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         function __construct( $db ) {
  34                 $this->db = $db;
  35         }
  36
  37         /**
  38          * Perform a full text search query via tsearch2 and return a result set.
  39          * Currently searches a page's current title (page.page_title) and
  40          * latest revision article text (pagecontent.old_text)
  41          *
  42          * @param $term String: raw search term
  43          * @return PostgresSearchResultSet
  44          */
  45         function searchTitle( $term ) {
  46                 $q = $this->searchQuery( $term , 'titlevector', 'page_title' );
  47                 $olderror = error_reporting(E_ERROR);
  48                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  49                 error_reporting($olderror);
  50                 if (!$resultSet) {
  51                         // Needed for "Query requires full scan, GIN doesn't support it"
  52                         return new SearchResultTooMany();
  53                 }
  54                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  55         }
  56         function searchText( $term ) {
  57                 $q = $this->searchQuery( $term, 'textvector', 'old_text' );
  58                 $olderror = error_reporting(E_ERROR);
  59                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  60                 error_reporting($olderror);
  61                 if (!$resultSet) {
  62                         return new SearchResultTooMany();
  63                 }
  64                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  65         }
  66
  67
  68         /*
  69          * Transform the user's search string into a better form for tsearch2
  70          * Returns an SQL fragment consisting of quoted text to search for.
  71         */
  72         function parseQuery( $term ) {
  73
  74                 wfDebug( "parseQuery received: $term \n" );
  75
  76                 ## No backslashes allowed
  77                 $term = preg_replace('/\\\/', '', $term);
  78
  79                 ## Collapse parens into nearby words:
  80                 $term = preg_replace('/\s*\(\s*/', ' (', $term);
  81                 $term = preg_replace('/\s*\)\s*/', ') ', $term);
  82
  83                 ## Treat colons as word separators:
  84                 $term = preg_replace('/:/', ' ', $term);
  85
  86                 $searchstring = '';
  87                 $m = array();
  88                 if( preg_match_all('/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  89                         foreach( $m as $terms ) {
  90                                 if (strlen($terms[1])) {
  91                                         $searchstring .= ' & !';
  92                                 }
  93                                 if (strtolower($terms[2]) === 'and') {
  94                                         $searchstring .= ' & ';
  95                                 }
  96                                 else if (strtolower($terms[2]) === 'or' or $terms[2] === '|') {
  97                                         $searchstring .= ' | ';
  98                                 }
  99                                 else if (strtolower($terms[2]) === 'not') {
 100                                         $searchstring .= ' & !';
 101                                 }
 102                                 else {
 103                                         $searchstring .= " & $terms[2]";
 104                                 }
 105                         }
 106                 }
 107
 108                 ## Strip out leading junk
 109                 $searchstring = preg_replace('/^[\s\&\|]+/', '', $searchstring);
 110
 111                 ## Remove any doubled-up operators
 112                 $searchstring = preg_replace('/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring);
 113
 114                 ## Remove any non-spaced operators (e.g. "Zounds!")
 115                 $searchstring = preg_replace('/([^ ])[\!\&\|]/', "$1", $searchstring);
 116
 117                 ## Remove any trailing whitespace or operators
 118                 $searchstring = preg_replace('/[\s\!\&\|]+$/', '', $searchstring);
 119
 120                 ## Remove unnecessary quotes around everything
 121                 $searchstring = preg_replace('/^[\'"](.*)[\'"]$/', "$1", $searchstring);
 122
 123                 ## Quote the whole thing
 124                 $searchstring = $this->db->addQuotes($searchstring);
 125
 126                 wfDebug( "parseQuery returned: $searchstring \n" );
 127
 128                 return $searchstring;
 129
 130         }
 131
 132         /**
 133          * Construct the full SQL query to do the search.
 134          * @param $term String
 135          * @param $fulltext String
 136          * @param $colname
 137          */
 138         function searchQuery( $term, $fulltext, $colname ) {
 139                 $postgresVersion = $this->db->getServerVersion();
 140
 141                 $prefix = $postgresVersion < 8.3 ? "'default'," : '';
 142
 143                 # Get the SQL fragment for the given term
 144                 $searchstring = $this->parseQuery( $term );
 145
 146                 ## We need a separate query here so gin does not complain about empty searches
 147                 $SQL = "SELECT to_tsquery($prefix $searchstring)";
 148                 $res = $this->db->doQuery($SQL);
 149                 if (!$res) {
 150                         ## TODO: Better output (example to catch: one 'two)
 151                         die ("Sorry, that was not a valid search string. Please go back and try again");
 152                 }
 153                 $top = pg_fetch_result($res,0,0);
 154
 155                 if ($top === "") { ## e.g. if only stopwords are used XXX return something better
 156                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score ".
 157                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 158                                 "AND r.rev_text_id = c.old_id AND 1=0";
 159                 }
 160                 else {
 161                         $m = array();
 162                         if( preg_match_all("/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 163                                 foreach( $m as $terms ) {
 164                                         $this->searchTerms[$terms[1]] = $terms[1];
 165                                 }
 166                         }
 167
 168                         $rankscore = $postgresVersion > 8.2 ? 5 : 1;
 169                         $rank = $postgresVersion < 8.3 ? 'rank' : 'ts_rank';
 170                         $query = "SELECT page_id, page_namespace, page_title, ".
 171                         "$rank($fulltext, to_tsquery($prefix $searchstring), $rankscore) AS score ".
 172                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 173                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery($prefix $searchstring)";
 174                 }
 175
 176                 ## Redirects
 177                 if (! $this->showRedirects)
 178                         $query .= ' AND page_is_redirect = 0';
 179
 180                 ## Namespaces - defaults to 0
 181                 if( !is_null($this->namespaces) ){ // null -> search all
 182                         if ( count($this->namespaces) < 1)
 183                                 $query .= ' AND page_namespace = 0';
 184                         else {
 185                                 $namespaces = $this->db->makeList( $this->namespaces );
 186                                 $query .= " AND page_namespace IN ($namespaces)";
 187                         }
 188                 }
 189
 190                 $query .= " ORDER BY score DESC, page_id DESC";
 191
 192                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 193
 194                 wfDebug( "searchQuery returned: $query \n" );
 195
 196                 return $query;
 197         }
 198
 199         ## Most of the work of these two functions are done automatically via triggers
 200
 201         function update( $pageid, $title, $text ) {
 202                 ## We don't want to index older revisions
 203                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id IN ".
 204                                 "(SELECT rev_text_id FROM revision WHERE rev_page = " . intval( $pageid ) .
 205                                 " ORDER BY rev_text_id DESC OFFSET 1)";
 206                 $this->db->doQuery($SQL);
 207                 return true;
 208         }
 209
 210         function updateTitle( $id, $title ) {
 211                 return true;
 212         }
 213
 214 } ## end of the SearchPostgres class
 215
 216 /**
 217  * @ingroup Search
 218  */
 219 class PostgresSearchResult extends SearchResult {
 220         function __construct( $row ) {
 221                 parent::__construct($row);
 222                 $this->score = $row->score;
 223         }
 224         function getScore() {
 225                 return $this->score;
 226         }
 227 }
 228
 229 /**
 230  * @ingroup Search
 231  */
 232 class PostgresSearchResultSet extends SqlSearchResultSet {
 233         function __construct( $resultSet, $terms ) {
 234                 parent::__construct( $resultSet, $terms );
 235         }
 236
 237         function next() {
 238                 $row = $this->mResultSet->fetchObject();
 239                 if( $row === false ) {
 240                         return false;
 241                 } else {
 242                         return new PostgresSearchResult( $row );
 243                 }
 244         }
 245 }