244          * @param Content $content
 245          * @param User $user User to get parser options from
 246          * @return stdClass|bool Returns false on cache miss
 247          */
 248         public static function checkCache( Title $title, Content $content, User $user ) {
 249                 global $wgMemc;
 250
 251                 $key = self::getStashKey( $title, $content, $user );
 252                 $editInfo = $wgMemc->get( $key );
 253                 if ( !is_object( $editInfo ) ) {
 254                         $start = microtime( true );
 255                         // We ignore user aborts and keep parsing. Block on any prior parsing
 256                         // so as to use it's results and make use of the time spent parsing.
 257                         if ( $wgMemc->lock( $key, 30, 30 ) ) {
 258                                 $editInfo = $wgMemc->get( $key );
 259                                 $wgMemc->unlock( $key );
 260                         }
 261                         $sec = microtime( true ) - $start;
 262                         if ( $sec > .01 ) {
 263                                 wfDebugLog( 'StashEdit', "Waited $sec seconds on '$key'." );
 264                         }
 265                 }
 266
 267                 if ( !is_object( $editInfo ) || !$editInfo->output ) {
 268                         wfDebugLog( 'StashEdit', "No cache value for key '$key'." );
 269                         return false;
 270                 }
 271
 272                 $time = wfTimestamp( TS_UNIX, $editInfo->output->getTimestamp() );
 273                 if ( ( time() - $time ) <= 3 ) {
 274                         wfDebugLog( 'StashEdit', "Timestamp-based cache hit for key '$key'." );
 275                         return $editInfo; // assume nothing changed
 276                 }
 277
 278                 $dbr = wfGetDB( DB_SLAVE );
 279
 280                 $templates = array(); // conditions to find changes/creations
 281                 $templateUses = 0; // expected existing templates
 282                 foreach ( $editInfo->output->getTemplateIds() as $ns => $stuff ) {
 283                         foreach ( $stuff as $dbkey => $revId ) {
 284                                 $templates[(string)$ns][$dbkey] = (int)$revId;
 285                                 ++$templateUses;
 286                         }
 287                 }
 288                 // Check that no templates used in the output changed...
 289                 if ( count( $templates ) ) {
 290                         $res = $dbr->select(
 291                                 'page',
 292                                 array( 'ns' => 'page_namespace', 'dbk' => 'page_title', 'page_latest' ),
 293                                 $dbr->makeWhereFrom2d( $templates, 'page_namespace', 'page_title' ),
 294                                 __METHOD__
 295                         );
 296                         $changed = false;
 297                         foreach ( $res as $row ) {
 298                                 $changed = $changed || ( $row->page_latest != $templates[$row->ns][$row->dbk] );
 299                         }
 300
 301                         if ( $changed || $res->numRows() != $templateUses ) {
 302                                 wfDebugLog( 'StashEdit', "Stale cache for key '$key'; template changed." );
 303                                 return false;
 304                         }
 305                 }
 306
 307                 $files = array(); // conditions to find changes/creations
 308                 foreach ( $editInfo->output->getFileSearchOptions() as $name => $options ) {
 309                         $files[$name] = (string)$options['sha1'];
 310                 }
 311                 // Check that no files used in the output changed...
 312                 if ( count( $files ) ) {
 313                         $res = $dbr->select(
 314                                 'image',
 315                                 array( 'name' => 'img_name', 'img_sha1' ),
 316                                 array( 'img_name' => array_keys( $files ) ),
 317                                 __METHOD__
 318                         );
 319                         $changed = false;
 320                         foreach ( $res as $row ) {
 321                                 $changed = $changed || ( $row->img_sha1 != $files[$row->name] );
 322                         }
 323
 324                         if ( $changed || $res->numRows() != count( $files ) ) {
 325                                 wfDebugLog( 'StashEdit', "Stale cache for key '$key'; file changed." );
 326                                 return false;
 327                         }
 328                 }
 329
 330                 wfDebugLog( 'StashEdit', "Cache hit for key '$key'." );
 331
 332                 return $editInfo;
 333         }
 334
 335         /**
 336          * Get the temporary prepared edit stash key for a user
 337          *
 338          * This key can be used for caching prepared edits provided:
 339          *   - a) The $user was used for PST options
 340          *   - b) The parser output was made from the PST using cannonical matching options
 341          *
 342          * @param Title $title
 343          * @param Content $content
 344          * @param User $user User to get parser options from
 345          * @return string
 346          */
 347         protected static function getStashKey( Title $title, Content $content, User $user ) {
 348                 $hash = sha1( implode( ':', array(
 349                         $content->getModel(),
 350                         $content->getDefaultFormat(),
 351                         sha1( $content->serialize( $content->getDefaultFormat() ) ),
 352                         $user->getId() ?: md5( $user->getName() ), // account for user parser options
 353                         $user->getId() ? $user->getDBTouched() : '-' // handle preference change races
 354                 ) ) );
 355
 356                 return wfMemcKey( 'prepared-edit', md5( $title->getPrefixedDBkey() ), $hash );
 357         }
 358
 359         /**
 360          * Build a value to store in memcached based on the PST content and parser output
 361          *
 362          * This makes a simple version of WikiPage::prepareContentForEdit() as stash info
 363          *
 364          * @param Content $pstContent
 365          * @param ParserOutput $parserOutput
 366          * @param string $timestamp TS_MW
 367          * @return array (stash info array, TTL in seconds) or (null, 0)
 368          */
 369         protected static function buildStashValue(
 370                 Content $pstContent, ParserOutput $parserOutput, $timestamp
 371         ) {
 372                 // If an item is renewed, mind the cache TTL determined by config and parser functions
 373                 $since = time() - wfTimestamp( TS_UNIX, $parserOutput->getTimestamp() );
 374                 $ttl = min( $parserOutput->getCacheExpiry() - $since, 5 * 60 );
 375
 376                 if ( $ttl > 0 && !$parserOutput->getFlag( 'vary-revision' ) ) {
 377                         // Only store what is actually needed
 378                         $stashInfo = (object)array(
 379                                 'pstContent' => $pstContent,
 380                                 'output'     => $parserOutput,
 381                                 'timestamp'  => $timestamp
 382                         );
 383                         return array( $stashInfo, $ttl );
 384                 }
 385
 386                 return array( null, 0 );
 387         }
 388
 389         public function getAllowedParams() {
 390                 return array(
 391                         'title' => array(
 392                                 ApiBase::PARAM_TYPE => 'string',
 393                                 ApiBase::PARAM_REQUIRED => true
 394                         ),
 395                         'section' => array(
 396                                 ApiBase::PARAM_TYPE => 'string',
 397                         ),
 398                         'sectiontitle' => array(
 399                                 ApiBase::PARAM_TYPE => 'string'
 400                         ),
 401                         'text' => array(
 402                                 ApiBase::PARAM_TYPE => 'text',
 403                                 ApiBase::PARAM_REQUIRED => true
 404                         ),
 405                         'contentmodel' => array(
 406                                 ApiBase::PARAM_TYPE => ContentHandler::getContentModels(),
 407                                 ApiBase::PARAM_REQUIRED => true
 408                         ),
 409                         'contentformat' => array(
 410                                 ApiBase::PARAM_TYPE => ContentHandler::getAllContentFormats(),
 411                                 ApiBase::PARAM_REQUIRED => true
 412                         ),
 413                         'baserevid' => array(
 414                                 ApiBase::PARAM_TYPE => 'integer',
 415                                 ApiBase::PARAM_REQUIRED => true
 416                         )
 417                 );
 418         }
 419
 420         function needsToken() {
 421                 return 'csrf';
 422         }
 423
 424         function mustBePosted() {
 425                 return true;
 426         }
 427
 428         function isInternal() {
 429                 return true;
 430         }
 431 }