maintenance/updateCollation.php

   1 <?php
   2 /**
   3  * Script will find all rows in the categorylinks table whose collation is
   4  * out-of-date (cl_collation != $wgCategoryCollation) and repopulate cl_sortkey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( dirname( __FILE__ ) . '/Maintenance.php' );
  30
  31 class UpdateCollation extends Maintenance {
  32         const BATCH_SIZE = 50; // Number of rows to process in one batch
  33         const SYNC_INTERVAL = 20; // Wait for slaves after this many batches
  34
  35         var $sizeHistogram = array();
  36
  37         public function __construct() {
  38                 parent::__construct();
  39
  40                 global $wgCategoryCollation;
  41                 $this->mDescription = <<<TEXT
  42 This script will find all rows in the categorylinks table whose collation is
  43 out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey
  44 using the page title and cl_sortkey_prefix.  If everything's collation is
  45 up-to-date, it will do nothing.
  46 TEXT;
  47
  48                 $this->addOption( 'force', 'Run on all rows, even if the collation is ' .
  49                         'supposed to be up-to-date.' );
  50                 $this->addOption( 'previous-collation', 'Set the previous value of ' .
  51                         '$wgCategoryCollation here to speed up this script, especially if your ' .
  52                         'categorylinks table is large. This will only update rows with that ' .
  53                         'collation, though, so it may miss out-of-date rows with a different, ' .
  54                         'even older collation.', false, true );
  55                 $this->addOption( 'target-collation', 'Set this to the new collation type to ' .
  56                         'use instead of $wgCategoryCollation. Usually you should not use this, ' .
  57                         'you should just update $wgCategoryCollation in LocalSettings.php.',
  58                         false, true );
  59                 $this->addOption( 'dry-run', 'Don\'t actually change the collations, just ' .
  60                         'compile statistics.' );
  61                 $this->addOption( 'verbose-stats', 'Show more statistics.' );
  62         }
  63
  64         public function execute() {
  65                 global $wgCategoryCollation, $wgMiserMode;
  66
  67                 $dbw = $this->getDB( DB_MASTER );
  68                 $force = $this->getOption( 'force' );
  69                 $dryRun = $this->getOption( 'dry-run' );
  70                 $verboseStats = $this->getOption( 'verbose-stats' );
  71                 if ( $this->hasOption( 'target-collation' ) ) {
  72                         $collationName = $this->getOption( 'target-collation' );
  73                         $collation = Collation::factory( $collationName );
  74                 } else {
  75                         $collationName = $wgCategoryCollation;
  76                         $collation = Collation::singleton();
  77                 }
  78
  79                 $options = array( 'LIMIT' => self::BATCH_SIZE, 'STRAIGHT_JOIN' );
  80
  81                 if ( $force || $dryRun ) {
  82                         $options['ORDER BY'] = 'cl_from, cl_to';
  83                         $collationConds = array();
  84                 } else {
  85                         if ( $this->hasOption( 'previous-collation' ) ) {
  86                                 $collationConds['cl_collation'] = $this->getOption( 'previous-collation' );
  87                         } else {
  88                                 $collationConds = array( 0 =>
  89                                         'cl_collation != ' . $dbw->addQuotes( $collationName )
  90                                 );
  91                         }
  92
  93                         if ( !$wgMiserMode ) {
  94                                 $count = $dbw->selectField(
  95                                         'categorylinks',
  96                                         'COUNT(*)',
  97                                         $collationConds,
  98                                         __METHOD__
  99                                 );
 100                         } else {
 101                                 $count = $dbw->estimateRowCount(
 102                                         'categorylinks',
 103                                         '*',
 104                                         $collationConds,
 105                                         __METHOD__
 106                                 );
 107                         }
 108                         if ( $count == 0 ) {
 109                                 $this->output( "Collations up-to-date.\n" );
 110                                 return;
 111                         }
 112                         $this->output( "Fixing collation for $count rows.\n" );
 113                 }
 114
 115                 $count = 0;
 116                 $batchCount = 0;
 117                 $batchConds = array();
 118                 do {
 119                         $this->output( "Selecting next " . self::BATCH_SIZE . " rows..." );
 120                         $res = $dbw->select(
 121                                 array( 'categorylinks', 'page' ),
 122                                 array( 'cl_from', 'cl_to', 'cl_sortkey_prefix', 'cl_collation',
 123                                         'cl_sortkey', 'page_namespace', 'page_title'
 124                                 ),
 125                                 array_merge( $collationConds, $batchConds, array( 'cl_from = page_id' ) ),
 126                                 __METHOD__,
 127                                 $options
 128                         );
 129                         $this->output( " processing..." );
 130
 131                         if ( !$dryRun ) {
 132                                 $dbw->begin( __METHOD__ );
 133                         }
 134                         foreach ( $res as $row ) {
 135                                 $title = Title::newFromRow( $row );
 136                                 if ( !$row->cl_collation ) {
 137                                         # This is an old-style row, so the sortkey needs to be
 138                                         # converted.
 139                                         if ( $row->cl_sortkey == $title->getText()
 140                                                 || $row->cl_sortkey == $title->getPrefixedText() ) {
 141                                                 $prefix = '';
 142                                         } else {
 143                                                 # Custom sortkey, use it as a prefix
 144                                                 $prefix = $row->cl_sortkey;
 145                                         }
 146                                 } else {
 147                                         $prefix = $row->cl_sortkey_prefix;
 148                                 }
 149                                 # cl_type will be wrong for lots of pages if cl_collation is 0,
 150                                 # so let's update it while we're here.
 151                                 if ( $title->getNamespace() == NS_CATEGORY ) {
 152                                         $type = 'subcat';
 153                                 } elseif ( $title->getNamespace() == NS_FILE ) {
 154                                         $type = 'file';
 155                                 } else {
 156                                         $type = 'page';
 157                                 }
 158                                 $newSortKey = $collation->getSortKey(
 159                                         $title->getCategorySortkey( $prefix ) );
 160                                 if ( $verboseStats ) {
 161                                         $this->updateSortKeySizeHistogram( $newSortKey );
 162                                 }
 163
 164                                 if ( !$dryRun ) {
 165                                         $dbw->update(
 166                                                 'categorylinks',
 167                                                 array(
 168                                                         'cl_sortkey' => $newSortKey,
 169                                                         'cl_sortkey_prefix' => $prefix,
 170                                                         'cl_collation' => $collationName,
 171                                                         'cl_type' => $type,
 172                                                         'cl_timestamp = cl_timestamp',
 173                                                 ),
 174                                                 array( 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ),
 175                                                 __METHOD__
 176                                         );
 177                                 }
 178                         }
 179                         if ( !$dryRun ) {
 180                                 $dbw->commit( __METHOD__ );
 181                         }
 182
 183                         if ( ( $force || $dryRun ) && $row ) {
 184                                 $encFrom = $dbw->addQuotes( $row->cl_from );
 185                                 $encTo = $dbw->addQuotes( $row->cl_to );
 186                                 $batchConds = array(
 187                                         "(cl_from = $encFrom AND cl_to > $encTo) " .
 188                                         " OR cl_from > $encFrom" );
 189                         }
 190
 191                         $count += $res->numRows();
 192                         $this->output( "$count done.\n" );
 193
 194                         if ( !$dryRun && ++$batchCount % self::SYNC_INTERVAL == 0 ) {
 195                                 $this->output( "Waiting for slaves ... " );
 196                                 wfWaitForSlaves();
 197                                 $this->output( "done\n" );
 198                         }
 199                 } while ( $res->numRows() == self::BATCH_SIZE );
 200
 201                 $this->output( "$count rows processed\n" );
 202
 203                 if ( $verboseStats ) {
 204                         $this->output( "\n" );
 205                         $this->showSortKeySizeHistogram();
 206                 }
 207         }
 208
 209         function updateSortKeySizeHistogram( $key ) {
 210                 $length = strlen( $key );
 211                 if ( !isset( $this->sizeHistogram[$length] ) ) {
 212                         $this->sizeHistogram[$length] = 0;
 213                 }
 214                 $this->sizeHistogram[$length]++;
 215         }
 216
 217         function showSortKeySizeHistogram() {
 218                 $maxLength = max( array_keys( $this->sizeHistogram ) );
 219                 if ( $maxLength == 0 ) {
 220                         return;
 221                 }
 222                 $numBins = 20;
 223                 $coarseHistogram = array_fill( 0, $numBins, 0 );
 224                 $coarseBoundaries = array();
 225                 $boundary = 0;
 226                 for ( $i = 0; $i < $numBins - 1; $i++ ) {
 227                         $boundary += $maxLength / $numBins;
 228                         $coarseBoundaries[$i] = round( $boundary );
 229                 }
 230                 $coarseBoundaries[$numBins - 1] = $maxLength + 1;
 231                 $raw = '';
 232                 for ( $i = 0; $i <= $maxLength; $i++ ) {
 233                         if ( $raw !== '' ) {
 234                                 $raw .= ', ';
 235                         }
 236                         if ( !isset( $this->sizeHistogram[$i] ) ) {
 237                                 $val = 0;
 238                         } else {
 239                                 $val = $this->sizeHistogram[$i];
 240                         }
 241                         for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++ ) {
 242                                 if ( $coarseBoundaries[$coarseIndex] > $i ) {
 243                                         $coarseHistogram[$coarseIndex] += $val;
 244                                         break;
 245                                 }
 246                         }
 247                         if ( $coarseIndex == $numBins - 1 ) {
 248                                 $coarseHistogram[$coarseIndex] += $val;
 249                         }
 250                         $raw .= $val;
 251                 }
 252
 253                 $this->output( "Sort key size histogram\nRaw data: $raw\n\n" );
 254
 255                 $maxBinVal = max( $coarseHistogram );
 256                 $scale = 60 / $maxBinVal;
 257                 $prevBoundary = 0;
 258                 for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++ ) {
 259                         if ( !isset( $coarseHistogram[$coarseIndex] ) ) {
 260                                 $val = 0;
 261                         } else {
 262                                 $val = $coarseHistogram[$coarseIndex];
 263                         }
 264                         $boundary = $coarseBoundaries[$coarseIndex];
 265                         $this->output( sprintf( "%-10s %-10d |%s\n",
 266                                 $prevBoundary . '-' . ( $boundary - 1 ) . ': ',
 267                                 $val,
 268                                 str_repeat( '*', $scale * $val ) ) );
 269                         $prevBoundary = $boundary;
 270                 }
 271         }
 272 }
 273
 274 $maintClass = "UpdateCollation";
 275 require_once( RUN_MAINTENANCE_IF_MAIN );