includes/search/SearchPostgres.php

   1 <?php
   2 /**
   3  * PostgreSQL search engine
   4  *
   5  * Copyright © 2006-2007 Greg Sabino Mullane <greg@turnstep.com>
   6  * https://www.mediawiki.org/
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with this program; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  21  * http://www.gnu.org/copyleft/gpl.html
  22  *
  23  * @file
  24  * @ingroup Search
  25  */
  26
  27 /**
  28  * Search engine hook base class for Postgres
  29  * @ingroup Search
  30  */
  31 class SearchPostgres extends SearchDatabase {
  32         /**
  33          * Perform a full text search query via tsearch2 and return a result set.
  34          * Currently searches a page's current title (page.page_title) and
  35          * latest revision article text (pagecontent.old_text)
  36          *
  37          * @param string $term Raw search term
  38          * @return SqlSearchResultSet
  39          */
  40         function searchTitle( $term ) {
  41                 $q = $this->searchQuery( $term, 'titlevector', 'page_title' );
  42                 $olderror = error_reporting( E_ERROR );
  43                 $resultSet = $this->db->query( $q, 'SearchPostgres', true );
  44                 error_reporting( $olderror );
  45                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  46         }
  47
  48         function searchText( $term ) {
  49                 $q = $this->searchQuery( $term, 'textvector', 'old_text' );
  50                 $olderror = error_reporting( E_ERROR );
  51                 $resultSet = $this->db->query( $q, 'SearchPostgres', true );
  52                 error_reporting( $olderror );
  53                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  54         }
  55
  56         /**
  57          * Transform the user's search string into a better form for tsearch2
  58          * Returns an SQL fragment consisting of quoted text to search for.
  59          *
  60          * @param string $term
  61          *
  62          * @return string
  63          */
  64         function parseQuery( $term ) {
  65                 wfDebug( "parseQuery received: $term \n" );
  66
  67                 # # No backslashes allowed
  68                 $term = preg_replace( '/\\\/', '', $term );
  69
  70                 # # Collapse parens into nearby words:
  71                 $term = preg_replace( '/\s*\(\s*/', ' (', $term );
  72                 $term = preg_replace( '/\s*\)\s*/', ') ', $term );
  73
  74                 # # Treat colons as word separators:
  75                 $term = preg_replace( '/:/', ' ', $term );
  76
  77                 $searchstring = '';
  78                 $m = [];
  79                 if ( preg_match_all( '/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  80                         foreach ( $m as $terms ) {
  81                                 if ( strlen( $terms[1] ) ) {
  82                                         $searchstring .= ' & !';
  83                                 }
  84                                 if ( strtolower( $terms[2] ) === 'and' ) {
  85                                         $searchstring .= ' & ';
  86                                 } elseif ( strtolower( $terms[2] ) === 'or' || $terms[2] === '|' ) {
  87                                         $searchstring .= ' | ';
  88                                 } elseif ( strtolower( $terms[2] ) === 'not' ) {
  89                                         $searchstring .= ' & !';
  90                                 } else {
  91                                         $searchstring .= " & $terms[2]";
  92                                 }
  93                         }
  94                 }
  95
  96                 # # Strip out leading junk
  97                 $searchstring = preg_replace( '/^[\s\&\|]+/', '', $searchstring );
  98
  99                 # # Remove any doubled-up operators
 100                 $searchstring = preg_replace( '/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring );
 101
 102                 # # Remove any non-spaced operators (e.g. "Zounds!")
 103                 $searchstring = preg_replace( '/([^ ])[\!\&\|]/', "$1", $searchstring );
 104
 105                 # # Remove any trailing whitespace or operators
 106                 $searchstring = preg_replace( '/[\s\!\&\|]+$/', '', $searchstring );
 107
 108                 # # Remove unnecessary quotes around everything
 109                 $searchstring = preg_replace( '/^[\'"](.*)[\'"]$/', "$1", $searchstring );
 110
 111                 # # Quote the whole thing
 112                 $searchstring = $this->db->addQuotes( $searchstring );
 113
 114                 wfDebug( "parseQuery returned: $searchstring \n" );
 115
 116                 return $searchstring;
 117         }
 118
 119         /**
 120          * Construct the full SQL query to do the search.
 121          * @param string $term
 122          * @param string $fulltext
 123          * @param string $colname
 124          * @return string
 125          */
 126         function searchQuery( $term, $fulltext, $colname ) {
 127                 # Get the SQL fragment for the given term
 128                 $searchstring = $this->parseQuery( $term );
 129
 130                 # # We need a separate query here so gin does not complain about empty searches
 131                 $sql = "SELECT to_tsquery($searchstring)";
 132                 $res = $this->db->query( $sql );
 133                 if ( !$res ) {
 134                         # # TODO: Better output (example to catch: one 'two)
 135                         die( "Sorry, that was not a valid search string. Please go back and try again" );
 136                 }
 137                 $top = $res->fetchRow()[0];
 138
 139                 $this->searchTerms = [];
 140                 if ( $top === "" ) { # # e.g. if only stopwords are used XXX return something better
 141                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score " .
 142                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 143                                 "AND r.rev_text_id = c.old_id AND 1=0";
 144                 } else {
 145                         $m = [];
 146                         if ( preg_match_all( "/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 147                                 foreach ( $m as $terms ) {
 148                                         $this->searchTerms[$terms[1]] = $terms[1];
 149                                 }
 150                         }
 151
 152                         $query = "SELECT page_id, page_namespace, page_title, " .
 153                                 "ts_rank($fulltext, to_tsquery($searchstring), 5) AS score " .
 154                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 155                                 "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery($searchstring)";
 156                 }
 157
 158                 # # Namespaces - defaults to 0
 159                 if ( !is_null( $this->namespaces ) ) { // null -> search all
 160                         if ( count( $this->namespaces ) < 1 ) {
 161                                 $query .= ' AND page_namespace = 0';
 162                         } else {
 163                                 $namespaces = $this->db->makeList( $this->namespaces );
 164                                 $query .= " AND page_namespace IN ($namespaces)";
 165                         }
 166                 }
 167
 168                 $query .= " ORDER BY score DESC, page_id DESC";
 169
 170                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 171
 172                 wfDebug( "searchQuery returned: $query \n" );
 173
 174                 return $query;
 175         }
 176
 177         # # Most of the work of these two functions are done automatically via triggers
 178
 179         function update( $pageid, $title, $text ) {
 180                 # # We don't want to index older revisions
 181                 $sql = "UPDATE pagecontent SET textvector = NULL WHERE textvector IS NOT NULL and old_id IN " .
 182                                 "(SELECT DISTINCT rev_text_id FROM revision WHERE rev_page = " . intval( $pageid ) .
 183                                 " ORDER BY rev_text_id DESC OFFSET 1)";
 184                 $this->db->query( $sql );
 185                 return true;
 186         }
 187
 188         function updateTitle( $id, $title ) {
 189                 return true;
 190         }
 191
 192 }