These results are based on our legacy PHP analysis, consider migrating to our new PHP analysis engine instead. Learn more
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3 | * Find all rows in the categorylinks table whose collation is out-of-date |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31 | * whose collation is out-of-date.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 = 100; // Number of rows to process in one batch |
||
37 | const SYNC_INTERVAL = 5; // Wait for replica DBs after this many batches |
||
38 | |||
39 | public $sizeHistogram = []; |
||
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47 | out-of-date (cl_collation != '$wgCategoryCollation') and repopulate cl_sortkey |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.', false, false, 'f' ); |
||
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 | $dbr = $this->getDB( DB_REPLICA ); |
||
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 | // Locally at least, (my local is a rather old version of mysql) |
||
90 | // mysql seems to filesort if there is both an equality |
||
91 | // (but not for an inequality) condition on cl_collation in the |
||
92 | // WHERE and it is also the first item in the ORDER BY. |
||
93 | if ( $this->hasOption( 'previous-collation' ) ) { |
||
94 | $orderBy = 'cl_to, cl_type, cl_from'; |
||
95 | } else { |
||
96 | $orderBy = 'cl_collation, cl_to, cl_type, cl_from'; |
||
97 | } |
||
98 | $options = [ |
||
99 | 'LIMIT' => self::BATCH_SIZE, |
||
100 | 'ORDER BY' => $orderBy, |
||
101 | 'STRAIGHT_JOIN' // per T58041 |
||
102 | ]; |
||
103 | |||
104 | if ( $force ) { |
||
105 | $collationConds = []; |
||
106 | } else { |
||
107 | if ( $this->hasOption( 'previous-collation' ) ) { |
||
108 | $collationConds['cl_collation'] = $this->getOption( 'previous-collation' ); |
||
109 | } else { |
||
110 | $collationConds = [ 0 => |
||
111 | 'cl_collation != ' . $dbw->addQuotes( $collationName ) |
||
112 | ]; |
||
113 | } |
||
114 | |||
115 | $count = $dbr->estimateRowCount( |
||
116 | 'categorylinks', |
||
117 | '*', |
||
118 | $collationConds, |
||
119 | __METHOD__ |
||
120 | ); |
||
121 | // Improve estimate if feasible |
||
122 | if ( $count < 1000000 ) { |
||
123 | $count = $dbr->selectField( |
||
124 | 'categorylinks', |
||
125 | 'COUNT(*)', |
||
126 | $collationConds, |
||
127 | __METHOD__ |
||
128 | ); |
||
129 | } |
||
130 | if ( $count == 0 ) { |
||
131 | $this->output( "Collations up-to-date.\n" ); |
||
132 | |||
133 | return; |
||
134 | } |
||
135 | if ( $dryRun ) { |
||
136 | $this->output( "$count rows would be updated.\n" ); |
||
137 | } else { |
||
138 | $this->output( "Fixing collation for $count rows.\n" ); |
||
139 | } |
||
140 | wfWaitForSlaves(); |
||
0 ignored issues
–
show
|
|||
141 | } |
||
142 | $count = 0; |
||
143 | $batchCount = 0; |
||
144 | $batchConds = []; |
||
145 | do { |
||
146 | $this->output( "Selecting next " . self::BATCH_SIZE . " rows..." ); |
||
147 | |||
148 | // cl_type must be selected as a number for proper paging because |
||
149 | // enums suck. |
||
150 | if ( $dbw->getType() === 'mysql' ) { |
||
151 | $clType = 'cl_type+0 AS "cl_type_numeric"'; |
||
152 | } else { |
||
153 | $clType = 'cl_type'; |
||
154 | } |
||
155 | $res = $dbw->select( |
||
156 | [ 'categorylinks', 'page' ], |
||
157 | [ 'cl_from', 'cl_to', 'cl_sortkey_prefix', 'cl_collation', |
||
158 | 'cl_sortkey', $clType, |
||
159 | 'page_namespace', 'page_title' |
||
160 | ], |
||
161 | array_merge( $collationConds, $batchConds, [ 'cl_from = page_id' ] ), |
||
162 | __METHOD__, |
||
163 | $options |
||
164 | ); |
||
165 | $this->output( " processing..." ); |
||
166 | |||
167 | if ( !$dryRun ) { |
||
168 | $this->beginTransaction( $dbw, __METHOD__ ); |
||
169 | } |
||
170 | foreach ( $res as $row ) { |
||
171 | $title = Title::newFromRow( $row ); |
||
172 | if ( !$row->cl_collation ) { |
||
173 | # This is an old-style row, so the sortkey needs to be |
||
174 | # converted. |
||
175 | if ( $row->cl_sortkey == $title->getText() |
||
176 | || $row->cl_sortkey == $title->getPrefixedText() |
||
177 | ) { |
||
178 | $prefix = ''; |
||
179 | } else { |
||
180 | # Custom sortkey, use it as a prefix |
||
181 | $prefix = $row->cl_sortkey; |
||
182 | } |
||
183 | } else { |
||
184 | $prefix = $row->cl_sortkey_prefix; |
||
185 | } |
||
186 | # cl_type will be wrong for lots of pages if cl_collation is 0, |
||
187 | # so let's update it while we're here. |
||
188 | if ( $title->getNamespace() == NS_CATEGORY ) { |
||
189 | $type = 'subcat'; |
||
190 | } elseif ( $title->getNamespace() == NS_FILE ) { |
||
191 | $type = 'file'; |
||
192 | } else { |
||
193 | $type = 'page'; |
||
194 | } |
||
195 | $newSortKey = $collation->getSortKey( |
||
196 | $title->getCategorySortkey( $prefix ) ); |
||
197 | if ( $verboseStats ) { |
||
198 | $this->updateSortKeySizeHistogram( $newSortKey ); |
||
199 | } |
||
200 | |||
201 | if ( !$dryRun ) { |
||
202 | $dbw->update( |
||
203 | 'categorylinks', |
||
204 | [ |
||
205 | 'cl_sortkey' => $newSortKey, |
||
206 | 'cl_sortkey_prefix' => $prefix, |
||
207 | 'cl_collation' => $collationName, |
||
208 | 'cl_type' => $type, |
||
209 | 'cl_timestamp = cl_timestamp', |
||
210 | ], |
||
211 | [ 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ], |
||
212 | __METHOD__ |
||
213 | ); |
||
214 | } |
||
215 | if ( $row ) { |
||
216 | $batchConds = [ $this->getBatchCondition( $row, $dbw ) ]; |
||
217 | } |
||
218 | } |
||
219 | if ( !$dryRun ) { |
||
220 | $this->commitTransaction( $dbw, __METHOD__ ); |
||
221 | } |
||
222 | |||
223 | $count += $res->numRows(); |
||
224 | $this->output( "$count done.\n" ); |
||
225 | |||
226 | if ( !$dryRun && ++$batchCount % self::SYNC_INTERVAL == 0 ) { |
||
227 | $this->output( "Waiting for replica DBs ... " ); |
||
228 | wfWaitForSlaves(); |
||
0 ignored issues
–
show
The function
wfWaitForSlaves() has been deprecated with message: since 1.27 Use LBFactory::waitForReplication
This function has been deprecated. The supplier of the file has supplied an explanatory message. The explanatory message should give you some clue as to whether and when the function will be removed from the class and what other function to use instead.
Loading history...
|
|||
229 | $this->output( "done\n" ); |
||
230 | } |
||
231 | } while ( $res->numRows() == self::BATCH_SIZE ); |
||
232 | |||
233 | $this->output( "$count rows processed\n" ); |
||
234 | |||
235 | if ( $verboseStats ) { |
||
236 | $this->output( "\n" ); |
||
237 | $this->showSortKeySizeHistogram(); |
||
238 | } |
||
239 | } |
||
240 | |||
241 | /** |
||
242 | * Return an SQL expression selecting rows which sort above the given row, |
||
243 | * assuming an ordering of cl_collation, cl_to, cl_type, cl_from |
||
244 | * @param stdClass $row |
||
245 | * @param Database $dbw |
||
246 | * @return string |
||
247 | */ |
||
248 | function getBatchCondition( $row, $dbw ) { |
||
249 | if ( $this->hasOption( 'previous-collation' ) ) { |
||
250 | $fields = [ 'cl_to', 'cl_type', 'cl_from' ]; |
||
251 | } else { |
||
252 | $fields = [ 'cl_collation', 'cl_to', 'cl_type', 'cl_from' ]; |
||
253 | } |
||
254 | $first = true; |
||
255 | $cond = false; |
||
256 | $prefix = false; |
||
257 | foreach ( $fields as $field ) { |
||
258 | if ( $dbw->getType() === 'mysql' && $field === 'cl_type' ) { |
||
259 | // Range conditions with enums are weird in mysql |
||
260 | // This must be a numeric literal, or it won't work. |
||
261 | $encValue = intval( $row->cl_type_numeric ); |
||
262 | } else { |
||
263 | $encValue = $dbw->addQuotes( $row->$field ); |
||
264 | } |
||
265 | $inequality = "$field > $encValue"; |
||
266 | $equality = "$field = $encValue"; |
||
267 | if ( $first ) { |
||
268 | $cond = $inequality; |
||
269 | $prefix = $equality; |
||
270 | $first = false; |
||
271 | } else { |
||
272 | $cond .= " OR ($prefix AND $inequality)"; |
||
273 | $prefix .= " AND $equality"; |
||
274 | } |
||
275 | } |
||
276 | |||
277 | return $cond; |
||
278 | } |
||
279 | |||
280 | function updateSortKeySizeHistogram( $key ) { |
||
281 | $length = strlen( $key ); |
||
282 | if ( !isset( $this->sizeHistogram[$length] ) ) { |
||
283 | $this->sizeHistogram[$length] = 0; |
||
284 | } |
||
285 | $this->sizeHistogram[$length]++; |
||
286 | } |
||
287 | |||
288 | function showSortKeySizeHistogram() { |
||
289 | $maxLength = max( array_keys( $this->sizeHistogram ) ); |
||
290 | if ( $maxLength == 0 ) { |
||
291 | return; |
||
292 | } |
||
293 | $numBins = 20; |
||
294 | $coarseHistogram = array_fill( 0, $numBins, 0 ); |
||
295 | $coarseBoundaries = []; |
||
296 | $boundary = 0; |
||
297 | for ( $i = 0; $i < $numBins - 1; $i++ ) { |
||
298 | $boundary += $maxLength / $numBins; |
||
299 | $coarseBoundaries[$i] = round( $boundary ); |
||
300 | } |
||
301 | $coarseBoundaries[$numBins - 1] = $maxLength + 1; |
||
302 | $raw = ''; |
||
303 | for ( $i = 0; $i <= $maxLength; $i++ ) { |
||
304 | if ( $raw !== '' ) { |
||
305 | $raw .= ', '; |
||
306 | } |
||
307 | if ( !isset( $this->sizeHistogram[$i] ) ) { |
||
308 | $val = 0; |
||
309 | } else { |
||
310 | $val = $this->sizeHistogram[$i]; |
||
311 | } |
||
312 | for ( $coarseIndex = 0; $coarseIndex < $numBins - 1; $coarseIndex++ ) { |
||
313 | if ( $coarseBoundaries[$coarseIndex] > $i ) { |
||
314 | $coarseHistogram[$coarseIndex] += $val; |
||
315 | break; |
||
316 | } |
||
317 | } |
||
318 | if ( $coarseIndex == $numBins - 1 ) { |
||
319 | $coarseHistogram[$coarseIndex] += $val; |
||
320 | } |
||
321 | $raw .= $val; |
||
322 | } |
||
323 | |||
324 | $this->output( "Sort key size histogram\nRaw data: $raw\n\n" ); |
||
325 | |||
326 | $maxBinVal = max( $coarseHistogram ); |
||
327 | $scale = 60 / $maxBinVal; |
||
328 | $prevBoundary = 0; |
||
329 | for ( $coarseIndex = 0; $coarseIndex < $numBins; $coarseIndex++ ) { |
||
330 | if ( !isset( $coarseHistogram[$coarseIndex] ) ) { |
||
331 | $val = 0; |
||
332 | } else { |
||
333 | $val = $coarseHistogram[$coarseIndex]; |
||
334 | } |
||
335 | $boundary = $coarseBoundaries[$coarseIndex]; |
||
336 | $this->output( sprintf( "%-10s %-10d |%s\n", |
||
337 | $prevBoundary . '-' . ( $boundary - 1 ) . ': ', |
||
338 | $val, |
||
339 | str_repeat( '*', $scale * $val ) ) ); |
||
340 | $prevBoundary = $boundary; |
||
341 | } |
||
342 | } |
||
343 | } |
||
344 | |||
345 | $maintClass = "UpdateCollation"; |
||
346 | require_once RUN_MAINTENANCE_IF_MAIN; |
||
347 |
This function has been deprecated. The supplier of the file has supplied an explanatory message.
The explanatory message should give you some clue as to whether and when the function will be removed from the class and what other function to use instead.