maintenance/populateContentTables.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  * @ingroup Maintenance
  20  */
  21
  22 use MediaWiki\MediaWikiServices;
  23 use MediaWiki\Revision\SlotRecord;
  24 use MediaWiki\Storage\BlobStore;
  25 use MediaWiki\Storage\NameTableStore;
  26 use MediaWiki\Storage\SqlBlobStore;
  27 use Wikimedia\Assert\Assert;
  28 use Wikimedia\Rdbms\IDatabase;
  29 use Wikimedia\Rdbms\IResultWrapper;
  30
  31 require_once __DIR__ . '/Maintenance.php';
  32
  33 /**
  34  * Populate the content and slot tables.
  35  * @since 1.32
  36  */
  37 class PopulateContentTables extends Maintenance {
  38
  39         /** @var IDatabase */
  40         private $dbw;
  41
  42         /** @var NameTableStore */
  43         private $contentModelStore;
  44
  45         /** @var BlobStore */
  46         private $blobStore;
  47
  48         /** @var int */
  49         private $mainRoleId;
  50
  51         /** @var array|null Map "{$modelId}:{$address}" to content_id */
  52         private $contentRowMap = null;
  53
  54         private $count = 0, $totalCount = 0;
  55
  56         public function __construct() {
  57                 parent::__construct();
  58
  59                 $this->addDescription( 'Populate content and slot tables' );
  60                 $this->addOption( 'table', 'revision or archive table, or `all` to populate both', false,
  61                         true );
  62                 $this->addOption( 'reuse-content',
  63                         'Reuse content table rows when the address and model are the same. '
  64                         . 'This will increase the script\'s time and memory usage, perhaps significantly.',
  65                         false, false );
  66                 $this->addOption( 'start-revision', 'The rev_id to start at', false, true );
  67                 $this->addOption( 'start-archive', 'The ar_rev_id to start at', false, true );
  68                 $this->setBatchSize( 500 );
  69         }
  70
  71         private function initServices() {
  72                 $this->dbw = $this->getDB( DB_MASTER );
  73                 $this->contentModelStore = MediaWikiServices::getInstance()->getContentModelStore();
  74                 $this->blobStore = MediaWikiServices::getInstance()->getBlobStore();
  75                 $this->mainRoleId = MediaWikiServices::getInstance()->getSlotRoleStore()
  76                         ->acquireId( SlotRecord::MAIN );
  77         }
  78
  79         public function execute() {
  80                 $multiContentRevisionSchemaMigrationStage =
  81                         $this->getConfig()->get( 'MultiContentRevisionSchemaMigrationStage' );
  82
  83                 $t0 = microtime( true );
  84
  85                 if ( ( $multiContentRevisionSchemaMigrationStage & SCHEMA_COMPAT_WRITE_NEW ) === 0 ) {
  86                         $this->writeln(
  87                                 '...cannot update while \$wgMultiContentRevisionSchemaMigrationStage '
  88                                 . 'does not have the SCHEMA_COMPAT_WRITE_NEW bit set.'
  89                         );
  90                         return false;
  91                 }
  92
  93                 $this->initServices();
  94
  95                 if ( $this->getOption( 'reuse-content', false ) ) {
  96                         $this->loadContentMap();
  97                 }
  98
  99                 foreach ( $this->getTables() as $table ) {
 100                         $this->populateTable( $table );
 101                 }
 102
 103                 $elapsed = microtime( true ) - $t0;
 104                 $this->writeln( "Done. Processed $this->totalCount rows in $elapsed seconds" );
 105                 return true;
 106         }
 107
 108         /**
 109          * @return string[]
 110          */
 111         private function getTables() {
 112                 $table = $this->getOption( 'table', 'all' );
 113                 $validTableOptions = [ 'all', 'revision', 'archive' ];
 114
 115                 if ( !in_array( $table, $validTableOptions ) ) {
 116                         $this->fatalError( 'Invalid table. Must be either `revision` or `archive` or `all`' );
 117                 }
 118
 119                 if ( $table === 'all' ) {
 120                         $tables = [ 'revision', 'archive' ];
 121                 } else {
 122                         $tables = [ $table ];
 123                 }
 124
 125                 return $tables;
 126         }
 127
 128         private function loadContentMap() {
 129                 $t0 = microtime( true );
 130                 $this->writeln( "Loading existing content table rows..." );
 131                 $this->contentRowMap = [];
 132                 $dbr = $this->getDB( DB_REPLICA );
 133                 $from = false;
 134                 while ( true ) {
 135                         $res = $dbr->select(
 136                                 'content',
 137                                 [ 'content_id', 'content_address', 'content_model' ],
 138                                 $from ? "content_id > $from" : '',
 139                                 __METHOD__,
 140                                 [ 'ORDER BY' => 'content_id', 'LIMIT' => $this->getBatchSize() ]
 141                         );
 142                         if ( !$res || !$res->numRows() ) {
 143                                 break;
 144                         }
 145                         foreach ( $res as $row ) {
 146                                 $from = $row->content_id;
 147                                 $this->contentRowMap["{$row->content_model}:{$row->content_address}"] = $row->content_id;
 148                         }
 149                 }
 150                 $elapsed = microtime( true ) - $t0;
 151                 $this->writeln( "Loaded " . count( $this->contentRowMap ) . " rows in $elapsed seconds" );
 152         }
 153
 154         /**
 155          * @param string $table
 156          */
 157         private function populateTable( $table ) {
 158                 $t0 = microtime( true );
 159                 $this->count = 0;
 160                 $this->writeln( "Populating $table..." );
 161
 162                 if ( $table === 'revision' ) {
 163                         $idField = 'rev_id';
 164                         $tables = [ 'revision', 'slots', 'page' ];
 165                         $fields = [
 166                                 'rev_id',
 167                                 'len' => 'rev_len',
 168                                 'sha1' => 'rev_sha1',
 169                                 'text_id' => 'rev_text_id',
 170                                 'content_model' => 'rev_content_model',
 171                                 'namespace' => 'page_namespace',
 172                                 'title' => 'page_title',
 173                         ];
 174                         $joins = [
 175                                 'slots' => [ 'LEFT JOIN', 'rev_id=slot_revision_id' ],
 176                                 'page' => [ 'LEFT JOIN', 'rev_page=page_id' ],
 177                         ];
 178                         $startOption = 'start-revision';
 179                 } else {
 180                         $idField = 'ar_rev_id';
 181                         $tables = [ 'archive', 'slots' ];
 182                         $fields = [
 183                                 'rev_id' => 'ar_rev_id',
 184                                 'len' => 'ar_len',
 185                                 'sha1' => 'ar_sha1',
 186                                 'text_id' => 'ar_text_id',
 187                                 'content_model' => 'ar_content_model',
 188                                 'namespace' => 'ar_namespace',
 189                                 'title' => 'ar_title',
 190                         ];
 191                         $joins = [
 192                                 'slots' => [ 'LEFT JOIN', 'ar_rev_id=slot_revision_id' ],
 193                         ];
 194                         $startOption = 'start-archive';
 195                 }
 196
 197                 if ( !$this->dbw->fieldExists( $table, $fields['text_id'], __METHOD__ ) ) {
 198                         $this->writeln( "No need to populate, $table.{$fields['text_id']} field does not exist" );
 199                         return;
 200                 }
 201
 202                 $minmax = $this->dbw->selectRow(
 203                         $table,
 204                         [ 'min' => "MIN( $idField )", 'max' => "MAX( $idField )" ],
 205                         '',
 206                         __METHOD__
 207                 );
 208                 if ( $this->hasOption( $startOption ) ) {
 209                         $minmax->min = (int)$this->getOption( $startOption );
 210                 }
 211                 if ( !$minmax || !is_numeric( $minmax->min ) || !is_numeric( $minmax->max ) ) {
 212                         // No rows?
 213                         $minmax = (object)[ 'min' => 1, 'max' => 0 ];
 214                 }
 215
 216                 $batchSize = $this->getBatchSize();
 217
 218                 for ( $startId = $minmax->min; $startId <= $minmax->max; $startId += $batchSize ) {
 219                         $endId = min( $startId + $batchSize - 1, $minmax->max );
 220                         $rows = $this->dbw->select(
 221                                 $tables,
 222                                 $fields,
 223                                 [
 224                                         "$idField >= $startId",
 225                                         "$idField <= $endId",
 226                                         'slot_revision_id IS NULL',
 227                                 ],
 228                                 __METHOD__,
 229                                 [ 'ORDER BY' => 'rev_id' ],
 230                                 $joins
 231                         );
 232                         if ( $rows->numRows() !== 0 ) {
 233                                 $this->populateContentTablesForRowBatch( $rows, $startId, $table );
 234                         }
 235
 236                         $elapsed = microtime( true ) - $t0;
 237                         $this->writeln(
 238                                 "... $table processed up to revision id $endId of {$minmax->max}"
 239                                 . " ($this->count rows in $elapsed seconds)"
 240                         );
 241                 }
 242
 243                 $elapsed = microtime( true ) - $t0;
 244                 $this->writeln( "Done populating $table table. Processed $this->count rows in $elapsed seconds" );
 245         }
 246
 247         /**
 248          * @param IResultWrapper $rows
 249          * @param int $startId
 250          * @param string $table
 251          * @return int|null
 252          */
 253         private function populateContentTablesForRowBatch( IResultWrapper $rows, $startId, $table ) {
 254                 $this->beginTransaction( $this->dbw, __METHOD__ );
 255
 256                 if ( $this->contentRowMap === null ) {
 257                         $map = [];
 258                 } else {
 259                         $map = &$this->contentRowMap;
 260                 }
 261                 $contentKeys = [];
 262
 263                 try {
 264                         // Step 1: Figure out content rows needing insertion.
 265                         $contentRows = [];
 266                         foreach ( $rows as $row ) {
 267                                 $revisionId = $row->rev_id;
 268
 269                                 Assert::invariant( $revisionId !== null, 'rev_id must not be null' );
 270
 271                                 $model = $this->getContentModel( $row );
 272                                 $modelId = $this->contentModelStore->acquireId( $model );
 273                                 $address = SqlBlobStore::makeAddressFromTextId( $row->text_id );
 274
 275                                 $key = "{$modelId}:{$address}";
 276                                 $contentKeys[$revisionId] = $key;
 277
 278                                 if ( !isset( $map[$key] ) ) {
 279                                         $this->fillMissingFields( $row, $model, $address );
 280
 281                                         $map[$key] = false;
 282                                         $contentRows[] = [
 283                                                 'content_size' => (int)$row->len,
 284                                                 'content_sha1' => $row->sha1,
 285                                                 'content_model' => $modelId,
 286                                                 'content_address' => $address,
 287                                         ];
 288                                 }
 289                         }
 290
 291                         // Step 2: Insert them, then read them back in for use in the next step.
 292                         if ( $contentRows ) {
 293                                 $id = $this->dbw->selectField( 'content', 'MAX(content_id)', '', __METHOD__ );
 294                                 $this->dbw->insert( 'content', $contentRows, __METHOD__ );
 295                                 $res = $this->dbw->select(
 296                                         'content',
 297                                         [ 'content_id', 'content_model', 'content_address' ],
 298                                         'content_id > ' . (int)$id,
 299                                         __METHOD__
 300                                 );
 301                                 foreach ( $res as $row ) {
 302                                         $key = $row->content_model . ':' . $row->content_address;
 303                                         $map[$key] = $row->content_id;
 304                                 }
 305                         }
 306
 307                         // Step 3: Insert the slot rows.
 308                         $slotRows = [];
 309                         foreach ( $rows as $row ) {
 310                                 $revisionId = $row->rev_id;
 311                                 $contentId = $map[$contentKeys[$revisionId]] ?? false;
 312                                 if ( $contentId === false ) {
 313                                         throw new \RuntimeException( "Content row for $revisionId not found after content insert" );
 314                                 }
 315                                 $slotRows[] = [
 316                                         'slot_revision_id' => $revisionId,
 317                                         'slot_role_id' => $this->mainRoleId,
 318                                         'slot_content_id' => $contentId,
 319                                         // There's no way to really know the previous revision, so assume no inheriting.
 320                                         // rev_parent_id can get changed on undeletions, and deletions can screw up
 321                                         // rev_timestamp ordering.
 322                                         'slot_origin' => $revisionId,
 323                                 ];
 324                         }
 325                         $this->dbw->insert( 'slots', $slotRows, __METHOD__ );
 326                         $this->count += count( $slotRows );
 327                         $this->totalCount += count( $slotRows );
 328                 } catch ( \Exception $e ) {
 329                         $this->rollbackTransaction( $this->dbw, __METHOD__ );
 330                         $this->fatalError( "Failed to populate content table $table row batch starting at $startId "
 331                                 . "due to exception: " . $e->__toString() );
 332                 }
 333
 334                 $this->commitTransaction( $this->dbw, __METHOD__ );
 335         }
 336
 337         /**
 338          * @param \stdClass $row
 339          * @return string
 340          */
 341         private function getContentModel( $row ) {
 342                 if ( isset( $row->content_model ) ) {
 343                         return $row->content_model;
 344                 }
 345
 346                 $title = Title::makeTitle( $row->namespace, $row->title );
 347
 348                 return ContentHandler::getDefaultModelFor( $title );
 349         }
 350
 351         /**
 352          * @param string $msg
 353          */
 354         private function writeln( $msg ) {
 355                 $this->output( "$msg\n" );
 356         }
 357
 358         /**
 359          * Compute any missing fields in $row.
 360          * The way the missing values are computed must correspond to the way this is done in SlotRecord.
 361          *
 362          * @param object $row to be modified
 363          * @param string $model
 364          * @param string $address
 365          */
 366         private function fillMissingFields( $row, $model, $address ) {
 367                 if ( !isset( $row->content_model ) ) {
 368                         // just for completeness
 369                         $row->content_model = $model;
 370                 }
 371
 372                 if ( isset( $row->len ) && isset( $row->sha1 ) && $row->sha1 !== '' ) {
 373                         // No need to load the content, quite now.
 374                         return;
 375                 }
 376
 377                 $blob = $this->blobStore->getBlob( $address );
 378
 379                 if ( !isset( $row->len ) ) {
 380                         // NOTE: The nominal size of the content may not be the length of the raw blob.
 381                         $handler = ContentHandler::getForModelID( $model );
 382                         $content = $handler->unserializeContent( $blob );
 383
 384                         $row->len = $content->getSize();
 385                 }
 386
 387                 if ( !isset( $row->sha1 ) || $row->sha1 === '' ) {
 388                         $row->sha1 = SlotRecord::base36Sha1( $blob );
 389                 }
 390         }
 391 }
 392
 393 $maintClass = 'PopulateContentTables';
 394 require_once RUN_MAINTENANCE_IF_MAIN;