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   3  * Find all rows in the categorylinks table whose collation is out-of-date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  33  * whose collation is out-of-date.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 = 50; // 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  49 out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey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, $wgMiserMode;
  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( 'LIMIT' => self::BATCH_SIZE, 'STRAIGHT_JOIN' );
  86
  87                 if ( $force || $dryRun ) {
  88                         $options['ORDER BY'] = 'cl_from, cl_to';
  89                         $collationConds = array();
  90                 } else {
  91                         if ( $this->hasOption( 'previous-collation' ) ) {
  92                                 $collationConds['cl_collation'] = $this->getOption( 'previous-collation' );
  93                         } else {
  94                                 $collationConds = array( 0 =>
  95                                         'cl_collation != ' . $dbw->addQuotes( $collationName )
  96                                 );
  97                         }
  98
  99                         if ( !$wgMiserMode ) {
 100                                 $count = $dbw->selectField(
 101                                         'categorylinks',
 102                                         'COUNT(*)',
 103                                         $collationConds,
 104                                         __METHOD__
 105                                 );
 106                         } else {
 107                                 $count = $dbw->estimateRowCount(
 108                                         'categorylinks',
 109                                         '*',
 110                                         $collationConds,
 111                                         __METHOD__
 112                                 );
 113                         }
 114                         if ( $count == 0 ) {
 115                                 $this->output( "Collations up-to-date.\n" );
 116                                 return;
 117                         }
 118                         $this->output( "Fixing collation for $count rows.\n" );
 119                 }
 120
 121                 $count = 0;
 122                 $batchCount = 0;
 123                 $batchConds = array();
 124                 do {
 125                         $this->output( "Selecting next " . self::BATCH_SIZE . " rows..." );
 126                         $res = $dbw->select(
 127                                 array( 'categorylinks', 'page' ),
 128                                 array( 'cl_from', 'cl_to', 'cl_sortkey_prefix', 'cl_collation',
 129                                         'cl_sortkey', 'page_namespace', 'page_title'
 130                                 ),
 131                                 array_merge( $collationConds, $batchConds, array( 'cl_from = page_id' ) ),
 132                                 __METHOD__,
 133                                 $options
 134                         );
 135                         $this->output( " processing..." );
 136
 137                         if ( !$dryRun ) {
 138                                 $dbw->begin( __METHOD__ );
 139                         }
 140                         foreach ( $res as $row ) {
 141                                 $title = Title::newFromRow( $row );
 142                                 if ( !$row->cl_collation ) {
 143                                         # This is an old-style row, so the sortkey needs to be
 144                                         # converted.
 145                                         if ( $row->cl_sortkey == $title->getText()
 146                                                 || $row->cl_sortkey == $title->getPrefixedText() ) {
 147                                                 $prefix = '';
 148                                         } else {
 149                                                 # Custom sortkey, use it as a prefix
 150                                                 $prefix = $row->cl_sortkey;
 151                                         }
 152                                 } else {
 153                                         $prefix = $row->cl_sortkey_prefix;
 154                                 }
 155                                 # cl_type will be wrong for lots of pages if cl_collation is 0,
 156                                 # so let's update it while we're here.
 157                                 if ( $title->getNamespace() == NS_CATEGORY ) {
 158                                         $type = 'subcat';
 159                                 } elseif ( $title->getNamespace() == NS_FILE ) {
 160                                         $type = 'file';
 161                                 } else {
 162                                         $type = 'page';
 163                                 }
 164                                 $newSortKey = $collation->getSortKey(
 165                                         $title->getCategorySortkey( $prefix ) );
 166                                 if ( $verboseStats ) {
 167                                         $this->updateSortKeySizeHistogram( $newSortKey );
 168                                 }
 169
 170                                 if ( !$dryRun ) {
 171                                         $dbw->update(
 172                                                 'categorylinks',
 173                                                 array(
 174                                                         'cl_sortkey' => $newSortKey,
 175                                                         'cl_sortkey_prefix' => $prefix,
 176                                                         'cl_collation' => $collationName,
 177                                                         'cl_type' => $type,
 178                                                         'cl_timestamp = cl_timestamp',
 179                                                 ),
 180                                                 array( 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ),
 181                                                 __METHOD__
 182                                         );
 183                                 }
 184                         }
 185                         if ( !$dryRun ) {
 186                                 $dbw->commit( __METHOD__ );
 187                         }
 188
 189                         if ( ( $force || $dryRun ) && $row ) {
 190                                 $encFrom = $dbw->addQuotes( $row->cl_from );
 191                                 $encTo = $dbw->addQuotes( $row->cl_to );
 192                                 $batchConds = array(
 193                                         "(cl_from = $encFrom AND cl_to > $encTo) " .
 194                                         " OR cl_from > $encFrom" );
 195                         }
 196
 197                         $count += $res->numRows();
 198                         $this->output( "$count done.\n" );
 199
 200                         if ( !$dryRun && ++$batchCount % self::SYNC_INTERVAL == 0 ) {
 201                                 $this->output( "Waiting for slaves ... " );
 202                                 wfWaitForSlaves();
 203                                 $this->output( "done\n" );
 204                         }
 205                 } while ( $res->numRows() == self::BATCH_SIZE );
 206
 207                 $this->output( "$count rows processed\n" );
 208
 209                 if ( $verboseStats ) {
 210                         $this->output( "\n" );
 211                         $this->showSortKeySizeHistogram();
 212                 }
 213         }
 214
 215         function updateSortKeySizeHistogram( $key ) {
 216                 $length = strlen( $key );
 217                 if ( !isset( $this->sizeHistogram[$length] ) ) {
 218                         $this->sizeHistogram[$length] = 0;
 219                 }
 220                 $this->sizeHistogram[$length]++;
 221         }
 222
 223         function showSortKeySizeHistogram() {
 224                 $maxLength = max( array_keys( $this->sizeHistogram ) );
 225                 if ( $maxLength == 0 ) {
 226                         return;
 227                 }
 228                 $numBins = 20;
 229                 $coarseHistogram = array_fill( 0, $numBins, 0 );
 230                 $coarseBoundaries = array();
 231                 $boundary = 0;
 232                 for ( $i = 0; $i < $numBins - 1; $i++ ) {
 233                         $boundary += $maxLength / $numBins;
 234                         $coarseBoundaries[$i] = round( $boundary );
 235                 }
 236                 $coarseBoundaries[$numBins - 1] = $maxLength + 1;
 237                 $raw = '';
 238                 for ( $i = 0; $i <= $maxLength; $i++ ) {
 239                         if ( $raw !== '' ) {
 240                                 $raw .= ', ';
 241                         }
 242                         if ( !isset( $this->sizeHistogram[$i] ) ) {
 243                                 $val = 0;
 244                         } else {
 245                                 $val = $this->sizeHistogram[$i];
 246                         }
 247                         for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++ ) {
 248                                 if ( $coarseBoundaries[$coarseIndex] > $i ) {
 249                                         $coarseHistogram[$coarseIndex] += $val;
 250                                         break;
 251                                 }
 252                         }
 253                         if ( $coarseIndex == $numBins - 1 ) {
 254                                 $coarseHistogram[$coarseIndex] += $val;
 255                         }
 256                         $raw .= $val;
 257                 }
 258
 259                 $this->output( "Sort key size histogram\nRaw data: $raw\n\n" );
 260
 261                 $maxBinVal = max( $coarseHistogram );
 262                 $scale = 60 / $maxBinVal;
 263                 $prevBoundary = 0;
 264                 for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++ ) {
 265                         if ( !isset( $coarseHistogram[$coarseIndex] ) ) {
 266                                 $val = 0;
 267                         } else {
 268                                 $val = $coarseHistogram[$coarseIndex];
 269                         }
 270                         $boundary = $coarseBoundaries[$coarseIndex];
 271                         $this->output( sprintf( "%-10s %-10d |%s\n",
 272                                 $prevBoundary . '-' . ( $boundary - 1 ) . ': ',
 273                                 $val,
 274                                 str_repeat( '*', $scale * $val ) ) );
 275                         $prevBoundary = $boundary;
 276                 }
 277         }
 278 }
 279
 280 $maintClass = "UpdateCollation";
 281 require_once( RUN_MAINTENANCE_IF_MAIN );