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 /**
  25  * Search engine hook for SQLite
  26  * @ingroup Search
  27  */
  28 class SearchSqlite extends SearchEngine {
  29
  30         /**
  31          * @var DatabaseSqlite
  32          */
  33         protected $db;
  34
  35         /**
  36          * Creates an instance of this class
  37          * @param $db DatabaseSqlite: database object
  38          */
  39         function __construct( $db ) {
  40                 parent::__construct( $db );
  41         }
  42
  43         /**
  44          * Whether fulltext search is supported by current schema
  45          * @return Boolean
  46          */
  47         function fulltextSearchSupported() {
  48                 return $this->db->checkForEnabledSearch();
  49         }
  50
  51         /**
  52          * Parse the user's query and transform it into an SQL fragment which will
  53          * become part of a WHERE clause
  54          *
  55          * @return string
  56          */
  57         function parseQuery( $filteredText, $fulltext ) {
  58                 global $wgContLang;
  59                 $lc = SearchEngine::legalSearchChars(); // Minus format chars
  60                 $searchon = '';
  61                 $this->searchTerms = array();
  62
  63                 $m = array();
  64                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
  65                                 $filteredText, $m, PREG_SET_ORDER ) ) {
  66                         foreach ( $m as $bits ) {
  67                                 @list( /* all */, $modifier, $term, $nonQuoted, $wildcard ) = $bits;
  68
  69                                 if ( $nonQuoted != '' ) {
  70                                         $term = $nonQuoted;
  71                                         $quote = '';
  72                                 } else {
  73                                         $term = str_replace( '"', '', $term );
  74                                         $quote = '"';
  75                                 }
  76
  77                                 if ( $searchon !== '' ) {
  78                                         $searchon .= ' ';
  79                                 }
  80
  81                                 // Some languages such as Serbian store the input form in the search index,
  82                                 // so we may need to search for matches in multiple writing system variants.
  83                                 $convertedVariants = $wgContLang->autoConvertToAllVariants( $term );
  84                                 if ( is_array( $convertedVariants ) ) {
  85                                         $variants = array_unique( array_values( $convertedVariants ) );
  86                                 } else {
  87                                         $variants = array( $term );
  88                                 }
  89
  90                                 // The low-level search index does some processing on input to work
  91                                 // around problems with minimum lengths and encoding in MySQL's
  92                                 // fulltext engine.
  93                                 // For Chinese this also inserts spaces between adjacent Han characters.
  94                                 $strippedVariants = array_map(
  95                                         array( $wgContLang, 'normalizeForSearch' ),
  96                                         $variants );
  97
  98                                 // Some languages such as Chinese force all variants to a canonical
  99                                 // form when stripping to the low-level search index, so to be sure
 100                                 // let's check our variants list for unique items after stripping.
 101                                 $strippedVariants = array_unique( $strippedVariants );
 102
 103                                 $searchon .= $modifier;
 104                                 if ( count( $strippedVariants ) > 1 ) {
 105                                         $searchon .= '(';
 106                                 }
 107                                 foreach ( $strippedVariants as $stripped ) {
 108                                         if ( $nonQuoted && strpos( $stripped, ' ' ) !== false ) {
 109                                                 // Hack for Chinese: we need to toss in quotes for
 110                                                 // multiple-character phrases since normalizeForSearch()
 111                                                 // added spaces between them to make word breaks.
 112                                                 $stripped = '"' . trim( $stripped ) . '"';
 113                                         }
 114                                         $searchon .= "$quote$stripped$quote$wildcard ";
 115                                 }
 116                                 if ( count( $strippedVariants ) > 1 ) {
 117                                         $searchon .= ')';
 118                                 }
 119
 120                                 // Match individual terms or quoted phrase in result highlighting...
 121                                 // Note that variants will be introduced in a later stage for highlighting!
 122                                 $regexp = $this->regexTerm( $term, $wildcard );
 123                                 $this->searchTerms[] = $regexp;
 124                         }
 125
 126                 } else {
 127                         wfDebug( __METHOD__ . ": Can't understand search query '{$filteredText}'\n" );
 128                 }
 129
 130                 $searchon = $this->db->strencode( $searchon );
 131                 $field = $this->getIndexField( $fulltext );
 132                 return " $field MATCH '$searchon' ";
 133         }
 134
 135         function regexTerm( $string, $wildcard ) {
 136                 global $wgContLang;
 137
 138                 $regex = preg_quote( $string, '/' );
 139                 if ( $wgContLang->hasWordBreaks() ) {
 140                         if ( $wildcard ) {
 141                                 // Don't cut off the final bit!
 142                                 $regex = "\b$regex";
 143                         } else {
 144                                 $regex = "\b$regex\b";
 145                         }
 146                 } else {
 147                         // For Chinese, words may legitimately abut other words in the text literal.
 148                         // Don't add \b boundary checks... note this could cause false positives
 149                         // for latin chars.
 150                 }
 151                 return $regex;
 152         }
 153
 154         public static function legalSearchChars() {
 155                 return "\"*" . parent::legalSearchChars();
 156         }
 157
 158         /**
 159          * Perform a full text search query and return a result set.
 160          *
 161          * @param string $term raw search term
 162          * @return SqliteSearchResultSet
 163          */
 164         function searchText( $term ) {
 165                 return $this->searchInternal( $term, true );
 166         }
 167
 168         /**
 169          * Perform a title-only search query and return a result set.
 170          *
 171          * @param string $term raw search term
 172          * @return SqliteSearchResultSet
 173          */
 174         function searchTitle( $term ) {
 175                 return $this->searchInternal( $term, false );
 176         }
 177
 178         protected function searchInternal( $term, $fulltext ) {
 179                 global $wgCountTotalSearchHits, $wgContLang;
 180
 181                 if ( !$this->fulltextSearchSupported() ) {
 182                         return null;
 183                 }
 184
 185                 $filteredTerm = $this->filter( $wgContLang->lc( $term ) );
 186                 $resultSet = $this->db->query( $this->getQuery( $filteredTerm, $fulltext ) );
 187
 188                 $total = null;
 189                 if ( $wgCountTotalSearchHits ) {
 190                         $totalResult = $this->db->query( $this->getCountQuery( $filteredTerm, $fulltext ) );
 191                         $row = $totalResult->fetchObject();
 192                         if ( $row ) {
 193                                 $total = intval( $row->c );
 194                         }
 195                         $totalResult->free();
 196                 }
 197
 198                 return new SqliteSearchResultSet( $resultSet, $this->searchTerms, $total );
 199         }
 200
 201         /**
 202          * Return a partial WHERE clause to exclude redirects, if so set
 203          * @return String
 204          */
 205         function queryRedirect() {
 206                 if ( $this->showRedirects ) {
 207                         return '';
 208                 } else {
 209                         return 'AND page_is_redirect=0';
 210                 }
 211         }
 212
 213         /**
 214          * Return a partial WHERE clause to limit the search to the given namespaces
 215          * @return String
 216          */
 217         function queryNamespaces() {
 218                 if ( is_null( $this->namespaces ) ) {
 219                         return '';  # search all
 220                 }
 221                 if ( !count( $this->namespaces ) ) {
 222                         $namespaces = '0';
 223                 } else {
 224                         $namespaces = $this->db->makeList( $this->namespaces );
 225                 }
 226                 return 'AND page_namespace IN (' . $namespaces . ')';
 227         }
 228
 229         /**
 230          * Returns a query with limit for number of results set.
 231          * @param $sql String:
 232          * @return String
 233          */
 234         function limitResult( $sql ) {
 235                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 236         }
 237
 238         /**
 239          * Construct the full SQL query to do the search.
 240          * The guts shoulds be constructed in queryMain()
 241          * @param $filteredTerm String
 242          * @param $fulltext Boolean
 243          * @return String
 244          */
 245         function getQuery( $filteredTerm, $fulltext ) {
 246                 return $this->limitResult(
 247                         $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 248                         $this->queryRedirect() . ' ' .
 249                         $this->queryNamespaces()
 250                 );
 251         }
 252
 253         /**
 254          * Picks which field to index on, depending on what type of query.
 255          * @param $fulltext Boolean
 256          * @return String
 257          */
 258         function getIndexField( $fulltext ) {
 259                 return $fulltext ? 'si_text' : 'si_title';
 260         }
 261
 262         /**
 263          * Get the base part of the search query.
 264          *
 265          * @param $filteredTerm String
 266          * @param $fulltext Boolean
 267          * @return String
 268          */
 269         function queryMain( $filteredTerm, $fulltext ) {
 270                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 271                 $page = $this->db->tableName( 'page' );
 272                 $searchindex = $this->db->tableName( 'searchindex' );
 273                 return "SELECT $searchindex.rowid, page_namespace, page_title " .
 274                         "FROM $page,$searchindex " .
 275                         "WHERE page_id=$searchindex.rowid AND $match";
 276         }
 277
 278         function getCountQuery( $filteredTerm, $fulltext ) {
 279                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 280                 $page = $this->db->tableName( 'page' );
 281                 $searchindex = $this->db->tableName( 'searchindex' );
 282                 return "SELECT COUNT(*) AS c " .
 283                         "FROM $page,$searchindex " .
 284                         "WHERE page_id=$searchindex.rowid AND $match" .
 285                         $this->queryRedirect() . ' ' .
 286                         $this->queryNamespaces();
 287         }
 288
 289         /**
 290          * Create or update the search index record for the given page.
 291          * Title and text should be pre-processed.
 292          *
 293          * @param $id Integer
 294          * @param $title String
 295          * @param $text String
 296          */
 297         function update( $id, $title, $text ) {
 298                 if ( !$this->fulltextSearchSupported() ) {
 299                         return;
 300                 }
 301                 // @todo find a method to do it in a single request,
 302                 // couldn't do it so far due to typelessness of FTS3 tables.
 303                 $dbw = wfGetDB( DB_MASTER );
 304
 305                 $dbw->delete( 'searchindex', array( 'rowid' => $id ), __METHOD__ );
 306
 307                 $dbw->insert( 'searchindex',
 308                         array(
 309                                 'rowid' => $id,
 310                                 'si_title' => $title,
 311                                 'si_text' => $text
 312                         ), __METHOD__ );
 313         }
 314
 315         /**
 316          * Update a search index record's title only.
 317          * Title should be pre-processed.
 318          *
 319          * @param $id Integer
 320          * @param $title String
 321          */
 322         function updateTitle( $id, $title ) {
 323                 if ( !$this->fulltextSearchSupported() ) {
 324                         return;
 325                 }
 326                 $dbw = wfGetDB( DB_MASTER );
 327
 328                 $dbw->update( 'searchindex',
 329                         array( 'si_title' => $title ),
 330                         array( 'rowid' => $id ),
 331                         __METHOD__ );
 332         }
 333 }
 334
 335 /**
 336  * @ingroup Search
 337  */
 338 class SqliteSearchResultSet extends SqlSearchResultSet {
 339         function __construct( $resultSet, $terms, $totalHits = null ) {
 340                 parent::__construct( $resultSet, $terms );
 341                 $this->mTotalHits = $totalHits;
 342         }
 343
 344         function getTotalHits() {
 345                 return $this->mTotalHits;
 346         }
 347 }