includes/api/ApiQueryExternalLinks.php

   1 <?php
   2 /**
   3  * Copyright © 2006 Yuri Astrakhan "<Firstname><Lastname>@gmail.com"
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  */
  22
  23 namespace MediaWiki\Api;
  24
  25 use MediaWiki\ExternalLinks\LinkFilter;
  26 use MediaWiki\Parser\Parser;
  27 use MediaWiki\Title\Title;
  28 use MediaWiki\Utils\UrlUtils;
  29 use Wikimedia\ParamValidator\ParamValidator;
  30 use Wikimedia\ParamValidator\TypeDef\IntegerDef;
  31 use Wikimedia\Rdbms\IExpression;
  32 use Wikimedia\Rdbms\LikeValue;
  33
  34 /**
  35  * A query module to list all external URLs found on a given set of pages.
  36  *
  37  * @ingroup API
  38  */
  39 class ApiQueryExternalLinks extends ApiQueryBase {
  40
  41         private UrlUtils $urlUtils;
  42
  43         public function __construct( ApiQuery $query, string $moduleName, UrlUtils $urlUtils ) {
  44                 parent::__construct( $query, $moduleName, 'el' );
  45
  46                 $this->urlUtils = $urlUtils;
  47         }
  48
  49         public function execute() {
  50                 $pages = $this->getPageSet()->getGoodPages();
  51                 if ( $pages === [] ) {
  52                         return;
  53                 }
  54
  55                 $params = $this->extractRequestParams();
  56                 $db = $this->getDB();
  57
  58                 $query = $params['query'];
  59                 $protocol = LinkFilter::getProtocolPrefix( $params['protocol'] );
  60
  61                 $fields = [ 'el_from' ];
  62                 $fields[] = 'el_to_domain_index';
  63                 $fields[] = 'el_to_path';
  64                 $continueField = 'el_to_domain_index';
  65                 $this->addFields( $fields );
  66
  67                 $this->addTables( 'externallinks' );
  68                 $this->addWhereFld( 'el_from', array_keys( $pages ) );
  69
  70                 if ( $query !== null && $query !== '' ) {
  71                         // Normalize query to match the normalization applied for the externallinks table
  72                         $query = Parser::normalizeLinkUrl( $query );
  73
  74                         $conds = LinkFilter::getQueryConditions( $query, [
  75                                 'protocol' => $protocol,
  76                                 'oneWildcard' => true,
  77                                 'db' => $db
  78                         ] );
  79                         if ( !$conds ) {
  80                                 $this->dieWithError( 'apierror-badquery' );
  81                         }
  82                         $this->addWhere( $conds );
  83                 } else {
  84                         if ( $protocol !== null ) {
  85                                 $this->addWhere(
  86                                         $db->expr( $continueField, IExpression::LIKE, new LikeValue( "$protocol", $db->anyString() ) )
  87                                 );
  88                         }
  89                 }
  90
  91                 $orderBy = [ 'el_id' ];
  92
  93                 $this->addOption( 'ORDER BY', $orderBy );
  94                 $this->addFields( $orderBy ); // Make sure
  95
  96                 $this->addOption( 'LIMIT', $params['limit'] + 1 );
  97
  98                 if ( $params['continue'] !== null ) {
  99                         $cont = $this->parseContinueParamOrDie( $params['continue'],
 100                                 array_fill( 0, count( $orderBy ), 'string' ) );
 101                         $conds = array_combine( $orderBy, array_map( 'rawurldecode', $cont ) );
 102                         $this->addWhere( $db->buildComparison( '>=', $conds ) );
 103                 }
 104
 105                 $res = $this->select( __METHOD__ );
 106
 107                 $count = 0;
 108                 foreach ( $res as $row ) {
 109                         if ( ++$count > $params['limit'] ) {
 110                                 // We've reached the one extra which shows that
 111                                 // there are additional pages to be had. Stop here...
 112                                 $this->setContinue( $orderBy, $row );
 113                                 break;
 114                         }
 115                         $entry = [];
 116                         $to = LinkFilter::reverseIndexes( $row->el_to_domain_index ) . $row->el_to_path;
 117                         // expand protocol-relative urls
 118                         if ( $params['expandurl'] ) {
 119                                 $to = (string)$this->urlUtils->expand( $to, PROTO_CANONICAL );
 120                         }
 121                         ApiResult::setContentValue( $entry, 'url', $to );
 122                         $fit = $this->addPageSubItem( $row->el_from, $entry );
 123                         if ( !$fit ) {
 124                                 $this->setContinue( $orderBy, $row );
 125                                 break;
 126                         }
 127                 }
 128         }
 129
 130         private function setContinue( $orderBy, $row ) {
 131                 $fields = [];
 132                 foreach ( $orderBy as $field ) {
 133                         $fields[] = strtr( $row->$field, [ '%' => '%25', '|' => '%7C' ] );
 134                 }
 135                 $this->setContinueEnumParameter( 'continue', implode( '|', $fields ) );
 136         }
 137
 138         public function getCacheMode( $params ) {
 139                 return 'public';
 140         }
 141
 142         public function getAllowedParams() {
 143                 return [
 144                         'limit' => [
 145                                 ParamValidator::PARAM_DEFAULT => 10,
 146                                 ParamValidator::PARAM_TYPE => 'limit',
 147                                 IntegerDef::PARAM_MIN => 1,
 148                                 IntegerDef::PARAM_MAX => ApiBase::LIMIT_BIG1,
 149                                 IntegerDef::PARAM_MAX2 => ApiBase::LIMIT_BIG2
 150                         ],
 151                         'continue' => [
 152                                 ApiBase::PARAM_HELP_MSG => 'api-help-param-continue',
 153                         ],
 154                         'protocol' => [
 155                                 ParamValidator::PARAM_TYPE => LinkFilter::prepareProtocols(),
 156                                 ParamValidator::PARAM_DEFAULT => '',
 157                         ],
 158                         'query' => null,
 159                         'expandurl' => [
 160                                 ParamValidator::PARAM_TYPE => 'boolean',
 161                                 ParamValidator::PARAM_DEFAULT => false,
 162                                 ParamValidator::PARAM_DEPRECATED => true,
 163                         ],
 164                 ];
 165         }
 166
 167         protected function getExamplesMessages() {
 168                 $title = Title::newMainPage()->getPrefixedText();
 169                 $mp = rawurlencode( $title );
 170
 171                 return [
 172                         "action=query&prop=extlinks&titles={$mp}"
 173                                 => 'apihelp-query+extlinks-example-simple',
 174                 ];
 175         }
 176
 177         public function getHelpUrls() {
 178                 return 'https://www.mediawiki.org/wiki/Special:MyLanguage/API:Extlinks';
 179         }
 180 }
 181
 182 /** @deprecated class alias since 1.43 */
 183 class_alias( ApiQueryExternalLinks::class, 'ApiQueryExternalLinks' );