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 PostgresSearchResultSet
  39          */
  40         function searchTitle( $term ) {
  41                 $q = $this->searchQuery( $term, 'titlevector', 'page_title' );
  42                 $olderror = error_reporting( E_ERROR );
  43                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  44                 error_reporting( $olderror );
  45                 if ( !$resultSet ) {
  46                         // Needed for "Query requires full scan, GIN doesn't support it"
  47                         return new SearchResultTooMany();
  48                 }
  49                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  50         }
  51
  52         function searchText( $term ) {
  53                 $q = $this->searchQuery( $term, 'textvector', 'old_text' );
  54                 $olderror = error_reporting( E_ERROR );
  55                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  56                 error_reporting( $olderror );
  57                 if ( !$resultSet ) {
  58                         return new SearchResultTooMany();
  59                 }
  60                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  61         }
  62
  63         /**
  64          * Transform the user's search string into a better form for tsearch2
  65          * Returns an SQL fragment consisting of quoted text to search for.
  66          *
  67          * @param string $term
  68          *
  69          * @return string
  70          */
  71         function parseQuery( $term ) {
  72
  73                 wfDebug( "parseQuery received: $term \n" );
  74
  75                 ## No backslashes allowed
  76                 $term = preg_replace( '/\\\/', '', $term );
  77
  78                 ## Collapse parens into nearby words:
  79                 $term = preg_replace( '/\s*\(\s*/', ' (', $term );
  80                 $term = preg_replace( '/\s*\)\s*/', ') ', $term );
  81
  82                 ## Treat colons as word separators:
  83                 $term = preg_replace( '/:/', ' ', $term );
  84
  85                 $searchstring = '';
  86                 $m = array();
  87                 if ( preg_match_all( '/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  88                         foreach ( $m as $terms ) {
  89                                 if ( strlen( $terms[1] ) ) {
  90                                         $searchstring .= ' & !';
  91                                 }
  92                                 if ( strtolower( $terms[2] ) === 'and' ) {
  93                                         $searchstring .= ' & ';
  94                                 }
  95                                 elseif ( strtolower( $terms[2] ) === 'or' or $terms[2] === '|' ) {
  96                                         $searchstring .= ' | ';
  97                                 }
  98                                 elseif ( strtolower( $terms[2] ) === 'not' ) {
  99                                         $searchstring .= ' & !';
 100                                 }
 101                                 else {
 102                                         $searchstring .= " & $terms[2]";
 103                                 }
 104                         }
 105                 }
 106
 107                 ## Strip out leading junk
 108                 $searchstring = preg_replace( '/^[\s\&\|]+/', '', $searchstring );
 109
 110                 ## Remove any doubled-up operators
 111                 $searchstring = preg_replace( '/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring );
 112
 113                 ## Remove any non-spaced operators (e.g. "Zounds!")
 114                 $searchstring = preg_replace( '/([^ ])[\!\&\|]/', "$1", $searchstring );
 115
 116                 ## Remove any trailing whitespace or operators
 117                 $searchstring = preg_replace( '/[\s\!\&\|]+$/', '', $searchstring );
 118
 119                 ## Remove unnecessary quotes around everything
 120                 $searchstring = preg_replace( '/^[\'"](.*)[\'"]$/', "$1", $searchstring );
 121
 122                 ## Quote the whole thing
 123                 $searchstring = $this->db->addQuotes( $searchstring );
 124
 125                 wfDebug( "parseQuery returned: $searchstring \n" );
 126
 127                 return $searchstring;
 128
 129         }
 130
 131         /**
 132          * Construct the full SQL query to do the search.
 133          * @param string $term
 134          * @param string $fulltext
 135          * @param string $colname
 136          * @return string
 137          */
 138         function searchQuery( $term, $fulltext, $colname ) {
 139                 # Get the SQL fragment for the given term
 140                 $searchstring = $this->parseQuery( $term );
 141
 142                 ## We need a separate query here so gin does not complain about empty searches
 143                 $sql = "SELECT to_tsquery($searchstring)";
 144                 $res = $this->db->query( $sql );
 145                 if ( !$res ) {
 146                         ## TODO: Better output (example to catch: one 'two)
 147                         die( "Sorry, that was not a valid search string. Please go back and try again" );
 148                 }
 149                 $top = $res->fetchRow();
 150                 $top = $top[0];
 151
 152                 if ( $top === "" ) { ## e.g. if only stopwords are used XXX return something better
 153                         $query = "SELECT page_id, page_namespace, page_title, 0 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 1=0";
 156                 }
 157                 else {
 158                         $m = array();
 159                         if ( preg_match_all( "/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 160                                 foreach ( $m as $terms ) {
 161                                         $this->searchTerms[$terms[1]] = $terms[1];
 162                                 }
 163                         }
 164
 165                         $query = "SELECT page_id, page_namespace, page_title, " .
 166                         "ts_rank($fulltext, to_tsquery($searchstring), 5) AS score " .
 167                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 168                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery($searchstring)";
 169                 }
 170
 171                 ## Namespaces - defaults to 0
 172                 if ( !is_null( $this->namespaces ) ) { // null -> search all
 173                         if ( count( $this->namespaces ) < 1 ) {
 174                                 $query .= ' AND page_namespace = 0';
 175                         } else {
 176                                 $namespaces = $this->db->makeList( $this->namespaces );
 177                                 $query .= " AND page_namespace IN ($namespaces)";
 178                         }
 179                 }
 180
 181                 $query .= " ORDER BY score DESC, page_id DESC";
 182
 183                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 184
 185                 wfDebug( "searchQuery returned: $query \n" );
 186
 187                 return $query;
 188         }
 189
 190         ## Most of the work of these two functions are done automatically via triggers
 191
 192         function update( $pageid, $title, $text ) {
 193                 ## We don't want to index older revisions
 194                 $sql = "UPDATE pagecontent SET textvector = NULL WHERE textvector IS NOT NULL and old_id IN " .
 195                                 "(SELECT rev_text_id FROM revision WHERE rev_page = " . intval( $pageid ) .
 196                                 " ORDER BY rev_text_id DESC OFFSET 1)";
 197                 $this->db->query( $sql );
 198                 return true;
 199         }
 200
 201         function updateTitle( $id, $title ) {
 202                 return true;
 203         }
 204
 205 } ## end of the SearchPostgres class
 206
 207 /**
 208  * @ingroup Search
 209  */
 210 class PostgresSearchResult extends SearchResult {
 211         function __construct( $row ) {
 212                 parent::__construct( $row );
 213                 $this->score = $row->score;
 214         }
 215
 216         function getScore() {
 217                 return $this->score;
 218         }
 219 }
 220
 221 /**
 222  * @ingroup Search
 223  */
 224 class PostgresSearchResultSet extends SqlSearchResultSet {
 225         function next() {
 226                 $row = $this->resultSet->fetchObject();
 227                 if ( $row === false ) {
 228                         return false;
 229                 } else {
 230                         return new PostgresSearchResult( $row );
 231                 }
 232         }
 233 }