includes/search/SearchResult.php

   1 <?php
   2 /**
   3  * Search engine result
   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  * @ingroup Search
  22  */
  23
  24 use MediaWiki\MediaWikiServices;
  25
  26 /**
  27  * @todo FIXME: This class is horribly factored. It would probably be better to
  28  * have a useful base class to which you pass some standard information, then
  29  * let the fancy self-highlighters extend that.
  30  * @ingroup Search
  31  */
  32 class SearchResult {
  33
  34         /**
  35          * @var Revision
  36          */
  37         protected $mRevision = null;
  38
  39         /**
  40          * @var File
  41          */
  42         protected $mImage = null;
  43
  44         /**
  45          * @var Title
  46          */
  47         protected $mTitle;
  48
  49         /**
  50          * @var string
  51          */
  52         protected $mText;
  53
  54         /**
  55          * @var SearchEngine
  56          */
  57         protected $searchEngine;
  58
  59         /**
  60          * A function returning a set of extension data.
  61          * @var Closure|null
  62          */
  63         protected $extensionData;
  64
  65         /**
  66          * Return a new SearchResult and initializes it with a title.
  67          *
  68          * @param Title $title
  69          * @param SearchResultSet|null $parentSet
  70          * @return SearchResult
  71          */
  72         public static function newFromTitle( $title, SearchResultSet $parentSet = null ) {
  73                 $result = new static();
  74                 $result->initFromTitle( $title );
  75                 if ( $parentSet ) {
  76                         $parentSet->augmentResult( $result );
  77                 }
  78                 return $result;
  79         }
  80
  81         /**
  82          * Initialize from a Title and if possible initializes a corresponding
  83          * Revision and File.
  84          *
  85          * @param Title $title
  86          */
  87         protected function initFromTitle( $title ) {
  88                 $this->mTitle = $title;
  89                 if ( !is_null( $this->mTitle ) ) {
  90                         $id = false;
  91                         Hooks::run( 'SearchResultInitFromTitle', [ $title, &$id ] );
  92                         $this->mRevision = Revision::newFromTitle(
  93                                 $this->mTitle, $id, Revision::READ_NORMAL );
  94                         if ( $this->mTitle->getNamespace() === NS_FILE ) {
  95                                 $this->mImage = wfFindFile( $this->mTitle );
  96                         }
  97                 }
  98                 $this->searchEngine = MediaWikiServices::getInstance()->newSearchEngine();
  99         }
 100
 101         /**
 102          * Check if this is result points to an invalid title
 103          *
 104          * @return bool
 105          */
 106         function isBrokenTitle() {
 107                 return is_null( $this->mTitle );
 108         }
 109
 110         /**
 111          * Check if target page is missing, happens when index is out of date
 112          *
 113          * @return bool
 114          */
 115         function isMissingRevision() {
 116                 return !$this->mRevision && !$this->mImage;
 117         }
 118
 119         /**
 120          * @return Title
 121          */
 122         function getTitle() {
 123                 return $this->mTitle;
 124         }
 125
 126         /**
 127          * Get the file for this page, if one exists
 128          * @return File|null
 129          */
 130         function getFile() {
 131                 return $this->mImage;
 132         }
 133
 134         /**
 135          * Lazy initialization of article text from DB
 136          */
 137         protected function initText() {
 138                 if ( !isset( $this->mText ) ) {
 139                         if ( $this->mRevision != null ) {
 140                                 $this->mText = $this->searchEngine->getTextFromContent(
 141                                                 $this->mTitle, $this->mRevision->getContent() );
 142                         } else { // TODO: can we fetch raw wikitext for commons images?
 143                                 $this->mText = '';
 144                         }
 145                 }
 146         }
 147
 148         /**
 149          * @param array $terms Terms to highlight
 150          * @return string Highlighted text snippet, null (and not '') if not supported
 151          */
 152         function getTextSnippet( $terms ) {
 153                 global $wgAdvancedSearchHighlighting;
 154                 $this->initText();
 155
 156                 // TODO: make highliter take a content object. Make ContentHandler a factory for SearchHighliter.
 157                 list( $contextlines, $contextchars ) = $this->searchEngine->userHighlightPrefs();
 158
 159                 $h = new SearchHighlighter();
 160                 if ( count( $terms ) > 0 ) {
 161                         if ( $wgAdvancedSearchHighlighting ) {
 162                                 return $h->highlightText( $this->mText, $terms, $contextlines, $contextchars );
 163                         } else {
 164                                 return $h->highlightSimple( $this->mText, $terms, $contextlines, $contextchars );
 165                         }
 166                 } else {
 167                         return $h->highlightNone( $this->mText, $contextlines, $contextchars );
 168                 }
 169         }
 170
 171         /**
 172          * @return string Highlighted title, '' if not supported
 173          */
 174         function getTitleSnippet() {
 175                 return '';
 176         }
 177
 178         /**
 179          * @return string Highlighted redirect name (redirect to this page), '' if none or not supported
 180          */
 181         function getRedirectSnippet() {
 182                 return '';
 183         }
 184
 185         /**
 186          * @return Title|null Title object for the redirect to this page, null if none or not supported
 187          */
 188         function getRedirectTitle() {
 189                 return null;
 190         }
 191
 192         /**
 193          * @return string Highlighted relevant section name, null if none or not supported
 194          */
 195         function getSectionSnippet() {
 196                 return '';
 197         }
 198
 199         /**
 200          * @return Title|null Title object (pagename+fragment) for the section,
 201          *  null if none or not supported
 202          */
 203         function getSectionTitle() {
 204                 return null;
 205         }
 206
 207         /**
 208          * @return string Highlighted relevant category name or '' if none or not supported
 209          */
 210         public function getCategorySnippet() {
 211                 return '';
 212         }
 213
 214         /**
 215          * @return string Timestamp
 216          */
 217         function getTimestamp() {
 218                 if ( $this->mRevision ) {
 219                         return $this->mRevision->getTimestamp();
 220                 } elseif ( $this->mImage ) {
 221                         return $this->mImage->getTimestamp();
 222                 }
 223                 return '';
 224         }
 225
 226         /**
 227          * @return int Number of words
 228          */
 229         function getWordCount() {
 230                 $this->initText();
 231                 return str_word_count( $this->mText );
 232         }
 233
 234         /**
 235          * @return int Size in bytes
 236          */
 237         function getByteSize() {
 238                 $this->initText();
 239                 return strlen( $this->mText );
 240         }
 241
 242         /**
 243          * @return string Interwiki prefix of the title (return iw even if title is broken)
 244          */
 245         function getInterwikiPrefix() {
 246                 return '';
 247         }
 248
 249         /**
 250          * @return string Interwiki namespace of the title (since we likely can't resolve it locally)
 251          */
 252         function getInterwikiNamespaceText() {
 253                 return '';
 254         }
 255
 256         /**
 257          * Did this match file contents (eg: PDF/DJVU)?
 258          * @return bool
 259          */
 260         function isFileMatch() {
 261                 return false;
 262         }
 263
 264         /**
 265          * Get the extension data as:
 266          * augmentor name => data
 267          * @return array[]
 268          */
 269         public function getExtensionData() {
 270                 if ( $this->extensionData ) {
 271                         return call_user_func( $this->extensionData );
 272                 } else {
 273                         return [];
 274                 }
 275         }
 276
 277         /**
 278          * Set extension data for this result.
 279          * The data is:
 280          * augmentor name => data
 281          * @param Closure|array $extensionData Takes no arguments, returns
 282          *  either array of extension data or null.
 283          */
 284         public function setExtensionData( $extensionData ) {
 285                 if ( $extensionData instanceof Closure ) {
 286                         $this->extensionData = $extensionData;
 287                 } elseif ( is_array( $extensionData ) ) {
 288                         wfDeprecated( __METHOD__ . ' with array argument', 1.32 );
 289                         $this->extensionData = function () use ( $extensionData ) {
 290                                 return $extensionData;
 291                         };
 292                 } else {
 293                         $type = is_object( $extensionData )
 294                                 ? get_class( $extensionData )
 295                                 : gettype( $extensionData );
 296                         throw new \InvalidArgumentException(
 297                                 __METHOD__ . " must be called with Closure|array, but received $type" );
 298                 }
 299         }
 300 }