includes/search/SearchSqlite.php

   1 <?php
   2 # SQLite search backend, based upon SearchMysql
   3 #
   4 # This program is free software; you can redistribute it and/or modify
   5 # it under the terms of the GNU General Public License as published by
   6 # the Free Software Foundation; either version 2 of the License, or
   7 # (at your option) any later version.
   8 #
   9 # This program is distributed in the hope that it will be useful,
  10 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  12 # GNU General Public License for more details.
  13 #
  14 # You should have received a copy of the GNU General Public License along
  15 # with this program; if not, write to the Free Software Foundation, Inc.,
  16 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  17 # http://www.gnu.org/copyleft/gpl.html
  18
  19 /**
  20  * @file
  21  * @ingroup Search
  22  */
  23
  24 /**
  25  * Search engine hook for SQLite
  26  * @ingroup Search
  27  */
  28 class SearchSqlite extends SearchEngine {
  29         /**
  30          * Creates an instance of this class
  31          * @param $db DatabaseSqlite: database object
  32          */
  33         function __construct( $db ) {
  34                 $this->db = $db;
  35         }
  36
  37         /**
  38          * Whether fulltext search is supported by current schema
  39          * @return Boolean
  40          */
  41         function fulltextSearchSupported() {
  42                 return $this->db->checkForEnabledSearch();
  43         }
  44
  45         /**
  46          * Parse the user's query and transform it into an SQL fragment which will
  47          * become part of a WHERE clause
  48          */
  49         function parseQuery( $filteredText, $fulltext ) {
  50                 global $wgContLang;
  51                 $lc = SearchEngine::legalSearchChars(); // Minus format chars
  52                 $searchon = '';
  53                 $this->searchTerms = array();
  54
  55                 $m = array();
  56                 if( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
  57                           $filteredText, $m, PREG_SET_ORDER ) ) {
  58                         foreach( $m as $bits ) {
  59                                 @list( /* all */, $modifier, $term, $nonQuoted, $wildcard ) = $bits;
  60
  61                                 if( $nonQuoted != '' ) {
  62                                         $term = $nonQuoted;
  63                                         $quote = '';
  64                                 } else {
  65                                         $term = str_replace( '"', '', $term );
  66                                         $quote = '"';
  67                                 }
  68
  69                                 if( $searchon !== '' ) $searchon .= ' ';
  70
  71                                 // Some languages such as Serbian store the input form in the search index,
  72                                 // so we may need to search for matches in multiple writing system variants.
  73                                 $convertedVariants = $wgContLang->autoConvertToAllVariants( $term );
  74                                 if( is_array( $convertedVariants ) ) {
  75                                         $variants = array_unique( array_values( $convertedVariants ) );
  76                                 } else {
  77                                         $variants = array( $term );
  78                                 }
  79
  80                                 // The low-level search index does some processing on input to work
  81                                 // around problems with minimum lengths and encoding in MySQL's
  82                                 // fulltext engine.
  83                                 // For Chinese this also inserts spaces between adjacent Han characters.
  84                                 $strippedVariants = array_map(
  85                                         array( $wgContLang, 'normalizeForSearch' ),
  86                                         $variants );
  87
  88                                 // Some languages such as Chinese force all variants to a canonical
  89                                 // form when stripping to the low-level search index, so to be sure
  90                                 // let's check our variants list for unique items after stripping.
  91                                 $strippedVariants = array_unique( $strippedVariants );
  92
  93                                 $searchon .= $modifier;
  94                                 if( count( $strippedVariants) > 1 )
  95                                         $searchon .= '(';
  96                                 foreach( $strippedVariants as $stripped ) {
  97                                         if( $nonQuoted && strpos( $stripped, ' ' ) !== false ) {
  98                                                 // Hack for Chinese: we need to toss in quotes for
  99                                                 // multiple-character phrases since normalizeForSearch()
 100                                                 // added spaces between them to make word breaks.
 101                                                 $stripped = '"' . trim( $stripped ) . '"';
 102                                         }
 103                                         $searchon .= "$quote$stripped$quote$wildcard ";
 104                                 }
 105                                 if( count( $strippedVariants) > 1 )
 106                                         $searchon .= ')';
 107
 108                                 // Match individual terms or quoted phrase in result highlighting...
 109                                 // Note that variants will be introduced in a later stage for highlighting!
 110                                 $regexp = $this->regexTerm( $term, $wildcard );
 111                                 $this->searchTerms[] = $regexp;
 112                         }
 113
 114                 } else {
 115                         wfDebug( __METHOD__ . ": Can't understand search query '{$filteredText}'\n" );
 116                 }
 117
 118                 $searchon = $this->db->strencode( $searchon );
 119                 $field = $this->getIndexField( $fulltext );
 120                 return " $field MATCH '$searchon' ";
 121         }
 122
 123         function regexTerm( $string, $wildcard ) {
 124                 global $wgContLang;
 125
 126                 $regex = preg_quote( $string, '/' );
 127                 if( $wgContLang->hasWordBreaks() ) {
 128                         if( $wildcard ) {
 129                                 // Don't cut off the final bit!
 130                                 $regex = "\b$regex";
 131                         } else {
 132                                 $regex = "\b$regex\b";
 133                         }
 134                 } else {
 135                         // For Chinese, words may legitimately abut other words in the text literal.
 136                         // Don't add \b boundary checks... note this could cause false positives
 137                         // for latin chars.
 138                 }
 139                 return $regex;
 140         }
 141
 142         public static function legalSearchChars() {
 143                 return "\"*" . parent::legalSearchChars();
 144         }
 145
 146         /**
 147          * Perform a full text search query and return a result set.
 148          *
 149          * @param $term String: raw search term
 150          * @return SqliteSearchResultSet
 151          */
 152         function searchText( $term ) {
 153                 return $this->searchInternal( $term, true );
 154         }
 155
 156         /**
 157          * Perform a title-only search query and return a result set.
 158          *
 159          * @param $term String: raw search term
 160          * @return SqliteSearchResultSet
 161          */
 162         function searchTitle( $term ) {
 163                 return $this->searchInternal( $term, false );
 164         }
 165
 166         protected function searchInternal( $term, $fulltext ) {
 167                 global $wgCountTotalSearchHits, $wgContLang;
 168
 169                 if ( !$this->fulltextSearchSupported() ) {
 170                         return null;
 171                 }
 172
 173                 $filteredTerm = $this->filter( $wgContLang->lc( $term ) );
 174                 $resultSet = $this->db->query( $this->getQuery( $filteredTerm, $fulltext ) );
 175
 176                 $total = null;
 177                 if( $wgCountTotalSearchHits ) {
 178                         $totalResult = $this->db->query( $this->getCountQuery( $filteredTerm, $fulltext ) );
 179                         $row = $totalResult->fetchObject();
 180                         if( $row ) {
 181                                 $total = intval( $row->c );
 182                         }
 183                         $totalResult->free();
 184                 }
 185
 186                 return new SqliteSearchResultSet( $resultSet, $this->searchTerms, $total );
 187         }
 188
 189
 190         /**
 191          * Return a partial WHERE clause to exclude redirects, if so set
 192          * @return String
 193          */
 194         function queryRedirect() {
 195                 if( $this->showRedirects ) {
 196                         return '';
 197                 } else {
 198                         return 'AND page_is_redirect=0';
 199                 }
 200         }
 201
 202         /**
 203          * Return a partial WHERE clause to limit the search to the given namespaces
 204          * @return String
 205          */
 206         function queryNamespaces() {
 207                 if( is_null($this->namespaces) )
 208                         return '';  # search all
 209                 if ( !count( $this->namespaces ) ) {
 210                         $namespaces = '0';
 211                 } else {
 212                         $namespaces = $this->db->makeList( $this->namespaces );
 213                 }
 214                 return 'AND page_namespace IN (' . $namespaces . ')';
 215         }
 216
 217         /**
 218          * Returns a query with limit for number of results set.
 219          * @param $sql String:
 220          * @return String
 221          */
 222         function limitResult( $sql ) {
 223                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 224         }
 225
 226         /**
 227          * Construct the full SQL query to do the search.
 228          * The guts shoulds be constructed in queryMain()
 229          * @param $filteredTerm String
 230          * @param $fulltext Boolean
 231          */
 232         function getQuery( $filteredTerm, $fulltext ) {
 233                 return $this->limitResult(
 234                         $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 235                         $this->queryRedirect() . ' ' .
 236                         $this->queryNamespaces()
 237                 );
 238         }
 239
 240         /**
 241          * Picks which field to index on, depending on what type of query.
 242          * @param $fulltext Boolean
 243          * @return String
 244          */
 245         function getIndexField( $fulltext ) {
 246                 return $fulltext ? 'si_text' : 'si_title';
 247         }
 248
 249         /**
 250          * Get the base part of the search query.
 251          *
 252          * @param $filteredTerm String
 253          * @param $fulltext Boolean
 254          * @return String
 255          */
 256         function queryMain( $filteredTerm, $fulltext ) {
 257                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 258                 $page        = $this->db->tableName( 'page' );
 259                 $searchindex = $this->db->tableName( 'searchindex' );
 260                 return "SELECT $searchindex.rowid, page_namespace, page_title " .
 261                         "FROM $page,$searchindex " .
 262                         "WHERE page_id=$searchindex.rowid AND $match";
 263         }
 264
 265         function getCountQuery( $filteredTerm, $fulltext ) {
 266                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 267                 $page        = $this->db->tableName( 'page' );
 268                 $searchindex = $this->db->tableName( 'searchindex' );
 269                 return "SELECT COUNT(*) AS c " .
 270                         "FROM $page,$searchindex " .
 271                         "WHERE page_id=$searchindex.rowid AND $match" .
 272                         $this->queryRedirect() . ' ' .
 273                         $this->queryNamespaces();
 274         }
 275
 276         /**
 277          * Create or update the search index record for the given page.
 278          * Title and text should be pre-processed.
 279          *
 280          * @param $id Integer
 281          * @param $title String
 282          * @param $text String
 283          */
 284         function update( $id, $title, $text ) {
 285                 if ( !$this->fulltextSearchSupported() ) {
 286                         return;
 287                 }
 288                 // @todo: find a method to do it in a single request,
 289                 // couldn't do it so far due to typelessness of FTS3 tables.
 290                 $dbw = wfGetDB( DB_MASTER );
 291
 292                 $dbw->delete( 'searchindex', array( 'rowid' => $id ), __METHOD__ );
 293
 294                 $dbw->insert( 'searchindex',
 295                         array(
 296                                 'rowid' => $id,
 297                                 'si_title' => $title,
 298                                 'si_text' => $text
 299                         ), __METHOD__ );
 300         }
 301
 302         /**
 303          * Update a search index record's title only.
 304          * Title should be pre-processed.
 305          *
 306          * @param $id Integer
 307          * @param $title String
 308          */
 309     function updateTitle( $id, $title ) {
 310                 if ( !$this->fulltextSearchSupported() ) {
 311                         return;
 312                 }
 313                 $dbw = wfGetDB( DB_MASTER );
 314
 315                 $dbw->update( 'searchindex',
 316                         array( 'si_title' => $title ),
 317                         array( 'rowid'  => $id ),
 318                         __METHOD__ );
 319         }
 320 }
 321
 322 /**
 323  * @ingroup Search
 324  */
 325 class SqliteSearchResultSet extends SqlSearchResultSet {
 326         function SqliteSearchResultSet( $resultSet, $terms, $totalHits=null ) {
 327                 parent::__construct( $resultSet, $terms );
 328                 $this->mTotalHits = $totalHits;
 329         }
 330
 331         function getTotalHits() {
 332                 return $this->mTotalHits;
 333         }
 334 }