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   3  * Find all rows in the categorylinks table whose collation is out-of-date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 require_once __DIR__ . '/Maintenance.php';
  28
  29 /**
  30  * Maintenance script that will find all rows in the categorylinks table
  31  * whose collation is out-of-date.
  32  *
  33  * @ingroup Maintenance
  34  */
  35 class UpdateCollation extends Maintenance {
  36         const BATCH_SIZE = 10000; // Number of rows to process in one batch
  37         const SYNC_INTERVAL = 20; // Wait for slaves after this many batches
  38
  39         public $sizeHistogram = array();
  40
  41         public function __construct() {
  42                 parent::__construct();
  43
  44                 global $wgCategoryCollation;
  45                 $this->addDescription( <<<TEXT
  46 This script will find all rows in the categorylinks table whose collation is
  47 out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey
  48 using the page title and cl_sortkey_prefix.  If all collations are
  49 up-to-date, it will do nothing.
  50 TEXT
  51                 );
  52
  53                 $this->addOption( 'force', 'Run on all rows, even if the collation is ' .
  54                         'supposed to be up-to-date.' );
  55                 $this->addOption( 'previous-collation', 'Set the previous value of ' .
  56                         '$wgCategoryCollation here to speed up this script, especially if your ' .
  57                         'categorylinks table is large. This will only update rows with that ' .
  58                         'collation, though, so it may miss out-of-date rows with a different, ' .
  59                         'even older collation.', false, true );
  60                 $this->addOption( 'target-collation', 'Set this to the new collation type to ' .
  61                         'use instead of $wgCategoryCollation. Usually you should not use this, ' .
  62                         'you should just update $wgCategoryCollation in LocalSettings.php.',
  63                         false, true );
  64                 $this->addOption( 'dry-run', 'Don\'t actually change the collations, just ' .
  65                         'compile statistics.' );
  66                 $this->addOption( 'verbose-stats', 'Show more statistics.' );
  67         }
  68
  69         public function execute() {
  70                 global $wgCategoryCollation;
  71
  72                 $dbw = $this->getDB( DB_MASTER );
  73                 $force = $this->getOption( 'force' );
  74                 $dryRun = $this->getOption( 'dry-run' );
  75                 $verboseStats = $this->getOption( 'verbose-stats' );
  76                 if ( $this->hasOption( 'target-collation' ) ) {
  77                         $collationName = $this->getOption( 'target-collation' );
  78                         $collation = Collation::factory( $collationName );
  79                 } else {
  80                         $collationName = $wgCategoryCollation;
  81                         $collation = Collation::singleton();
  82                 }
  83
  84                 // Collation sanity check: in some cases the constructor will work,
  85                 // but this will raise an exception, breaking all category pages
  86                 $collation->getFirstLetter( 'MediaWiki' );
  87
  88                 $options = array(
  89                         'LIMIT' => self::BATCH_SIZE,
  90                         'ORDER BY' => 'cl_from, cl_to',
  91                         'STRAIGHT_JOIN',
  92                 );
  93
  94                 if ( $force || $dryRun ) {
  95                         $collationConds = array();
  96                 } else {
  97                         if ( $this->hasOption( 'previous-collation' ) ) {
  98                                 $collationConds['cl_collation'] = $this->getOption( 'previous-collation' );
  99                         } else {
 100                                 $collationConds = array( 0 =>
 101                                         'cl_collation != ' . $dbw->addQuotes( $collationName )
 102                                 );
 103                         }
 104
 105                         $count = $dbw->estimateRowCount(
 106                                 'categorylinks',
 107                                 '*',
 108                                 $collationConds,
 109                                 __METHOD__
 110                         );
 111                         // Improve estimate if feasible
 112                         if ( $count < 1000000 ) {
 113                                 $count = $dbw->selectField(
 114                                         'categorylinks',
 115                                         'COUNT(*)',
 116                                         $collationConds,
 117                                         __METHOD__
 118                                 );
 119                         }
 120                         if ( $count == 0 ) {
 121                                 $this->output( "Collations up-to-date.\n" );
 122
 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', '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                                 $this->beginTransaction( $dbw, __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                                         ) {
 155                                                 $prefix = '';
 156                                         } else {
 157                                                 # Custom sortkey, use it as a prefix
 158                                                 $prefix = $row->cl_sortkey;
 159                                         }
 160                                 } else {
 161                                         $prefix = $row->cl_sortkey_prefix;
 162                                 }
 163                                 # cl_type will be wrong for lots of pages if cl_collation is 0,
 164                                 # so let's update it while we're here.
 165                                 if ( $title->getNamespace() == NS_CATEGORY ) {
 166                                         $type = 'subcat';
 167                                 } elseif ( $title->getNamespace() == NS_FILE ) {
 168                                         $type = 'file';
 169                                 } else {
 170                                         $type = 'page';
 171                                 }
 172                                 $newSortKey = $collation->getSortKey(
 173                                         $title->getCategorySortkey( $prefix ) );
 174                                 if ( $verboseStats ) {
 175                                         $this->updateSortKeySizeHistogram( $newSortKey );
 176                                 }
 177
 178                                 if ( !$dryRun ) {
 179                                         $dbw->update(
 180                                                 'categorylinks',
 181                                                 array(
 182                                                         'cl_sortkey' => $newSortKey,
 183                                                         'cl_sortkey_prefix' => $prefix,
 184                                                         'cl_collation' => $collationName,
 185                                                         'cl_type' => $type,
 186                                                         'cl_timestamp = cl_timestamp',
 187                                                 ),
 188                                                 array( 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ),
 189                                                 __METHOD__
 190                                         );
 191                                 }
 192                                 if ( $row ) {
 193                                         $batchConds = array( $this->getBatchCondition( $row, $dbw ) );
 194                                 }
 195                         }
 196                         if ( !$dryRun ) {
 197                                 $this->commitTransaction( $dbw, __METHOD__ );
 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_from, cl_to
 221          * @param stdClass $row
 222          * @param DatabaseBase $dbw
 223          * @return string
 224          */
 225         function getBatchCondition( $row, $dbw ) {
 226                 $fields = array( 'cl_from', 'cl_to' );
 227                 $first = true;
 228                 $cond = false;
 229                 $prefix = false;
 230                 foreach ( $fields as $field ) {
 231                         $encValue = $dbw->addQuotes( $row->$field );
 232                         $inequality = "$field > $encValue";
 233                         $equality = "$field = $encValue";
 234                         if ( $first ) {
 235                                 $cond = $inequality;
 236                                 $prefix = $equality;
 237                                 $first = false;
 238                         } else {
 239                                 $cond .= " OR ($prefix AND $inequality)";
 240                                 $prefix .= " AND $equality";
 241                         }
 242                 }
 243
 244                 return $cond;
 245         }
 246
 247         function updateSortKeySizeHistogram( $key ) {
 248                 $length = strlen( $key );
 249                 if ( !isset( $this->sizeHistogram[$length] ) ) {
 250                         $this->sizeHistogram[$length] = 0;
 251                 }
 252                 $this->sizeHistogram[$length]++;
 253         }
 254
 255         function showSortKeySizeHistogram() {
 256                 $maxLength = max( array_keys( $this->sizeHistogram ) );
 257                 if ( $maxLength == 0 ) {
 258                         return;
 259                 }
 260                 $numBins = 20;
 261                 $coarseHistogram = array_fill( 0, $numBins, 0 );
 262                 $coarseBoundaries = array();
 263                 $boundary = 0;
 264                 for ( $i = 0; $i < $numBins - 1; $i++ ) {
 265                         $boundary += $maxLength / $numBins;
 266                         $coarseBoundaries[$i] = round( $boundary );
 267                 }
 268                 $coarseBoundaries[$numBins - 1] = $maxLength + 1;
 269                 $raw = '';
 270                 for ( $i = 0; $i <= $maxLength; $i++ ) {
 271                         if ( $raw !== '' ) {
 272                                 $raw .= ', ';
 273                         }
 274                         if ( !isset( $this->sizeHistogram[$i] ) ) {
 275                                 $val = 0;
 276                         } else {
 277                                 $val = $this->sizeHistogram[$i];
 278                         }
 279                         for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++ ) {
 280                                 if ( $coarseBoundaries[$coarseIndex] > $i ) {
 281                                         $coarseHistogram[$coarseIndex] += $val;
 282                                         break;
 283                                 }
 284                         }
 285                         if ( $coarseIndex == $numBins - 1 ) {
 286                                 $coarseHistogram[$coarseIndex] += $val;
 287                         }
 288                         $raw .= $val;
 289                 }
 290
 291                 $this->output( "Sort key size histogram\nRaw data: $raw\n\n" );
 292
 293                 $maxBinVal = max( $coarseHistogram );
 294                 $scale = 60 / $maxBinVal;
 295                 $prevBoundary = 0;
 296                 for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++ ) {
 297                         if ( !isset( $coarseHistogram[$coarseIndex] ) ) {
 298                                 $val = 0;
 299                         } else {
 300                                 $val = $coarseHistogram[$coarseIndex];
 301                         }
 302                         $boundary = $coarseBoundaries[$coarseIndex];
 303                         $this->output( sprintf( "%-10s %-10d |%s\n",
 304                                 $prevBoundary . '-' . ( $boundary - 1 ) . ': ',
 305                                 $val,
 306                                 str_repeat( '*', $scale * $val ) ) );
 307                         $prevBoundary = $boundary;
 308                 }
 309         }
 310 }
 311
 312 $maintClass = "UpdateCollation";
 313 require_once RUN_MAINTENANCE_IF_MAIN;