SearchOracle.php

Go to the documentation of this file.

<?php
class SearchOracle extends SearchEngine {

    private $reservedWords = array(
        'ABOUT' => 1,
        'ACCUM' => 1,
        'AND' => 1,
        'BT' => 1,
        'BTG' => 1,
        'BTI' => 1,
        'BTP' => 1,
        'FUZZY' => 1,
        'HASPATH' => 1,
        'INPATH' => 1,
        'MINUS' => 1,
        'NEAR' => 1,
        'NOT' => 1,
        'NT' => 1,
        'NTG' => 1,
        'NTI' => 1,
        'NTP' => 1,
        'OR' => 1,
        'PT' => 1,
        'RT' => 1,
        'SQE' => 1,
        'SYN' => 1,
        'TR' => 1,
        'TRSYN' => 1,
        'TT' => 1,
        'WITHIN' => 1,
    );

    function __construct( $db ) {
        parent::__construct( $db );
    }

    function searchText( $term ) {
        if ( $term == '' ) {
            return new SqlSearchResultSet( false, '' );
        }

        $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
        return new SqlSearchResultSet( $resultSet, $this->searchTerms );
    }

    function searchTitle( $term ) {
        if ( $term == '' ) {
            return new SqlSearchResultSet( false, '' );
        }

        $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
        return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
    }

    function queryRedirect() {
        if ( $this->showRedirects ) {
            return '';
        } else {
            return 'AND page_is_redirect=0';
        }
    }

    function queryNamespaces() {
        if ( is_null( $this->namespaces ) ) {
            return '';
        }
        if ( !count( $this->namespaces ) ) {
            $namespaces = '0';
        } else {
            $namespaces = $this->db->makeList( $this->namespaces );
        }
        return 'AND page_namespace IN (' . $namespaces . ')';
    }

    function queryLimit( $sql ) {
        return $this->db->limitResult( $sql, $this->limit, $this->offset );
    }

    function queryRanking( $filteredTerm, $fulltext ) {
        return ' ORDER BY score(1)';
    }

    function getQuery( $filteredTerm, $fulltext ) {
        return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
            $this->queryRedirect() . ' ' .
            $this->queryNamespaces() . ' ' .
            $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
    }

    function getIndexField( $fulltext ) {
        return $fulltext ? 'si_text' : 'si_title';
    }

    function queryMain( $filteredTerm, $fulltext ) {
        $match = $this->parseQuery( $filteredTerm, $fulltext );
        $page = $this->db->tableName( 'page' );
        $searchindex = $this->db->tableName( 'searchindex' );
        return 'SELECT page_id, page_namespace, page_title ' .
            "FROM $page,$searchindex " .
            'WHERE page_id=si_page AND ' . $match;
    }

    function parseQuery( $filteredText, $fulltext ) {
        global $wgContLang;
        $lc = SearchEngine::legalSearchChars();
        $this->searchTerms = array();

        # @todo FIXME: This doesn't handle parenthetical expressions.
        $m = array();
        $searchon = '';
        if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
                $filteredText, $m, PREG_SET_ORDER ) ) {
            foreach ( $m as $terms ) {
                // Search terms in all variant forms, only
                // apply on wiki with LanguageConverter
                $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
                if ( is_array( $temp_terms ) ) {
                    $temp_terms = array_unique( array_values( $temp_terms ) );
                    foreach ( $temp_terms as $t ) {
                        $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $t );
                    }
                }
                else {
                    $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $terms[2] );
                }
                if ( !empty( $terms[3] ) ) {
                    $regexp = preg_quote( $terms[3], '/' );
                    if ( $terms[4] ) {
                        $regexp .= "[0-9A-Za-z_]+";
                    }
                } else {
                    $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
                }
                $this->searchTerms[] = $regexp;
            }
        }

        $searchon = $this->db->addQuotes( ltrim( $searchon, ' &' ) );
        $field = $this->getIndexField( $fulltext );
        return " CONTAINS($field, $searchon, 1) > 0 ";
    }

    private function escapeTerm( $t ) {
        global $wgContLang;
        $t = $wgContLang->normalizeForSearch( $t );
        $t = isset( $this->reservedWords[strtoupper( $t )] ) ? '{' . $t . '}' : $t;
        $t = preg_replace( '/^"(.*)"$/', '($1)', $t );
        $t = preg_replace( '/([-&|])/', '\\\\$1', $t );
        return $t;
    }
    function update( $id, $title, $text ) {
        $dbw = wfGetDB( DB_MASTER );
        $dbw->replace( 'searchindex',
            array( 'si_page' ),
            array(
                'si_page' => $id,
                'si_title' => $title,
                'si_text' => $text
            ), 'SearchOracle::update' );

        // Sync the index
        // We need to specify the DB name (i.e. user/schema) here so that
        // it can work from the installer, where
        //     ALTER SESSION SET CURRENT_SCHEMA = ...
        // was used.
        $dbw->query( "CALL ctx_ddl.sync_index(" .
            $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_text_idx', 'raw' ) ) . ")" );
        $dbw->query( "CALL ctx_ddl.sync_index(" .
            $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_title_idx', 'raw' ) ) . ")" );
    }

    function updateTitle( $id, $title ) {
        $dbw = wfGetDB( DB_MASTER );

        $dbw->update( 'searchindex',
            array( 'si_title' => $title ),
            array( 'si_page' => $id ),
            'SearchOracle::updateTitle',
            array() );
    }

    public static function legalSearchChars() {
        return "\"" . parent::legalSearchChars();
    }
}