maintenance/rebuildrecentchanges.php

   1 <?php
   2 /**
   3  * Rebuild recent changes from scratch.  This takes several hours,
   4  * depending on the database size and server configuration.
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License along
  17  * with this program; if not, write to the Free Software Foundation, Inc.,
  18  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @file
  22  * @ingroup Maintenance
  23  * @todo Document
  24  */
  25
  26 require_once __DIR__ . '/Maintenance.php';
  27
  28 /**
  29  * Maintenance script that rebuilds recent changes from scratch.
  30  *
  31  * @ingroup Maintenance
  32  */
  33 class RebuildRecentchanges extends Maintenance {
  34         public function __construct() {
  35                 parent::__construct();
  36                 $this->addDescription( 'Rebuild recent changes' );
  37
  38                 $this->addOption(
  39                         'from',
  40                         "Don't empty the table, only insert rows in requested time range (in YYYYMMDDHHMMSS format)",
  41                         false,
  42                         true
  43                 );
  44                 $this->addOption(
  45                         'to',
  46                         "Don't empty the table, only insert rows in requested time range (in YYYYMMDDHHMMSS format)",
  47                         false,
  48                         true
  49                 );
  50         }
  51
  52         public function execute() {
  53                 if (
  54                         ( $this->hasOption( 'from' ) && !$this->hasOption( 'to' ) ) ||
  55                         ( !$this->hasOption( 'from' ) && $this->hasOption( 'to' ) )
  56                 ) {
  57                         $this->error( "Both 'from' and 'to' must be given, or neither", 1 );
  58                 }
  59
  60                 $this->rebuildRecentChangesTablePass1();
  61                 $this->rebuildRecentChangesTablePass2();
  62                 $this->rebuildRecentChangesTablePass3();
  63                 $this->rebuildRecentChangesTablePass4();
  64                 $this->rebuildRecentChangesTablePass5();
  65                 if ( !( $this->hasOption( 'from' ) && $this->hasOption( 'to' ) ) ) {
  66                         $this->purgeFeeds();
  67                 }
  68                 $this->output( "Done.\n" );
  69         }
  70
  71         /**
  72          * Rebuild pass 1: Insert `recentchanges` entries for page revisions.
  73          */
  74         private function rebuildRecentChangesTablePass1() {
  75                 $dbw = $this->getDB( DB_MASTER );
  76
  77                 if ( $this->hasOption( 'from' ) && $this->hasOption( 'to' ) ) {
  78                         $this->cutoffFrom = wfTimestamp( TS_UNIX, $this->getOption( 'from' ) );
  79                         $this->cutoffTo = wfTimestamp( TS_UNIX, $this->getOption( 'to' ) );
  80
  81                         $sec = $this->cutoffTo - $this->cutoffFrom;
  82                         $days = $sec / 24 / 3600;
  83                         $this->output( "Rebuilding range of $sec seconds ($days days)\n" );
  84
  85                 } else {
  86                         global $wgRCMaxAge;
  87
  88                         $days = $wgRCMaxAge / 24 / 3600;
  89                         $this->output( "Rebuilding \$wgRCMaxAge=$wgRCMaxAge seconds ($days days)\n" );
  90
  91                         $this->cutoffFrom = time() - $wgRCMaxAge;
  92                         $this->cutoffTo = time();
  93
  94                         $this->output( "Clearing recentchanges table...\n" );
  95                         $dbw->delete( 'recentchanges', '*' );
  96                 }
  97
  98                 $this->output( "Loading from page and revision tables...\n" );
  99
 100                 $dbw->insertSelect( 'recentchanges', [ 'page', 'revision' ],
 101                         [
 102                                 'rc_timestamp' => 'rev_timestamp',
 103                                 'rc_user' => 'rev_user',
 104                                 'rc_user_text' => 'rev_user_text',
 105                                 'rc_namespace' => 'page_namespace',
 106                                 'rc_title' => 'page_title',
 107                                 'rc_comment' => 'rev_comment',
 108                                 'rc_minor' => 'rev_minor_edit',
 109                                 'rc_bot' => 0,
 110                                 'rc_new' => 'page_is_new',
 111                                 'rc_cur_id' => 'page_id',
 112                                 'rc_this_oldid' => 'rev_id',
 113                                 'rc_last_oldid' => 0, // is this ok?
 114                                 'rc_type' => $dbw->conditional( 'page_is_new != 0', RC_NEW, RC_EDIT ),
 115                                 'rc_source' => $dbw->conditional(
 116                                                 'page_is_new != 0',
 117                                                 $dbw->addQuotes( RecentChange::SRC_NEW ),
 118                                                 $dbw->addQuotes( RecentChange::SRC_EDIT )
 119                                 ),
 120                                 'rc_deleted' => 'rev_deleted'
 121                         ],
 122                         [
 123                                 'rev_timestamp > ' . $dbw->addQuotes( $dbw->timestamp( $this->cutoffFrom ) ),
 124                                 'rev_timestamp < ' . $dbw->addQuotes( $dbw->timestamp( $this->cutoffTo ) ),
 125                                 'rev_page=page_id'
 126                         ],
 127                         __METHOD__,
 128                         [], // INSERT options
 129                         [ 'ORDER BY' => 'rev_timestamp DESC', 'LIMIT' => 5000 ] // SELECT options
 130                 );
 131         }
 132
 133         /**
 134          * Rebuild pass 2: Enhance entries for page revisions with references to the previous revision
 135          * (rc_last_oldid, rc_new etc.) and size differences (rc_old_len, rc_new_len).
 136          */
 137         private function rebuildRecentChangesTablePass2() {
 138                 $dbw = $this->getDB( DB_MASTER );
 139                 list( $recentchanges, $revision ) = $dbw->tableNamesN( 'recentchanges', 'revision' );
 140
 141                 $this->output( "Updating links and size differences...\n" );
 142
 143                 # Fill in the rc_last_oldid field, which points to the previous edit
 144                 $sql = "SELECT rc_cur_id,rc_this_oldid,rc_timestamp FROM $recentchanges " .
 145                         "WHERE rc_timestamp > " . $dbw->addQuotes( $dbw->timestamp( $this->cutoffFrom ) ) . ' ' .
 146                         "AND rc_timestamp < " . $dbw->addQuotes( $dbw->timestamp( $this->cutoffTo ) ) . ' ' .
 147                         "ORDER BY rc_cur_id,rc_timestamp";
 148                 $res = $dbw->query( $sql, DB_MASTER );
 149
 150                 $lastCurId = 0;
 151                 $lastOldId = 0;
 152                 foreach ( $res as $obj ) {
 153                         $new = 0;
 154                         if ( $obj->rc_cur_id != $lastCurId ) {
 155                                 # Switch! Look up the previous last edit, if any
 156                                 $lastCurId = intval( $obj->rc_cur_id );
 157                                 $emit = $obj->rc_timestamp;
 158                                 $sql2 = "SELECT rev_id,rev_len FROM $revision " .
 159                                         "WHERE rev_page={$lastCurId} " .
 160                                         "AND rev_timestamp<'{$emit}' ORDER BY rev_timestamp DESC";
 161                                 $sql2 = $dbw->limitResult( $sql2, 1, false );
 162                                 $res2 = $dbw->query( $sql2 );
 163                                 $row = $dbw->fetchObject( $res2 );
 164                                 if ( $row ) {
 165                                         $lastOldId = intval( $row->rev_id );
 166                                         # Grab the last text size if available
 167                                         $lastSize = !is_null( $row->rev_len ) ? intval( $row->rev_len ) : null;
 168                                 } else {
 169                                         # No previous edit
 170                                         $lastOldId = 0;
 171                                         $lastSize = null;
 172                                         $new = 1; // probably true
 173                                 }
 174                         }
 175                         if ( $lastCurId == 0 ) {
 176                                 $this->output( "Uhhh, something wrong? No curid\n" );
 177                         } else {
 178                                 # Grab the entry's text size
 179                                 $size = $dbw->selectField( 'revision', 'rev_len', [ 'rev_id' => $obj->rc_this_oldid ] );
 180
 181                                 $dbw->update( 'recentchanges',
 182                                         [
 183                                                 'rc_last_oldid' => $lastOldId,
 184                                                 'rc_new' => $new,
 185                                                 'rc_type' => $new,
 186                                                 'rc_source' => $new === 1 ? RecentChange::SRC_NEW : RecentChange::SRC_EDIT,
 187                                                 'rc_old_len' => $lastSize,
 188                                                 'rc_new_len' => $size,
 189                                         ], [
 190                                                 'rc_cur_id' => $lastCurId,
 191                                                 'rc_this_oldid' => $obj->rc_this_oldid,
 192                                         ],
 193                                         __METHOD__
 194                                 );
 195
 196                                 $lastOldId = intval( $obj->rc_this_oldid );
 197                                 $lastSize = $size;
 198                         }
 199                 }
 200         }
 201
 202         /**
 203          * Rebuild pass 3: Insert `recentchanges` entries for action logs.
 204          */
 205         private function rebuildRecentChangesTablePass3() {
 206                 $dbw = $this->getDB( DB_MASTER );
 207
 208                 $this->output( "Loading from user, page, and logging tables...\n" );
 209
 210                 global $wgLogTypes, $wgLogRestrictions;
 211                 // Some logs don't go in RC. This should check for that
 212                 $basicRCLogs = array_diff( $wgLogTypes, array_keys( $wgLogRestrictions ) );
 213
 214                 list( $logging, $page ) = $dbw->tableNamesN( 'logging', 'page' );
 215                 $dbw->insertSelect(
 216                         'recentchanges',
 217                         [
 218                                 'user',
 219                                 "$logging LEFT JOIN $page ON (log_namespace=page_namespace AND log_title=page_title)"
 220                         ],
 221                         [
 222                                 'rc_timestamp' => 'log_timestamp',
 223                                 'rc_user' => 'log_user',
 224                                 'rc_user_text' => 'user_name',
 225                                 'rc_namespace' => 'log_namespace',
 226                                 'rc_title' => 'log_title',
 227                                 'rc_comment' => 'log_comment',
 228                                 'rc_minor' => 0,
 229                                 'rc_bot' => 0,
 230                                 'rc_patrolled' => 1,
 231                                 'rc_new' => 0,
 232                                 'rc_this_oldid' => 0,
 233                                 'rc_last_oldid' => 0,
 234                                 'rc_type' => RC_LOG,
 235                                 'rc_source' => $dbw->addQuotes( RecentChange::SRC_LOG ),
 236                                 'rc_cur_id' => $dbw->cascadingDeletes() ? 'page_id' : 'COALESCE(page_id, 0)',
 237                                 'rc_log_type' => 'log_type',
 238                                 'rc_log_action' => 'log_action',
 239                                 'rc_logid' => 'log_id',
 240                                 'rc_params' => 'log_params',
 241                                 'rc_deleted' => 'log_deleted'
 242                         ],
 243                         [
 244                                 'log_timestamp > ' . $dbw->addQuotes( $dbw->timestamp( $this->cutoffFrom ) ),
 245                                 'log_timestamp < ' . $dbw->addQuotes( $dbw->timestamp( $this->cutoffTo ) ),
 246                                 'log_user=user_id',
 247                                 'log_type' => $basicRCLogs,
 248                         ],
 249                         __METHOD__,
 250                         [], // INSERT options
 251                         [ 'ORDER BY' => 'log_timestamp DESC', 'LIMIT' => 5000 ] // SELECT options
 252                 );
 253         }
 254
 255         /**
 256          * Rebuild pass 4: Mark bot and autopatrolled entries.
 257          */
 258         private function rebuildRecentChangesTablePass4() {
 259                 global $wgUseRCPatrol;
 260
 261                 $dbw = $this->getDB( DB_MASTER );
 262
 263                 list( $recentchanges, $usergroups, $user ) =
 264                         $dbw->tableNamesN( 'recentchanges', 'user_groups', 'user' );
 265
 266                 $botgroups = User::getGroupsWithPermission( 'bot' );
 267                 $autopatrolgroups = $wgUseRCPatrol ? User::getGroupsWithPermission( 'autopatrol' ) : [];
 268                 # Flag our recent bot edits
 269                 if ( !empty( $botgroups ) ) {
 270                         $botwhere = $dbw->makeList( $botgroups );
 271                         $botusers = [];
 272
 273                         $this->output( "Flagging bot account edits...\n" );
 274
 275                         # Find all users that are bots
 276                         $sql = "SELECT DISTINCT user_name FROM $usergroups, $user " .
 277                                 "WHERE ug_group IN($botwhere) AND user_id = ug_user";
 278                         $res = $dbw->query( $sql, DB_MASTER );
 279
 280                         foreach ( $res as $obj ) {
 281                                 $botusers[] = $dbw->addQuotes( $obj->user_name );
 282                         }
 283                         # Fill in the rc_bot field
 284                         if ( !empty( $botusers ) ) {
 285                                 $botwhere = implode( ',', $botusers );
 286                                 $sql2 = "UPDATE $recentchanges SET rc_bot=1 " .
 287                                         "WHERE rc_user_text IN($botwhere) " .
 288                                         "AND rc_timestamp > " . $dbw->addQuotes( $dbw->timestamp( $this->cutoffFrom ) ) . ' ' .
 289                                         "AND rc_timestamp < " . $dbw->addQuotes( $dbw->timestamp( $this->cutoffTo ) );
 290                                 $dbw->query( $sql2 );
 291                         }
 292                 }
 293                 global $wgMiserMode;
 294                 # Flag our recent autopatrolled edits
 295                 if ( !$wgMiserMode && !empty( $autopatrolgroups ) ) {
 296                         $patrolwhere = $dbw->makeList( $autopatrolgroups );
 297                         $patrolusers = [];
 298
 299                         $this->output( "Flagging auto-patrolled edits...\n" );
 300
 301                         # Find all users in RC with autopatrol rights
 302                         $sql = "SELECT DISTINCT user_name FROM $usergroups, $user " .
 303                                 "WHERE ug_group IN($patrolwhere) AND user_id = ug_user";
 304                         $res = $dbw->query( $sql, DB_MASTER );
 305
 306                         foreach ( $res as $obj ) {
 307                                 $patrolusers[] = $dbw->addQuotes( $obj->user_name );
 308                         }
 309
 310                         # Fill in the rc_patrolled field
 311                         if ( !empty( $patrolusers ) ) {
 312                                 $patrolwhere = implode( ',', $patrolusers );
 313                                 $sql2 = "UPDATE $recentchanges SET rc_patrolled=1 " .
 314                                         "WHERE rc_user_text IN($patrolwhere) " .
 315                                         "AND rc_timestamp > " . $dbw->addQuotes( $dbw->timestamp( $this->cutoffFrom ) ) . ' ' .
 316                                         "AND rc_timestamp < " . $dbw->addQuotes( $dbw->timestamp( $this->cutoffTo ) );
 317                                 $dbw->query( $sql2 );
 318                         }
 319                 }
 320         }
 321
 322         /**
 323          * Rebuild pass 5: Delete duplicate entries where we generate both a page revision and a log entry
 324          * for a single action (upload only, at the moment, but potentially also move, protect, ...).
 325          */
 326         private function rebuildRecentChangesTablePass5() {
 327                 $dbw = wfGetDB( DB_MASTER );
 328
 329                 $this->output( "Removing duplicate revision and logging entries...\n" );
 330
 331                 $res = $dbw->select(
 332                         [ 'logging', 'log_search' ],
 333                         [ 'ls_value', 'ls_log_id' ],
 334                         [
 335                                 'ls_log_id = log_id',
 336                                 'ls_field' => 'associated_rev_id',
 337                                 'log_type' => 'upload',
 338                                 'log_timestamp > ' . $dbw->addQuotes( $dbw->timestamp( $this->cutoffFrom ) ),
 339                                 'log_timestamp < ' . $dbw->addQuotes( $dbw->timestamp( $this->cutoffTo ) ),
 340                         ],
 341                         __METHOD__
 342                 );
 343                 foreach ( $res as $obj ) {
 344                         $rev_id = $obj->ls_value;
 345                         $log_id = $obj->ls_log_id;
 346
 347                         // Mark the logging row as having an associated rev id
 348                         $dbw->update(
 349                                 'recentchanges',
 350                                 /*SET*/ [ 'rc_this_oldid' => $rev_id ],
 351                                 /*WHERE*/ [ 'rc_logid' => $log_id ],
 352                                 __METHOD__
 353                         );
 354
 355                         // Delete the revision row
 356                         $dbw->delete(
 357                                 'recentchanges',
 358                                 /*WHERE*/ [ 'rc_this_oldid' => $rev_id, 'rc_logid' => 0 ],
 359                                 __METHOD__
 360                         );
 361                 }
 362         }
 363
 364         /**
 365          * Purge cached feeds in $messageMemc
 366          */
 367         private function purgeFeeds() {
 368                 global $wgFeedClasses, $messageMemc;
 369
 370                 $this->output( "Deleting feed timestamps.\n" );
 371
 372                 foreach ( $wgFeedClasses as $feed => $className ) {
 373                         $messageMemc->delete( wfMemcKey( 'rcfeed', $feed, 'timestamp' ) ); # Good enough for now.
 374                 }
 375         }
 376 }
 377
 378 $maintClass = "RebuildRecentchanges";
 379 require_once RUN_MAINTENANCE_IF_MAIN;