maintenance/Maintenance.php

   1 <?php
   2 /**
   3  * This program is free software; you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation; either version 2 of the License, or
   6  * (at your option) any later version.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License along
  14  * with this program; if not, write to the Free Software Foundation, Inc.,
  15  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  16  * http://www.gnu.org/copyleft/gpl.html
  17  *
  18  * @file
  19  * @ingroup Maintenance
  20  * @defgroup Maintenance Maintenance
  21  */
  22
  23 // Make sure we're on PHP5.3.2 or better
  24 if ( !function_exists( 'version_compare' ) || version_compare( PHP_VERSION, '5.3.2' ) < 0 ) {
  25         // We need to use dirname( __FILE__ ) here cause __DIR__ is PHP5.3+
  26         require_once( dirname( __FILE__ ) . '/../includes/PHPVersionError.php' );
  27         wfPHPVersionError( 'cli' );
  28 }
  29
  30 /**
  31  * @defgroup MaintenanceArchive Maintenance archives
  32  * @ingroup Maintenance
  33  */
  34
  35 // Define this so scripts can easily find doMaintenance.php
  36 define( 'RUN_MAINTENANCE_IF_MAIN', __DIR__ . '/doMaintenance.php' );
  37 define( 'DO_MAINTENANCE', RUN_MAINTENANCE_IF_MAIN ); // original name, harmless
  38
  39 $maintClass = false;
  40
  41 /**
  42  * Abstract maintenance class for quickly writing and churning out
  43  * maintenance scripts with minimal effort. All that _must_ be defined
  44  * is the execute() method. See docs/maintenance.txt for more info
  45  * and a quick demo of how to use it.
  46  *
  47  * @author Chad Horohoe <chad@anyonecanedit.org>
  48  * @since 1.16
  49  * @ingroup Maintenance
  50  */
  51 abstract class Maintenance {
  52
  53         /**
  54          * Constants for DB access type
  55          * @see Maintenance::getDbType()
  56          */
  57         const DB_NONE  = 0;
  58         const DB_STD   = 1;
  59         const DB_ADMIN = 2;
  60
  61         // Const for getStdin()
  62         const STDIN_ALL = 'all';
  63
  64         // This is the desired params
  65         protected $mParams = array();
  66
  67         // Array of mapping short parameters to long ones
  68         protected $mShortParamsMap = array();
  69
  70         // Array of desired args
  71         protected $mArgList = array();
  72
  73         // This is the list of options that were actually passed
  74         protected $mOptions = array();
  75
  76         // This is the list of arguments that were actually passed
  77         protected $mArgs = array();
  78
  79         // Name of the script currently running
  80         protected $mSelf;
  81
  82         // Special vars for params that are always used
  83         protected $mQuiet = false;
  84         protected $mDbUser, $mDbPass;
  85
  86         // A description of the script, children should change this
  87         protected $mDescription = '';
  88
  89         // Have we already loaded our user input?
  90         protected $mInputLoaded = false;
  91
  92         /**
  93          * Batch size. If a script supports this, they should set
  94          * a default with setBatchSize()
  95          *
  96          * @var int
  97          */
  98         protected $mBatchSize = null;
  99
 100         // Generic options added by addDefaultParams()
 101         private $mGenericParameters = array();
 102         // Generic options which might or not be supported by the script
 103         private $mDependantParameters = array();
 104
 105         /**
 106          * Used by getDD() / setDB()
 107          * @var DatabaseBase
 108          */
 109         private $mDb = null;
 110
 111         /**
 112          * List of all the core maintenance scripts. This is added
 113          * to scripts added by extensions in $wgMaintenanceScripts
 114          * and returned by getMaintenanceScripts()
 115          */
 116         protected static $mCoreScripts = null;
 117
 118         /**
 119          * Default constructor. Children should call this *first* if implementing
 120          * their own constructors
 121          */
 122         public function __construct() {
 123                 // Setup $IP, using MW_INSTALL_PATH if it exists
 124                 global $IP;
 125                 $IP = strval( getenv( 'MW_INSTALL_PATH' ) ) !== ''
 126                         ? getenv( 'MW_INSTALL_PATH' )
 127                         : realpath( __DIR__ . '/..' );
 128
 129                 $this->addDefaultParams();
 130                 register_shutdown_function( array( $this, 'outputChanneled' ), false );
 131         }
 132
 133         /**
 134          * Should we execute the maintenance script, or just allow it to be included
 135          * as a standalone class? It checks that the call stack only includes this
 136          * function and "requires" (meaning was called from the file scope)
 137          *
 138          * @return Boolean
 139          */
 140         public static function shouldExecute() {
 141                 $bt = debug_backtrace();
 142                 $count = count( $bt );
 143                 if ( $count < 2 ) {
 144                         return false; // sanity
 145                 }
 146                 if ( $bt[0]['class'] !== 'Maintenance' || $bt[0]['function'] !== 'shouldExecute' ) {
 147                         return false; // last call should be to this function
 148                 }
 149                 $includeFuncs = array( 'require_once', 'require', 'include', 'include_once' );
 150                 for( $i=1; $i < $count; $i++ ) {
 151                         if ( !in_array( $bt[$i]['function'], $includeFuncs ) ) {
 152                                 return false; // previous calls should all be "requires"
 153                         }
 154                 }
 155                 return true;
 156         }
 157
 158         /**
 159          * Do the actual work. All child classes will need to implement this
 160          */
 161         abstract public function execute();
 162
 163         /**
 164          * Add a parameter to the script. Will be displayed on --help
 165          * with the associated description
 166          *
 167          * @param $name String: the name of the param (help, version, etc)
 168          * @param $description String: the description of the param to show on --help
 169          * @param $required Boolean: is the param required?
 170          * @param $withArg Boolean: is an argument required with this option?
 171          * @param $shortName String: character to use as short name
 172          */
 173         protected function addOption( $name, $description, $required = false, $withArg = false, $shortName = false ) {
 174                 $this->mParams[$name] = array( 'desc' => $description, 'require' => $required, 'withArg' => $withArg, 'shortName' => $shortName );
 175                 if ( $shortName !== false ) {
 176                         $this->mShortParamsMap[$shortName] = $name;
 177                 }
 178         }
 179
 180         /**
 181          * Checks to see if a particular param exists.
 182          * @param $name String: the name of the param
 183          * @return Boolean
 184          */
 185         protected function hasOption( $name ) {
 186                 return isset( $this->mOptions[$name] );
 187         }
 188
 189         /**
 190          * Get an option, or return the default
 191          * @param $name String: the name of the param
 192          * @param $default Mixed: anything you want, default null
 193          * @return Mixed
 194          */
 195         protected function getOption( $name, $default = null ) {
 196                 if ( $this->hasOption( $name ) ) {
 197                         return $this->mOptions[$name];
 198                 } else {
 199                         // Set it so we don't have to provide the default again
 200                         $this->mOptions[$name] = $default;
 201                         return $this->mOptions[$name];
 202                 }
 203         }
 204
 205         /**
 206          * Add some args that are needed
 207          * @param $arg String: name of the arg, like 'start'
 208          * @param $description String: short description of the arg
 209          * @param $required Boolean: is this required?
 210          */
 211         protected function addArg( $arg, $description, $required = true ) {
 212                 $this->mArgList[] = array(
 213                         'name' => $arg,
 214                         'desc' => $description,
 215                         'require' => $required
 216                 );
 217         }
 218
 219         /**
 220          * Remove an option.  Useful for removing options that won't be used in your script.
 221          * @param $name String: the option to remove.
 222          */
 223         protected function deleteOption( $name ) {
 224                 unset( $this->mParams[$name] );
 225         }
 226
 227         /**
 228          * Set the description text.
 229          * @param $text String: the text of the description
 230          */
 231         protected function addDescription( $text ) {
 232                 $this->mDescription = $text;
 233         }
 234
 235         /**
 236          * Does a given argument exist?
 237          * @param $argId Integer: the integer value (from zero) for the arg
 238          * @return Boolean
 239          */
 240         protected function hasArg( $argId = 0 ) {
 241                 return isset( $this->mArgs[$argId] );
 242         }
 243
 244         /**
 245          * Get an argument.
 246          * @param $argId Integer: the integer value (from zero) for the arg
 247          * @param $default Mixed: the default if it doesn't exist
 248          * @return mixed
 249          */
 250         protected function getArg( $argId = 0, $default = null ) {
 251                 return $this->hasArg( $argId ) ? $this->mArgs[$argId] : $default;
 252         }
 253
 254         /**
 255          * Set the batch size.
 256          * @param $s Integer: the number of operations to do in a batch
 257          */
 258         protected function setBatchSize( $s = 0 ) {
 259                 $this->mBatchSize = $s;
 260
 261                 // If we support $mBatchSize, show the option.
 262                 // Used to be in addDefaultParams, but in order for that to
 263                 // work, subclasses would have to call this function in the constructor
 264                 // before they called parent::__construct which is just weird
 265                 // (and really wasn't done).
 266                 if ( $this->mBatchSize ) {
 267                         $this->addOption( 'batch-size', 'Run this many operations ' .
 268                                 'per batch, default: ' . $this->mBatchSize, false, true );
 269                         if ( isset( $this->mParams['batch-size'] ) ) {
 270                                 // This seems a little ugly...
 271                                 $this->mDependantParameters['batch-size'] = $this->mParams['batch-size'];
 272                         }
 273                 }
 274         }
 275
 276         /**
 277          * Get the script's name
 278          * @return String
 279          */
 280         public function getName() {
 281                 return $this->mSelf;
 282         }
 283
 284         /**
 285          * Return input from stdin.
 286          * @param $len Integer: the number of bytes to read. If null,
 287          *        just return the handle. Maintenance::STDIN_ALL returns
 288          *        the full length
 289          * @return Mixed
 290          */
 291         protected function getStdin( $len = null ) {
 292                 if ( $len == Maintenance::STDIN_ALL ) {
 293                         return file_get_contents( 'php://stdin' );
 294                 }
 295                 $f = fopen( 'php://stdin', 'rt' );
 296                 if ( !$len ) {
 297                         return $f;
 298                 }
 299                 $input = fgets( $f, $len );
 300                 fclose( $f );
 301                 return rtrim( $input );
 302         }
 303
 304         /**
 305          * @return bool
 306          */
 307         public function isQuiet() {
 308                 return $this->mQuiet;
 309         }
 310
 311         /**
 312          * Throw some output to the user. Scripts can call this with no fears,
 313          * as we handle all --quiet stuff here
 314          * @param $out String: the text to show to the user
 315          * @param $channel Mixed: unique identifier for the channel. See
 316          *     function outputChanneled.
 317          */
 318         protected function output( $out, $channel = null ) {
 319                 if ( $this->mQuiet ) {
 320                         return;
 321                 }
 322                 if ( $channel === null ) {
 323                         $this->cleanupChanneled();
 324                         print( $out );
 325                 } else {
 326                         $out = preg_replace( '/\n\z/', '', $out );
 327                         $this->outputChanneled( $out, $channel );
 328                 }
 329         }
 330
 331         /**
 332          * Throw an error to the user. Doesn't respect --quiet, so don't use
 333          * this for non-error output
 334          * @param $err String: the error to display
 335          * @param $die Int: if > 0, go ahead and die out using this int as the code
 336          */
 337         protected function error( $err, $die = 0 ) {
 338                 $this->outputChanneled( false );
 339                 if ( php_sapi_name() == 'cli' ) {
 340                         fwrite( STDERR, $err . "\n" );
 341                 } else {
 342                         print $err;
 343                 }
 344                 $die = intval( $die );
 345                 if ( $die > 0 ) {
 346                         die( $die );
 347                 }
 348         }
 349
 350         private $atLineStart = true;
 351         private $lastChannel = null;
 352
 353         /**
 354          * Clean up channeled output.  Output a newline if necessary.
 355          */
 356         public function cleanupChanneled() {
 357                 if ( !$this->atLineStart ) {
 358                         print "\n";
 359                         $this->atLineStart = true;
 360                 }
 361         }
 362
 363         /**
 364          * Message outputter with channeled message support. Messages on the
 365          * same channel are concatenated, but any intervening messages in another
 366          * channel start a new line.
 367          * @param $msg String: the message without trailing newline
 368          * @param $channel string Channel identifier or null for no
 369          *     channel. Channel comparison uses ===.
 370          */
 371         public function outputChanneled( $msg, $channel = null ) {
 372                 if ( $msg === false ) {
 373                         $this->cleanupChanneled();
 374                         return;
 375                 }
 376
 377                 // End the current line if necessary
 378                 if ( !$this->atLineStart && $channel !== $this->lastChannel ) {
 379                         print "\n";
 380                 }
 381
 382                 print $msg;
 383
 384                 $this->atLineStart = false;
 385                 if ( $channel === null ) {
 386                         // For unchanneled messages, output trailing newline immediately
 387                         print "\n";
 388                         $this->atLineStart = true;
 389                 }
 390                 $this->lastChannel = $channel;
 391         }
 392
 393         /**
 394          * Does the script need different DB access? By default, we give Maintenance
 395          * scripts normal rights to the DB. Sometimes, a script needs admin rights
 396          * access for a reason and sometimes they want no access. Subclasses should
 397          * override and return one of the following values, as needed:
 398          *    Maintenance::DB_NONE  -  For no DB access at all
 399          *    Maintenance::DB_STD   -  For normal DB access, default
 400          *    Maintenance::DB_ADMIN -  For admin DB access
 401          * @return Integer
 402          */
 403         public function getDbType() {
 404                 return Maintenance::DB_STD;
 405         }
 406
 407         /**
 408          * Add the default parameters to the scripts
 409          */
 410         protected function addDefaultParams() {
 411
 412                 # Generic (non script dependant) options:
 413
 414                 $this->addOption( 'help', 'Display this help message', false, false, 'h' );
 415                 $this->addOption( 'quiet', 'Whether to supress non-error output', false, false, 'q' );
 416                 $this->addOption( 'conf', 'Location of LocalSettings.php, if not default', false, true );
 417                 $this->addOption( 'wiki', 'For specifying the wiki ID', false, true );
 418                 $this->addOption( 'globals', 'Output globals at the end of processing for debugging' );
 419                 $this->addOption( 'memory-limit', 'Set a specific memory limit for the script, "max" for no limit or "default" to avoid changing it' );
 420                 $this->addOption( 'server', "The protocol and server name to use in URLs, e.g. " .
 421                                 "http://en.wikipedia.org. This is sometimes necessary because " .
 422                                 "server name detection may fail in command line scripts.", false, true );
 423
 424                 # Save generic options to display them separately in help
 425                 $this->mGenericParameters = $this->mParams ;
 426
 427                 # Script dependant options:
 428
 429                 // If we support a DB, show the options
 430                 if ( $this->getDbType() > 0 ) {
 431                         $this->addOption( 'dbuser', 'The DB user to use for this script', false, true );
 432                         $this->addOption( 'dbpass', 'The password to use for this script', false, true );
 433                 }
 434
 435                 # Save additional script dependant options to display
 436                 # them separately in help
 437                 $this->mDependantParameters = array_diff_key( $this->mParams, $this->mGenericParameters );
 438         }
 439
 440         /**
 441          * Run a child maintenance script. Pass all of the current arguments
 442          * to it.
 443          * @param $maintClass String: a name of a child maintenance class
 444          * @param $classFile String: full path of where the child is
 445          * @return Maintenance child
 446          */
 447         public function runChild( $maintClass, $classFile = null ) {
 448                 // Make sure the class is loaded first
 449                 if ( !MWInit::classExists( $maintClass ) ) {
 450                         if ( $classFile ) {
 451                                 require_once( $classFile );
 452                         }
 453                         if ( !MWInit::classExists( $maintClass ) ) {
 454                                 $this->error( "Cannot spawn child: $maintClass" );
 455                         }
 456                 }
 457
 458                 /**
 459                  * @var $child Maintenance
 460                  */
 461                 $child = new $maintClass();
 462                 $child->loadParamsAndArgs( $this->mSelf, $this->mOptions, $this->mArgs );
 463                 if ( !is_null( $this->mDb ) ) {
 464                         $child->setDB( $this->mDb );
 465                 }
 466                 return $child;
 467         }
 468
 469         /**
 470          * Do some sanity checking and basic setup
 471          */
 472         public function setup() {
 473                 global $wgCommandLineMode, $wgRequestTime;
 474
 475                 # Abort if called from a web server
 476                 if ( isset( $_SERVER ) && isset( $_SERVER['REQUEST_METHOD'] ) ) {
 477                         $this->error( 'This script must be run from the command line', true );
 478                 }
 479
 480                 # Make sure we can handle script parameters
 481                 if ( !function_exists( 'hphp_thread_set_warmup_enabled' ) && !ini_get( 'register_argc_argv' ) ) {
 482                         $this->error( 'Cannot get command line arguments, register_argc_argv is set to false', true );
 483                 }
 484
 485                 if ( version_compare( phpversion(), '5.2.4' ) >= 0 ) {
 486                         // Send PHP warnings and errors to stderr instead of stdout.
 487                         // This aids in diagnosing problems, while keeping messages
 488                         // out of redirected output.
 489                         if ( ini_get( 'display_errors' ) ) {
 490                                 ini_set( 'display_errors', 'stderr' );
 491                         }
 492
 493                         // Don't touch the setting on earlier versions of PHP,
 494                         // as setting it would disable output if you'd wanted it.
 495
 496                         // Note that exceptions are also sent to stderr when
 497                         // command-line mode is on, regardless of PHP version.
 498                 }
 499
 500                 $this->loadParamsAndArgs();
 501                 $this->maybeHelp();
 502
 503                 # Set the memory limit
 504                 # Note we need to set it again later in cache LocalSettings changed it
 505                 $this->adjustMemoryLimit();
 506
 507                 # Set max execution time to 0 (no limit). PHP.net says that
 508                 # "When running PHP from the command line the default setting is 0."
 509                 # But sometimes this doesn't seem to be the case.
 510                 ini_set( 'max_execution_time', 0 );
 511
 512                 $wgRequestTime = microtime( true );
 513
 514                 # Define us as being in MediaWiki
 515                 define( 'MEDIAWIKI', true );
 516
 517                 $wgCommandLineMode = true;
 518                 # Turn off output buffering if it's on
 519                 @ob_end_flush();
 520
 521                 $this->validateParamsAndArgs();
 522         }
 523
 524         /**
 525          * Normally we disable the memory_limit when running admin scripts.
 526          * Some scripts may wish to actually set a limit, however, to avoid
 527          * blowing up unexpectedly. We also support a --memory-limit option,
 528          * to allow sysadmins to explicitly set one if they'd prefer to override
 529          * defaults (or for people using Suhosin which yells at you for trying
 530          * to disable the limits)
 531          * @return string
 532          */
 533         public function memoryLimit() {
 534                 $limit = $this->getOption( 'memory-limit', 'max' );
 535                 $limit = trim( $limit, "\" '" ); // trim quotes in case someone misunderstood
 536                 return $limit;
 537         }
 538
 539         /**
 540          * Adjusts PHP's memory limit to better suit our needs, if needed.
 541          */
 542         protected function adjustMemoryLimit() {
 543                 $limit = $this->memoryLimit();
 544                 if ( $limit == 'max' ) {
 545                         $limit = -1; // no memory limit
 546                 }
 547                 if ( $limit != 'default' ) {
 548                         ini_set( 'memory_limit', $limit );
 549                 }
 550         }
 551
 552         /**
 553          * Clear all params and arguments.
 554          */
 555         public function clearParamsAndArgs() {
 556                 $this->mOptions = array();
 557                 $this->mArgs = array();
 558                 $this->mInputLoaded = false;
 559         }
 560
 561         /**
 562          * Process command line arguments
 563          * $mOptions becomes an array with keys set to the option names
 564          * $mArgs becomes a zero-based array containing the non-option arguments
 565          *
 566          * @param $self String The name of the script, if any
 567          * @param $opts Array An array of options, in form of key=>value
 568          * @param $args Array An array of command line arguments
 569          */
 570         public function loadParamsAndArgs( $self = null, $opts = null, $args = null ) {
 571                 # If we were given opts or args, set those and return early
 572                 if ( $self ) {
 573                         $this->mSelf = $self;
 574                         $this->mInputLoaded = true;
 575                 }
 576                 if ( $opts ) {
 577                         $this->mOptions = $opts;
 578                         $this->mInputLoaded = true;
 579                 }
 580                 if ( $args ) {
 581                         $this->mArgs = $args;
 582                         $this->mInputLoaded = true;
 583                 }
 584
 585                 # If we've already loaded input (either by user values or from $argv)
 586                 # skip on loading it again. The array_shift() will corrupt values if
 587                 # it's run again and again
 588                 if ( $this->mInputLoaded ) {
 589                         $this->loadSpecialVars();
 590                         return;
 591                 }
 592
 593                 global $argv;
 594                 $this->mSelf = array_shift( $argv );
 595
 596                 $options = array();
 597                 $args = array();
 598
 599                 # Parse arguments
 600                 for ( $arg = reset( $argv ); $arg !== false; $arg = next( $argv ) ) {
 601                         if ( $arg == '--' ) {
 602                                 # End of options, remainder should be considered arguments
 603                                 $arg = next( $argv );
 604                                 while ( $arg !== false ) {
 605                                         $args[] = $arg;
 606                                         $arg = next( $argv );
 607                                 }
 608                                 break;
 609                         } elseif ( substr( $arg, 0, 2 ) == '--' ) {
 610                                 # Long options
 611                                 $option = substr( $arg, 2 );
 612                                 if ( array_key_exists( $option, $options ) ) {
 613                                         $this->error( "\nERROR: $option parameter given twice\n" );
 614                                         $this->maybeHelp( true );
 615                                 }
 616                                 if ( isset( $this->mParams[$option] ) && $this->mParams[$option]['withArg'] ) {
 617                                         $param = next( $argv );
 618                                         if ( $param === false ) {
 619                                                 $this->error( "\nERROR: $option parameter needs a value after it\n" );
 620                                                 $this->maybeHelp( true );
 621                                         }
 622                                         $options[$option] = $param;
 623                                 } else {
 624                                         $bits = explode( '=', $option, 2 );
 625                                         if ( count( $bits ) > 1 ) {
 626                                                 $option = $bits[0];
 627                                                 $param = $bits[1];
 628                                         } else {
 629                                                 $param = 1;
 630                                         }
 631                                         $options[$option] = $param;
 632                                 }
 633                         } elseif ( substr( $arg, 0, 1 ) == '-' ) {
 634                                 # Short options
 635                                 for ( $p = 1; $p < strlen( $arg ); $p++ ) {
 636                                         $option = $arg { $p } ;
 637                                         if ( !isset( $this->mParams[$option] ) && isset( $this->mShortParamsMap[$option] ) ) {
 638                                                 $option = $this->mShortParamsMap[$option];
 639                                         }
 640                                         if ( array_key_exists( $option, $options ) ) {
 641                                                 $this->error( "\nERROR: $option parameter given twice\n" );
 642                                                 $this->maybeHelp( true );
 643                                         }
 644                                         if ( isset( $this->mParams[$option]['withArg'] ) && $this->mParams[$option]['withArg'] ) {
 645                                                 $param = next( $argv );
 646                                                 if ( $param === false ) {
 647                                                         $this->error( "\nERROR: $option parameter needs a value after it\n" );
 648                                                         $this->maybeHelp( true );
 649                                                 }
 650                                                 $options[$option] = $param;
 651                                         } else {
 652                                                 $options[$option] = 1;
 653                                         }
 654                                 }
 655                         } else {
 656                                 $args[] = $arg;
 657                         }
 658                 }
 659
 660                 $this->mOptions = $options;
 661                 $this->mArgs = $args;
 662                 $this->loadSpecialVars();
 663                 $this->mInputLoaded = true;
 664         }
 665
 666         /**
 667          * Run some validation checks on the params, etc
 668          */
 669         protected function validateParamsAndArgs() {
 670                 $die = false;
 671                 # Check to make sure we've got all the required options
 672                 foreach ( $this->mParams as $opt => $info ) {
 673                         if ( $info['require'] && !$this->hasOption( $opt ) ) {
 674                                 $this->error( "Param $opt required!" );
 675                                 $die = true;
 676                         }
 677                 }
 678                 # Check arg list too
 679                 foreach ( $this->mArgList as $k => $info ) {
 680                         if ( $info['require'] && !$this->hasArg( $k ) ) {
 681                                 $this->error( 'Argument <' . $info['name'] . '> required!' );
 682                                 $die = true;
 683                         }
 684                 }
 685
 686                 if ( $die ) {
 687                         $this->maybeHelp( true );
 688                 }
 689         }
 690
 691         /**
 692          * Handle the special variables that are global to all scripts
 693          */
 694         protected function loadSpecialVars() {
 695                 if ( $this->hasOption( 'dbuser' ) ) {
 696                         $this->mDbUser = $this->getOption( 'dbuser' );
 697                 }
 698                 if ( $this->hasOption( 'dbpass' ) ) {
 699                         $this->mDbPass = $this->getOption( 'dbpass' );
 700                 }
 701                 if ( $this->hasOption( 'quiet' ) ) {
 702                         $this->mQuiet = true;
 703                 }
 704                 if ( $this->hasOption( 'batch-size' ) ) {
 705                         $this->mBatchSize = intval( $this->getOption( 'batch-size' ) );
 706                 }
 707         }
 708
 709         /**
 710          * Maybe show the help.
 711          * @param $force boolean Whether to force the help to show, default false
 712          */
 713         protected function maybeHelp( $force = false ) {
 714                 if( !$force && !$this->hasOption( 'help' ) ) {
 715                         return;
 716                 }
 717
 718                 $screenWidth = 80; // TODO: Caculate this!
 719                 $tab = "    ";
 720                 $descWidth = $screenWidth - ( 2 * strlen( $tab ) );
 721
 722                 ksort( $this->mParams );
 723                 $this->mQuiet = false;
 724
 725                 // Description ...
 726                 if ( $this->mDescription ) {
 727                         $this->output( "\n" . $this->mDescription . "\n" );
 728                 }
 729                 $output = "\nUsage: php " . basename( $this->mSelf );
 730
 731                 // ... append parameters ...
 732                 if ( $this->mParams ) {
 733                         $output .= " [--" . implode( array_keys( $this->mParams ), "|--" ) . "]";
 734                 }
 735
 736                 // ... and append arguments.
 737                 if ( $this->mArgList ) {
 738                         $output .= ' ';
 739                         foreach ( $this->mArgList as $k => $arg ) {
 740                                 if ( $arg['require'] ) {
 741                                         $output .= '<' . $arg['name'] . '>';
 742                                 } else {
 743                                         $output .= '[' . $arg['name'] . ']';
 744                                 }
 745                                 if ( $k < count( $this->mArgList ) - 1 )
 746                                         $output .= ' ';
 747                         }
 748                 }
 749                 $this->output( "$output\n\n" );
 750
 751                 # TODO abstract some repetitive code below
 752
 753                 // Generic parameters
 754                 $this->output( "Generic maintenance parameters:\n" );
 755                 foreach ( $this->mGenericParameters as $par => $info ) {
 756                         if ( $info['shortName'] !== false ) {
 757                                 $par .= " (-{$info['shortName']})";
 758                         }
 759                         $this->output(
 760                                 wordwrap( "$tab--$par: " . $info['desc'], $descWidth,
 761                                                 "\n$tab$tab" ) . "\n"
 762                         );
 763                 }
 764                 $this->output( "\n" );
 765
 766                 $scriptDependantParams = $this->mDependantParameters;
 767                 if( count($scriptDependantParams) > 0 ) {
 768                         $this->output( "Script dependant parameters:\n" );
 769                         // Parameters description
 770                         foreach ( $scriptDependantParams as $par => $info ) {
 771                                 if ( $info['shortName'] !== false ) {
 772                                         $par .= " (-{$info['shortName']})";
 773                                 }
 774                                 $this->output(
 775                                         wordwrap( "$tab--$par: " . $info['desc'], $descWidth,
 776                                                         "\n$tab$tab" ) . "\n"
 777                                 );
 778                         }
 779                         $this->output( "\n" );
 780                 }
 781
 782
 783                 // Script specific parameters not defined on construction by
 784                 // Maintenance::addDefaultParams()
 785                 $scriptSpecificParams = array_diff_key(
 786                         # all script parameters:
 787                         $this->mParams,
 788                         # remove the Maintenance default parameters:
 789                         $this->mGenericParameters,
 790                         $this->mDependantParameters
 791                 );
 792                 if( count($scriptSpecificParams) > 0 ) {
 793                         $this->output( "Script specific parameters:\n" );
 794                         // Parameters description
 795                         foreach ( $scriptSpecificParams as $par => $info ) {
 796                                 if ( $info['shortName'] !== false ) {
 797                                         $par .= " (-{$info['shortName']})";
 798                                 }
 799                                 $this->output(
 800                                         wordwrap( "$tab--$par: " . $info['desc'], $descWidth,
 801                                                         "\n$tab$tab" ) . "\n"
 802                                 );
 803                         }
 804                         $this->output( "\n" );
 805                 }
 806
 807                 // Print arguments
 808                 if( count( $this->mArgList ) > 0 ) {
 809                         $this->output( "Arguments:\n" );
 810                         // Arguments description
 811                         foreach ( $this->mArgList as $info ) {
 812                                 $openChar = $info['require'] ? '<' : '[';
 813                                 $closeChar = $info['require'] ? '>' : ']';
 814                                 $this->output(
 815                                         wordwrap( "$tab$openChar" . $info['name'] . "$closeChar: " .
 816                                                 $info['desc'], $descWidth, "\n$tab$tab" ) . "\n"
 817                                 );
 818                         }
 819                         $this->output( "\n" );
 820                 }
 821
 822                 die( 1 );
 823         }
 824
 825         /**
 826          * Handle some last-minute setup here.
 827          */
 828         public function finalSetup() {
 829                 global $wgCommandLineMode, $wgShowSQLErrors, $wgServer;
 830                 global $wgDBadminuser, $wgDBadminpassword;
 831                 global $wgDBuser, $wgDBpassword, $wgDBservers, $wgLBFactoryConf;
 832
 833                 # Turn off output buffering again, it might have been turned on in the settings files
 834                 if ( ob_get_level() ) {
 835                         ob_end_flush();
 836                 }
 837                 # Same with these
 838                 $wgCommandLineMode = true;
 839
 840                 # Override $wgServer
 841                 if( $this->hasOption( 'server') ) {
 842                         $wgServer = $this->getOption( 'server', $wgServer );
 843                 }
 844
 845                 # If these were passed, use them
 846                 if ( $this->mDbUser ) {
 847                         $wgDBadminuser = $this->mDbUser;
 848                 }
 849                 if ( $this->mDbPass ) {
 850                         $wgDBadminpassword = $this->mDbPass;
 851                 }
 852
 853                 if ( $this->getDbType() == self::DB_ADMIN && isset( $wgDBadminuser ) ) {
 854                         $wgDBuser = $wgDBadminuser;
 855                         $wgDBpassword = $wgDBadminpassword;
 856
 857                         if ( $wgDBservers ) {
 858                                 /**
 859                                  * @var $wgDBservers array
 860                                  */
 861                                 foreach ( $wgDBservers as $i => $server ) {
 862                                         $wgDBservers[$i]['user'] = $wgDBuser;
 863                                         $wgDBservers[$i]['password'] = $wgDBpassword;
 864                                 }
 865                         }
 866                         if ( isset( $wgLBFactoryConf['serverTemplate'] ) ) {
 867                                 $wgLBFactoryConf['serverTemplate']['user'] = $wgDBuser;
 868                                 $wgLBFactoryConf['serverTemplate']['password'] = $wgDBpassword;
 869                         }
 870                         LBFactory::destroyInstance();
 871                 }
 872
 873                 $this->afterFinalSetup();
 874
 875                 $wgShowSQLErrors = true;
 876                 @set_time_limit( 0 );
 877                 $this->adjustMemoryLimit();
 878         }
 879
 880         /**
 881          * Execute a callback function at the end of initialisation
 882          */
 883         protected function afterFinalSetup() {
 884                 if ( defined( 'MW_CMDLINE_CALLBACK' ) ) {
 885                         call_user_func( MW_CMDLINE_CALLBACK );
 886                 }
 887         }
 888
 889         /**
 890          * Potentially debug globals. Originally a feature only
 891          * for refreshLinks
 892          */
 893         public function globals() {
 894                 if ( $this->hasOption( 'globals' ) ) {
 895                         print_r( $GLOBALS );
 896                 }
 897         }
 898
 899         /**
 900          * Generic setup for most installs. Returns the location of LocalSettings
 901          * @return String
 902          */
 903         public function loadSettings() {
 904                 global $wgCommandLineMode, $IP;
 905
 906                 if ( isset( $this->mOptions['conf'] ) ) {
 907                         $settingsFile = $this->mOptions['conf'];
 908                 } elseif ( defined("MW_CONFIG_FILE") ) {
 909                         $settingsFile = MW_CONFIG_FILE;
 910                 } else {
 911                         $settingsFile = "$IP/LocalSettings.php";
 912                 }
 913                 if ( isset( $this->mOptions['wiki'] ) ) {
 914                         $bits = explode( '-', $this->mOptions['wiki'] );
 915                         if ( count( $bits ) == 1 ) {
 916                                 $bits[] = '';
 917                         }
 918                         define( 'MW_DB', $bits[0] );
 919                         define( 'MW_PREFIX', $bits[1] );
 920                 }
 921
 922                 if ( !is_readable( $settingsFile ) ) {
 923                         $this->error( "A copy of your installation's LocalSettings.php\n" .
 924                                                 "must exist and be readable in the source directory.\n" .
 925                                                 "Use --conf to specify it." , true );
 926                 }
 927                 $wgCommandLineMode = true;
 928                 return $settingsFile;
 929         }
 930
 931         /**
 932          * Support function for cleaning up redundant text records
 933          * @param $delete Boolean: whether or not to actually delete the records
 934          * @author Rob Church <robchur@gmail.com>
 935          */
 936         public function purgeRedundantText( $delete = true ) {
 937                 # Data should come off the master, wrapped in a transaction
 938                 $dbw = $this->getDB( DB_MASTER );
 939                 $dbw->begin( __METHOD__ );
 940
 941                 $tbl_arc = $dbw->tableName( 'archive' );
 942                 $tbl_rev = $dbw->tableName( 'revision' );
 943                 $tbl_txt = $dbw->tableName( 'text' );
 944
 945                 # Get "active" text records from the revisions table
 946                 $this->output( 'Searching for active text records in revisions table...' );
 947                 $res = $dbw->query( "SELECT DISTINCT rev_text_id FROM $tbl_rev" );
 948                 foreach ( $res as $row ) {
 949                         $cur[] = $row->rev_text_id;
 950                 }
 951                 $this->output( "done.\n" );
 952
 953                 # Get "active" text records from the archive table
 954                 $this->output( 'Searching for active text records in archive table...' );
 955                 $res = $dbw->query( "SELECT DISTINCT ar_text_id FROM $tbl_arc" );
 956                 foreach ( $res as $row ) {
 957                         $cur[] = $row->ar_text_id;
 958                 }
 959                 $this->output( "done.\n" );
 960
 961                 # Get the IDs of all text records not in these sets
 962                 $this->output( 'Searching for inactive text records...' );
 963                 $set = implode( ', ', $cur );
 964                 $res = $dbw->query( "SELECT old_id FROM $tbl_txt WHERE old_id NOT IN ( $set )" );
 965                 $old = array();
 966                 foreach ( $res as $row ) {
 967                         $old[] = $row->old_id;
 968                 }
 969                 $this->output( "done.\n" );
 970
 971                 # Inform the user of what we're going to do
 972                 $count = count( $old );
 973                 $this->output( "$count inactive items found.\n" );
 974
 975                 # Delete as appropriate
 976                 if ( $delete && $count ) {
 977                         $this->output( 'Deleting...' );
 978                         $set = implode( ', ', $old );
 979                         $dbw->query( "DELETE FROM $tbl_txt WHERE old_id IN ( $set )" );
 980                         $this->output( "done.\n" );
 981                 }
 982
 983                 # Done
 984                 $dbw->commit( __METHOD__ );
 985         }
 986
 987         /**
 988          * Get the maintenance directory.
 989          * @return string
 990          */
 991         protected function getDir() {
 992                 return __DIR__;
 993         }
 994
 995         /**
 996          * Get the list of available maintenance scripts. Note
 997          * that if you call this _before_ calling doMaintenance
 998          * you won't have any extensions in it yet
 999          * @return Array
1000          */
1001         public static function getMaintenanceScripts() {
1002                 global $wgMaintenanceScripts;
1003                 return $wgMaintenanceScripts + self::getCoreScripts();
1004         }
1005
1006         /**
1007          * Return all of the core maintenance scripts
1008          * @return array
1009          */
1010         protected static function getCoreScripts() {
1011                 if ( !self::$mCoreScripts ) {
1012                         $paths = array(
1013                                 __DIR__,
1014                                 __DIR__ . '/language',
1015                                 __DIR__ . '/storage',
1016                         );
1017                         self::$mCoreScripts = array();
1018                         foreach ( $paths as $p ) {
1019                                 $handle = opendir( $p );
1020                                 while ( ( $file = readdir( $handle ) ) !== false ) {
1021                                         if ( $file == 'Maintenance.php' ) {
1022                                                 continue;
1023                                         }
1024                                         $file = $p . '/' . $file;
1025                                         if ( is_dir( $file ) || !strpos( $file, '.php' ) ||
1026                                                 ( strpos( file_get_contents( $file ), '$maintClass' ) === false ) ) {
1027                                                 continue;
1028                                         }
1029                                         require( $file );
1030                                         $vars = get_defined_vars();
1031                                         if ( array_key_exists( 'maintClass', $vars ) ) {
1032                                                 self::$mCoreScripts[$vars['maintClass']] = $file;
1033                                         }
1034                                 }
1035                                 closedir( $handle );
1036                         }
1037                 }
1038                 return self::$mCoreScripts;
1039         }
1040
1041         /**
1042          * Returns a database to be used by current maintenance script. It can be set by setDB().
1043          * If not set, wfGetDB() will be used.
1044          * This function has the same parameters as wfGetDB()
1045          *
1046          * @return DatabaseBase
1047          */
1048         protected function &getDB( $db, $groups = array(), $wiki = false ) {
1049                 if ( is_null( $this->mDb ) ) {
1050                         return wfGetDB( $db, $groups, $wiki );
1051                 } else {
1052                         return $this->mDb;
1053                 }
1054         }
1055
1056         /**
1057          * Sets database object to be returned by getDB().
1058          *
1059          * @param $db DatabaseBase: Database object to be used
1060          */
1061         public function setDB( &$db ) {
1062                 $this->mDb = $db;
1063         }
1064
1065         /**
1066          * Lock the search index
1067          * @param &$db DatabaseBase object
1068          */
1069         private function lockSearchindex( &$db ) {
1070                 $write = array( 'searchindex' );
1071                 $read = array( 'page', 'revision', 'text', 'interwiki', 'l10n_cache', 'user' );
1072                 $db->lockTables( $read, $write, __CLASS__ . '::' . __METHOD__ );
1073         }
1074
1075         /**
1076          * Unlock the tables
1077          * @param &$db DatabaseBase object
1078          */
1079         private function unlockSearchindex( &$db ) {
1080                 $db->unlockTables(  __CLASS__ . '::' . __METHOD__ );
1081         }
1082
1083         /**
1084          * Unlock and lock again
1085          * Since the lock is low-priority, queued reads will be able to complete
1086          * @param &$db DatabaseBase object
1087          */
1088         private function relockSearchindex( &$db ) {
1089                 $this->unlockSearchindex( $db );
1090                 $this->lockSearchindex( $db );
1091         }
1092
1093         /**
1094          * Perform a search index update with locking
1095          * @param $maxLockTime Integer: the maximum time to keep the search index locked.
1096          * @param $callback callback String: the function that will update the function.
1097          * @param $dbw DatabaseBase object
1098          * @param $results
1099          */
1100         public function updateSearchIndex( $maxLockTime, $callback, $dbw, $results ) {
1101                 $lockTime = time();
1102
1103                 # Lock searchindex
1104                 if ( $maxLockTime ) {
1105                         $this->output( "   --- Waiting for lock ---" );
1106                         $this->lockSearchindex( $dbw );
1107                         $lockTime = time();
1108                         $this->output( "\n" );
1109                 }
1110
1111                 # Loop through the results and do a search update
1112                 foreach ( $results as $row ) {
1113                         # Allow reads to be processed
1114                         if ( $maxLockTime && time() > $lockTime + $maxLockTime ) {
1115                                 $this->output( "    --- Relocking ---" );
1116                                 $this->relockSearchindex( $dbw );
1117                                 $lockTime = time();
1118                                 $this->output( "\n" );
1119                         }
1120                         call_user_func( $callback, $dbw, $row );
1121                 }
1122
1123                 # Unlock searchindex
1124                 if ( $maxLockTime ) {
1125                         $this->output( "    --- Unlocking --" );
1126                         $this->unlockSearchindex( $dbw );
1127                         $this->output( "\n" );
1128                 }
1129
1130         }
1131
1132         /**
1133          * Update the searchindex table for a given pageid
1134          * @param $dbw DatabaseBase a database write handle
1135          * @param $pageId Integer: the page ID to update.
1136          * @return null|string
1137          */
1138         public function updateSearchIndexForPage( $dbw, $pageId ) {
1139                 // Get current revision
1140                 $rev = Revision::loadFromPageId( $dbw, $pageId );
1141                 $title = null;
1142                 if ( $rev ) {
1143                         $titleObj = $rev->getTitle();
1144                         $title = $titleObj->getPrefixedDBkey();
1145                         $this->output( "$title..." );
1146                         # Update searchindex
1147                         # TODO: pass the Content object to SearchUpdate, let the search engine decide how to deal with it.
1148                         $u = new SearchUpdate( $pageId, $titleObj->getText(), $rev->getContent()->getTextForSearchIndex() );
1149                         $u->doUpdate();
1150                         $this->output( "\n" );
1151                 }
1152                 return $title;
1153         }
1154
1155         /**
1156          * Wrapper for posix_isatty()
1157          * We default as considering stdin a tty (for nice readline methods)
1158          * but treating stout as not a tty to avoid color codes
1159          *
1160          * @param $fd int File descriptor
1161          * @return bool
1162          */
1163         public static function posix_isatty( $fd ) {
1164                 if ( !MWInit::functionExists( 'posix_isatty' ) ) {
1165                         return !$fd;
1166                 } else {
1167                         return posix_isatty( $fd );
1168                 }
1169         }
1170
1171         /**
1172          * Prompt the console for input
1173          * @param $prompt String what to begin the line with, like '> '
1174          * @return String response
1175          */
1176         public static function readconsole( $prompt = '> ' ) {
1177                 static $isatty = null;
1178                 if ( is_null( $isatty ) ) {
1179                         $isatty = self::posix_isatty( 0 /*STDIN*/ );
1180                 }
1181
1182                 if ( $isatty && function_exists( 'readline' ) ) {
1183                         return readline( $prompt );
1184                 } else {
1185                         if ( $isatty ) {
1186                                 $st = self::readlineEmulation( $prompt );
1187                         } else {
1188                                 if ( feof( STDIN ) ) {
1189                                         $st = false;
1190                                 } else {
1191                                         $st = fgets( STDIN, 1024 );
1192                                 }
1193                         }
1194                         if ( $st === false ) return false;
1195                         $resp = trim( $st );
1196                         return $resp;
1197                 }
1198         }
1199
1200         /**
1201          * Emulate readline()
1202          * @param $prompt String what to begin the line with, like '> '
1203          * @return String
1204          */
1205         private static function readlineEmulation( $prompt ) {
1206                 $bash = Installer::locateExecutableInDefaultPaths( array( 'bash' ) );
1207                 if ( !wfIsWindows() && $bash ) {
1208                         $retval = false;
1209                         $encPrompt = wfEscapeShellArg( $prompt );
1210                         $command = "read -er -p $encPrompt && echo \"\$REPLY\"";
1211                         $encCommand = wfEscapeShellArg( $command );
1212                         $line = wfShellExec( "$bash -c $encCommand", $retval );
1213
1214                         if ( $retval == 0 ) {
1215                                 return $line;
1216                         } elseif ( $retval == 127 ) {
1217                                 // Couldn't execute bash even though we thought we saw it.
1218                                 // Shell probably spit out an error message, sorry :(
1219                                 // Fall through to fgets()...
1220                         } else {
1221                                 // EOF/ctrl+D
1222                                 return false;
1223                         }
1224                 }
1225
1226                 // Fallback... we'll have no editing controls, EWWW
1227                 if ( feof( STDIN ) ) {
1228                         return false;
1229                 }
1230                 print $prompt;
1231                 return fgets( STDIN, 1024 );
1232         }
1233 }
1234
1235 /**
1236  * Fake maintenance wrapper, mostly used for the web installer/updater
1237  */
1238 class FakeMaintenance extends Maintenance {
1239         protected $mSelf = "FakeMaintenanceScript";
1240         public function execute() {
1241                 return;
1242         }
1243 }
1244
1245 /**
1246  * Class for scripts that perform database maintenance and want to log the
1247  * update in `updatelog` so we can later skip it
1248  */
1249 abstract class LoggedUpdateMaintenance extends Maintenance {
1250         public function __construct() {
1251                 parent::__construct();
1252                 $this->addOption( 'force', 'Run the update even if it was completed already' );
1253                 $this->setBatchSize( 200 );
1254         }
1255
1256         public function execute() {
1257                 $db = $this->getDB( DB_MASTER );
1258                 $key = $this->getUpdateKey();
1259
1260                 if ( !$this->hasOption( 'force' ) &&
1261                         $db->selectRow( 'updatelog', '1', array( 'ul_key' => $key ), __METHOD__ ) )
1262                 {
1263                         $this->output( "..." . $this->updateSkippedMessage() . "\n" );
1264                         return true;
1265                 }
1266
1267                 if ( !$this->doDBUpdates() ) {
1268                         return false;
1269                 }
1270
1271                 if (
1272                         $db->insert( 'updatelog', array( 'ul_key' => $key ), __METHOD__, 'IGNORE' ) )
1273                 {
1274                         return true;
1275                 } else {
1276                         $this->output( $this->updatelogFailedMessage() . "\n" );
1277                         return false;
1278                 }
1279         }
1280
1281         /**
1282          * Message to show that the update was done already and was just skipped
1283          * @return String
1284          */
1285         protected function updateSkippedMessage() {
1286                 $key = $this->getUpdateKey();
1287                 return "Update '{$key}' already logged as completed.";
1288         }
1289
1290         /**
1291          * Message to show the the update log was unable to log the completion of this update
1292          * @return String
1293          */
1294         protected function updatelogFailedMessage() {
1295                 $key = $this->getUpdateKey();
1296                 return "Unable to log update '{$key}' as completed.";
1297         }
1298
1299         /**
1300          * Do the actual work. All child classes will need to implement this.
1301          * Return true to log the update as done or false (usually on failure).
1302          * @return Bool
1303          */
1304         abstract protected function doDBUpdates();
1305
1306         /**
1307          * Get the update key name to go in the update log table
1308          * @return String
1309          */
1310         abstract protected function getUpdateKey();
1311 }