includes/api/ApiQueryCategoryMembers.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on June 14, 2007
   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 if ( !defined( 'MEDIAWIKI' ) ) {
  28         // Eclipse helper - will be ignored in production
  29         require_once( "ApiQueryBase.php" );
  30 }
  31
  32 /**
  33  * A query module to enumerate pages that belong to a category.
  34  *
  35  * @ingroup API
  36  */
  37 class ApiQueryCategoryMembers extends ApiQueryGeneratorBase {
  38
  39         public function __construct( $query, $moduleName ) {
  40                 parent::__construct( $query, $moduleName, 'cm' );
  41         }
  42
  43         public function execute() {
  44                 $this->run();
  45         }
  46
  47         public function getCacheMode( $params ) {
  48                 return 'public';
  49         }
  50
  51         public function executeGenerator( $resultPageSet ) {
  52                 $this->run( $resultPageSet );
  53         }
  54
  55         /**
  56          * @param $resultPageSet ApiPageSet
  57          * @return void
  58          */
  59         private function run( $resultPageSet = null ) {
  60                 $params = $this->extractRequestParams();
  61
  62                 $this->requireOnlyOneParameter( $params, 'title', 'pageid' );
  63
  64                 if ( isset( $params['title'] ) ) {
  65                         $categoryTitle = Title::newFromText( $params['title'] );
  66
  67                         if ( is_null( $categoryTitle ) || $categoryTitle->getNamespace() != NS_CATEGORY ) {
  68                                 $this->dieUsage( 'The category name you entered is not valid', 'invalidcategory' );
  69                         }
  70                 } elseif( isset( $params['pageid'] ) ) {
  71                         $categoryTitle = Title::newFromID( $params['pageid'] );
  72
  73                         if ( !$categoryTitle ) {
  74                                 $this->dieUsageMsg( array( 'nosuchpageid', $params['pageid'] ) );
  75                         } elseif ( $categoryTitle->getNamespace() != NS_CATEGORY ) {
  76                                 $this->dieUsage( 'The category name you entered is not valid', 'invalidcategory' );
  77                         }
  78                 }
  79
  80                 $prop = array_flip( $params['prop'] );
  81                 $fld_ids = isset( $prop['ids'] );
  82                 $fld_title = isset( $prop['title'] );
  83                 $fld_sortkey = isset( $prop['sortkey'] );
  84                 $fld_sortkeyprefix = isset( $prop['sortkeyprefix'] );
  85                 $fld_timestamp = isset( $prop['timestamp'] );
  86                 $fld_type = isset( $prop['type'] );
  87
  88                 if ( is_null( $resultPageSet ) ) {
  89                         $this->addFields( array( 'cl_from', 'cl_sortkey', 'cl_type', 'page_namespace', 'page_title' ) );
  90                         $this->addFieldsIf( 'page_id', $fld_ids );
  91                         $this->addFieldsIf( 'cl_sortkey_prefix', $fld_sortkeyprefix );
  92                 } else {
  93                         $this->addFields( $resultPageSet->getPageTableFields() ); // will include page_ id, ns, title
  94                         $this->addFields( array( 'cl_from', 'cl_sortkey', 'cl_type' ) );
  95                 }
  96
  97                 $this->addFieldsIf( 'cl_timestamp', $fld_timestamp || $params['sort'] == 'timestamp' );
  98
  99                 $this->addTables( array( 'page', 'categorylinks' ) );   // must be in this order for 'USE INDEX'
 100
 101                 $this->addWhereFld( 'cl_to', $categoryTitle->getDBkey() );
 102                 $queryTypes = $params['type'];
 103                 $contWhere = false;
 104
 105                 // Scanning large datasets for rare categories sucks, and I already told
 106                 // how to have efficient subcategory access :-) ~~~~ (oh well, domas)
 107                 global $wgMiserMode;
 108                 $miser_ns = array();
 109                 if ( $wgMiserMode ) {
 110                         $miser_ns = $params['namespace'];
 111                 } else {
 112                         $this->addWhereFld( 'page_namespace', $params['namespace'] );
 113                 }
 114
 115                 $dir = $params['dir'] == 'asc' ? 'newer' : 'older';
 116
 117                 if ( $params['sort'] == 'timestamp' ) {
 118                         $this->addWhereRange( 'cl_timestamp',
 119                                 $dir,
 120                                 $params['start'],
 121                                 $params['end'] );
 122
 123                         $this->addOption( 'USE INDEX', 'cl_timestamp' );
 124                 } else {
 125                         if ( $params['continue'] ) {
 126                                 $cont = explode( '|', $params['continue'], 3 );
 127                                 if ( count( $cont ) != 3 ) {
 128                                         $this->dieUsage( 'Invalid continue param. You should pass the original value returned '.
 129                                                 'by the previous query', '_badcontinue'
 130                                         );
 131                                 }
 132
 133                                 // Remove the types to skip from $queryTypes
 134                                 $contTypeIndex = array_search( $cont[0], $queryTypes );
 135                                 $queryTypes = array_slice( $queryTypes, $contTypeIndex );
 136
 137                                 // Add a WHERE clause for sortkey and from
 138                                 // pack( "H*", $foo ) is used to convert hex back to binary
 139                                 $escSortkey = $this->getDB()->addQuotes( pack( "H*", $cont[1] ) );
 140                                 $from = intval( $cont[2] );
 141                                 $op = $dir == 'newer' ? '>' : '<';
 142                                 // $contWhere is used further down
 143                                 $contWhere = "cl_sortkey $op $escSortkey OR " .
 144                                         "(cl_sortkey = $escSortkey AND " .
 145                                         "cl_from $op= $from)";
 146
 147                         } else {
 148                                 // The below produces ORDER BY cl_sortkey, cl_from, possibly with DESC added to each of them
 149                                 $this->addWhereRange( 'cl_sortkey',
 150                                         $dir,
 151                                         $params['startsortkey'],
 152                                         $params['endsortkey'] );
 153                                 $this->addWhereRange( 'cl_from', $dir, null, null );
 154                         }
 155                         $this->addOption( 'USE INDEX', 'cl_sortkey' );
 156                 }
 157
 158                 $this->addWhere( 'cl_from=page_id' );
 159
 160                 $limit = $params['limit'];
 161                 $this->addOption( 'LIMIT', $limit + 1 );
 162
 163                 if ( $params['sort'] == 'sortkey' ) {
 164                         // Run a separate SELECT query for each value of cl_type.
 165                         // This is needed because cl_type is an enum, and MySQL has
 166                         // inconsistencies between ORDER BY cl_type and
 167                         // WHERE cl_type >= 'foo' making proper paging impossible
 168                         // and unindexed.
 169                         $rows = array();
 170                         $first = true;
 171                         foreach ( $queryTypes as $type ) {
 172                                 $extraConds = array( 'cl_type' => $type );
 173                                 if ( $first && $contWhere ) {
 174                                         // Continuation condition. Only added to the
 175                                         // first query, otherwise we'll skip things
 176                                         $extraConds[] = $contWhere;
 177                                 }
 178                                 $res = $this->select( __METHOD__, array( 'where' => $extraConds ) );
 179                                 $rows = array_merge( $rows, iterator_to_array( $res ) );
 180                                 if ( count( $rows ) >= $limit + 1 ) {
 181                                         break;
 182                                 }
 183                                 $first = false;
 184                         }
 185                 } else {
 186                         // Sorting by timestamp
 187                         // No need to worry about per-type queries because we
 188                         // aren't sorting or filtering by type anyway
 189                         $res = $this->select( __METHOD__ );
 190                         $rows = iterator_to_array( $res );
 191                 }
 192                 $count = 0;
 193                 foreach ( $rows as $row ) {
 194                         if ( ++ $count > $limit ) {
 195                                 // We've reached the one extra which shows that there are additional pages to be had. Stop here...
 196                                 // TODO: Security issue - if the user has no right to view next title, it will still be shown
 197                                 if ( $params['sort'] == 'timestamp' ) {
 198                                         $this->setContinueEnumParameter( 'start', wfTimestamp( TS_ISO_8601, $row->cl_timestamp ) );
 199                                 } else {
 200                                         $sortkey = bin2hex( $row->cl_sortkey );
 201                                         $this->setContinueEnumParameter( 'continue',
 202                                                 "{$row->cl_type}|$sortkey|{$row->cl_from}"
 203                                         );
 204                                 }
 205                                 break;
 206                         }
 207
 208                         // Since domas won't tell anyone what he told long ago, apply
 209                         // cmnamespace here. This means the query may return 0 actual
 210                         // results, but on the other hand it could save returning 5000
 211                         // useless results to the client. ~~~~
 212                         if ( count( $miser_ns ) && !in_array( $row->page_namespace, $miser_ns ) ) {
 213                                 continue;
 214                         }
 215
 216                         if ( is_null( $resultPageSet ) ) {
 217                                 $vals = array();
 218                                 if ( $fld_ids ) {
 219                                         $vals['pageid'] = intval( $row->page_id );
 220                                 }
 221                                 if ( $fld_title ) {
 222                                         $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 223                                         ApiQueryBase::addTitleInfo( $vals, $title );
 224                                 }
 225                                 if ( $fld_sortkey ) {
 226                                         $vals['sortkey'] = bin2hex( $row->cl_sortkey );
 227                                 }
 228                                 if ( $fld_sortkeyprefix ) {
 229                                         $vals['sortkeyprefix'] = $row->cl_sortkey_prefix;
 230                                 }
 231                                 if ( $fld_type  ) {
 232                                         $vals['type'] = $row->cl_type;
 233                                 }
 234                                 if ( $fld_timestamp ) {
 235                                         $vals['timestamp'] = wfTimestamp( TS_ISO_8601, $row->cl_timestamp );
 236                                 }
 237                                 $fit = $this->getResult()->addValue( array( 'query', $this->getModuleName() ),
 238                                                 null, $vals );
 239                                 if ( !$fit ) {
 240                                         if ( $params['sort'] == 'timestamp' ) {
 241                                                 $this->setContinueEnumParameter( 'start', wfTimestamp( TS_ISO_8601, $row->cl_timestamp ) );
 242                                         } else {
 243                                                 $sortkey = bin2hex( $row->cl_sortkey );
 244                                                 $this->setContinueEnumParameter( 'continue',
 245                                                         "{$row->cl_type}|$sortkey|{$row->cl_from}"
 246                                                 );
 247                                         }
 248                                         break;
 249                                 }
 250                         } else {
 251                                 $resultPageSet->processDbRow( $row );
 252                         }
 253                 }
 254
 255                 if ( is_null( $resultPageSet ) ) {
 256                         $this->getResult()->setIndexedTagName_internal(
 257                                          array( 'query', $this->getModuleName() ), 'cm' );
 258                 }
 259         }
 260
 261         public function getAllowedParams() {
 262                 return array(
 263                         'title' => array(
 264                                 ApiBase::PARAM_TYPE => 'string',
 265                         ),
 266                         'pageid' => array(
 267                                 ApiBase::PARAM_TYPE => 'integer'
 268                         ),
 269                         'prop' => array(
 270                                 ApiBase::PARAM_DFLT => 'ids|title',
 271                                 ApiBase::PARAM_ISMULTI => true,
 272                                 ApiBase::PARAM_TYPE => array (
 273                                         'ids',
 274                                         'title',
 275                                         'sortkey',
 276                                         'sortkeyprefix',
 277                                         'type',
 278                                         'timestamp',
 279                                 )
 280                         ),
 281                         'namespace' => array (
 282                                 ApiBase::PARAM_ISMULTI => true,
 283                                 ApiBase::PARAM_TYPE => 'namespace',
 284                         ),
 285                         'type' => array(
 286                                 ApiBase::PARAM_ISMULTI => true,
 287                                 ApiBase::PARAM_DFLT => 'page|subcat|file',
 288                                 ApiBase::PARAM_TYPE => array(
 289                                         'page',
 290                                         'subcat',
 291                                         'file'
 292                                 )
 293                         ),
 294                         'continue' => null,
 295                         'limit' => array(
 296                                 ApiBase::PARAM_TYPE => 'limit',
 297                                 ApiBase::PARAM_DFLT => 10,
 298                                 ApiBase::PARAM_MIN => 1,
 299                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_BIG1,
 300                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_BIG2
 301                         ),
 302                         'sort' => array(
 303                                 ApiBase::PARAM_DFLT => 'sortkey',
 304                                 ApiBase::PARAM_TYPE => array(
 305                                         'sortkey',
 306                                         'timestamp'
 307                                 )
 308                         ),
 309                         'dir' => array(
 310                                 ApiBase::PARAM_DFLT => 'asc',
 311                                 ApiBase::PARAM_TYPE => array(
 312                                         'asc',
 313                                         'desc'
 314                                 )
 315                         ),
 316                         'start' => array(
 317                                 ApiBase::PARAM_TYPE => 'timestamp'
 318                         ),
 319                         'end' => array(
 320                                 ApiBase::PARAM_TYPE => 'timestamp'
 321                         ),
 322                         'startsortkey' => null,
 323                         'endsortkey' => null,
 324                 );
 325         }
 326
 327         public function getParamDescription() {
 328                 global $wgMiserMode;
 329                 $p = $this->getModulePrefix();
 330                 $desc = array(
 331                         'title' => "Which category to enumerate (required). Must include Category: prefix. Cannot be used together with {$p}pageid",
 332                         'pageid' => "Page ID of the category to enumerate. Cannot be used together with {$p}title",
 333                         'prop' => array(
 334                                 'What pieces of information to include',
 335                                 ' ids           - Adds the page ID',
 336                                 ' title         - Adds the title and namespace ID of the page',
 337                                 ' sortkey       - Adds the sortkey used for sorting in the category (hexadecimal string)',
 338                                 ' sortkeyprefix - Adds the sortkey prefix used for sorting in the category (human-readable part of the sortkey)',
 339                                 ' type          - Adds the type that the page has been categorised as (page, subcat or file)',
 340                                 ' timestamp     - Adds the timestamp of when the page was included',
 341                         ),
 342                         'namespace' => 'Only include pages in these namespaces',
 343                         'type' => "What type of category members to include. Ignored when {$p}sort=timestamp is set",
 344                         'sort' => 'Property to sort by',
 345                         'dir' => 'In which direction to sort',
 346                         'start' => "Timestamp to start listing from. Can only be used with {$p}sort=timestamp",
 347                         'end' => "Timestamp to end listing at. Can only be used with {$p}sort=timestamp",
 348                         'startsortkey' => "Sortkey to start listing from. Must be given in binary format. Can only be used with {$p}sort=sortkey",
 349                         'endsortkey' => "Sortkey to end listing at. Must be given in binary format. Can only be used with {$p}sort=sortkey",
 350                         'continue' => 'For large categories, give the value retured from previous query',
 351                         'limit' => 'The maximum number of pages to return.',
 352                 );
 353
 354                 if ( $wgMiserMode ) {
 355                         $desc['namespace'] = array(
 356                                 $desc['namespace'],
 357                                 "NOTE: Due to \$wgMiserMode, using this may result in fewer than \"{$p}limit\" results",
 358                                 'returned before continuing; in extreme cases, zero results may be returned.',
 359                                 "Note that you can use {$p}type=subcat or {$p}type=file instead of {$p}namespace=14 or 6.",
 360                         );
 361                 }
 362                 return $desc;
 363         }
 364
 365         public function getDescription() {
 366                 return 'List all pages in a given category';
 367         }
 368
 369         public function getPossibleErrors() {
 370                 return array_merge( parent::getPossibleErrors(),
 371                         $this->getRequireOnlyOneParameterErrorMessages( array( 'title', 'pageid' ) ),
 372                         array(
 373                                 array( 'code' => 'invalidcategory', 'info' => 'The category name you entered is not valid' ),
 374                                 array( 'code' => 'badcontinue', 'info' => 'Invalid continue param. You should pass the original value returned by the previous query' ),
 375                                 array( 'nosuchpageid', 'pageid' ),
 376                         )
 377                 );
 378         }
 379
 380         protected function getExamples() {
 381                 return array(
 382                         'Get first 10 pages in [[Category:Physics]]:',
 383                         '  api.php?action=query&list=categorymembers&cmtitle=Category:Physics',
 384                         'Get page info about first 10 pages in [[Category:Physics]]:',
 385                         '  api.php?action=query&generator=categorymembers&gcmtitle=Category:Physics&prop=info',
 386                 );
 387         }
 388
 389         public function getVersion() {
 390                 return __CLASS__ . ': $Id$';
 391         }
 392 }