maintenance/updateCollation.php

   1 <?php
   2 /**
   3  * Find all rows in the categorylinks table whose collation is out-of-date
   4  * (cl_collation != $wgCategoryCollation) and repopulate cl_sortkey
   5  * using the page title and cl_sortkey_prefix.
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License along
  18  * with this program; if not, write to the Free Software Foundation, Inc.,
  19  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  20  * http://www.gnu.org/copyleft/gpl.html
  21  *
  22  * @file
  23  * @ingroup Maintenance
  24  * @author Aryeh Gregor (Simetrical)
  25  */
  26
  27 #$optionsWithArgs = array( 'begin', 'max-slave-lag' );
  28
  29 require_once __DIR__ . '/Maintenance.php';
  30
  31 /**
  32  * Maintenance script that will find all rows in the categorylinks table
  33  * whose collation is out-of-date.
  34  *
  35  * @ingroup Maintenance
  36  */
  37 class UpdateCollation extends Maintenance {
  38         const BATCH_SIZE = 10000; // Number of rows to process in one batch
  39         const SYNC_INTERVAL = 20; // Wait for slaves after this many batches
  40
  41         public $sizeHistogram = array();
  42
  43         public function __construct() {
  44                 parent::__construct();
  45
  46                 global $wgCategoryCollation;
  47                 $this->mDescription = <<<TEXT
  48 This script will find all rows in the categorylinks table whose collation is
  49 out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey
  50 using the page title and cl_sortkey_prefix.  If everything's collation is
  51 up-to-date, it will do nothing.
  52 TEXT;
  53
  54                 $this->addOption( 'force', 'Run on all rows, even if the collation is ' .
  55                         'supposed to be up-to-date.' );
  56                 $this->addOption( 'previous-collation', 'Set the previous value of ' .
  57                         '$wgCategoryCollation here to speed up this script, especially if your ' .
  58                         'categorylinks table is large. This will only update rows with that ' .
  59                         'collation, though, so it may miss out-of-date rows with a different, ' .
  60                         'even older collation.', false, true );
  61                 $this->addOption( 'target-collation', 'Set this to the new collation type to ' .
  62                         'use instead of $wgCategoryCollation. Usually you should not use this, ' .
  63                         'you should just update $wgCategoryCollation in LocalSettings.php.',
  64                         false, true );
  65                 $this->addOption( 'dry-run', 'Don\'t actually change the collations, just ' .
  66                         'compile statistics.' );
  67                 $this->addOption( 'verbose-stats', 'Show more statistics.' );
  68         }
  69
  70         public function execute() {
  71                 global $wgCategoryCollation;
  72
  73                 $dbw = $this->getDB( DB_MASTER );
  74                 $force = $this->getOption( 'force' );
  75                 $dryRun = $this->getOption( 'dry-run' );
  76                 $verboseStats = $this->getOption( 'verbose-stats' );
  77                 if ( $this->hasOption( 'target-collation' ) ) {
  78                         $collationName = $this->getOption( 'target-collation' );
  79                         $collation = Collation::factory( $collationName );
  80                 } else {
  81                         $collationName = $wgCategoryCollation;
  82                         $collation = Collation::singleton();
  83                 }
  84
  85                 $options = array(
  86                         'LIMIT' => self::BATCH_SIZE,
  87                         'ORDER BY' => 'cl_to, cl_type, cl_from',
  88                         'STRAIGHT_JOIN',
  89                 );
  90
  91                 if ( $force || $dryRun ) {
  92                         $collationConds = array();
  93                 } else {
  94                         if ( $this->hasOption( 'previous-collation' ) ) {
  95                                 $collationConds['cl_collation'] = $this->getOption( 'previous-collation' );
  96                         } else {
  97                                 $collationConds = array( 0 =>
  98                                         'cl_collation != ' . $dbw->addQuotes( $collationName )
  99                                 );
 100                         }
 101
 102                         $count = $dbw->estimateRowCount(
 103                                 'categorylinks',
 104                                 '*',
 105                                 $collationConds,
 106                                 __METHOD__
 107                         );
 108                         // Improve estimate if feasible
 109                         if ( $count < 1000000 ) {
 110                                 $count = $dbw->selectField(
 111                                         'categorylinks',
 112                                         'COUNT(*)',
 113                                         $collationConds,
 114                                         __METHOD__
 115                                 );
 116                         }
 117                         if ( $count == 0 ) {
 118                                 $this->output( "Collations up-to-date.\n" );
 119                                 return;
 120                         }
 121                         $this->output( "Fixing collation for $count rows.\n" );
 122                 }
 123
 124                 $count = 0;
 125                 $batchCount = 0;
 126                 $batchConds = array();
 127                 do {
 128                         $this->output( "Selecting next " . self::BATCH_SIZE . " rows..." );
 129                         $res = $dbw->select(
 130                                 array( 'categorylinks', 'page' ),
 131                                 array( 'cl_from', 'cl_to', 'cl_sortkey_prefix', 'cl_collation',
 132                                         'cl_sortkey', 'cl_type', 'page_namespace', 'page_title'
 133                                 ),
 134                                 array_merge( $collationConds, $batchConds, array( 'cl_from = page_id' ) ),
 135                                 __METHOD__,
 136                                 $options
 137                         );
 138                         $this->output( " processing..." );
 139
 140                         if ( !$dryRun ) {
 141                                 $dbw->begin( __METHOD__ );
 142                         }
 143                         foreach ( $res as $row ) {
 144                                 $title = Title::newFromRow( $row );
 145                                 if ( !$row->cl_collation ) {
 146                                         # This is an old-style row, so the sortkey needs to be
 147                                         # converted.
 148                                         if ( $row->cl_sortkey == $title->getText()
 149                                                 || $row->cl_sortkey == $title->getPrefixedText() ) {
 150                                                 $prefix = '';
 151                                         } else {
 152                                                 # Custom sortkey, use it as a prefix
 153                                                 $prefix = $row->cl_sortkey;
 154                                         }
 155                                 } else {
 156                                         $prefix = $row->cl_sortkey_prefix;
 157                                 }
 158                                 # cl_type will be wrong for lots of pages if cl_collation is 0,
 159                                 # so let's update it while we're here.
 160                                 if ( $title->getNamespace() == NS_CATEGORY ) {
 161                                         $type = 'subcat';
 162                                 } elseif ( $title->getNamespace() == NS_FILE ) {
 163                                         $type = 'file';
 164                                 } else {
 165                                         $type = 'page';
 166                                 }
 167                                 $newSortKey = $collation->getSortKey(
 168                                         $title->getCategorySortkey( $prefix ) );
 169                                 if ( $verboseStats ) {
 170                                         $this->updateSortKeySizeHistogram( $newSortKey );
 171                                 }
 172
 173                                 if ( !$dryRun ) {
 174                                         $dbw->update(
 175                                                 'categorylinks',
 176                                                 array(
 177                                                         'cl_sortkey' => $newSortKey,
 178                                                         'cl_sortkey_prefix' => $prefix,
 179                                                         'cl_collation' => $collationName,
 180                                                         'cl_type' => $type,
 181                                                         'cl_timestamp = cl_timestamp',
 182                                                 ),
 183                                                 array( 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ),
 184                                                 __METHOD__
 185                                         );
 186                                 }
 187                         }
 188                         if ( !$dryRun ) {
 189                                 $dbw->commit( __METHOD__ );
 190                         }
 191
 192                         if ( $row ) {
 193                                 $batchConds = array( $this->getBatchCondition( $row ) );
 194                         }
 195
 196                         $count += $res->numRows();
 197                         $this->output( "$count done.\n" );
 198
 199                         if ( !$dryRun && ++$batchCount % self::SYNC_INTERVAL == 0 ) {
 200                                 $this->output( "Waiting for slaves ... " );
 201                                 wfWaitForSlaves();
 202                                 $this->output( "done\n" );
 203                         }
 204                 } while ( $res->numRows() == self::BATCH_SIZE );
 205
 206                 $this->output( "$count rows processed\n" );
 207
 208                 if ( $verboseStats ) {
 209                         $this->output( "\n" );
 210                         $this->showSortKeySizeHistogram();
 211                 }
 212         }
 213
 214         /**
 215          * Return an SQL expression selecting rows which sort above the given row,
 216          * assuming an ordering of cl_to, cl_type, cl_from
 217          */
 218         function getBatchCondition( $row ) {
 219                 $dbw = $this->getDB( DB_MASTER );
 220                 $fields = array( 'cl_to', 'cl_type', 'cl_from' );
 221                 $first = true;
 222                 $cond = false;
 223                 $prefix = false;
 224                 foreach ( $fields as $field ) {
 225                         $encValue = $dbw->addQuotes( $row->$field );
 226                         $inequality = "$field > $encValue";
 227                         $equality = "$field = $encValue";
 228                         if ( $first ) {
 229                                 $cond = $inequality;
 230                                 $prefix = $equality;
 231                                 $first = false;
 232                         } else {
 233                                 $cond .= " OR ($prefix AND $inequality)";
 234                                 $prefix .= " AND $equality";
 235                         }
 236                 }
 237                 return $cond;
 238         }
 239
 240         function updateSortKeySizeHistogram( $key ) {
 241                 $length = strlen( $key );
 242                 if ( !isset( $this->sizeHistogram[$length] ) ) {
 243                         $this->sizeHistogram[$length] = 0;
 244                 }
 245                 $this->sizeHistogram[$length]++;
 246         }
 247
 248         function showSortKeySizeHistogram() {
 249                 $maxLength = max( array_keys( $this->sizeHistogram ) );
 250                 if ( $maxLength == 0 ) {
 251                         return;
 252                 }
 253                 $numBins = 20;
 254                 $coarseHistogram = array_fill( 0, $numBins, 0 );
 255                 $coarseBoundaries = array();
 256                 $boundary = 0;
 257                 for ( $i = 0; $i < $numBins - 1; $i++ ) {
 258                         $boundary += $maxLength / $numBins;
 259                         $coarseBoundaries[$i] = round( $boundary );
 260                 }
 261                 $coarseBoundaries[$numBins - 1] = $maxLength + 1;
 262                 $raw = '';
 263                 for ( $i = 0; $i <= $maxLength; $i++ ) {
 264                         if ( $raw !== '' ) {
 265                                 $raw .= ', ';
 266                         }
 267                         if ( !isset( $this->sizeHistogram[$i] ) ) {
 268                                 $val = 0;
 269                         } else {
 270                                 $val = $this->sizeHistogram[$i];
 271                         }
 272                         for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++ ) {
 273                                 if ( $coarseBoundaries[$coarseIndex] > $i ) {
 274                                         $coarseHistogram[$coarseIndex] += $val;
 275                                         break;
 276                                 }
 277                         }
 278                         if ( $coarseIndex == $numBins - 1 ) {
 279                                 $coarseHistogram[$coarseIndex] += $val;
 280                         }
 281                         $raw .= $val;
 282                 }
 283
 284                 $this->output( "Sort key size histogram\nRaw data: $raw\n\n" );
 285
 286                 $maxBinVal = max( $coarseHistogram );
 287                 $scale = 60 / $maxBinVal;
 288                 $prevBoundary = 0;
 289                 for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++ ) {
 290                         if ( !isset( $coarseHistogram[$coarseIndex] ) ) {
 291                                 $val = 0;
 292                         } else {
 293                                 $val = $coarseHistogram[$coarseIndex];
 294                         }
 295                         $boundary = $coarseBoundaries[$coarseIndex];
 296                         $this->output( sprintf( "%-10s %-10d |%s\n",
 297                                 $prevBoundary . '-' . ( $boundary - 1 ) . ': ',
 298                                 $val,
 299                                 str_repeat( '*', $scale * $val ) ) );
 300                         $prevBoundary = $boundary;
 301                 }
 302         }
 303 }
 304
 305 $maintClass = "UpdateCollation";
 306 require_once RUN_MAINTENANCE_IF_MAIN;