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( $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 $resultPageSet ApiPageSet
  48          * @return void
  49          */
  50         public function executeGenerator( $resultPageSet ) {
  51                 if ( $resultPageSet->isResolvingRedirects() ) {
  52                         $this->dieUsage( 'Use "gapfilterredir=nonredirects" option instead of "redirects" when using allpages as a generator', 'params' );
  53                 }
  54
  55                 $this->run( $resultPageSet );
  56         }
  57
  58         /**
  59          * @param $resultPageSet ApiPageSet
  60          * @return void
  61          */
  62         private function run( $resultPageSet = null ) {
  63                 $db = $this->getDB();
  64
  65                 $params = $this->extractRequestParams();
  66
  67                 // Page filters
  68                 $this->addTables( 'page' );
  69
  70                 if ( !is_null( $params['continue'] ) ) {
  71                         $cont = explode( '|', $params['continue'] );
  72                         $this->dieContinueUsageIf( count( $cont ) != 1 );
  73                         $op = $params['dir'] == 'descending' ? '<' : '>';
  74                         $cont_from = $db->addQuotes( $cont[0] );
  75                         $this->addWhere( "page_title $op= $cont_from" );
  76                 }
  77
  78                 if ( $params['filterredir'] == 'redirects' ) {
  79                         $this->addWhereFld( 'page_is_redirect', 1 );
  80                 } elseif ( $params['filterredir'] == 'nonredirects' ) {
  81                         $this->addWhereFld( 'page_is_redirect', 0 );
  82                 }
  83
  84                 $this->addWhereFld( 'page_namespace', $params['namespace'] );
  85                 $dir = ( $params['dir'] == 'descending' ? 'older' : 'newer' );
  86                 $from = ( is_null( $params['from'] ) ? null : $this->titlePartToKey( $params['from'] ) );
  87                 $to = ( is_null( $params['to'] ) ? null : $this->titlePartToKey( $params['to'] ) );
  88                 $this->addWhereRange( 'page_title', $dir, $from, $to );
  89
  90                 if ( isset( $params['prefix'] ) ) {
  91                         $this->addWhere( 'page_title' . $db->buildLike( $this->titlePartToKey( $params['prefix'] ), $db->anyString() ) );
  92                 }
  93
  94                 if ( is_null( $resultPageSet ) ) {
  95                         $selectFields = array(
  96                                 'page_namespace',
  97                                 'page_title',
  98                                 'page_id'
  99                         );
 100                 } else {
 101                         $selectFields = $resultPageSet->getPageTableFields();
 102                 }
 103
 104                 $this->addFields( $selectFields );
 105                 $forceNameTitleIndex = true;
 106                 if ( isset( $params['minsize'] ) ) {
 107                         $this->addWhere( 'page_len>=' . intval( $params['minsize'] ) );
 108                         $forceNameTitleIndex = false;
 109                 }
 110
 111                 if ( isset( $params['maxsize'] ) ) {
 112                         $this->addWhere( 'page_len<=' . intval( $params['maxsize'] ) );
 113                         $forceNameTitleIndex = false;
 114                 }
 115
 116                 // Page protection filtering
 117                 if ( count( $params['prtype'] ) || $params['prexpiry'] != 'all' ) {
 118                         $this->addTables( 'page_restrictions' );
 119                         $this->addWhere( 'page_id=pr_page' );
 120                         $this->addWhere( "pr_expiry > {$db->addQuotes( $db->timestamp() )} OR pr_expiry IS NULL" );
 121
 122                         if ( count( $params['prtype'] ) ) {
 123                                 $this->addWhereFld( 'pr_type', $params['prtype'] );
 124
 125                                 if ( isset( $params['prlevel'] ) ) {
 126                                         // Remove the empty string and '*' from the prlevel array
 127                                         $prlevel = array_diff( $params['prlevel'], array( '', '*' ) );
 128
 129                                         if ( count( $prlevel ) ) {
 130                                                 $this->addWhereFld( 'pr_level', $prlevel );
 131                                         }
 132                                 }
 133                                 if ( $params['prfiltercascade'] == 'cascading' ) {
 134                                         $this->addWhereFld( 'pr_cascade', 1 );
 135                                 } elseif ( $params['prfiltercascade'] == 'noncascading' ) {
 136                                         $this->addWhereFld( 'pr_cascade', 0 );
 137                                 }
 138                         }
 139                         $forceNameTitleIndex = false;
 140
 141                         if ( $params['prexpiry'] == 'indefinite' ) {
 142                                 $this->addWhere( "pr_expiry = {$db->addQuotes( $db->getInfinity() )} OR pr_expiry IS NULL" );
 143                         } elseif ( $params['prexpiry'] == 'definite' ) {
 144                                 $this->addWhere( "pr_expiry != {$db->addQuotes( $db->getInfinity() )}" );
 145                         }
 146
 147                         $this->addOption( 'DISTINCT' );
 148
 149                 } elseif ( isset( $params['prlevel'] ) ) {
 150                         $this->dieUsage( 'prlevel may not be used without prtype', 'params' );
 151                 }
 152
 153                 if ( $params['filterlanglinks'] == 'withoutlanglinks' ) {
 154                         $this->addTables( 'langlinks' );
 155                         $this->addJoinConds( array( 'langlinks' => array( 'LEFT JOIN', 'page_id=ll_from' ) ) );
 156                         $this->addWhere( 'll_from IS NULL' );
 157                         $forceNameTitleIndex = false;
 158                 } elseif ( $params['filterlanglinks'] == 'withlanglinks' ) {
 159                         $this->addTables( 'langlinks' );
 160                         $this->addWhere( 'page_id=ll_from' );
 161                         $this->addOption( 'STRAIGHT_JOIN' );
 162                         // We have to GROUP BY all selected fields to stop
 163                         // PostgreSQL from whining
 164                         $this->addOption( 'GROUP BY', $selectFields );
 165                         $forceNameTitleIndex = false;
 166                 }
 167
 168                 if ( $forceNameTitleIndex ) {
 169                         $this->addOption( 'USE INDEX', 'name_title' );
 170                 }
 171
 172                 $limit = $params['limit'];
 173                 $this->addOption( 'LIMIT', $limit + 1 );
 174                 $res = $this->select( __METHOD__ );
 175
 176                 //Get gender information
 177                 if ( MWNamespace::hasGenderDistinction( $params['namespace'] ) ) {
 178                         $users = array();
 179                         foreach ( $res as $row ) {
 180                                 $users[] = $row->page_title;
 181                         }
 182                         GenderCache::singleton()->doQuery( $users, __METHOD__ );
 183                         $res->rewind(); //reset
 184                 }
 185
 186                 $count = 0;
 187                 $result = $this->getResult();
 188                 foreach ( $res as $row ) {
 189                         if ( ++ $count > $limit ) {
 190                                 // We've reached the one extra which shows that there are additional pages to be had. Stop here...
 191                                 $this->setContinueEnumParameter( 'continue', $row->page_title );
 192                                 break;
 193                         }
 194
 195                         if ( is_null( $resultPageSet ) ) {
 196                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 197                                 $vals = array(
 198                                         'pageid' => intval( $row->page_id ),
 199                                         'ns' => intval( $title->getNamespace() ),
 200                                         'title' => $title->getPrefixedText()
 201                                 );
 202                                 $fit = $result->addValue( array( 'query', $this->getModuleName() ), null, $vals );
 203                                 if ( !$fit ) {
 204                                         $this->setContinueEnumParameter( 'continue', $row->page_title );
 205                                         break;
 206                                 }
 207                         } else {
 208                                 $resultPageSet->processDbRow( $row );
 209                         }
 210                 }
 211
 212                 if ( is_null( $resultPageSet ) ) {
 213                         $result->setIndexedTagName_internal( array( 'query', $this->getModuleName() ), 'p' );
 214                 }
 215         }
 216
 217         public function getAllowedParams() {
 218                 global $wgRestrictionLevels;
 219
 220                 return array(
 221                         'from' => null,
 222                         'continue' => null,
 223                         'to' => null,
 224                         'prefix' => null,
 225                         'namespace' => array(
 226                                 ApiBase::PARAM_DFLT => NS_MAIN,
 227                                 ApiBase::PARAM_TYPE => 'namespace',
 228                         ),
 229                         'filterredir' => array(
 230                                 ApiBase::PARAM_DFLT => 'all',
 231                                 ApiBase::PARAM_TYPE => array(
 232                                         'all',
 233                                         'redirects',
 234                                         'nonredirects'
 235                                 )
 236                         ),
 237                         'minsize' => array(
 238                                 ApiBase::PARAM_TYPE => 'integer',
 239                         ),
 240                         'maxsize' => array(
 241                                 ApiBase::PARAM_TYPE => 'integer',
 242                         ),
 243                         'prtype' => array(
 244                                 ApiBase::PARAM_TYPE => Title::getFilteredRestrictionTypes( true ),
 245                                 ApiBase::PARAM_ISMULTI => true
 246                         ),
 247                         'prlevel' => array(
 248                                 ApiBase::PARAM_TYPE => $wgRestrictionLevels,
 249                                 ApiBase::PARAM_ISMULTI => true
 250                         ),
 251                         'prfiltercascade' => array(
 252                                 ApiBase::PARAM_DFLT => 'all',
 253                                 ApiBase::PARAM_TYPE => array(
 254                                         'cascading',
 255                                         'noncascading',
 256                                         'all'
 257                                 ),
 258                         ),
 259                         'limit' => array(
 260                                 ApiBase::PARAM_DFLT => 10,
 261                                 ApiBase::PARAM_TYPE => 'limit',
 262                                 ApiBase::PARAM_MIN => 1,
 263                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_BIG1,
 264                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_BIG2
 265                         ),
 266                         'dir' => array(
 267                                 ApiBase::PARAM_DFLT => 'ascending',
 268                                 ApiBase::PARAM_TYPE => array(
 269                                         'ascending',
 270                                         'descending'
 271                                 )
 272                         ),
 273                         'filterlanglinks' => array(
 274                                 ApiBase::PARAM_TYPE => array(
 275                                         'withlanglinks',
 276                                         'withoutlanglinks',
 277                                         'all'
 278                                 ),
 279                                 ApiBase::PARAM_DFLT => 'all'
 280                         ),
 281                         'prexpiry' => array(
 282                                 ApiBase::PARAM_TYPE => array(
 283                                         'indefinite',
 284                                         'definite',
 285                                         'all'
 286                                 ),
 287                                 ApiBase::PARAM_DFLT => 'all'
 288                         ),
 289                 );
 290         }
 291
 292         public function getParamDescription() {
 293                 $p = $this->getModulePrefix();
 294                 return array(
 295                         'from' => 'The page title to start enumerating from',
 296                         'continue' => 'When more results are available, use this to continue',
 297                         'to' => 'The page title to stop enumerating at',
 298                         'prefix' => 'Search for all page titles that begin with this value',
 299                         'namespace' => 'The namespace to enumerate',
 300                         'filterredir' => 'Which pages to list',
 301                         'dir' => 'The direction in which to list',
 302                         'minsize' => 'Limit to pages with at least this many bytes',
 303                         'maxsize' => 'Limit to pages with at most this many bytes',
 304                         'prtype' => 'Limit to protected pages only',
 305                         'prlevel' => "The protection level (must be used with {$p}prtype= parameter)",
 306                         'prfiltercascade' => "Filter protections based on cascadingness (ignored when {$p}prtype isn't set)",
 307                         'filterlanglinks' => 'Filter based on whether a page has langlinks',
 308                         'limit' => 'How many total pages to return.',
 309                         'prexpiry' => array(
 310                                 'Which protection expiry to filter the page on',
 311                                 ' indefinite - Get only pages with indefinite protection expiry',
 312                                 ' definite - Get only pages with a definite (specific) protection expiry',
 313                                 ' all - Get pages with any protections expiry'
 314                         ),
 315                 );
 316         }
 317
 318         public function getResultProperties() {
 319                 return array(
 320                         '' => array(
 321                                 'pageid' => 'integer',
 322                                 'ns' => 'namespace',
 323                                 'title' => 'string'
 324                         )
 325                 );
 326         }
 327
 328         public function getDescription() {
 329                 return 'Enumerate all pages sequentially in a given namespace';
 330         }
 331
 332         public function getPossibleErrors() {
 333                 return array_merge( parent::getPossibleErrors(), array(
 334                         array( 'code' => 'params', 'info' => 'Use "gapfilterredir=nonredirects" option instead of "redirects" when using allpages as a generator' ),
 335                         array( 'code' => 'params', 'info' => 'prlevel may not be used without prtype' ),
 336                 ) );
 337         }
 338
 339         public function getExamples() {
 340                 return array(
 341                         'api.php?action=query&list=allpages&apfrom=B' => array(
 342                                 'Simple Use',
 343                                 'Show a list of pages starting at the letter "B"',
 344                         ),
 345                         'api.php?action=query&generator=allpages&gaplimit=4&gapfrom=T&prop=info' => array(
 346                                 'Using as Generator',
 347                                 'Show info about 4 pages starting at the letter "T"',
 348                         ),
 349                         'api.php?action=query&generator=allpages&gaplimit=2&gapfilterredir=nonredirects&gapfrom=Re&prop=revisions&rvprop=content' => array(
 350                                 'Show content of first 2 non-redirect pages beginning at "Re"',
 351                         )
 352                 );
 353         }
 354
 355         public function getHelpUrls() {
 356                 return 'https://www.mediawiki.org/wiki/API:Allpages';
 357         }
 358 }