includes/api/ApiQueryAllPages.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on Sep 25, 2006
   6  *
   7  * Copyright © 2006 Yuri Astrakhan "<Firstname><Lastname>@gmail.com"
   8  *
   9  * This program is free software; you can redistribute it and/or modify
  10  * it under the terms of the GNU General Public License as published by
  11  * the Free Software Foundation; either version 2 of the License, or
  12  * (at your option) any later version.
  13  *
  14  * This program is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  17  * GNU General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU General Public License along
  20  * with this program; if not, write to the Free Software Foundation, Inc.,
  21  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  22  * http://www.gnu.org/copyleft/gpl.html
  23  *
  24  * @file
  25  */
  26
  27 /**
  28  * Query module to enumerate all available pages.
  29  *
  30  * @ingroup API
  31  */
  32 class ApiQueryAllPages extends ApiQueryGeneratorBase {
  33
  34         public function __construct( ApiQuery $query, $moduleName ) {
  35                 parent::__construct( $query, $moduleName, 'ap' );
  36         }
  37
  38         public function execute() {
  39                 $this->run();
  40         }
  41
  42         public function getCacheMode( $params ) {
  43                 return 'public';
  44         }
  45
  46         /**
  47          * @param ApiPageSet $resultPageSet
  48          * @return void
  49          */
  50         public function executeGenerator( $resultPageSet ) {
  51                 if ( $resultPageSet->isResolvingRedirects() ) {
  52                         $this->dieUsage(
  53                                 'Use "gapfilterredir=nonredirects" option instead of "redirects" ' .
  54                                         'when using allpages as a generator',
  55                                 'params'
  56                         );
  57                 }
  58
  59                 $this->run( $resultPageSet );
  60         }
  61
  62         /**
  63          * @param ApiPageSet $resultPageSet
  64          * @return void
  65          */
  66         private function run( $resultPageSet = null ) {
  67                 $db = $this->getDB();
  68
  69                 $params = $this->extractRequestParams();
  70
  71                 // Page filters
  72                 $this->addTables( 'page' );
  73
  74                 if ( !is_null( $params['continue'] ) ) {
  75                         $cont = explode( '|', $params['continue'] );
  76                         $this->dieContinueUsageIf( count( $cont ) != 1 );
  77                         $op = $params['dir'] == 'descending' ? '<' : '>';
  78                         $cont_from = $db->addQuotes( $cont[0] );
  79                         $this->addWhere( "page_title $op= $cont_from" );
  80                 }
  81
  82                 if ( $params['filterredir'] == 'redirects' ) {
  83                         $this->addWhereFld( 'page_is_redirect', 1 );
  84                 } elseif ( $params['filterredir'] == 'nonredirects' ) {
  85                         $this->addWhereFld( 'page_is_redirect', 0 );
  86                 }
  87
  88                 $this->addWhereFld( 'page_namespace', $params['namespace'] );
  89                 $dir = ( $params['dir'] == 'descending' ? 'older' : 'newer' );
  90                 $from = ( $params['from'] === null
  91                         ? null
  92                         : $this->titlePartToKey( $params['from'], $params['namespace'] ) );
  93                 $to = ( $params['to'] === null
  94                         ? null
  95                         : $this->titlePartToKey( $params['to'], $params['namespace'] ) );
  96                 $this->addWhereRange( 'page_title', $dir, $from, $to );
  97
  98                 if ( isset( $params['prefix'] ) ) {
  99                         $this->addWhere( 'page_title' . $db->buildLike(
 100                                 $this->titlePartToKey( $params['prefix'], $params['namespace'] ),
 101                                 $db->anyString() ) );
 102                 }
 103
 104                 if ( is_null( $resultPageSet ) ) {
 105                         $selectFields = array(
 106                                 'page_namespace',
 107                                 'page_title',
 108                                 'page_id'
 109                         );
 110                 } else {
 111                         $selectFields = $resultPageSet->getPageTableFields();
 112                 }
 113
 114                 $this->addFields( $selectFields );
 115                 $forceNameTitleIndex = true;
 116                 if ( isset( $params['minsize'] ) ) {
 117                         $this->addWhere( 'page_len>=' . intval( $params['minsize'] ) );
 118                         $forceNameTitleIndex = false;
 119                 }
 120
 121                 if ( isset( $params['maxsize'] ) ) {
 122                         $this->addWhere( 'page_len<=' . intval( $params['maxsize'] ) );
 123                         $forceNameTitleIndex = false;
 124                 }
 125
 126                 // Page protection filtering
 127                 if ( count( $params['prtype'] ) || $params['prexpiry'] != 'all' ) {
 128                         $this->addTables( 'page_restrictions' );
 129                         $this->addWhere( 'page_id=pr_page' );
 130                         $this->addWhere( "pr_expiry > {$db->addQuotes( $db->timestamp() )} OR pr_expiry IS NULL" );
 131
 132                         if ( count( $params['prtype'] ) ) {
 133                                 $this->addWhereFld( 'pr_type', $params['prtype'] );
 134
 135                                 if ( isset( $params['prlevel'] ) ) {
 136                                         // Remove the empty string and '*' from the prlevel array
 137                                         $prlevel = array_diff( $params['prlevel'], array( '', '*' ) );
 138
 139                                         if ( count( $prlevel ) ) {
 140                                                 $this->addWhereFld( 'pr_level', $prlevel );
 141                                         }
 142                                 }
 143                                 if ( $params['prfiltercascade'] == 'cascading' ) {
 144                                         $this->addWhereFld( 'pr_cascade', 1 );
 145                                 } elseif ( $params['prfiltercascade'] == 'noncascading' ) {
 146                                         $this->addWhereFld( 'pr_cascade', 0 );
 147                                 }
 148                         }
 149                         $forceNameTitleIndex = false;
 150
 151                         if ( $params['prexpiry'] == 'indefinite' ) {
 152                                 $this->addWhere( "pr_expiry = {$db->addQuotes( $db->getInfinity() )} OR pr_expiry IS NULL" );
 153                         } elseif ( $params['prexpiry'] == 'definite' ) {
 154                                 $this->addWhere( "pr_expiry != {$db->addQuotes( $db->getInfinity() )}" );
 155                         }
 156
 157                         $this->addOption( 'DISTINCT' );
 158                 } elseif ( isset( $params['prlevel'] ) ) {
 159                         $this->dieUsage( 'prlevel may not be used without prtype', 'params' );
 160                 }
 161
 162                 if ( $params['filterlanglinks'] == 'withoutlanglinks' ) {
 163                         $this->addTables( 'langlinks' );
 164                         $this->addJoinConds( array( 'langlinks' => array( 'LEFT JOIN', 'page_id=ll_from' ) ) );
 165                         $this->addWhere( 'll_from IS NULL' );
 166                         $forceNameTitleIndex = false;
 167                 } elseif ( $params['filterlanglinks'] == 'withlanglinks' ) {
 168                         $this->addTables( 'langlinks' );
 169                         $this->addWhere( 'page_id=ll_from' );
 170                         $this->addOption( 'STRAIGHT_JOIN' );
 171                         // We have to GROUP BY all selected fields to stop
 172                         // PostgreSQL from whining
 173                         $this->addOption( 'GROUP BY', $selectFields );
 174                         $forceNameTitleIndex = false;
 175                 }
 176
 177                 if ( $forceNameTitleIndex ) {
 178                         $this->addOption( 'USE INDEX', 'name_title' );
 179                 }
 180
 181                 $limit = $params['limit'];
 182                 $this->addOption( 'LIMIT', $limit + 1 );
 183                 $res = $this->select( __METHOD__ );
 184
 185                 //Get gender information
 186                 if ( MWNamespace::hasGenderDistinction( $params['namespace'] ) ) {
 187                         $users = array();
 188                         foreach ( $res as $row ) {
 189                                 $users[] = $row->page_title;
 190                         }
 191                         GenderCache::singleton()->doQuery( $users, __METHOD__ );
 192                         $res->rewind(); //reset
 193                 }
 194
 195                 $count = 0;
 196                 $result = $this->getResult();
 197                 foreach ( $res as $row ) {
 198                         if ( ++$count > $limit ) {
 199                                 // We've reached the one extra which shows that there are
 200                                 // additional pages to be had. Stop here...
 201                                 $this->setContinueEnumParameter( 'continue', $row->page_title );
 202                                 break;
 203                         }
 204
 205                         if ( is_null( $resultPageSet ) ) {
 206                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 207                                 $vals = array(
 208                                         'pageid' => intval( $row->page_id ),
 209                                         'ns' => intval( $title->getNamespace() ),
 210                                         'title' => $title->getPrefixedText()
 211                                 );
 212                                 $fit = $result->addValue( array( 'query', $this->getModuleName() ), null, $vals );
 213                                 if ( !$fit ) {
 214                                         $this->setContinueEnumParameter( 'continue', $row->page_title );
 215                                         break;
 216                                 }
 217                         } else {
 218                                 $resultPageSet->processDbRow( $row );
 219                         }
 220                 }
 221
 222                 if ( is_null( $resultPageSet ) ) {
 223                         $result->setIndexedTagName_internal( array( 'query', $this->getModuleName() ), 'p' );
 224                 }
 225         }
 226
 227         public function getAllowedParams() {
 228                 return array(
 229                         'from' => null,
 230                         'continue' => null,
 231                         'to' => null,
 232                         'prefix' => null,
 233                         'namespace' => array(
 234                                 ApiBase::PARAM_DFLT => NS_MAIN,
 235                                 ApiBase::PARAM_TYPE => 'namespace',
 236                         ),
 237                         'filterredir' => array(
 238                                 ApiBase::PARAM_DFLT => 'all',
 239                                 ApiBase::PARAM_TYPE => array(
 240                                         'all',
 241                                         'redirects',
 242                                         'nonredirects'
 243                                 )
 244                         ),
 245                         'minsize' => array(
 246                                 ApiBase::PARAM_TYPE => 'integer',
 247                         ),
 248                         'maxsize' => array(
 249                                 ApiBase::PARAM_TYPE => 'integer',
 250                         ),
 251                         'prtype' => array(
 252                                 ApiBase::PARAM_TYPE => Title::getFilteredRestrictionTypes( true ),
 253                                 ApiBase::PARAM_ISMULTI => true
 254                         ),
 255                         'prlevel' => array(
 256                                 ApiBase::PARAM_TYPE => $this->getConfig()->get( 'RestrictionLevels' ),
 257                                 ApiBase::PARAM_ISMULTI => true
 258                         ),
 259                         'prfiltercascade' => array(
 260                                 ApiBase::PARAM_DFLT => 'all',
 261                                 ApiBase::PARAM_TYPE => array(
 262                                         'cascading',
 263                                         'noncascading',
 264                                         'all'
 265                                 ),
 266                         ),
 267                         'limit' => array(
 268                                 ApiBase::PARAM_DFLT => 10,
 269                                 ApiBase::PARAM_TYPE => 'limit',
 270                                 ApiBase::PARAM_MIN => 1,
 271                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_BIG1,
 272                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_BIG2
 273                         ),
 274                         'dir' => array(
 275                                 ApiBase::PARAM_DFLT => 'ascending',
 276                                 ApiBase::PARAM_TYPE => array(
 277                                         'ascending',
 278                                         'descending'
 279                                 )
 280                         ),
 281                         'filterlanglinks' => array(
 282                                 ApiBase::PARAM_TYPE => array(
 283                                         'withlanglinks',
 284                                         'withoutlanglinks',
 285                                         'all'
 286                                 ),
 287                                 ApiBase::PARAM_DFLT => 'all'
 288                         ),
 289                         'prexpiry' => array(
 290                                 ApiBase::PARAM_TYPE => array(
 291                                         'indefinite',
 292                                         'definite',
 293                                         'all'
 294                                 ),
 295                                 ApiBase::PARAM_DFLT => 'all'
 296                         ),
 297                 );
 298         }
 299
 300         public function getParamDescription() {
 301                 $p = $this->getModulePrefix();
 302
 303                 return array(
 304                         'from' => 'The page title to start enumerating from',
 305                         'continue' => 'When more results are available, use this to continue',
 306                         'to' => 'The page title to stop enumerating at',
 307                         'prefix' => 'Search for all page titles that begin with this value',
 308                         'namespace' => 'The namespace to enumerate',
 309                         'filterredir' => 'Which pages to list',
 310                         'dir' => 'The direction in which to list',
 311                         'minsize' => 'Limit to pages with at least this many bytes',
 312                         'maxsize' => 'Limit to pages with at most this many bytes',
 313                         'prtype' => 'Limit to protected pages only',
 314                         'prlevel' => "The protection level (must be used with {$p}prtype= parameter)",
 315                         'prfiltercascade'
 316                                 => "Filter protections based on cascadingness (ignored when {$p}prtype isn't set)",
 317                         'filterlanglinks' => array(
 318                                 'Filter based on whether a page has langlinks',
 319                                 'Note that this may not consider langlinks added by extensions.',
 320                         ),
 321                         'limit' => 'How many total pages to return.',
 322                         'prexpiry' => array(
 323                                 'Which protection expiry to filter the page on',
 324                                 ' indefinite - Get only pages with indefinite protection expiry',
 325                                 ' definite - Get only pages with a definite (specific) protection expiry',
 326                                 ' all - Get pages with any protections expiry'
 327                         ),
 328                 );
 329         }
 330
 331         public function getDescription() {
 332                 return 'Enumerate all pages sequentially in a given namespace.';
 333         }
 334
 335         public function getExamples() {
 336                 return array(
 337                         'api.php?action=query&list=allpages&apfrom=B' => array(
 338                                 'Simple Use',
 339                                 'Show a list of pages starting at the letter "B"',
 340                         ),
 341                         'api.php?action=query&generator=allpages&gaplimit=4&gapfrom=T&prop=info' => array(
 342                                 'Using as Generator',
 343                                 'Show info about 4 pages starting at the letter "T"',
 344                         ),
 345                         'api.php?action=query&generator=allpages&gaplimit=2&' .
 346                                 'gapfilterredir=nonredirects&gapfrom=Re&prop=revisions&rvprop=content'
 347                                 => array( 'Show content of first 2 non-redirect pages beginning at "Re"' )
 348                 );
 349         }
 350
 351         public function getHelpUrls() {
 352                 return 'https://www.mediawiki.org/wiki/API:Allpages';
 353         }
 354 }