includes/api/ApiQueryDuplicateFiles.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on Sep 27, 2008
   6  *
   7  * Copyright © 2008 Roan Kattouw "<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  * A query module to list duplicates of the given file(s)
  29  *
  30  * @ingroup API
  31  */
  32 class ApiQueryDuplicateFiles extends ApiQueryGeneratorBase {
  33
  34         public function __construct( $query, $moduleName ) {
  35                 parent::__construct( $query, $moduleName, 'df' );
  36         }
  37
  38         public function execute() {
  39                 $this->run();
  40         }
  41
  42         public function getCacheMode( $params ) {
  43                 return 'public';
  44         }
  45
  46         public function executeGenerator( $resultPageSet ) {
  47                 $this->run( $resultPageSet );
  48         }
  49
  50         /**
  51          * @param ApiPageSet $resultPageSet
  52          */
  53         private function run( $resultPageSet = null ) {
  54                 $params = $this->extractRequestParams();
  55                 $namespaces = $this->getPageSet()->getAllTitlesByNamespace();
  56                 if ( empty( $namespaces[NS_FILE] ) ) {
  57                         return;
  58                 }
  59                 $images = $namespaces[NS_FILE];
  60
  61                 if ( $params['dir'] == 'descending' ) {
  62                         $images = array_reverse( $images );
  63                 }
  64
  65                 $skipUntilThisDup = false;
  66                 if ( isset( $params['continue'] ) ) {
  67                         $cont = explode( '|', $params['continue'] );
  68                         $this->dieContinueUsageIf( count( $cont ) != 2 );
  69                         $fromImage = $cont[0];
  70                         $skipUntilThisDup = $cont[1];
  71                         // Filter out any images before $fromImage
  72                         foreach ( $images as $image => $pageId ) {
  73                                 if ( $image < $fromImage ) {
  74                                         unset( $images[$image] );
  75                                 } else {
  76                                         break;
  77                                 }
  78                         }
  79                 }
  80
  81                 $filesToFind = array_keys( $images );
  82                 if ( $params['localonly'] ) {
  83                         $files = RepoGroup::singleton()->getLocalRepo()->findFiles( $filesToFind );
  84                 } else {
  85                         $files = RepoGroup::singleton()->findFiles( $filesToFind );
  86                 }
  87
  88                 $fit = true;
  89                 $count = 0;
  90                 $titles = array();
  91
  92                 $sha1s = array();
  93                 foreach ( $files as $file ) {
  94                         /** @var $file File */
  95                         $sha1s[$file->getName()] = $file->getSha1();
  96                 }
  97
  98                 // find all files with the hashes, result format is: array( hash => array( dup1, dup2 ), hash1 => ... )
  99                 $filesToFindBySha1s = array_unique( array_values( $sha1s ) );
 100                 if ( $params['localonly'] ) {
 101                         $filesBySha1s = RepoGroup::singleton()->getLocalRepo()->findBySha1s( $filesToFindBySha1s );
 102                 } else {
 103                         $filesBySha1s = RepoGroup::singleton()->findBySha1s( $filesToFindBySha1s );
 104                 }
 105
 106                 // iterate over $images to handle continue param correct
 107                 foreach ( $images as $image => $pageId ) {
 108                         if ( !isset( $sha1s[$image] ) ) {
 109                                 continue; //file does not exist
 110                         }
 111                         $sha1 = $sha1s[$image];
 112                         $dupFiles = $filesBySha1s[$sha1];
 113                         if ( $params['dir'] == 'descending' ) {
 114                                 $dupFiles = array_reverse( $dupFiles );
 115                         }
 116                         /** @var $dupFile File */
 117                         foreach ( $dupFiles as $dupFile ) {
 118                                 $dupName = $dupFile->getName();
 119                                 if ( $image == $dupName && $dupFile->isLocal() ) {
 120                                         continue; //ignore the local file itself
 121                                 }
 122                                 if ( $skipUntilThisDup !== false && $dupName < $skipUntilThisDup ) {
 123                                         continue; //skip to pos after the image from continue param
 124                                 }
 125                                 $skipUntilThisDup = false;
 126                                 if ( ++$count > $params['limit'] ) {
 127                                         $fit = false; //break outer loop
 128                                         // We're one over limit which shows that
 129                                         // there are additional images to be had. Stop here...
 130                                         $this->setContinueEnumParameter( 'continue', $image . '|' . $dupName );
 131                                         break;
 132                                 }
 133                                 if ( !is_null( $resultPageSet ) ) {
 134                                         $titles[] = $dupFile->getTitle();
 135                                 } else {
 136                                         $r = array(
 137                                                 'name' => $dupName,
 138                                                 'user' => $dupFile->getUser( 'text' ),
 139                                                 'timestamp' => wfTimestamp( TS_ISO_8601, $dupFile->getTimestamp() )
 140                                         );
 141                                         if ( !$dupFile->isLocal() ) {
 142                                                 $r['shared'] = '';
 143                                         }
 144                                         $fit = $this->addPageSubItem( $pageId, $r );
 145                                         if ( !$fit ) {
 146                                                 $this->setContinueEnumParameter( 'continue', $image . '|' . $dupName );
 147                                                 break;
 148                                         }
 149                                 }
 150                         }
 151                         if ( !$fit ) {
 152                                 break;
 153                         }
 154                 }
 155                 if ( !is_null( $resultPageSet ) ) {
 156                         $resultPageSet->populateFromTitles( $titles );
 157                 }
 158         }
 159
 160         public function getAllowedParams() {
 161                 return array(
 162                         'limit' => array(
 163                                 ApiBase::PARAM_DFLT => 10,
 164                                 ApiBase::PARAM_TYPE => 'limit',
 165                                 ApiBase::PARAM_MIN => 1,
 166                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_BIG1,
 167                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_BIG2
 168                         ),
 169                         'continue' => null,
 170                         'dir' => array(
 171                                 ApiBase::PARAM_DFLT => 'ascending',
 172                                 ApiBase::PARAM_TYPE => array(
 173                                         'ascending',
 174                                         'descending'
 175                                 )
 176                         ),
 177                         'localonly' => false,
 178                 );
 179         }
 180
 181         public function getParamDescription() {
 182                 return array(
 183                         'limit' => 'How many duplicate files to return',
 184                         'continue' => 'When more results are available, use this to continue',
 185                         'dir' => 'The direction in which to list',
 186                         'localonly' => 'Look only for files in the local repository',
 187                 );
 188         }
 189
 190         public function getResultProperties() {
 191                 return array(
 192                         '' => array(
 193                                 'name' => 'string',
 194                                 'user' => 'string',
 195                                 'timestamp' => 'timestamp',
 196                                 'shared' => 'boolean',
 197                         )
 198                 );
 199         }
 200
 201         public function getDescription() {
 202                 return 'List all files that are duplicates of the given file(s) based on hash values';
 203         }
 204
 205         public function getExamples() {
 206                 return array(
 207                         'api.php?action=query&titles=File:Albert_Einstein_Head.jpg&prop=duplicatefiles',
 208                         'api.php?action=query&generator=allimages&prop=duplicatefiles',
 209                 );
 210         }
 211
 212         public function getHelpUrls() {
 213                 return 'https://www.mediawiki.org/wiki/API:Properties#duplicatefiles_.2F_df';
 214         }
 215 }