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
  39         public function execute() {
  40                 $this->rebuildRecentChangesTablePass1();
  41                 $this->rebuildRecentChangesTablePass2();
  42                 $this->rebuildRecentChangesTablePass3();
  43                 $this->rebuildRecentChangesTablePass4();
  44                 $this->rebuildRecentChangesTablePass5();
  45                 $this->purgeFeeds();
  46                 $this->output( "Done.\n" );
  47         }
  48
  49         /**
  50          * Rebuild pass 1: Insert `recentchanges` entries for page revisions.
  51          */
  52         private function rebuildRecentChangesTablePass1() {
  53                 $dbw = $this->getDB( DB_MASTER );
  54
  55                 $dbw->delete( 'recentchanges', '*' );
  56
  57                 $this->output( "Loading from page and revision tables...\n" );
  58
  59                 global $wgRCMaxAge;
  60
  61                 $this->output( '$wgRCMaxAge=' . $wgRCMaxAge );
  62                 $days = $wgRCMaxAge / 24 / 3600;
  63                 if ( intval( $days ) == $days ) {
  64                         $this->output( " (" . $days . " days)\n" );
  65                 } else {
  66                         $this->output( " (approx. " . intval( $days ) . " days)\n" );
  67                 }
  68
  69                 $cutoff = time() - $wgRCMaxAge;
  70                 $dbw->insertSelect( 'recentchanges', array( 'page', 'revision' ),
  71                         array(
  72                                 'rc_timestamp' => 'rev_timestamp',
  73                                 'rc_user' => 'rev_user',
  74                                 'rc_user_text' => 'rev_user_text',
  75                                 'rc_namespace' => 'page_namespace',
  76                                 'rc_title' => 'page_title',
  77                                 'rc_comment' => 'rev_comment',
  78                                 'rc_minor' => 'rev_minor_edit',
  79                                 'rc_bot' => 0,
  80                                 'rc_new' => 'page_is_new',
  81                                 'rc_cur_id' => 'page_id',
  82                                 'rc_this_oldid' => 'rev_id',
  83                                 'rc_last_oldid' => 0, // is this ok?
  84                                 'rc_type' => $dbw->conditional( 'page_is_new != 0', RC_NEW, RC_EDIT ),
  85                                 'rc_source' => $dbw->conditional(
  86                                                 'page_is_new != 0',
  87                                                 $dbw->addQuotes( RecentChange::SRC_NEW ),
  88                                                 $dbw->addQuotes( RecentChange::SRC_EDIT )
  89                                 ),
  90                                 'rc_deleted' => 'rev_deleted'
  91                         ),
  92                         array(
  93                                 'rev_timestamp > ' . $dbw->addQuotes( $dbw->timestamp( $cutoff ) ),
  94                                 'rev_page=page_id'
  95                         ),
  96                         __METHOD__,
  97                         array(), // INSERT options
  98                         array( 'ORDER BY' => 'rev_timestamp DESC', 'LIMIT' => 5000 ) // SELECT options
  99                 );
 100         }
 101
 102         /**
 103          * Rebuild pass 2: Enhance entries for page revisions with references to the previous revision
 104          * (rc_last_oldid, rc_new etc.) and size differences (rc_old_len, rc_new_len).
 105          */
 106         private function rebuildRecentChangesTablePass2() {
 107                 $dbw = $this->getDB( DB_MASTER );
 108                 list( $recentchanges, $revision ) = $dbw->tableNamesN( 'recentchanges', 'revision' );
 109
 110                 $this->output( "Updating links and size differences...\n" );
 111
 112                 # Fill in the rc_last_oldid field, which points to the previous edit
 113                 $sql = "SELECT rc_cur_id,rc_this_oldid,rc_timestamp FROM $recentchanges " .
 114                         "ORDER BY rc_cur_id,rc_timestamp";
 115                 $res = $dbw->query( $sql, DB_MASTER );
 116
 117                 $lastCurId = 0;
 118                 $lastOldId = 0;
 119                 foreach ( $res as $obj ) {
 120                         $new = 0;
 121                         if ( $obj->rc_cur_id != $lastCurId ) {
 122                                 # Switch! Look up the previous last edit, if any
 123                                 $lastCurId = intval( $obj->rc_cur_id );
 124                                 $emit = $obj->rc_timestamp;
 125                                 $sql2 = "SELECT rev_id,rev_len FROM $revision " .
 126                                         "WHERE rev_page={$lastCurId} " .
 127                                         "AND rev_timestamp<'{$emit}' ORDER BY rev_timestamp DESC";
 128                                 $sql2 = $dbw->limitResult( $sql2, 1, false );
 129                                 $res2 = $dbw->query( $sql2 );
 130                                 $row = $dbw->fetchObject( $res2 );
 131                                 if ( $row ) {
 132                                         $lastOldId = intval( $row->rev_id );
 133                                         # Grab the last text size if available
 134                                         $lastSize = !is_null( $row->rev_len ) ? intval( $row->rev_len ) : null;
 135                                 } else {
 136                                         # No previous edit
 137                                         $lastOldId = 0;
 138                                         $lastSize = null;
 139                                         $new = 1; // probably true
 140                                 }
 141                         }
 142                         if ( $lastCurId == 0 ) {
 143                                 $this->output( "Uhhh, something wrong? No curid\n" );
 144                         } else {
 145                                 # Grab the entry's text size
 146                                 $size = $dbw->selectField( 'revision', 'rev_len', array( 'rev_id' => $obj->rc_this_oldid ) );
 147
 148                                 $dbw->update( 'recentchanges',
 149                                         array(
 150                                                 'rc_last_oldid' => $lastOldId,
 151                                                 'rc_new' => $new,
 152                                                 'rc_type' => $new,
 153                                                 'rc_source' => $new === 1 ? RecentChange::SRC_NEW : RecentChange::SRC_EDIT,
 154                                                 'rc_old_len' => $lastSize,
 155                                                 'rc_new_len' => $size,
 156                                         ), array(
 157                                                 'rc_cur_id' => $lastCurId,
 158                                                 'rc_this_oldid' => $obj->rc_this_oldid,
 159                                         ),
 160                                         __METHOD__
 161                                 );
 162
 163                                 $lastOldId = intval( $obj->rc_this_oldid );
 164                                 $lastSize = $size;
 165                         }
 166                 }
 167         }
 168
 169         /**
 170          * Rebuild pass 3: Insert `recentchanges` entries for action logs.
 171          */
 172         private function rebuildRecentChangesTablePass3() {
 173                 $dbw = $this->getDB( DB_MASTER );
 174
 175                 $this->output( "Loading from user, page, and logging tables...\n" );
 176
 177                 global $wgRCMaxAge, $wgLogTypes, $wgLogRestrictions;
 178                 // Some logs don't go in RC. This should check for that
 179                 $basicRCLogs = array_diff( $wgLogTypes, array_keys( $wgLogRestrictions ) );
 180
 181                 $cutoff = time() - $wgRCMaxAge;
 182                 list( $logging, $page ) = $dbw->tableNamesN( 'logging', 'page' );
 183                 $dbw->insertSelect(
 184                         'recentchanges',
 185                         array(
 186                                 'user',
 187                                 "$logging LEFT JOIN $page ON (log_namespace=page_namespace AND log_title=page_title)"
 188                         ),
 189                         array(
 190                                 'rc_timestamp' => 'log_timestamp',
 191                                 'rc_user' => 'log_user',
 192                                 'rc_user_text' => 'user_name',
 193                                 'rc_namespace' => 'log_namespace',
 194                                 'rc_title' => 'log_title',
 195                                 'rc_comment' => 'log_comment',
 196                                 'rc_minor' => 0,
 197                                 'rc_bot' => 0,
 198                                 'rc_patrolled' => 1,
 199                                 'rc_new' => 0,
 200                                 'rc_this_oldid' => 0,
 201                                 'rc_last_oldid' => 0,
 202                                 'rc_type' => RC_LOG,
 203                                 'rc_source' => $dbw->addQuotes( RecentChange::SRC_LOG ),
 204                                 'rc_cur_id' => $dbw->cascadingDeletes() ? 'page_id' : 'COALESCE(page_id, 0)',
 205                                 'rc_log_type' => 'log_type',
 206                                 'rc_log_action' => 'log_action',
 207                                 'rc_logid' => 'log_id',
 208                                 'rc_params' => 'log_params',
 209                                 'rc_deleted' => 'log_deleted'
 210                         ),
 211                         array(
 212                                 'log_timestamp > ' . $dbw->addQuotes( $dbw->timestamp( $cutoff ) ),
 213                                 'log_user=user_id',
 214                                 'log_type' => $basicRCLogs,
 215                         ),
 216                         __METHOD__,
 217                         array(), // INSERT options
 218                         array( 'ORDER BY' => 'log_timestamp DESC', 'LIMIT' => 5000 ) // SELECT options
 219                 );
 220         }
 221
 222         /**
 223          * Rebuild pass 4: Mark bot and autopatrolled entries.
 224          */
 225         private function rebuildRecentChangesTablePass4() {
 226                 global $wgUseRCPatrol;
 227
 228                 $dbw = $this->getDB( DB_MASTER );
 229
 230                 list( $recentchanges, $usergroups, $user ) =
 231                         $dbw->tableNamesN( 'recentchanges', 'user_groups', 'user' );
 232
 233                 $botgroups = User::getGroupsWithPermission( 'bot' );
 234                 $autopatrolgroups = $wgUseRCPatrol ? User::getGroupsWithPermission( 'autopatrol' ) : array();
 235                 # Flag our recent bot edits
 236                 if ( !empty( $botgroups ) ) {
 237                         $botwhere = $dbw->makeList( $botgroups );
 238                         $botusers = array();
 239
 240                         $this->output( "Flagging bot account edits...\n" );
 241
 242                         # Find all users that are bots
 243                         $sql = "SELECT DISTINCT user_name FROM $usergroups, $user " .
 244                                 "WHERE ug_group IN($botwhere) AND user_id = ug_user";
 245                         $res = $dbw->query( $sql, DB_MASTER );
 246
 247                         foreach ( $res as $obj ) {
 248                                 $botusers[] = $dbw->addQuotes( $obj->user_name );
 249                         }
 250                         # Fill in the rc_bot field
 251                         if ( !empty( $botusers ) ) {
 252                                 $botwhere = implode( ',', $botusers );
 253                                 $sql2 = "UPDATE $recentchanges SET rc_bot=1 " .
 254                                         "WHERE rc_user_text IN($botwhere)";
 255                                 $dbw->query( $sql2 );
 256                         }
 257                 }
 258                 global $wgMiserMode;
 259                 # Flag our recent autopatrolled edits
 260                 if ( !$wgMiserMode && !empty( $autopatrolgroups ) ) {
 261                         $patrolwhere = $dbw->makeList( $autopatrolgroups );
 262                         $patrolusers = array();
 263
 264                         $this->output( "Flagging auto-patrolled edits...\n" );
 265
 266                         # Find all users in RC with autopatrol rights
 267                         $sql = "SELECT DISTINCT user_name FROM $usergroups, $user " .
 268                                 "WHERE ug_group IN($patrolwhere) AND user_id = ug_user";
 269                         $res = $dbw->query( $sql, DB_MASTER );
 270
 271                         foreach ( $res as $obj ) {
 272                                 $patrolusers[] = $dbw->addQuotes( $obj->user_name );
 273                         }
 274
 275                         # Fill in the rc_patrolled field
 276                         if ( !empty( $patrolusers ) ) {
 277                                 $patrolwhere = implode( ',', $patrolusers );
 278                                 $sql2 = "UPDATE $recentchanges SET rc_patrolled=1 " .
 279                                         "WHERE rc_user_text IN($patrolwhere)";
 280                                 $dbw->query( $sql2 );
 281                         }
 282                 }
 283         }
 284
 285         /**
 286          * Rebuild pass 5: Delete duplicate entries where we generate both a page revision and a log entry
 287          * for a single action (upload only, at the moment, but potentially also move, protect, ...).
 288          */
 289         private function rebuildRecentChangesTablePass5() {
 290                 $dbw = wfGetDB( DB_MASTER );
 291
 292                 $this->output( "Removing duplicate revision and logging entries...\n" );
 293
 294                 $res = $dbw->select(
 295                         array( 'logging', 'log_search' ),
 296                         array( 'ls_value', 'ls_log_id' ),
 297                         array(
 298                                 'ls_log_id = log_id',
 299                                 'ls_field' => 'associated_rev_id',
 300                                 'log_type' => 'upload',
 301                         ),
 302                         __METHOD__
 303                 );
 304                 foreach ( $res as $obj ) {
 305                         $rev_id = $obj->ls_value;
 306                         $log_id = $obj->ls_log_id;
 307
 308                         // Mark the logging row as having an associated rev id
 309                         $dbw->update(
 310                                 'recentchanges',
 311                                 /*SET*/ array( 'rc_this_oldid' => $rev_id ),
 312                                 /*WHERE*/ array( 'rc_logid' => $log_id ),
 313                                 __METHOD__
 314                         );
 315
 316                         // Delete the revision row
 317                         $dbw->delete(
 318                                 'recentchanges',
 319                                 /*WHERE*/ array( 'rc_this_oldid' => $rev_id, 'rc_logid' => 0 ),
 320                                 __METHOD__
 321                         );
 322                 }
 323         }
 324
 325         /**
 326          * Purge cached feeds in $messageMemc
 327          */
 328         private function purgeFeeds() {
 329                 global $wgFeedClasses, $messageMemc;
 330
 331                 $this->output( "Deleting feed timestamps.\n" );
 332
 333                 foreach ( $wgFeedClasses as $feed => $className ) {
 334                         $messageMemc->delete( wfMemcKey( 'rcfeed', $feed, 'timestamp' ) ); # Good enough for now.
 335                 }
 336         }
 337 }
 338
 339 $maintClass = "RebuildRecentchanges";
 340 require_once RUN_MAINTENANCE_IF_MAIN;