maintenance/migrateComments.php

   1 <?php
   2 /**
   3  * Migrate comments from pre-1.30 columns to the 'comment' table
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Maintenance
  22  */
  23
  24 use Wikimedia\Rdbms\IDatabase;
  25
  26 require_once __DIR__ . '/Maintenance.php';
  27
  28 /**
  29  * Maintenance script that migrates comments from pre-1.30 columns to the
  30  * 'comment' table
  31  *
  32  * @ingroup Maintenance
  33  */
  34 class MigrateComments extends LoggedUpdateMaintenance {
  35         public function __construct() {
  36                 parent::__construct();
  37                 $this->addDescription( 'Migrates comments from pre-1.30 columns to the \'comment\' table' );
  38                 $this->setBatchSize( 100 );
  39         }
  40
  41         protected function getUpdateKey() {
  42                 return __CLASS__;
  43         }
  44
  45         protected function updateSkippedMessage() {
  46                 return 'comments already migrated.';
  47         }
  48
  49         protected function doDBUpdates() {
  50                 global $wgCommentTableSchemaMigrationStage;
  51
  52                 if ( $wgCommentTableSchemaMigrationStage < MIGRATION_WRITE_NEW ) {
  53                         $this->output(
  54                                 "...cannot update while \$wgCommentTableSchemaMigrationStage < MIGRATION_WRITE_NEW\n"
  55                         );
  56                         return false;
  57                 }
  58
  59                 $this->migrateToTemp(
  60                         'revision', 'rev_id', 'rev_comment', 'revcomment_rev', 'revcomment_comment_id'
  61                 );
  62                 $this->migrate( 'archive', 'ar_id', 'ar_comment' );
  63                 $this->migrate( 'ipblocks', 'ipb_id', 'ipb_reason' );
  64                 $this->migrate( 'image', 'img_name', 'img_description' );
  65                 $this->migrate( 'oldimage', [ 'oi_name', 'oi_timestamp' ], 'oi_description' );
  66                 $this->migrate( 'filearchive', 'fa_id', 'fa_deleted_reason' );
  67                 $this->migrate( 'filearchive', 'fa_id', 'fa_description' );
  68                 $this->migrate( 'recentchanges', 'rc_id', 'rc_comment' );
  69                 $this->migrate( 'logging', 'log_id', 'log_comment' );
  70                 $this->migrate( 'protected_titles', [ 'pt_namespace', 'pt_title' ], 'pt_reason' );
  71                 return true;
  72         }
  73
  74         /**
  75          * Fetch comment IDs for a set of comments
  76          * @param IDatabase $dbw
  77          * @param array &$comments Keys are comment names, values will be set to IDs.
  78          * @return int Count of added comments
  79          */
  80         private function loadCommentIDs( IDatabase $dbw, array &$comments ) {
  81                 $count = 0;
  82                 $needComments = $comments;
  83
  84                 while ( true ) {
  85                         $where = [];
  86                         foreach ( $needComments as $need => $dummy ) {
  87                                 $where[] = $dbw->makeList(
  88                                         [
  89                                                 'comment_hash' => CommentStore::hash( $need, null ),
  90                                                 'comment_text' => $need,
  91                                         ],
  92                                         LIST_AND
  93                                 );
  94                         }
  95
  96                         $res = $dbw->select(
  97                                 'comment',
  98                                 [ 'comment_id', 'comment_text' ],
  99                                 [
 100                                         $dbw->makeList( $where, LIST_OR ),
 101                                         'comment_data' => null,
 102                                 ],
 103                                 __METHOD__
 104                         );
 105                         foreach ( $res as $row ) {
 106                                 $comments[$row->comment_text] = $row->comment_id;
 107                                 unset( $needComments[$row->comment_text] );
 108                         }
 109
 110                         if ( !$needComments ) {
 111                                 break;
 112                         }
 113
 114                         $dbw->insert(
 115                                 'comment',
 116                                 array_map( function ( $v ) {
 117                                         return [
 118                                                 'comment_hash' => CommentStore::hash( $v, null ),
 119                                                 'comment_text' => $v,
 120                                         ];
 121                                 }, array_keys( $needComments ) ),
 122                                 __METHOD__
 123                         );
 124                         $count += $dbw->affectedRows();
 125                 }
 126                 return $count;
 127         }
 128
 129         /**
 130          * Migrate comments in a table.
 131          *
 132          * Assumes any row with the ID field non-zero have already been migrated.
 133          * Assumes the new field name is the same as the old with '_id' appended.
 134          * Blanks the old fields while migrating.
 135          *
 136          * @param string $table Table to migrate
 137          * @param string|string[] $primaryKey Primary key of the table.
 138          * @param string $oldField Old comment field name
 139          */
 140         protected function migrate( $table, $primaryKey, $oldField ) {
 141                 $newField = $oldField . '_id';
 142                 $primaryKey = (array)$primaryKey;
 143                 $pkFilter = array_flip( $primaryKey );
 144                 $this->output( "Beginning migration of $table.$oldField to $table.$newField\n" );
 145                 wfWaitForSlaves();
 146
 147                 $dbw = $this->getDB( DB_MASTER );
 148                 $next = '1=1';
 149                 $countUpdated = 0;
 150                 $countComments = 0;
 151                 while ( true ) {
 152                         // Fetch the rows needing update
 153                         $res = $dbw->select(
 154                                 $table,
 155                                 array_merge( $primaryKey, [ $oldField ] ),
 156                                 [
 157                                         $newField => 0,
 158                                         $next,
 159                                 ],
 160                                 __METHOD__,
 161                                 [
 162                                         'ORDER BY' => $primaryKey,
 163                                         'LIMIT' => $this->getBatchSize(),
 164                                 ]
 165                         );
 166                         if ( !$res->numRows() ) {
 167                                 break;
 168                         }
 169
 170                         // Collect the distinct comments from those rows
 171                         $comments = [];
 172                         foreach ( $res as $row ) {
 173                                 $comments[$row->$oldField] = 0;
 174                         }
 175                         $countComments += $this->loadCommentIDs( $dbw, $comments );
 176
 177                         // Update the existing rows
 178                         foreach ( $res as $row ) {
 179                                 $dbw->update(
 180                                         $table,
 181                                         [
 182                                                 $newField => $comments[$row->$oldField],
 183                                                 $oldField => '',
 184                                         ],
 185                                         array_intersect_key( (array)$row, $pkFilter ) + [
 186                                                 $newField => 0
 187                                         ],
 188                                         __METHOD__
 189                                 );
 190                                 $countUpdated += $dbw->affectedRows();
 191                         }
 192
 193                         // Calculate the "next" condition
 194                         $next = '';
 195                         $prompt = [];
 196                         for ( $i = count( $primaryKey ) - 1; $i >= 0; $i-- ) {
 197                                 $field = $primaryKey[$i];
 198                                 $prompt[] = $row->$field;
 199                                 $value = $dbw->addQuotes( $row->$field );
 200                                 if ( $next === '' ) {
 201                                         $next = "$field > $value";
 202                                 } else {
 203                                         $next = "$field > $value OR $field = $value AND ($next)";
 204                                 }
 205                         }
 206                         $prompt = implode( ' ', array_reverse( $prompt ) );
 207                         $this->output( "... $prompt\n" );
 208                         wfWaitForSlaves();
 209                 }
 210
 211                 $this->output(
 212                         "Completed migration, updated $countUpdated row(s) with $countComments new comment(s)\n"
 213                 );
 214         }
 215
 216         /**
 217          * Migrate comments in a table to a temporary table.
 218          *
 219          * Assumes any row with the ID field non-zero have already been migrated.
 220          * Assumes the new table is named "{$table}_comment_temp", and it has two
 221          * columns, in order, being the primary key of the original table and the
 222          * comment ID field.
 223          * Blanks the old fields while migrating.
 224          *
 225          * @param string $table Table to migrate
 226          * @param string $primaryKey Primary key of the table.
 227          * @param string $oldField Old comment field name
 228          * @param string $newPrimaryKey Primary key of the new table.
 229          * @param string $newField New comment field name
 230          */
 231         protected function migrateToTemp( $table, $primaryKey, $oldField, $newPrimaryKey, $newField ) {
 232                 $newTable = $table . '_comment_temp';
 233                 $this->output( "Beginning migration of $table.$oldField to $newTable.$newField\n" );
 234                 wfWaitForSlaves();
 235
 236                 $dbw = $this->getDB( DB_MASTER );
 237                 $next = [];
 238                 $countUpdated = 0;
 239                 $countComments = 0;
 240                 while ( true ) {
 241                         // Fetch the rows needing update
 242                         $res = $dbw->select(
 243                                 [ $table, $newTable ],
 244                                 [ $primaryKey, $oldField ],
 245                                 [ $newPrimaryKey => null ] + $next,
 246                                 __METHOD__,
 247                                 [
 248                                         'ORDER BY' => $primaryKey,
 249                                         'LIMIT' => $this->getBatchSize(),
 250                                 ],
 251                                 [ $newTable => [ 'LEFT JOIN', "{$primaryKey}={$newPrimaryKey}" ] ]
 252                         );
 253                         if ( !$res->numRows() ) {
 254                                 break;
 255                         }
 256
 257                         // Collect the distinct comments from those rows
 258                         $comments = [];
 259                         foreach ( $res as $row ) {
 260                                 $comments[$row->$oldField] = 0;
 261                         }
 262                         $countComments += $this->loadCommentIDs( $dbw, $comments );
 263
 264                         // Update rows
 265                         $inserts = [];
 266                         $updates = [];
 267                         foreach ( $res as $row ) {
 268                                 $inserts[] = [
 269                                         $newPrimaryKey => $row->$primaryKey,
 270                                         $newField => $comments[$row->$oldField]
 271                                 ];
 272                                 $updates[] = $row->$primaryKey;
 273                         }
 274                         $this->beginTransaction( $dbw, __METHOD__ );
 275                         $dbw->insert( $newTable, $inserts, __METHOD__ );
 276                         $dbw->update( $table, [ $oldField => '' ], [ $primaryKey => $updates ], __METHOD__ );
 277                         $countUpdated += $dbw->affectedRows();
 278                         $this->commitTransaction( $dbw, __METHOD__ );
 279
 280                         // Calculate the "next" condition
 281                         $next = [ $primaryKey . ' > ' . $dbw->addQuotes( $row->$primaryKey ) ];
 282                         $this->output( "... {$row->$primaryKey}\n" );
 283                 }
 284
 285                 $this->output(
 286                         "Completed migration, updated $countUpdated row(s) with $countComments new comment(s)\n"
 287                 );
 288         }
 289 }
 290
 291 $maintClass = MigrateComments::class;
 292 require_once RUN_MAINTENANCE_IF_MAIN;