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