includes/search/SearchPostgres.php

   1 <?php
   2 /**
   3  * PostgreSQL search engine
   4  *
   5  * Copyright © 2006-2007 Greg Sabino Mullane <greg@turnstep.com>
   6  * http://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 SearchEngine {
  32
  33         /**
  34          * @var DatabasePostgres
  35          */
  36         protected $db;
  37         /**
  38          * Creates an instance of this class
  39          * @param $db DatabaseSqlite: database object
  40          */
  41         function __construct( $db ) {
  42                 parent::__construct( $db );
  43         }
  44
  45         /**
  46          * Perform a full text search query via tsearch2 and return a result set.
  47          * Currently searches a page's current title (page.page_title) and
  48          * latest revision article text (pagecontent.old_text)
  49          *
  50          * @param string $term raw search term
  51          * @return PostgresSearchResultSet
  52          */
  53         function searchTitle( $term ) {
  54                 $q = $this->searchQuery( $term, 'titlevector', 'page_title' );
  55                 $olderror = error_reporting( E_ERROR );
  56                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  57                 error_reporting( $olderror );
  58                 if ( !$resultSet ) {
  59                         // Needed for "Query requires full scan, GIN doesn't support it"
  60                         return new SearchResultTooMany();
  61                 }
  62                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  63         }
  64
  65         function searchText( $term ) {
  66                 $q = $this->searchQuery( $term, 'textvector', 'old_text' );
  67                 $olderror = error_reporting( E_ERROR );
  68                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  69                 error_reporting( $olderror );
  70                 if ( !$resultSet ) {
  71                         return new SearchResultTooMany();
  72                 }
  73                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  74         }
  75
  76         /**
  77          * Transform the user's search string into a better form for tsearch2
  78          * Returns an SQL fragment consisting of quoted text to search for.
  79          *
  80          * @param $term string
  81          *
  82          * @return string
  83          */
  84         function parseQuery( $term ) {
  85
  86                 wfDebug( "parseQuery received: $term \n" );
  87
  88                 ## No backslashes allowed
  89                 $term = preg_replace( '/\\\/', '', $term );
  90
  91                 ## Collapse parens into nearby words:
  92                 $term = preg_replace( '/\s*\(\s*/', ' (', $term );
  93                 $term = preg_replace( '/\s*\)\s*/', ') ', $term );
  94
  95                 ## Treat colons as word separators:
  96                 $term = preg_replace( '/:/', ' ', $term );
  97
  98                 $searchstring = '';
  99                 $m = array();
 100                 if ( preg_match_all( '/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
 101                         foreach ( $m as $terms ) {
 102                                 if ( strlen( $terms[1] ) ) {
 103                                         $searchstring .= ' & !';
 104                                 }
 105                                 if ( strtolower( $terms[2] ) === 'and' ) {
 106                                         $searchstring .= ' & ';
 107                                 }
 108                                 elseif ( strtolower( $terms[2] ) === 'or' or $terms[2] === '|' ) {
 109                                         $searchstring .= ' | ';
 110                                 }
 111                                 elseif ( strtolower( $terms[2] ) === 'not' ) {
 112                                         $searchstring .= ' & !';
 113                                 }
 114                                 else {
 115                                         $searchstring .= " & $terms[2]";
 116                                 }
 117                         }
 118                 }
 119
 120                 ## Strip out leading junk
 121                 $searchstring = preg_replace( '/^[\s\&\|]+/', '', $searchstring );
 122
 123                 ## Remove any doubled-up operators
 124                 $searchstring = preg_replace( '/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring );
 125
 126                 ## Remove any non-spaced operators (e.g. "Zounds!")
 127                 $searchstring = preg_replace( '/([^ ])[\!\&\|]/', "$1", $searchstring );
 128
 129                 ## Remove any trailing whitespace or operators
 130                 $searchstring = preg_replace( '/[\s\!\&\|]+$/', '', $searchstring );
 131
 132                 ## Remove unnecessary quotes around everything
 133                 $searchstring = preg_replace( '/^[\'"](.*)[\'"]$/', "$1", $searchstring );
 134
 135                 ## Quote the whole thing
 136                 $searchstring = $this->db->addQuotes( $searchstring );
 137
 138                 wfDebug( "parseQuery returned: $searchstring \n" );
 139
 140                 return $searchstring;
 141
 142         }
 143
 144         /**
 145          * Construct the full SQL query to do the search.
 146          * @param $term String
 147          * @param $fulltext String
 148          * @param $colname
 149          * @return string
 150          */
 151         function searchQuery( $term, $fulltext, $colname ) {
 152                 # Get the SQL fragment for the given term
 153                 $searchstring = $this->parseQuery( $term );
 154
 155                 ## We need a separate query here so gin does not complain about empty searches
 156                 $SQL = "SELECT to_tsquery($searchstring)";
 157                 $res = $this->db->query( $SQL );
 158                 if ( !$res ) {
 159                         ## TODO: Better output (example to catch: one 'two)
 160                         die( "Sorry, that was not a valid search string. Please go back and try again" );
 161                 }
 162                 $top = $res->fetchRow();
 163                 $top = $top[0];
 164
 165                 if ( $top === "" ) { ## e.g. if only stopwords are used XXX return something better
 166                         $query = "SELECT page_id, page_namespace, page_title, 0 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 1=0";
 169                 }
 170                 else {
 171                         $m = array();
 172                         if ( preg_match_all( "/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 173                                 foreach ( $m as $terms ) {
 174                                         $this->searchTerms[$terms[1]] = $terms[1];
 175                                 }
 176                         }
 177
 178                         $query = "SELECT page_id, page_namespace, page_title, " .
 179                         "ts_rank($fulltext, to_tsquery($searchstring), 5) AS score " .
 180                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 181                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery($searchstring)";
 182                 }
 183
 184                 ## Redirects
 185                 if ( !$this->showRedirects ) {
 186                         $query .= ' AND page_is_redirect = 0';
 187                 }
 188
 189                 ## Namespaces - defaults to 0
 190                 if ( !is_null( $this->namespaces ) ) { // null -> search all
 191                         if ( count( $this->namespaces ) < 1 ) {
 192                                 $query .= ' AND page_namespace = 0';
 193                         } else {
 194                                 $namespaces = $this->db->makeList( $this->namespaces );
 195                                 $query .= " AND page_namespace IN ($namespaces)";
 196                         }
 197                 }
 198
 199                 $query .= " ORDER BY score DESC, page_id DESC";
 200
 201                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 202
 203                 wfDebug( "searchQuery returned: $query \n" );
 204
 205                 return $query;
 206         }
 207
 208         ## Most of the work of these two functions are done automatically via triggers
 209
 210         function update( $pageid, $title, $text ) {
 211                 ## We don't want to index older revisions
 212                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id IN " .
 213                                 "(SELECT rev_text_id FROM revision WHERE rev_page = " . intval( $pageid ) .
 214                                 " ORDER BY rev_text_id DESC OFFSET 1)";
 215                 $this->db->query( $SQL );
 216                 return true;
 217         }
 218
 219         function updateTitle( $id, $title ) {
 220                 return true;
 221         }
 222
 223 } ## end of the SearchPostgres class
 224
 225 /**
 226  * @ingroup Search
 227  */
 228 class PostgresSearchResult extends SearchResult {
 229         function __construct( $row ) {
 230                 parent::__construct( $row );
 231                 $this->score = $row->score;
 232         }
 233         function getScore() {
 234                 return $this->score;
 235         }
 236 }
 237
 238 /**
 239  * @ingroup Search
 240  */
 241 class PostgresSearchResultSet extends SqlSearchResultSet {
 242         function __construct( $resultSet, $terms ) {
 243                 parent::__construct( $resultSet, $terms );
 244         }
 245
 246         function next() {
 247                 $row = $this->mResultSet->fetchObject();
 248                 if ( $row === false ) {
 249                         return false;
 250                 } else {
 251                         return new PostgresSearchResult( $row );
 252                 }
 253         }
 254 }