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                 $this->migrateToTemp(
  51                         'revision', 'rev_id', 'rev_comment', 'revcomment_rev', 'revcomment_comment_id'
  52                 );
  53                 $this->migrate( 'archive', 'ar_id', 'ar_comment' );
  54                 $this->migrate( 'ipblocks', 'ipb_id', 'ipb_reason' );
  55                 $this->migrate( 'image', 'img_name', 'img_description' );
  56                 $this->migrate( 'oldimage', [ 'oi_name', 'oi_timestamp' ], 'oi_description' );
  57                 $this->migrate( 'filearchive', 'fa_id', 'fa_deleted_reason' );
  58                 $this->migrate( 'filearchive', 'fa_id', 'fa_description' );
  59                 $this->migrate( 'recentchanges', 'rc_id', 'rc_comment' );
  60                 $this->migrate( 'logging', 'log_id', 'log_comment' );
  61                 $this->migrate( 'protected_titles', [ 'pt_namespace', 'pt_title' ], 'pt_reason' );
  62                 return true;
  63         }
  64
  65         /**
  66          * Fetch comment IDs for a set of comments
  67          * @param IDatabase $dbw
  68          * @param array &$comments Keys are comment names, values will be set to IDs.
  69          * @return int Count of added comments
  70          */
  71         private function loadCommentIDs( IDatabase $dbw, array &$comments ) {
  72                 $count = 0;
  73                 $needComments = $comments;
  74
  75                 while ( true ) {
  76                         $where = [];
  77                         foreach ( $needComments as $need => $dummy ) {
  78                                 $where[] = $dbw->makeList(
  79                                         [
  80                                                 'comment_hash' => CommentStore::hash( $need, null ),
  81                                                 'comment_text' => $need,
  82                                         ],
  83                                         LIST_AND
  84                                 );
  85                         }
  86
  87                         $res = $dbw->select(
  88                                 'comment',
  89                                 [ 'comment_id', 'comment_text' ],
  90                                 [
  91                                         $dbw->makeList( $where, LIST_OR ),
  92                                         'comment_data' => null,
  93                                 ],
  94                                 __METHOD__
  95                         );
  96                         foreach ( $res as $row ) {
  97                                 $comments[$row->comment_text] = $row->comment_id;
  98                                 unset( $needComments[$row->comment_text] );
  99                         }
 100
 101                         if ( !$needComments ) {
 102                                 break;
 103                         }
 104
 105                         $dbw->insert(
 106                                 'comment',
 107                                 array_map( function ( $v ) {
 108                                         return [
 109                                                 'comment_hash' => CommentStore::hash( $v, null ),
 110                                                 'comment_text' => $v,
 111                                         ];
 112                                 }, array_keys( $needComments ) ),
 113                                 __METHOD__
 114                         );
 115                         $count += $dbw->affectedRows();
 116                 }
 117                 return $count;
 118         }
 119
 120         /**
 121          * Migrate comments in a table.
 122          *
 123          * Assumes any row with the ID field non-zero have already been migrated.
 124          * Assumes the new field name is the same as the old with '_id' appended.
 125          * Blanks the old fields while migrating.
 126          *
 127          * @param string $table Table to migrate
 128          * @param string|string[] $primaryKey Primary key of the table.
 129          * @param string $oldField Old comment field name
 130          */
 131         protected function migrate( $table, $primaryKey, $oldField ) {
 132                 $dbw = $this->getDB( DB_MASTER );
 133                 if ( !$dbw->fieldExists( $table, $oldField, __METHOD__ ) ) {
 134                         $this->output( "No need to migrate $table.$oldField, field does not exist\n" );
 135                         return;
 136                 }
 137
 138                 $newField = $oldField . '_id';
 139                 $primaryKey = (array)$primaryKey;
 140                 $pkFilter = array_flip( $primaryKey );
 141                 $this->output( "Beginning migration of $table.$oldField to $table.$newField\n" );
 142                 wfWaitForSlaves();
 143
 144                 $next = '1=1';
 145                 $countUpdated = 0;
 146                 $countComments = 0;
 147                 while ( true ) {
 148                         // Fetch the rows needing update
 149                         $res = $dbw->select(
 150                                 $table,
 151                                 array_merge( $primaryKey, [ $oldField ] ),
 152                                 [
 153                                         $newField => 0,
 154                                         $next,
 155                                 ],
 156                                 __METHOD__,
 157                                 [
 158                                         'ORDER BY' => $primaryKey,
 159                                         'LIMIT' => $this->getBatchSize(),
 160                                 ]
 161                         );
 162                         if ( !$res->numRows() ) {
 163                                 break;
 164                         }
 165
 166                         // Collect the distinct comments from those rows
 167                         $comments = [];
 168                         foreach ( $res as $row ) {
 169                                 $comments[$row->$oldField] = 0;
 170                         }
 171                         $countComments += $this->loadCommentIDs( $dbw, $comments );
 172
 173                         // Update the existing rows
 174                         foreach ( $res as $row ) {
 175                                 $dbw->update(
 176                                         $table,
 177                                         [
 178                                                 $newField => $comments[$row->$oldField],
 179                                                 $oldField => '',
 180                                         ],
 181                                         array_intersect_key( (array)$row, $pkFilter ) + [
 182                                                 $newField => 0
 183                                         ],
 184                                         __METHOD__
 185                                 );
 186                                 $countUpdated += $dbw->affectedRows();
 187                         }
 188
 189                         // Calculate the "next" condition
 190                         $next = '';
 191                         $prompt = [];
 192                         for ( $i = count( $primaryKey ) - 1; $i >= 0; $i-- ) {
 193                                 $field = $primaryKey[$i];
 194                                 $prompt[] = $row->$field;
 195                                 $value = $dbw->addQuotes( $row->$field );
 196                                 if ( $next === '' ) {
 197                                         $next = "$field > $value";
 198                                 } else {
 199                                         $next = "$field > $value OR $field = $value AND ($next)";
 200                                 }
 201                         }
 202                         $prompt = implode( ' ', array_reverse( $prompt ) );
 203                         $this->output( "... $prompt\n" );
 204                         wfWaitForSlaves();
 205                 }
 206
 207                 $this->output(
 208                         "Completed migration, updated $countUpdated row(s) with $countComments new comment(s)\n"
 209                 );
 210         }
 211
 212         /**
 213          * Migrate comments in a table to a temporary table.
 214          *
 215          * Assumes any row with the ID field non-zero have already been migrated.
 216          * Assumes the new table is named "{$table}_comment_temp", and it has two
 217          * columns, in order, being the primary key of the original table and the
 218          * comment ID field.
 219          * Blanks the old fields while migrating.
 220          *
 221          * @param string $table Table to migrate
 222          * @param string $primaryKey Primary key of the table.
 223          * @param string $oldField Old comment field name
 224          * @param string $newPrimaryKey Primary key of the new table.
 225          * @param string $newField New comment field name
 226          */
 227         protected function migrateToTemp( $table, $primaryKey, $oldField, $newPrimaryKey, $newField ) {
 228                 $dbw = $this->getDB( DB_MASTER );
 229                 if ( !$dbw->fieldExists( $table, $oldField, __METHOD__ ) ) {
 230                         $this->output( "No need to migrate $table.$oldField, field does not exist\n" );
 231                         return;
 232                 }
 233
 234                 $newTable = $table . '_comment_temp';
 235                 $this->output( "Beginning migration of $table.$oldField to $newTable.$newField\n" );
 236                 wfWaitForSlaves();
 237
 238                 $dbw = $this->getDB( DB_MASTER );
 239                 $next = [];
 240                 $countUpdated = 0;
 241                 $countComments = 0;
 242                 while ( true ) {
 243                         // Fetch the rows needing update
 244                         $res = $dbw->select(
 245                                 [ $table, $newTable ],
 246                                 [ $primaryKey, $oldField ],
 247                                 [ $newPrimaryKey => null ] + $next,
 248                                 __METHOD__,
 249                                 [
 250                                         'ORDER BY' => $primaryKey,
 251                                         'LIMIT' => $this->getBatchSize(),
 252                                 ],
 253                                 [ $newTable => [ 'LEFT JOIN', "{$primaryKey}={$newPrimaryKey}" ] ]
 254                         );
 255                         if ( !$res->numRows() ) {
 256                                 break;
 257                         }
 258
 259                         // Collect the distinct comments from those rows
 260                         $comments = [];
 261                         foreach ( $res as $row ) {
 262                                 $comments[$row->$oldField] = 0;
 263                         }
 264                         $countComments += $this->loadCommentIDs( $dbw, $comments );
 265
 266                         // Update rows
 267                         $inserts = [];
 268                         $updates = [];
 269                         foreach ( $res as $row ) {
 270                                 $inserts[] = [
 271                                         $newPrimaryKey => $row->$primaryKey,
 272                                         $newField => $comments[$row->$oldField]
 273                                 ];
 274                                 $updates[] = $row->$primaryKey;
 275                         }
 276                         $this->beginTransaction( $dbw, __METHOD__ );
 277                         $dbw->insert( $newTable, $inserts, __METHOD__ );
 278                         $dbw->update( $table, [ $oldField => '' ], [ $primaryKey => $updates ], __METHOD__ );
 279                         $countUpdated += $dbw->affectedRows();
 280                         $this->commitTransaction( $dbw, __METHOD__ );
 281
 282                         // Calculate the "next" condition
 283                         $next = [ $primaryKey . ' > ' . $dbw->addQuotes( $row->$primaryKey ) ];
 284                         $this->output( "... {$row->$primaryKey}\n" );
 285                 }
 286
 287                 $this->output(
 288                         "Completed migration, updated $countUpdated row(s) with $countComments new comment(s)\n"
 289                 );
 290         }
 291 }
 292
 293 $maintClass = MigrateComments::class;
 294 require_once RUN_MAINTENANCE_IF_MAIN;