includes/search/SearchSqlite.php

   1 <?php
   2 /**
   3  * SQLite search backend, based upon SearchMysql
   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  * Search engine hook for SQLite
  28  * @ingroup Search
  29  */
  30 class SearchSqlite extends SearchDatabase {
  31         /**
  32          * Whether fulltext search is supported by current schema
  33          * @return bool
  34          */
  35         function fulltextSearchSupported() {
  36                 return $this->db->checkForEnabledSearch();
  37         }
  38
  39         /**
  40          * Parse the user's query and transform it into an SQL fragment which will
  41          * become part of a WHERE clause
  42          *
  43          * @param string $filteredText
  44          * @param bool $fulltext
  45          * @return string
  46          */
  47         private function parseQuery( $filteredText, $fulltext ) {
  48                 $lc = $this->legalSearchChars( self::CHARS_NO_SYNTAX ); // Minus syntax chars (" and *)
  49                 $searchon = '';
  50                 $this->searchTerms = [];
  51
  52                 $m = [];
  53                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
  54                                 $filteredText, $m, PREG_SET_ORDER ) ) {
  55                         foreach ( $m as $bits ) {
  56                                 Wikimedia\suppressWarnings();
  57                                 list( /* all */, $modifier, $term, $nonQuoted, $wildcard ) = $bits;
  58                                 Wikimedia\restoreWarnings();
  59
  60                                 if ( $nonQuoted != '' ) {
  61                                         $term = $nonQuoted;
  62                                         $quote = '';
  63                                 } else {
  64                                         $term = str_replace( '"', '', $term );
  65                                         $quote = '"';
  66                                 }
  67
  68                                 if ( $searchon !== '' ) {
  69                                         $searchon .= ' ';
  70                                 }
  71
  72                                 // Some languages such as Serbian store the input form in the search index,
  73                                 // so we may need to search for matches in multiple writing system variants.
  74                                 $convertedVariants = MediaWikiServices::getInstance()->getContentLanguage()->
  75                                         autoConvertToAllVariants( $term );
  76                                 if ( is_array( $convertedVariants ) ) {
  77                                         $variants = array_unique( array_values( $convertedVariants ) );
  78                                 } else {
  79                                         $variants = [ $term ];
  80                                 }
  81
  82                                 // The low-level search index does some processing on input to work
  83                                 // around problems with minimum lengths and encoding in MySQL's
  84                                 // fulltext engine.
  85                                 // For Chinese this also inserts spaces between adjacent Han characters.
  86                                 $strippedVariants = array_map(
  87                                         [ MediaWikiServices::getInstance()->getContentLanguage(),
  88                                                 'normalizeForSearch' ],
  89                                         $variants );
  90
  91                                 // Some languages such as Chinese force all variants to a canonical
  92                                 // form when stripping to the low-level search index, so to be sure
  93                                 // let's check our variants list for unique items after stripping.
  94                                 $strippedVariants = array_unique( $strippedVariants );
  95
  96                                 $searchon .= $modifier;
  97                                 if ( count( $strippedVariants ) > 1 ) {
  98                                         $searchon .= '(';
  99                                 }
 100                                 foreach ( $strippedVariants as $stripped ) {
 101                                         if ( $nonQuoted && strpos( $stripped, ' ' ) !== false ) {
 102                                                 // Hack for Chinese: we need to toss in quotes for
 103                                                 // multiple-character phrases since normalizeForSearch()
 104                                                 // added spaces between them to make word breaks.
 105                                                 $stripped = '"' . trim( $stripped ) . '"';
 106                                         }
 107                                         $searchon .= "$quote$stripped$quote$wildcard ";
 108                                 }
 109                                 if ( count( $strippedVariants ) > 1 ) {
 110                                         $searchon .= ')';
 111                                 }
 112
 113                                 // Match individual terms or quoted phrase in result highlighting...
 114                                 // Note that variants will be introduced in a later stage for highlighting!
 115                                 $regexp = $this->regexTerm( $term, $wildcard );
 116                                 $this->searchTerms[] = $regexp;
 117                         }
 118
 119                 } else {
 120                         wfDebug( __METHOD__ . ": Can't understand search query '{$filteredText}'\n" );
 121                 }
 122
 123                 $searchon = $this->db->addQuotes( $searchon );
 124                 $field = $this->getIndexField( $fulltext );
 125                 return " $field MATCH $searchon ";
 126         }
 127
 128         private function regexTerm( $string, $wildcard ) {
 129                 $regex = preg_quote( $string, '/' );
 130                 if ( MediaWikiServices::getInstance()->getContentLanguage()->hasWordBreaks() ) {
 131                         if ( $wildcard ) {
 132                                 // Don't cut off the final bit!
 133                                 $regex = "\b$regex";
 134                         } else {
 135                                 $regex = "\b$regex\b";
 136                         }
 137                 } else {
 138                         // For Chinese, words may legitimately abut other words in the text literal.
 139                         // Don't add \b boundary checks... note this could cause false positives
 140                         // for Latin chars.
 141                 }
 142                 return $regex;
 143         }
 144
 145         public static function legalSearchChars( $type = self::CHARS_ALL ) {
 146                 $searchChars = parent::legalSearchChars( $type );
 147                 if ( $type === self::CHARS_ALL ) {
 148                         // " for phrase, * for wildcard
 149                         $searchChars = "\"*" . $searchChars;
 150                 }
 151                 return $searchChars;
 152         }
 153
 154         /**
 155          * Perform a full text search query and return a result set.
 156          *
 157          * @param string $term Raw search term
 158          * @return SqlSearchResultSet
 159          */
 160         protected function doSearchTextInDB( $term ) {
 161                 return $this->searchInternal( $term, true );
 162         }
 163
 164         /**
 165          * Perform a title-only search query and return a result set.
 166          *
 167          * @param string $term Raw search term
 168          * @return SqlSearchResultSet
 169          */
 170         protected function doSearchTitleInDB( $term ) {
 171                 return $this->searchInternal( $term, false );
 172         }
 173
 174         protected function searchInternal( $term, $fulltext ) {
 175                 if ( !$this->fulltextSearchSupported() ) {
 176                         return null;
 177                 }
 178
 179                 $filteredTerm =
 180                         $this->filter( MediaWikiServices::getInstance()->getContentLanguage()->lc( $term ) );
 181                 $resultSet = $this->db->query( $this->getQuery( $filteredTerm, $fulltext ) );
 182
 183                 $total = null;
 184                 $totalResult = $this->db->query( $this->getCountQuery( $filteredTerm, $fulltext ) );
 185                 $row = $totalResult->fetchObject();
 186                 if ( $row ) {
 187                         $total = intval( $row->c );
 188                 }
 189                 $totalResult->free();
 190
 191                 return new SqlSearchResultSet( $resultSet, $this->searchTerms, $total );
 192         }
 193
 194         /**
 195          * Return a partial WHERE clause to limit the search to the given namespaces
 196          * @return string
 197          */
 198         private function queryNamespaces() {
 199                 if ( is_null( $this->namespaces ) ) {
 200                         return '';  # search all
 201                 }
 202                 if ( $this->namespaces === [] ) {
 203                         $namespaces = '0';
 204                 } else {
 205                         $namespaces = $this->db->makeList( $this->namespaces );
 206                 }
 207                 return 'AND page_namespace IN (' . $namespaces . ')';
 208         }
 209
 210         /**
 211          * Returns a query with limit for number of results set.
 212          * @param string $sql
 213          * @return string
 214          */
 215         private function limitResult( $sql ) {
 216                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 217         }
 218
 219         /**
 220          * Construct the full SQL query to do the search.
 221          * The guts shoulds be constructed in queryMain()
 222          * @param string $filteredTerm
 223          * @param bool $fulltext
 224          * @return string
 225          */
 226         private function getQuery( $filteredTerm, $fulltext ) {
 227                 return $this->limitResult(
 228                         $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 229                         $this->queryNamespaces()
 230                 );
 231         }
 232
 233         /**
 234          * Picks which field to index on, depending on what type of query.
 235          * @param bool $fulltext
 236          * @return string
 237          */
 238         private function getIndexField( $fulltext ) {
 239                 return $fulltext ? 'si_text' : 'si_title';
 240         }
 241
 242         /**
 243          * Get the base part of the search query.
 244          *
 245          * @param string $filteredTerm
 246          * @param bool $fulltext
 247          * @return string
 248          */
 249         private function queryMain( $filteredTerm, $fulltext ) {
 250                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 251                 $page = $this->db->tableName( 'page' );
 252                 $searchindex = $this->db->tableName( 'searchindex' );
 253                 return "SELECT $searchindex.rowid, page_namespace, page_title " .
 254                         "FROM $page,$searchindex " .
 255                         "WHERE page_id=$searchindex.rowid AND $match";
 256         }
 257
 258         private function getCountQuery( $filteredTerm, $fulltext ) {
 259                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 260                 $page = $this->db->tableName( 'page' );
 261                 $searchindex = $this->db->tableName( 'searchindex' );
 262                 return "SELECT COUNT(*) AS c " .
 263                         "FROM $page,$searchindex " .
 264                         "WHERE page_id=$searchindex.rowid AND $match " .
 265                         $this->queryNamespaces();
 266         }
 267
 268         /**
 269          * Create or update the search index record for the given page.
 270          * Title and text should be pre-processed.
 271          *
 272          * @param int $id
 273          * @param string $title
 274          * @param string $text
 275          */
 276         function update( $id, $title, $text ) {
 277                 if ( !$this->fulltextSearchSupported() ) {
 278                         return;
 279                 }
 280                 // @todo find a method to do it in a single request,
 281                 // couldn't do it so far due to typelessness of FTS3 tables.
 282                 $dbw = wfGetDB( DB_MASTER );
 283
 284                 $dbw->delete( 'searchindex', [ 'rowid' => $id ], __METHOD__ );
 285
 286                 $dbw->insert( 'searchindex',
 287                         [
 288                                 'rowid' => $id,
 289                                 'si_title' => $title,
 290                                 'si_text' => $text
 291                         ], __METHOD__ );
 292         }
 293
 294         /**
 295          * Update a search index record's title only.
 296          * Title should be pre-processed.
 297          *
 298          * @param int $id
 299          * @param string $title
 300          */
 301         function updateTitle( $id, $title ) {
 302                 if ( !$this->fulltextSearchSupported() ) {
 303                         return;
 304                 }
 305                 $dbw = wfGetDB( DB_MASTER );
 306
 307                 $dbw->update( 'searchindex',
 308                         [ 'si_title' => $title ],
 309                         [ 'rowid' => $id ],
 310                         __METHOD__ );
 311         }
 312 }