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   3  * Find all rows in the categorylinks table whose collation is out-of-date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  33  * whose collation is out-of-date.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  49 out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey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                 // Collation sanity check: in some cases the constructor will work,
  86                 // but this will raise an exception, breaking all category pages
  87                 $collation->getFirstLetter( 'MediaWiki' );
  88
  89                 $options = array(
  90                         'LIMIT' => self::BATCH_SIZE,
  91                         'ORDER BY' => 'cl_to, cl_type, cl_from',
  92                         'STRAIGHT_JOIN',
  93                 );
  94
  95                 if ( $force || $dryRun ) {
  96                         $collationConds = array();
  97                 } else {
  98                         if ( $this->hasOption( 'previous-collation' ) ) {
  99                                 $collationConds['cl_collation'] = $this->getOption( 'previous-collation' );
 100                         } else {
 101                                 $collationConds = array( 0 =>
 102                                         'cl_collation != ' . $dbw->addQuotes( $collationName )
 103                                 );
 104                         }
 105
 106                         $count = $dbw->estimateRowCount(
 107                                 'categorylinks',
 108                                 '*',
 109                                 $collationConds,
 110                                 __METHOD__
 111                         );
 112                         // Improve estimate if feasible
 113                         if ( $count < 1000000 ) {
 114                                 $count = $dbw->selectField(
 115                                         'categorylinks',
 116                                         'COUNT(*)',
 117                                         $collationConds,
 118                                         __METHOD__
 119                                 );
 120                         }
 121                         if ( $count == 0 ) {
 122                                 $this->output( "Collations up-to-date.\n" );
 123                                 return;
 124                         }
 125                         $this->output( "Fixing collation for $count rows.\n" );
 126                 }
 127
 128                 $count = 0;
 129                 $batchCount = 0;
 130                 $batchConds = array();
 131                 do {
 132                         $this->output( "Selecting next " . self::BATCH_SIZE . " rows..." );
 133                         $res = $dbw->select(
 134                                 array( 'categorylinks', 'page' ),
 135                                 array( 'cl_from', 'cl_to', 'cl_sortkey_prefix', 'cl_collation',
 136                                         'cl_sortkey', 'cl_type', 'page_namespace', 'page_title'
 137                                 ),
 138                                 array_merge( $collationConds, $batchConds, array( 'cl_from = page_id' ) ),
 139                                 __METHOD__,
 140                                 $options
 141                         );
 142                         $this->output( " processing..." );
 143
 144                         if ( !$dryRun ) {
 145                                 $dbw->begin( __METHOD__ );
 146                         }
 147                         foreach ( $res as $row ) {
 148                                 $title = Title::newFromRow( $row );
 149                                 if ( !$row->cl_collation ) {
 150                                         # This is an old-style row, so the sortkey needs to be
 151                                         # converted.
 152                                         if ( $row->cl_sortkey == $title->getText()
 153                                                 || $row->cl_sortkey == $title->getPrefixedText() ) {
 154                                                 $prefix = '';
 155                                         } else {
 156                                                 # Custom sortkey, use it as a prefix
 157                                                 $prefix = $row->cl_sortkey;
 158                                         }
 159                                 } else {
 160                                         $prefix = $row->cl_sortkey_prefix;
 161                                 }
 162                                 # cl_type will be wrong for lots of pages if cl_collation is 0,
 163                                 # so let's update it while we're here.
 164                                 if ( $title->getNamespace() == NS_CATEGORY ) {
 165                                         $type = 'subcat';
 166                                 } elseif ( $title->getNamespace() == NS_FILE ) {
 167                                         $type = 'file';
 168                                 } else {
 169                                         $type = 'page';
 170                                 }
 171                                 $newSortKey = $collation->getSortKey(
 172                                         $title->getCategorySortkey( $prefix ) );
 173                                 if ( $verboseStats ) {
 174                                         $this->updateSortKeySizeHistogram( $newSortKey );
 175                                 }
 176
 177                                 if ( !$dryRun ) {
 178                                         $dbw->update(
 179                                                 'categorylinks',
 180                                                 array(
 181                                                         'cl_sortkey' => $newSortKey,
 182                                                         'cl_sortkey_prefix' => $prefix,
 183                                                         'cl_collation' => $collationName,
 184                                                         'cl_type' => $type,
 185                                                         'cl_timestamp = cl_timestamp',
 186                                                 ),
 187                                                 array( 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ),
 188                                                 __METHOD__
 189                                         );
 190                                 }
 191                         }
 192                         if ( !$dryRun ) {
 193                                 $dbw->commit( __METHOD__ );
 194                         }
 195
 196                         if ( $row ) {
 197                                 $batchConds = array( $this->getBatchCondition( $row ) );
 198                         }
 199
 200                         $count += $res->numRows();
 201                         $this->output( "$count done.\n" );
 202
 203                         if ( !$dryRun && ++$batchCount % self::SYNC_INTERVAL == 0 ) {
 204                                 $this->output( "Waiting for slaves ... " );
 205                                 wfWaitForSlaves();
 206                                 $this->output( "done\n" );
 207                         }
 208                 } while ( $res->numRows() == self::BATCH_SIZE );
 209
 210                 $this->output( "$count rows processed\n" );
 211
 212                 if ( $verboseStats ) {
 213                         $this->output( "\n" );
 214                         $this->showSortKeySizeHistogram();
 215                 }
 216         }
 217
 218         /**
 219          * Return an SQL expression selecting rows which sort above the given row,
 220          * assuming an ordering of cl_to, cl_type, cl_from
 221          */
 222         function getBatchCondition( $row ) {
 223                 $dbw = $this->getDB( DB_MASTER );
 224                 $fields = array( 'cl_to', 'cl_type', 'cl_from' );
 225                 $first = true;
 226                 $cond = false;
 227                 $prefix = false;
 228                 foreach ( $fields as $field ) {
 229                         $encValue = $dbw->addQuotes( $row->$field );
 230                         $inequality = "$field > $encValue";
 231                         $equality = "$field = $encValue";
 232                         if ( $first ) {
 233                                 $cond = $inequality;
 234                                 $prefix = $equality;
 235                                 $first = false;
 236                         } else {
 237                                 $cond .= " OR ($prefix AND $inequality)";
 238                                 $prefix .= " AND $equality";
 239                         }
 240                 }
 241                 return $cond;
 242         }
 243
 244         function updateSortKeySizeHistogram( $key ) {
 245                 $length = strlen( $key );
 246                 if ( !isset( $this->sizeHistogram[$length] ) ) {
 247                         $this->sizeHistogram[$length] = 0;
 248                 }
 249                 $this->sizeHistogram[$length]++;
 250         }
 251
 252         function showSortKeySizeHistogram() {
 253                 $maxLength = max( array_keys( $this->sizeHistogram ) );
 254                 if ( $maxLength == 0 ) {
 255                         return;
 256                 }
 257                 $numBins = 20;
 258                 $coarseHistogram = array_fill( 0, $numBins, 0 );
 259                 $coarseBoundaries = array();
 260                 $boundary = 0;
 261                 for ( $i = 0; $i < $numBins - 1; $i++ ) {
 262                         $boundary += $maxLength / $numBins;
 263                         $coarseBoundaries[$i] = round( $boundary );
 264                 }
 265                 $coarseBoundaries[$numBins - 1] = $maxLength + 1;
 266                 $raw = '';
 267                 for ( $i = 0; $i <= $maxLength; $i++ ) {
 268                         if ( $raw !== '' ) {
 269                                 $raw .= ', ';
 270                         }
 271                         if ( !isset( $this->sizeHistogram[$i] ) ) {
 272                                 $val = 0;
 273                         } else {
 274                                 $val = $this->sizeHistogram[$i];
 275                         }
 276                         for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++ ) {
 277                                 if ( $coarseBoundaries[$coarseIndex] > $i ) {
 278                                         $coarseHistogram[$coarseIndex] += $val;
 279                                         break;
 280                                 }
 281                         }
 282                         if ( $coarseIndex == $numBins - 1 ) {
 283                                 $coarseHistogram[$coarseIndex] += $val;
 284                         }
 285                         $raw .= $val;
 286                 }
 287
 288                 $this->output( "Sort key size histogram\nRaw data: $raw\n\n" );
 289
 290                 $maxBinVal = max( $coarseHistogram );
 291                 $scale = 60 / $maxBinVal;
 292                 $prevBoundary = 0;
 293                 for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++ ) {
 294                         if ( !isset( $coarseHistogram[$coarseIndex] ) ) {
 295                                 $val = 0;
 296                         } else {
 297                                 $val = $coarseHistogram[$coarseIndex];
 298                         }
 299                         $boundary = $coarseBoundaries[$coarseIndex];
 300                         $this->output( sprintf( "%-10s %-10d |%s\n",
 301                                 $prevBoundary . '-' . ( $boundary - 1 ) . ': ',
 302                                 $val,
 303                                 str_repeat( '*', $scale * $val ) ) );
 304                         $prevBoundary = $boundary;
 305                 }
 306         }
 307 }
 308
 309 $maintClass = "UpdateCollation";
 310 require_once RUN_MAINTENANCE_IF_MAIN;