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   3  * Find all rows in the categorylinks table whose collation is out-of-date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  33  * whose collation is out-of-date.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  49 out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey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 all collations are
  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
 124                                 return;
 125                         }
 126                         $this->output( "Fixing collation for $count rows.\n" );
 127                 }
 128
 129                 $count = 0;
 130                 $batchCount = 0;
 131                 $batchConds = array();
 132                 do {
 133                         $this->output( "Selecting next " . self::BATCH_SIZE . " rows..." );
 134                         $res = $dbw->select(
 135                                 array( 'categorylinks', 'page' ),
 136                                 array( 'cl_from', 'cl_to', 'cl_sortkey_prefix', 'cl_collation',
 137                                         'cl_sortkey', 'cl_type', 'page_namespace', 'page_title'
 138                                 ),
 139                                 array_merge( $collationConds, $batchConds, array( 'cl_from = page_id' ) ),
 140                                 __METHOD__,
 141                                 $options
 142                         );
 143                         $this->output( " processing..." );
 144
 145                         if ( !$dryRun ) {
 146                                 $dbw->begin( __METHOD__ );
 147                         }
 148                         foreach ( $res as $row ) {
 149                                 $title = Title::newFromRow( $row );
 150                                 if ( !$row->cl_collation ) {
 151                                         # This is an old-style row, so the sortkey needs to be
 152                                         # converted.
 153                                         if ( $row->cl_sortkey == $title->getText()
 154                                                 || $row->cl_sortkey == $title->getPrefixedText()
 155                                         ) {
 156                                                 $prefix = '';
 157                                         } else {
 158                                                 # Custom sortkey, use it as a prefix
 159                                                 $prefix = $row->cl_sortkey;
 160                                         }
 161                                 } else {
 162                                         $prefix = $row->cl_sortkey_prefix;
 163                                 }
 164                                 # cl_type will be wrong for lots of pages if cl_collation is 0,
 165                                 # so let's update it while we're here.
 166                                 if ( $title->getNamespace() == NS_CATEGORY ) {
 167                                         $type = 'subcat';
 168                                 } elseif ( $title->getNamespace() == NS_FILE ) {
 169                                         $type = 'file';
 170                                 } else {
 171                                         $type = 'page';
 172                                 }
 173                                 $newSortKey = $collation->getSortKey(
 174                                         $title->getCategorySortkey( $prefix ) );
 175                                 if ( $verboseStats ) {
 176                                         $this->updateSortKeySizeHistogram( $newSortKey );
 177                                 }
 178
 179                                 if ( !$dryRun ) {
 180                                         $dbw->update(
 181                                                 'categorylinks',
 182                                                 array(
 183                                                         'cl_sortkey' => $newSortKey,
 184                                                         'cl_sortkey_prefix' => $prefix,
 185                                                         'cl_collation' => $collationName,
 186                                                         'cl_type' => $type,
 187                                                         'cl_timestamp = cl_timestamp',
 188                                                 ),
 189                                                 array( 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ),
 190                                                 __METHOD__
 191                                         );
 192                                 }
 193                                 if ( $row ) {
 194                                         $batchConds = array( $this->getBatchCondition( $row, $dbw ) );
 195                                 }
 196                         }
 197                         if ( !$dryRun ) {
 198                                 $dbw->commit( __METHOD__ );
 199                         }
 200
 201                         $count += $res->numRows();
 202                         $this->output( "$count done.\n" );
 203
 204                         if ( !$dryRun && ++$batchCount % self::SYNC_INTERVAL == 0 ) {
 205                                 $this->output( "Waiting for slaves ... " );
 206                                 wfWaitForSlaves();
 207                                 $this->output( "done\n" );
 208                         }
 209                 } while ( $res->numRows() == self::BATCH_SIZE );
 210
 211                 $this->output( "$count rows processed\n" );
 212
 213                 if ( $verboseStats ) {
 214                         $this->output( "\n" );
 215                         $this->showSortKeySizeHistogram();
 216                 }
 217         }
 218
 219         /**
 220          * Return an SQL expression selecting rows which sort above the given row,
 221          * assuming an ordering of cl_to, cl_type, cl_from
 222          * @param stdClass $row
 223          * @param DatabaseBase $dbw
 224          * @return string
 225          */
 226         function getBatchCondition( $row, $dbw ) {
 227                 $fields = array( 'cl_to', 'cl_type', 'cl_from' );
 228                 $first = true;
 229                 $cond = false;
 230                 $prefix = false;
 231                 foreach ( $fields as $field ) {
 232                         $encValue = $dbw->addQuotes( $row->$field );
 233                         $inequality = "$field > $encValue";
 234                         $equality = "$field = $encValue";
 235                         if ( $first ) {
 236                                 $cond = $inequality;
 237                                 $prefix = $equality;
 238                                 $first = false;
 239                         } else {
 240                                 $cond .= " OR ($prefix AND $inequality)";
 241                                 $prefix .= " AND $equality";
 242                         }
 243                 }
 244
 245                 return $cond;
 246         }
 247
 248         function updateSortKeySizeHistogram( $key ) {
 249                 $length = strlen( $key );
 250                 if ( !isset( $this->sizeHistogram[$length] ) ) {
 251                         $this->sizeHistogram[$length] = 0;
 252                 }
 253                 $this->sizeHistogram[$length]++;
 254         }
 255
 256         function showSortKeySizeHistogram() {
 257                 $maxLength = max( array_keys( $this->sizeHistogram ) );
 258                 if ( $maxLength == 0 ) {
 259                         return;
 260                 }
 261                 $numBins = 20;
 262                 $coarseHistogram = array_fill( 0, $numBins, 0 );
 263                 $coarseBoundaries = array();
 264                 $boundary = 0;
 265                 for ( $i = 0; $i < $numBins - 1; $i++ ) {
 266                         $boundary += $maxLength / $numBins;
 267                         $coarseBoundaries[$i] = round( $boundary );
 268                 }
 269                 $coarseBoundaries[$numBins - 1] = $maxLength + 1;
 270                 $raw = '';
 271                 for ( $i = 0; $i <= $maxLength; $i++ ) {
 272                         if ( $raw !== '' ) {
 273                                 $raw .= ', ';
 274                         }
 275                         if ( !isset( $this->sizeHistogram[$i] ) ) {
 276                                 $val = 0;
 277                         } else {
 278                                 $val = $this->sizeHistogram[$i];
 279                         }
 280                         for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++ ) {
 281                                 if ( $coarseBoundaries[$coarseIndex] > $i ) {
 282                                         $coarseHistogram[$coarseIndex] += $val;
 283                                         break;
 284                                 }
 285                         }
 286                         if ( $coarseIndex == $numBins - 1 ) {
 287                                 $coarseHistogram[$coarseIndex] += $val;
 288                         }
 289                         $raw .= $val;
 290                 }
 291
 292                 $this->output( "Sort key size histogram\nRaw data: $raw\n\n" );
 293
 294                 $maxBinVal = max( $coarseHistogram );
 295                 $scale = 60 / $maxBinVal;
 296                 $prevBoundary = 0;
 297                 for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++ ) {
 298                         if ( !isset( $coarseHistogram[$coarseIndex] ) ) {
 299                                 $val = 0;
 300                         } else {
 301                                 $val = $coarseHistogram[$coarseIndex];
 302                         }
 303                         $boundary = $coarseBoundaries[$coarseIndex];
 304                         $this->output( sprintf( "%-10s %-10d |%s\n",
 305                                 $prevBoundary . '-' . ( $boundary - 1 ) . ': ',
 306                                 $val,
 307                                 str_repeat( '*', $scale * $val ) ) );
 308                         $prevBoundary = $boundary;
 309                 }
 310         }
 311 }
 312
 313 $maintClass = "UpdateCollation";
 314 require_once RUN_MAINTENANCE_IF_MAIN;