3 * Find all rows in the categorylinks table whose collation is out-of-date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.
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.
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.
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
23 * @ingroup Maintenance
24 * @author Aryeh Gregor (Simetrical)
27 require_once __DIR__
. '/Maintenance.php';
29 use Wikimedia\Rdbms\IDatabase
;
32 * Maintenance script that will find all rows in the categorylinks table
33 * whose collation is out-of-date.
35 * @ingroup Maintenance
37 class UpdateCollation
extends Maintenance
{
38 const BATCH_SIZE
= 100; // Number of rows to process in one batch
39 const SYNC_INTERVAL
= 5; // Wait for replica DBs after this many batches
41 public $sizeHistogram = [];
43 public function __construct() {
44 parent
::__construct();
46 global $wgCategoryCollation;
47 $this->addDescription( <<<TEXT
48 This script will find all rows in the categorylinks table whose collation is
49 out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey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.
55 $this->addOption( 'force', 'Run on all rows, even if the collation is ' .
56 'supposed to be up-to-date.', false, false, 'f' );
57 $this->addOption( 'previous-collation', 'Set the previous value of ' .
58 '$wgCategoryCollation here to speed up this script, especially if your ' .
59 'categorylinks table is large. This will only update rows with that ' .
60 'collation, though, so it may miss out-of-date rows with a different, ' .
61 'even older collation.', false, true );
62 $this->addOption( 'target-collation', 'Set this to the new collation type to ' .
63 'use instead of $wgCategoryCollation. Usually you should not use this, ' .
64 'you should just update $wgCategoryCollation in LocalSettings.php.',
66 $this->addOption( 'dry-run', 'Don\'t actually change the collations, just ' .
67 'compile statistics.' );
68 $this->addOption( 'verbose-stats', 'Show more statistics.' );
71 public function execute() {
72 global $wgCategoryCollation;
74 $dbw = $this->getDB( DB_MASTER
);
75 $dbr = $this->getDB( DB_REPLICA
);
76 $force = $this->getOption( 'force' );
77 $dryRun = $this->getOption( 'dry-run' );
78 $verboseStats = $this->getOption( 'verbose-stats' );
79 if ( $this->hasOption( 'target-collation' ) ) {
80 $collationName = $this->getOption( 'target-collation' );
81 $collation = Collation
::factory( $collationName );
83 $collationName = $wgCategoryCollation;
84 $collation = Collation
::singleton();
87 // Collation sanity check: in some cases the constructor will work,
88 // but this will raise an exception, breaking all category pages
89 $collation->getFirstLetter( 'MediaWiki' );
91 // Locally at least, (my local is a rather old version of mysql)
92 // mysql seems to filesort if there is both an equality
93 // (but not for an inequality) condition on cl_collation in the
94 // WHERE and it is also the first item in the ORDER BY.
95 if ( $this->hasOption( 'previous-collation' ) ) {
96 $orderBy = 'cl_to, cl_type, cl_from';
98 $orderBy = 'cl_collation, cl_to, cl_type, cl_from';
101 'LIMIT' => self
::BATCH_SIZE
,
102 'ORDER BY' => $orderBy,
103 'STRAIGHT_JOIN' // per T58041
107 $collationConds = [];
109 if ( $this->hasOption( 'previous-collation' ) ) {
110 $collationConds['cl_collation'] = $this->getOption( 'previous-collation' );
112 $collationConds = [ 0 =>
113 'cl_collation != ' . $dbw->addQuotes( $collationName )
117 $count = $dbr->estimateRowCount(
123 // Improve estimate if feasible
124 if ( $count < 1000000 ) {
125 $count = $dbr->selectField(
133 $this->output( "Collations up-to-date.\n" );
138 $this->output( "$count rows would be updated.\n" );
140 $this->output( "Fixing collation for $count rows.\n" );
147 $this->output( "Selecting next " . self
::BATCH_SIZE
. " rows..." );
149 // cl_type must be selected as a number for proper paging because
151 if ( $dbw->getType() === 'mysql' ) {
152 $clType = 'cl_type+0 AS "cl_type_numeric"';
157 [ 'categorylinks', 'page' ],
158 [ 'cl_from', 'cl_to', 'cl_sortkey_prefix', 'cl_collation',
159 'cl_sortkey', $clType,
160 'page_namespace', 'page_title'
162 array_merge( $collationConds, $batchConds, [ 'cl_from = page_id' ] ),
166 $this->output( " processing..." );
169 $this->beginTransaction( $dbw, __METHOD__
);
171 foreach ( $res as $row ) {
172 $title = Title
::newFromRow( $row );
173 if ( !$row->cl_collation
) {
174 # This is an old-style row, so the sortkey needs to be
176 if ( $row->cl_sortkey
== $title->getText()
177 ||
$row->cl_sortkey
== $title->getPrefixedText()
181 # Custom sortkey, use it as a prefix
182 $prefix = $row->cl_sortkey
;
185 $prefix = $row->cl_sortkey_prefix
;
187 # cl_type will be wrong for lots of pages if cl_collation is 0,
188 # so let's update it while we're here.
189 $type = MWNamespace
::getCategoryLinkType( $title->getNamespace() );
190 $newSortKey = $collation->getSortKey(
191 $title->getCategorySortkey( $prefix ) );
192 if ( $verboseStats ) {
193 $this->updateSortKeySizeHistogram( $newSortKey );
197 // Add 1 to the count if the sortkey was changed. (Note that this doesn't count changes in
198 // other fields, if any, those usually only happen when upgrading old MediaWikis.)
199 $count +
= ( $row->cl_sortkey
!== $newSortKey );
204 'cl_sortkey' => $newSortKey,
205 'cl_sortkey_prefix' => $prefix,
206 'cl_collation' => $collationName,
208 'cl_timestamp = cl_timestamp',
210 [ 'cl_from' => $row->cl_from
, 'cl_to' => $row->cl_to
],
216 $batchConds = [ $this->getBatchCondition( $row, $dbw ) ];
220 $this->commitTransaction( $dbw, __METHOD__
);
224 $this->output( "$count rows would be updated so far.\n" );
226 $this->output( "$count done.\n" );
228 } while ( $res->numRows() == self
::BATCH_SIZE
);
231 $this->output( "$count rows processed\n" );
234 if ( $verboseStats ) {
235 $this->output( "\n" );
236 $this->showSortKeySizeHistogram();
241 * Return an SQL expression selecting rows which sort above the given row,
242 * assuming an ordering of cl_collation, cl_to, cl_type, cl_from
243 * @param stdClass $row
244 * @param IDatabase $dbw
247 function getBatchCondition( $row, $dbw ) {
248 if ( $this->hasOption( 'previous-collation' ) ) {
249 $fields = [ 'cl_to', 'cl_type', 'cl_from' ];
251 $fields = [ 'cl_collation', 'cl_to', 'cl_type', 'cl_from' ];
256 foreach ( $fields as $field ) {
257 if ( $dbw->getType() === 'mysql' && $field === 'cl_type' ) {
258 // Range conditions with enums are weird in mysql
259 // This must be a numeric literal, or it won't work.
260 $encValue = intval( $row->cl_type_numeric
);
262 $encValue = $dbw->addQuotes( $row->$field );
264 $inequality = "$field > $encValue";
265 $equality = "$field = $encValue";
271 $cond .= " OR ($prefix AND $inequality)";
272 $prefix .= " AND $equality";
279 function updateSortKeySizeHistogram( $key ) {
280 $length = strlen( $key );
281 if ( !isset( $this->sizeHistogram
[$length] ) ) {
282 $this->sizeHistogram
[$length] = 0;
284 $this->sizeHistogram
[$length]++
;
287 function showSortKeySizeHistogram() {
288 $maxLength = max( array_keys( $this->sizeHistogram
) );
289 if ( $maxLength == 0 ) {
293 $coarseHistogram = array_fill( 0, $numBins, 0 );
294 $coarseBoundaries = [];
296 for ( $i = 0; $i < $numBins - 1; $i++
) {
297 $boundary +
= $maxLength / $numBins;
298 $coarseBoundaries[$i] = round( $boundary );
300 $coarseBoundaries[$numBins - 1] = $maxLength +
1;
302 for ( $i = 0; $i <= $maxLength; $i++
) {
306 $val = $this->sizeHistogram
[$i] ??
0;
307 for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++
) {
308 if ( $coarseBoundaries[$coarseIndex] > $i ) {
309 $coarseHistogram[$coarseIndex] +
= $val;
313 if ( $coarseIndex == $numBins - 1 ) {
314 $coarseHistogram[$coarseIndex] +
= $val;
319 $this->output( "Sort key size histogram\nRaw data: $raw\n\n" );
321 $maxBinVal = max( $coarseHistogram );
322 $scale = 60 / $maxBinVal;
324 for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++
) {
325 $val = $coarseHistogram[$coarseIndex] ??
0;
326 $boundary = $coarseBoundaries[$coarseIndex];
327 $this->output( sprintf( "%-10s %-10d |%s\n",
328 $prevBoundary . '-' . ( $boundary - 1 ) . ': ',
330 str_repeat( '*', $scale * $val ) ) );
331 $prevBoundary = $boundary;
336 $maintClass = UpdateCollation
::class;
337 require_once RUN_MAINTENANCE_IF_MAIN
;