MediaWiki  1.30.0
refreshLinks.php
Go to the documentation of this file.
1 <?php
25 
26 require_once __DIR__ . '/Maintenance.php';
27 
33 class RefreshLinks extends Maintenance {
34  const REPORTING_INTERVAL = 100;
35 
37  protected $namespace = false;
38 
39  public function __construct() {
40  parent::__construct();
41  $this->addDescription( 'Refresh link tables' );
42  $this->addOption( 'dfn-only', 'Delete links from nonexistent articles only' );
43  $this->addOption( 'new-only', 'Only affect articles with just a single edit' );
44  $this->addOption( 'redirects-only', 'Only fix redirects, not all links' );
45  $this->addOption( 'old-redirects-only', 'Only fix redirects with no redirect table entry' );
46  $this->addOption( 'e', 'Last page id to refresh', false, true );
47  $this->addOption( 'dfn-chunk-size', 'Maximum number of existent IDs to check per ' .
48  'query, default 100000', false, true );
49  $this->addOption( 'namespace', 'Only fix pages in this namespace', false, true );
50  $this->addOption( 'category', 'Only fix pages in this category', false, true );
51  $this->addOption( 'tracking-category', 'Only fix pages in this tracking category', false, true );
52  $this->addArg( 'start', 'Page_id to start from, default 1', false );
53  $this->setBatchSize( 100 );
54  }
55 
56  public function execute() {
57  // Note that there is a difference between not specifying the start
58  // and end IDs and using the minimum and maximum values from the page
59  // table. In the latter case, deleteLinksFromNonexistent() will not
60  // delete entries for nonexistent IDs that fall outside the range.
61  $start = (int)$this->getArg( 0 ) ?: null;
62  $end = (int)$this->getOption( 'e' ) ?: null;
63  $dfnChunkSize = (int)$this->getOption( 'dfn-chunk-size', 100000 );
64  $ns = $this->getOption( 'namespace' );
65  if ( $ns === null ) {
66  $this->namespace = false;
67  } else {
68  $this->namespace = (int)$ns;
69  }
70  if ( ( $category = $this->getOption( 'category', false ) ) !== false ) {
71  $title = Title::makeTitleSafe( NS_CATEGORY, $category );
72  if ( !$title ) {
73  $this->error( "'$category' is an invalid category name!\n", true );
74  }
75  $this->refreshCategory( $title );
76  } elseif ( ( $category = $this->getOption( 'tracking-category', false ) ) !== false ) {
77  $this->refreshTrackingCategory( $category );
78  } elseif ( !$this->hasOption( 'dfn-only' ) ) {
79  $new = $this->hasOption( 'new-only' );
80  $redir = $this->hasOption( 'redirects-only' );
81  $oldRedir = $this->hasOption( 'old-redirects-only' );
82  $this->doRefreshLinks( $start, $new, $end, $redir, $oldRedir );
83  $this->deleteLinksFromNonexistent( null, null, $this->mBatchSize, $dfnChunkSize );
84  } else {
85  $this->deleteLinksFromNonexistent( $start, $end, $this->mBatchSize, $dfnChunkSize );
86  }
87  }
88 
89  private function namespaceCond() {
90  return $this->namespace !== false
91  ? [ 'page_namespace' => $this->namespace ]
92  : [];
93  }
94 
103  private function doRefreshLinks( $start, $newOnly = false,
104  $end = null, $redirectsOnly = false, $oldRedirectsOnly = false
105  ) {
106  $dbr = $this->getDB( DB_REPLICA, [ 'vslow' ] );
107 
108  if ( $start === null ) {
109  $start = 1;
110  }
111 
112  // Give extensions a chance to optimize settings
113  Hooks::run( 'MaintenanceRefreshLinksInit', [ $this ] );
114 
115  $what = $redirectsOnly ? "redirects" : "links";
116 
117  if ( $oldRedirectsOnly ) {
118  # This entire code path is cut-and-pasted from below. Hurrah.
119 
120  $conds = [
121  "page_is_redirect=1",
122  "rd_from IS NULL",
123  self::intervalCond( $dbr, 'page_id', $start, $end ),
124  ] + $this->namespaceCond();
125 
126  $res = $dbr->select(
127  [ 'page', 'redirect' ],
128  'page_id',
129  $conds,
130  __METHOD__,
131  [],
132  [ 'redirect' => [ "LEFT JOIN", "page_id=rd_from" ] ]
133  );
134  $num = $res->numRows();
135  $this->output( "Refreshing $num old redirects from $start...\n" );
136 
137  $i = 0;
138 
139  foreach ( $res as $row ) {
140  if ( !( ++$i % self::REPORTING_INTERVAL ) ) {
141  $this->output( "$i\n" );
142  wfWaitForSlaves();
143  }
144  $this->fixRedirect( $row->page_id );
145  }
146  } elseif ( $newOnly ) {
147  $this->output( "Refreshing $what from " );
148  $res = $dbr->select( 'page',
149  [ 'page_id' ],
150  [
151  'page_is_new' => 1,
152  self::intervalCond( $dbr, 'page_id', $start, $end ),
153  ] + $this->namespaceCond(),
154  __METHOD__
155  );
156  $num = $res->numRows();
157  $this->output( "$num new articles...\n" );
158 
159  $i = 0;
160  foreach ( $res as $row ) {
161  if ( !( ++$i % self::REPORTING_INTERVAL ) ) {
162  $this->output( "$i\n" );
163  wfWaitForSlaves();
164  }
165  if ( $redirectsOnly ) {
166  $this->fixRedirect( $row->page_id );
167  } else {
168  self::fixLinksFromArticle( $row->page_id, $this->namespace );
169  }
170  }
171  } else {
172  if ( !$end ) {
173  $maxPage = $dbr->selectField( 'page', 'max(page_id)', false );
174  $maxRD = $dbr->selectField( 'redirect', 'max(rd_from)', false );
175  $end = max( $maxPage, $maxRD );
176  }
177  $this->output( "Refreshing redirects table.\n" );
178  $this->output( "Starting from page_id $start of $end.\n" );
179 
180  for ( $id = $start; $id <= $end; $id++ ) {
181  if ( !( $id % self::REPORTING_INTERVAL ) ) {
182  $this->output( "$id\n" );
183  wfWaitForSlaves();
184  }
185  $this->fixRedirect( $id );
186  }
187 
188  if ( !$redirectsOnly ) {
189  $this->output( "Refreshing links tables.\n" );
190  $this->output( "Starting from page_id $start of $end.\n" );
191 
192  for ( $id = $start; $id <= $end; $id++ ) {
193  if ( !( $id % self::REPORTING_INTERVAL ) ) {
194  $this->output( "$id\n" );
195  wfWaitForSlaves();
196  }
197  self::fixLinksFromArticle( $id, $this->namespace );
198  }
199  }
200  }
201  }
202 
215  private function fixRedirect( $id ) {
216  $page = WikiPage::newFromID( $id );
217  $dbw = $this->getDB( DB_MASTER );
218 
219  if ( $page === null ) {
220  // This page doesn't exist (any more)
221  // Delete any redirect table entry for it
222  $dbw->delete( 'redirect', [ 'rd_from' => $id ],
223  __METHOD__ );
224 
225  return;
226  } elseif ( $this->namespace !== false
227  && !$page->getTitle()->inNamespace( $this->namespace )
228  ) {
229  return;
230  }
231 
232  $rt = null;
233  $content = $page->getContent( Revision::RAW );
234  if ( $content !== null ) {
235  $rt = $content->getUltimateRedirectTarget();
236  }
237 
238  if ( $rt === null ) {
239  // The page is not a redirect
240  // Delete any redirect table entry for it
241  $dbw->delete( 'redirect', [ 'rd_from' => $id ], __METHOD__ );
242  $fieldValue = 0;
243  } else {
244  $page->insertRedirectEntry( $rt );
245  $fieldValue = 1;
246  }
247 
248  // Update the page table to be sure it is an a consistent state
249  $dbw->update( 'page', [ 'page_is_redirect' => $fieldValue ],
250  [ 'page_id' => $id ], __METHOD__ );
251  }
252 
258  public static function fixLinksFromArticle( $id, $ns = false ) {
259  $page = WikiPage::newFromID( $id );
260 
261  LinkCache::singleton()->clear();
262 
263  if ( $page === null ) {
264  return;
265  } elseif ( $ns !== false
266  && !$page->getTitle()->inNamespace( $ns ) ) {
267  return;
268  }
269 
270  $content = $page->getContent( Revision::RAW );
271  if ( $content === null ) {
272  return;
273  }
274 
275  $updates = $content->getSecondaryDataUpdates(
276  $page->getTitle(), /* $old = */ null, /* $recursive = */ false );
277  foreach ( $updates as $update ) {
278  DeferredUpdates::addUpdate( $update );
280  }
281  }
282 
294  private function deleteLinksFromNonexistent( $start = null, $end = null, $batchSize = 100,
295  $chunkSize = 100000
296  ) {
297  wfWaitForSlaves();
298  $this->output( "Deleting illegal entries from the links tables...\n" );
299  $dbr = $this->getDB( DB_REPLICA, [ 'vslow' ] );
300  do {
301  // Find the start of the next chunk. This is based only
302  // on existent page_ids.
303  $nextStart = $dbr->selectField(
304  'page',
305  'page_id',
306  [ self::intervalCond( $dbr, 'page_id', $start, $end ) ]
307  + $this->namespaceCond(),
308  __METHOD__,
309  [ 'ORDER BY' => 'page_id', 'OFFSET' => $chunkSize ]
310  );
311 
312  if ( $nextStart !== false ) {
313  // To find the end of the current chunk, subtract one.
314  // This will serve to limit the number of rows scanned in
315  // dfnCheckInterval(), per query, to at most the sum of
316  // the chunk size and deletion batch size.
317  $chunkEnd = $nextStart - 1;
318  } else {
319  // This is the last chunk. Check all page_ids up to $end.
320  $chunkEnd = $end;
321  }
322 
323  $fmtStart = $start !== null ? "[$start" : '(-INF';
324  $fmtChunkEnd = $chunkEnd !== null ? "$chunkEnd]" : 'INF)';
325  $this->output( " Checking interval $fmtStart, $fmtChunkEnd\n" );
326  $this->dfnCheckInterval( $start, $chunkEnd, $batchSize );
327 
328  $start = $nextStart;
329 
330  } while ( $nextStart !== false );
331  }
332 
339  private function dfnCheckInterval( $start = null, $end = null, $batchSize = 100 ) {
340  $dbw = $this->getDB( DB_MASTER );
341  $dbr = $this->getDB( DB_REPLICA, [ 'vslow' ] );
342 
343  $linksTables = [ // table name => page_id field
344  'pagelinks' => 'pl_from',
345  'imagelinks' => 'il_from',
346  'categorylinks' => 'cl_from',
347  'templatelinks' => 'tl_from',
348  'externallinks' => 'el_from',
349  'iwlinks' => 'iwl_from',
350  'langlinks' => 'll_from',
351  'redirect' => 'rd_from',
352  'page_props' => 'pp_page',
353  ];
354 
355  foreach ( $linksTables as $table => $field ) {
356  $this->output( " $table: 0" );
357  $tableStart = $start;
358  $counter = 0;
359  do {
360  $ids = $dbr->selectFieldValues(
361  $table,
362  $field,
363  [
364  self::intervalCond( $dbr, $field, $tableStart, $end ),
365  "$field NOT IN ({$dbr->selectSQLText( 'page', 'page_id' )})",
366  ],
367  __METHOD__,
368  [ 'DISTINCT', 'ORDER BY' => $field, 'LIMIT' => $batchSize ]
369  );
370 
371  $numIds = count( $ids );
372  if ( $numIds ) {
373  $counter += $numIds;
374  $dbw->delete( $table, [ $field => $ids ], __METHOD__ );
375  $this->output( ", $counter" );
376  $tableStart = $ids[$numIds - 1] + 1;
377  wfWaitForSlaves();
378  }
379 
380  } while ( $numIds >= $batchSize && ( $end === null || $tableStart <= $end ) );
381 
382  $this->output( " deleted.\n" );
383  }
384  }
385 
398  private static function intervalCond( IDatabase $db, $var, $start, $end ) {
399  if ( $start === null && $end === null ) {
400  return "$var IS NOT NULL";
401  } elseif ( $end === null ) {
402  return "$var >= {$db->addQuotes( $start )}";
403  } elseif ( $start === null ) {
404  return "$var <= {$db->addQuotes( $end )}";
405  } else {
406  return "$var BETWEEN {$db->addQuotes( $start )} AND {$db->addQuotes( $end )}";
407  }
408  }
409 
415  private function refreshTrackingCategory( $category ) {
416  $cats = $this->getPossibleCategories( $category );
417 
418  if ( !$cats ) {
419  $this->error( "Tracking category '$category' is disabled\n" );
420  // Output to stderr but don't bail out,
421  }
422 
423  foreach ( $cats as $cat ) {
424  $this->refreshCategory( $cat );
425  }
426  }
427 
433  private function refreshCategory( Title $category ) {
434  $this->output( "Refreshing pages in category '{$category->getText()}'...\n" );
435 
436  $dbr = $this->getDB( DB_REPLICA );
437  $conds = [
438  'page_id=cl_from',
439  'cl_to' => $category->getDBkey(),
440  ];
441  if ( $this->namespace !== false ) {
442  $conds['page_namespace'] = $this->namespace;
443  }
444 
445  $i = 0;
446  $timestamp = '';
447  $lastId = 0;
448  do {
449  $finalConds = $conds;
450  $timestamp = $dbr->addQuotes( $timestamp );
451  $finalConds [] =
452  "(cl_timestamp > $timestamp OR (cl_timestamp = $timestamp AND cl_from > $lastId))";
453  $res = $dbr->select( [ 'page', 'categorylinks' ],
454  [ 'page_id', 'cl_timestamp' ],
455  $finalConds,
456  __METHOD__,
457  [
458  'ORDER BY' => [ 'cl_timestamp', 'cl_from' ],
459  'LIMIT' => $this->mBatchSize,
460  ]
461  );
462 
463  foreach ( $res as $row ) {
464  if ( !( ++$i % self::REPORTING_INTERVAL ) ) {
465  $this->output( "$i\n" );
466  wfWaitForSlaves();
467  }
468  $lastId = $row->page_id;
469  $timestamp = $row->cl_timestamp;
470  self::fixLinksFromArticle( $row->page_id );
471  }
472 
473  } while ( $res->numRows() == $this->mBatchSize );
474  }
475 
482  private function getPossibleCategories( $categoryKey ) {
483  $trackingCategories = new TrackingCategories( $this->getConfig() );
484  $cats = $trackingCategories->getTrackingCategories();
485  if ( isset( $cats[$categoryKey] ) ) {
486  return $cats[$categoryKey]['cats'];
487  }
488  $this->error( "Unknown tracking category {$categoryKey}\n", true );
489  }
490 }
491 
492 $maintClass = 'RefreshLinks';
493 require_once RUN_MAINTENANCE_IF_MAIN;
Maintenance\$mBatchSize
int $mBatchSize
Batch size.
Definition: Maintenance.php:103
false
processing should stop and the error should be shown to the user * false
Definition: hooks.txt:187
captcha-old.count
count
Definition: captcha-old.py:249
Maintenance\addDescription
addDescription( $text)
Set the description text.
Definition: Maintenance.php:287
use
as see the revision history and available at free of to any person obtaining a copy of this software and associated documentation to deal in the Software without including without limitation the rights to use
Definition: MIT-LICENSE.txt:10
DeferredUpdates\addUpdate
static addUpdate(DeferrableUpdate $update, $stage=self::POSTSEND)
Add an update to the deferred list to be run later by execute()
Definition: DeferredUpdates.php:76
RUN_MAINTENANCE_IF_MAIN
require_once RUN_MAINTENANCE_IF_MAIN
Definition: maintenance.txt:50
$res
$res
Definition: database.txt:21
Maintenance
Abstract maintenance class for quickly writing and churning out maintenance scripts with minimal effo...
Definition: maintenance.txt:39
wfWaitForSlaves
wfWaitForSlaves( $ifWritesSince=null, $wiki=false, $cluster=false, $timeout=null)
Waits for the replica DBs to catch up to the master position.
Definition: GlobalFunctions.php:3010
php
injection txt This is an overview of how MediaWiki makes use of dependency injection The design described here grew from the discussion of RFC T384 The term dependency this means that anything an object needs to operate should be injected from the the object itself should only know narrow no concrete implementation of the logic it relies on The requirement to inject everything typically results in an architecture that based on two main types of and essentially stateless service objects that use other service objects to operate on the value objects As of the beginning MediaWiki is only starting to use the DI approach Much of the code still relies on global state or direct resulting in a highly cyclical dependency which acts as the top level factory for services in MediaWiki which can be used to gain access to default instances of various services MediaWikiServices however also allows new services to be defined and default services to be redefined Services are defined or redefined by providing a callback the instantiator that will return a new instance of the service When it will create an instance of MediaWikiServices and populate it with the services defined in the files listed by thereby bootstrapping the DI framework Per $wgServiceWiringFiles lists includes ServiceWiring php
Definition: injection.txt:35
Wikimedia\Rdbms\IDatabase
Basic database interface for live and lazy-loaded relation database handles.
Definition: IDatabase.php:40
REPORTING_INTERVAL
const REPORTING_INTERVAL
Definition: moveToExternal.php:24
Title\getDBkey
getDBkey()
Get the main part with underscores.
Definition: Title.php:955
$title
namespace and then decline to actually register it file or subcat img or subcat $title
Definition: hooks.txt:932
Maintenance\getConfig
getConfig()
Definition: Maintenance.php:504
Maintenance\addOption
addOption( $name, $description, $required=false, $withArg=false, $shortName=false, $multiOccurrence=false)
Add a parameter to the script.
Definition: Maintenance.php:215
DB_REPLICA
const DB_REPLICA
Definition: defines.php:25
NS_CATEGORY
const NS_CATEGORY
Definition: Defines.php:79
DB_MASTER
const DB_MASTER
Definition: defines.php:26
Title\makeTitleSafe
static makeTitleSafe( $ns, $title, $fragment='', $interwiki='')
Create a new Title from a namespace index and a DB key.
Definition: Title.php:557
WikiPage\newFromID
static newFromID( $id, $from='fromdb')
Constructor from a page id.
Definition: WikiPage.php:159
Revision\RAW
const RAW
Definition: Revision.php:100
DeferredUpdates\doUpdates
static doUpdates( $mode='run', $stage=self::ALL)
Do any deferred updates and clear the list.
Definition: DeferredUpdates.php:123
TrackingCategories
This class performs some operations related to tracking categories, such as creating a list of all su...
Definition: TrackingCategories.php:27
Title
Represents a title within MediaWiki.
Definition: Title.php:39
Maintenance\getOption
getOption( $name, $default=null)
Get an option, or return the default.
Definition: Maintenance.php:250
$dbr
if(! $regexes) $dbr
Definition: cleanup.php:94
LinkCache\singleton
static singleton()
Get an instance of this class.
Definition: LinkCache.php:67
Maintenance\addArg
addArg( $arg, $description, $required=true)
Add some args that are needed.
Definition: Maintenance.php:267
as
This document is intended to provide useful advice for parties seeking to redistribute MediaWiki to end users It s targeted particularly at maintainers for Linux since it s been observed that distribution packages of MediaWiki often break We ve consistently had to recommend that users seeking support use official tarballs instead of their distribution s and this often solves whatever problem the user is having It would be nice if this could such as
Definition: distributors.txt:9
Maintenance\getDB
getDB( $db, $groups=[], $wiki=false)
Returns a database to be used by current maintenance script.
Definition: Maintenance.php:1251
Maintenance\error
error( $err, $die=0)
Throw an error to the user.
Definition: Maintenance.php:392
Maintenance\output
output( $out, $channel=null)
Throw some output to the user.
Definition: Maintenance.php:373
Maintenance\hasOption
hasOption( $name)
Checks to see if a particular param exists.
Definition: Maintenance.php:236
Maintenance\getArg
getArg( $argId=0, $default=null)
Get an argument.
Definition: Maintenance.php:306
Hooks\run
static run( $event, array $args=[], $deprecatedVersion=null)
Call hook functions defined in Hooks::register and $wgHooks.
Definition: Hooks.php:203
Maintenance\setBatchSize
setBatchSize( $s=0)
Set the batch size.
Definition: Maintenance.php:314