includes/deferred/SearchUpdate.php

   1 <?php
   2 /**
   3  * Search index updater
   4  *
   5  * See deferred.txt
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License along
  18  * with this program; if not, write to the Free Software Foundation, Inc.,
  19  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  20  * http://www.gnu.org/copyleft/gpl.html
  21  *
  22  * @file
  23  * @ingroup Search
  24  */
  25
  26 /**
  27  * Database independant search index updater
  28  *
  29  * @ingroup Search
  30  */
  31 class SearchUpdate implements DeferrableUpdate {
  32         /**
  33          * Page id being updated
  34          * @var int
  35          */
  36         private $id = 0;
  37
  38         /**
  39          * Title we're updating
  40          * @var Title
  41          */
  42         private $title;
  43
  44         /**
  45          * Content of the page (not text)
  46          * @var Content|false
  47          */
  48         private $content;
  49
  50         /**
  51          * Constructor
  52          *
  53          * @param int $id Page id to update
  54          * @param Title|string $title Title of page to update
  55          * @param Content|string|false $c Content of the page to update.
  56          *  If a Content object, text will be gotten from it. String is for back-compat.
  57          *  Passing false tells the backend to just update the title, not the content
  58          */
  59         public function __construct( $id, $title, $c = false ) {
  60                 if ( is_string( $title ) ) {
  61                         $nt = Title::newFromText( $title );
  62                 } else {
  63                         $nt = $title;
  64                 }
  65
  66                 if ( $nt ) {
  67                         $this->id = $id;
  68                         // is_string() check is back-compat for ApprovedRevs
  69                         if ( is_string( $c ) ) {
  70                                 $this->content = new TextContent( $c );
  71                         } else {
  72                                 $this->content = $c ?: false;
  73                         }
  74                         $this->title = $nt;
  75                 } else {
  76                         wfDebug( "SearchUpdate object created with invalid title '$title'\n" );
  77                 }
  78         }
  79
  80         /**
  81          * Perform actual update for the entry
  82          */
  83         public function doUpdate() {
  84                 global $wgDisableSearchUpdate;
  85
  86                 if ( $wgDisableSearchUpdate || !$this->id ) {
  87                         return;
  88                 }
  89
  90                 wfProfileIn( __METHOD__ );
  91
  92                 $page = WikiPage::newFromId( $this->id, WikiPage::READ_LATEST );
  93                 $indexTitle = Title::indexTitle( $this->title->getNamespace(), $this->title->getText() );
  94
  95                 foreach ( SearchEngine::getSearchTypes() as $type ) {
  96                         $search = SearchEngine::create( $type );
  97                         if ( !$search->supports( 'search-update' ) ) {
  98                                 continue;
  99                         }
 100
 101                         $normalTitle = $search->normalizeText( $indexTitle );
 102
 103                         if ( $page === null ) {
 104                                 $search->delete( $this->id, $normalTitle );
 105                                 continue;
 106                         } elseif ( $this->content === false ) {
 107                                 $search->updateTitle( $this->id, $normalTitle );
 108                                 continue;
 109                         }
 110
 111                         $text = $search->getTextFromContent( $this->title, $this->content );
 112                         if ( !$search->textAlreadyUpdatedForIndex() ) {
 113                                 $text = self::updateText( $text );
 114                         }
 115
 116                         # Perform the actual update
 117                         $search->update( $this->id, $normalTitle, $search->normalizeText( $text ) );
 118                 }
 119
 120                 wfProfileOut( __METHOD__ );
 121         }
 122
 123         /**
 124          * Clean text for indexing. Only really suitable for indexing in databases.
 125          * If you're using a real search engine, you'll probably want to override
 126          * this behavior and do something nicer with the original wikitext.
 127          */
 128         public static function updateText( $text ) {
 129                 global $wgContLang;
 130
 131                 # Language-specific strip/conversion
 132                 $text = $wgContLang->normalizeForSearch( $text );
 133                 $lc = SearchEngine::legalSearchChars() . '&#;';
 134
 135                 wfProfileIn( __METHOD__ . '-regexps' );
 136                 $text = preg_replace( "/<\\/?\\s*[A-Za-z][^>]*?>/",
 137                         ' ', $wgContLang->lc( " " . $text . " " ) ); # Strip HTML markup
 138                 $text = preg_replace( "/(^|\\n)==\\s*([^\\n]+)\\s*==(\\s)/sD",
 139                         "\\1\\2 \\2 \\2\\3", $text ); # Emphasize headings
 140
 141                 # Strip external URLs
 142                 $uc = "A-Za-z0-9_\\/:.,~%\\-+&;#?!=()@\\x80-\\xFF";
 143                 $protos = "http|https|ftp|mailto|news|gopher";
 144                 $pat = "/(^|[^\\[])({$protos}):[{$uc}]+([^{$uc}]|$)/";
 145                 $text = preg_replace( $pat, "\\1 \\3", $text );
 146
 147                 $p1 = "/([^\\[])\\[({$protos}):[{$uc}]+]/";
 148                 $p2 = "/([^\\[])\\[({$protos}):[{$uc}]+\\s+([^\\]]+)]/";
 149                 $text = preg_replace( $p1, "\\1 ", $text );
 150                 $text = preg_replace( $p2, "\\1 \\3 ", $text );
 151
 152                 # Internal image links
 153                 $pat2 = "/\\[\\[image:([{$uc}]+)\\.(gif|png|jpg|jpeg)([^{$uc}])/i";
 154                 $text = preg_replace( $pat2, " \\1 \\3", $text );
 155
 156                 $text = preg_replace( "/([^{$lc}])([{$lc}]+)]]([a-z]+)/",
 157                         "\\1\\2 \\2\\3", $text ); # Handle [[game]]s
 158
 159                 # Strip all remaining non-search characters
 160                 $text = preg_replace( "/[^{$lc}]+/", " ", $text );
 161
 162                 # Handle 's, s'
 163                 #
 164                 #   $text = preg_replace( "/([{$lc}]+)'s /", "\\1 \\1's ", $text );
 165                 #   $text = preg_replace( "/([{$lc}]+)s' /", "\\1s ", $text );
 166                 #
 167                 # These tail-anchored regexps are insanely slow. The worst case comes
 168                 # when Japanese or Chinese text (ie, no word spacing) is written on
 169                 # a wiki configured for Western UTF-8 mode. The Unicode characters are
 170                 # expanded to hex codes and the "words" are very long paragraph-length
 171                 # monstrosities. On a large page the above regexps may take over 20
 172                 # seconds *each* on a 1GHz-level processor.
 173                 #
 174                 # Following are reversed versions which are consistently fast
 175                 # (about 3 milliseconds on 1GHz-level processor).
 176                 #
 177                 $text = strrev( preg_replace( "/ s'([{$lc}]+)/", " s'\\1 \\1", strrev( $text ) ) );
 178                 $text = strrev( preg_replace( "/ 's([{$lc}]+)/", " s\\1", strrev( $text ) ) );
 179
 180                 # Strip wiki '' and '''
 181                 $text = preg_replace( "/''[']*/", " ", $text );
 182                 wfProfileOut( __METHOD__ . '-regexps' );
 183
 184                 return $text;
 185         }
 186 }