includes/api/ApiQueryExtLinksUsage.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on July 7, 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 /**
  28  * @ingroup API
  29  */
  30 class ApiQueryExtLinksUsage extends ApiQueryGeneratorBase {
  31
  32         public function __construct( $query, $moduleName ) {
  33                 parent::__construct( $query, $moduleName, 'eu' );
  34         }
  35
  36         public function execute() {
  37                 $this->run();
  38         }
  39
  40         public function getCacheMode( $params ) {
  41                 return 'public';
  42         }
  43
  44         public function executeGenerator( $resultPageSet ) {
  45                 $this->run( $resultPageSet );
  46         }
  47
  48         /**
  49          * @param $resultPageSet ApiPageSet
  50          * @return void
  51          */
  52         private function run( $resultPageSet = null ) {
  53                 $params = $this->extractRequestParams();
  54
  55                 $query = $params['query'];
  56                 $protocol = self::getProtocolPrefix( $params['protocol'] );
  57
  58                 $this->addTables( array( 'page', 'externallinks' ) ); // must be in this order for 'USE INDEX'
  59                 $this->addOption( 'USE INDEX', 'el_index' );
  60                 $this->addWhere( 'page_id=el_from' );
  61
  62                 global $wgMiserMode;
  63                 $miser_ns = array();
  64                 if ( $wgMiserMode ) {
  65                         $miser_ns = $params['namespace'];
  66                 } else {
  67                         $this->addWhereFld( 'page_namespace', $params['namespace'] );
  68                 }
  69
  70                 $whereQuery = $this->prepareUrlQuerySearchString( $query, $protocol );
  71
  72                 if ( $whereQuery !== null ) {
  73                         $this->addWhere( $whereQuery );
  74                 }
  75
  76                 $prop = array_flip( $params['prop'] );
  77                 $fld_ids = isset( $prop['ids'] );
  78                 $fld_title = isset( $prop['title'] );
  79                 $fld_url = isset( $prop['url'] );
  80
  81                 if ( is_null( $resultPageSet ) ) {
  82                         $this->addFields( array(
  83                                 'page_id',
  84                                 'page_namespace',
  85                                 'page_title'
  86                         ) );
  87                         $this->addFieldsIf( 'el_to', $fld_url );
  88                 } else {
  89                         $this->addFields( $resultPageSet->getPageTableFields() );
  90                 }
  91
  92                 $limit = $params['limit'];
  93                 $offset = $params['offset'];
  94                 $this->addOption( 'LIMIT', $limit + 1 );
  95                 if ( isset( $offset ) ) {
  96                         $this->addOption( 'OFFSET', $offset );
  97                 }
  98
  99                 $res = $this->select( __METHOD__ );
 100
 101                 $result = $this->getResult();
 102                 $count = 0;
 103                 foreach ( $res as $row ) {
 104                         if ( ++ $count > $limit ) {
 105                                 // We've reached the one extra which shows that there are additional pages to be had. Stop here...
 106                                 $this->setContinueEnumParameter( 'offset', $offset + $limit );
 107                                 break;
 108                         }
 109
 110                         if ( count( $miser_ns ) && !in_array( $row->page_namespace, $miser_ns ) ) {
 111                                 continue;
 112                         }
 113
 114                         if ( is_null( $resultPageSet ) ) {
 115                                 $vals = array();
 116                                 if ( $fld_ids ) {
 117                                         $vals['pageid'] = intval( $row->page_id );
 118                                 }
 119                                 if ( $fld_title ) {
 120                                         $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 121                                         ApiQueryBase::addTitleInfo( $vals, $title );
 122                                 }
 123                                 if ( $fld_url ) {
 124                                         $to = $row->el_to;
 125                                         // expand protocol-relative urls
 126                                         if ( $params['expandurl'] ) {
 127                                                 $to = wfExpandUrl( $to, PROTO_CANONICAL );
 128                                         }
 129                                         $vals['url'] = $to;
 130                                 }
 131                                 $fit = $result->addValue( array( 'query', $this->getModuleName() ), null, $vals );
 132                                 if ( !$fit ) {
 133                                         $this->setContinueEnumParameter( 'offset', $offset + $count - 1 );
 134                                         break;
 135                                 }
 136                         } else {
 137                                 $resultPageSet->processDbRow( $row );
 138                         }
 139                 }
 140
 141                 if ( is_null( $resultPageSet ) ) {
 142                         $result->setIndexedTagName_internal( array( 'query', $this->getModuleName() ),
 143                                         $this->getModulePrefix() );
 144                 }
 145         }
 146
 147         public function getAllowedParams() {
 148                 return array(
 149                         'prop' => array(
 150                                 ApiBase::PARAM_ISMULTI => true,
 151                                 ApiBase::PARAM_DFLT => 'ids|title|url',
 152                                 ApiBase::PARAM_TYPE => array(
 153                                         'ids',
 154                                         'title',
 155                                         'url'
 156                                 )
 157                         ),
 158                         'offset' => array(
 159                                 ApiBase::PARAM_TYPE => 'integer'
 160                         ),
 161                         'protocol' => array(
 162                                 ApiBase::PARAM_TYPE => self::prepareProtocols(),
 163                                 ApiBase::PARAM_DFLT => '',
 164                         ),
 165                         'query' => null,
 166                         'namespace' => array(
 167                                 ApiBase::PARAM_ISMULTI => true,
 168                                 ApiBase::PARAM_TYPE => 'namespace'
 169                         ),
 170                         'limit' => array(
 171                                 ApiBase::PARAM_DFLT => 10,
 172                                 ApiBase::PARAM_TYPE => 'limit',
 173                                 ApiBase::PARAM_MIN => 1,
 174                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_BIG1,
 175                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_BIG2
 176                         ),
 177                         'expandurl' => false,
 178                 );
 179         }
 180
 181         public static function prepareProtocols() {
 182                 global $wgUrlProtocols;
 183                 $protocols = array( '' );
 184                 foreach ( $wgUrlProtocols as $p ) {
 185                         if ( $p !== '//' ) {
 186                                 $protocols[] = substr( $p, 0, strpos( $p, ':' ) );
 187                         }
 188                 }
 189                 return $protocols;
 190         }
 191
 192         public static function getProtocolPrefix( $protocol ) {
 193                 // Find the right prefix
 194                 global $wgUrlProtocols;
 195                 if ( $protocol && !in_array( $protocol, $wgUrlProtocols ) ) {
 196                         foreach ( $wgUrlProtocols as $p ) {
 197                                 if ( substr( $p, 0, strlen( $protocol ) ) === $protocol ) {
 198                                         $protocol = $p;
 199                                         break;
 200                                 }
 201                         }
 202
 203                         return $protocol;
 204                 } else {
 205                         return null;
 206                 }
 207         }
 208
 209         public function getParamDescription() {
 210                 global $wgMiserMode;
 211                 $p = $this->getModulePrefix();
 212                 $desc = array(
 213                         'prop' => array(
 214                                 'What pieces of information to include',
 215                                 ' ids    - Adds the ID of page',
 216                                 ' title  - Adds the title and namespace ID of the page',
 217                                 ' url    - Adds the URL used in the page',
 218                         ),
 219                         'offset' => 'Used for paging. Use the value returned for "continue"',
 220                         'protocol' => array(
 221                                 "Protocol of the URL. If empty and {$p}query set, the protocol is http.",
 222                                 "Leave both this and {$p}query empty to list all external links"
 223                         ),
 224                         'query' => 'Search string without protocol. See [[Special:LinkSearch]]. Leave empty to list all external links',
 225                         'namespace' => 'The page namespace(s) to enumerate.',
 226                         'limit' => 'How many pages to return.',
 227                         'expandurl' => 'Expand protocol-relative URLs with the canonical protocol',
 228                 );
 229
 230                 if ( $wgMiserMode ) {
 231                         $desc['namespace'] = array(
 232                                 $desc['namespace'],
 233                                 "NOTE: Due to \$wgMiserMode, using this may result in fewer than \"{$p}limit\" results",
 234                                 'returned before continuing; in extreme cases, zero results may be returned',
 235                         );
 236                 }
 237
 238                 return $desc;
 239         }
 240
 241         public function getResultProperties() {
 242                 return array(
 243                         'ids' => array(
 244                                 'pageid' => 'integer'
 245                         ),
 246                         'title' => array(
 247                                 'ns' => 'namespace',
 248                                 'title' => 'string'
 249                         ),
 250                         'url' => array(
 251                                 'url' => 'string'
 252                         )
 253                 );
 254         }
 255
 256         public function getDescription() {
 257                 return 'Enumerate pages that contain a given URL';
 258         }
 259
 260         public function getPossibleErrors() {
 261                 return array_merge( parent::getPossibleErrors(), array(
 262                         array( 'code' => 'bad_query', 'info' => 'Invalid query' ),
 263                 ) );
 264         }
 265
 266         public function getExamples() {
 267                 return array(
 268                         'api.php?action=query&list=exturlusage&euquery=www.mediawiki.org'
 269                 );
 270         }
 271
 272         public function getHelpUrls() {
 273                 return 'https://www.mediawiki.org/wiki/API:Exturlusage';
 274         }
 275 }