includes/SearchPostgres.php

   1 <?php
   2 # Copyright (C) 2006-2007 Greg Sabino Mullane <greg@turnstep.com>
   3 # http://www.mediawiki.org/
   4 #
   5 # This program is free software; you can redistribute it and/or modify
   6 # it under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 2 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # This program is distributed in the hope that it will be useful,
  11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License along
  16 # with this program; if not, write to the Free Software Foundation, Inc.,
  17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18 # http://www.gnu.org/copyleft/gpl.html
  19
  20 /**
  21  * @file
  22  * @ingroup Search
  23  */
  24
  25 /**
  26  * Search engine hook base class for Postgres
  27  * @ingroup Search
  28  */
  29 class SearchPostgres extends SearchEngine {
  30
  31         function __construct( $db ) {
  32                 $this->db = $db;
  33         }
  34
  35         /**
  36          * Perform a full text search query via tsearch2 and return a result set.
  37          * Currently searches a page's current title (page.page_title) and
  38          * latest revision article text (pagecontent.old_text)
  39          *
  40          * @param $term String: raw search term
  41          * @return PostgresSearchResultSet
  42          */
  43         function searchTitle( $term ) {
  44                 $q = $this->searchQuery( $term , 'titlevector', 'page_title' );
  45                 $olderror = error_reporting(E_ERROR);
  46                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  47                 error_reporting($olderror);
  48                 if (!$resultSet) {
  49                         // Needed for "Query requires full scan, GIN doesn't support it"
  50                         return new SearchResultTooMany();
  51                 }
  52                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  53         }
  54         function searchText( $term ) {
  55                 $q = $this->searchQuery( $term, 'textvector', 'old_text' );
  56                 $olderror = error_reporting(E_ERROR);
  57                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  58                 error_reporting($olderror);
  59                 if (!$resultSet) {
  60                         return new SearchResultTooMany();
  61                 }
  62                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  63         }
  64
  65
  66         /*
  67          * Transform the user's search string into a better form for tsearch2
  68          * Returns an SQL fragment consisting of quoted text to search for.
  69         */
  70         function parseQuery( $term ) {
  71
  72                 wfDebug( "parseQuery received: $term \n" );
  73
  74                 ## No backslashes allowed
  75                 $term = preg_replace('/\\\/', '', $term);
  76
  77                 ## Collapse parens into nearby words:
  78                 $term = preg_replace('/\s*\(\s*/', ' (', $term);
  79                 $term = preg_replace('/\s*\)\s*/', ') ', $term);
  80
  81                 ## Treat colons as word separators:
  82                 $term = preg_replace('/:/', ' ', $term);
  83
  84                 $searchstring = '';
  85                 $m = array();
  86                 if( preg_match_all('/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  87                         foreach( $m as $terms ) {
  88                                 if (strlen($terms[1])) {
  89                                         $searchstring .= ' & !';
  90                                 }
  91                                 if (strtolower($terms[2]) === 'and') {
  92                                         $searchstring .= ' & ';
  93                                 }
  94                                 else if (strtolower($terms[2]) === 'or' or $terms[2] === '|') {
  95                                         $searchstring .= ' | ';
  96                                 }
  97                                 else if (strtolower($terms[2]) === 'not') {
  98                                         $searchstring .= ' & !';
  99                                 }
 100                                 else {
 101                                         $searchstring .= " & $terms[2]";
 102                                 }
 103                         }
 104                 }
 105
 106                 ## Strip out leading junk
 107                 $searchstring = preg_replace('/^[\s\&\|]+/', '', $searchstring);
 108
 109                 ## Remove any doubled-up operators
 110                 $searchstring = preg_replace('/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring);
 111
 112                 ## Remove any non-spaced operators (e.g. "Zounds!")
 113                 $searchstring = preg_replace('/([^ ])[\!\&\|]/', "$1", $searchstring);
 114
 115                 ## Remove any trailing whitespace or operators
 116                 $searchstring = preg_replace('/[\s\!\&\|]+$/', '', $searchstring);
 117
 118                 ## Remove unnecessary quotes around everything
 119                 $searchstring = preg_replace('/^[\'"](.*)[\'"]$/', "$1", $searchstring);
 120
 121                 ## Quote the whole thing
 122                 $searchstring = $this->db->addQuotes($searchstring);
 123
 124                 wfDebug( "parseQuery returned: $searchstring \n" );
 125
 126                 return $searchstring;
 127
 128         }
 129
 130         /**
 131          * Construct the full SQL query to do the search.
 132          * @param $filteredTerm String
 133          * @param $fulltext String
 134          */
 135         function searchQuery( $term, $fulltext, $colname ) {
 136                 global $wgDBversion;
 137
 138                 if ( !isset( $wgDBversion ) ) {
 139                         $this->db->getServerVersion();
 140                         $wgDBversion = $this->db->numeric_version;
 141                 }
 142                 $prefix = $wgDBversion < 8.3 ? "'default'," : '';
 143
 144                 # Get the SQL fragment for the given term
 145                 $searchstring = $this->parseQuery( $term );
 146
 147                 ## We need a separate query here so gin does not complain about empty searches
 148                 $SQL = "SELECT to_tsquery($prefix $searchstring)";
 149                 $res = $this->db->doQuery($SQL);
 150                 if (!$res) {
 151                         ## TODO: Better output (example to catch: one 'two)
 152                         die ("Sorry, that was not a valid search string. Please go back and try again");
 153                 }
 154                 $top = pg_fetch_result($res,0,0);
 155
 156                 if ($top === "") { ## e.g. if only stopwords are used XXX return something better
 157                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score ".
 158                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 159                                 "AND r.rev_text_id = c.old_id AND 1=0";
 160                 }
 161                 else {
 162                         $m = array();
 163                         if( preg_match_all("/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 164                                 foreach( $m as $terms ) {
 165                                         $this->searchTerms[$terms[1]] = $terms[1];
 166                                 }
 167                         }
 168
 169                         $rankscore = $wgDBversion > 8.2 ? 5 : 1;
 170                         $rank = $wgDBversion < 8.3 ? 'rank' : 'ts_rank';
 171                         $query = "SELECT page_id, page_namespace, page_title, ".
 172                         "$rank($fulltext, to_tsquery($prefix $searchstring), $rankscore) AS score ".
 173                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 174                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery($prefix $searchstring)";
 175                 }
 176
 177                 ## Redirects
 178                 if (! $this->showRedirects)
 179                         $query .= ' AND page_is_redirect = 0';
 180
 181                 ## Namespaces - defaults to 0
 182                 if( !is_null($this->namespaces) ){ // null -> search all
 183                         if ( count($this->namespaces) < 1)
 184                                 $query .= ' AND page_namespace = 0';
 185                         else {
 186                                 $namespaces = $this->db->makeList( $this->namespaces );
 187                                 $query .= " AND page_namespace IN ($namespaces)";
 188                         }
 189                 }
 190
 191                 $query .= " ORDER BY score DESC, page_id DESC";
 192
 193                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 194
 195                 wfDebug( "searchQuery returned: $query \n" );
 196
 197                 return $query;
 198         }
 199
 200         ## Most of the work of these two functions are done automatically via triggers
 201
 202         function update( $pageid, $title, $text ) {
 203                 ## We don't want to index older revisions
 204                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id IN ".
 205                                 "(SELECT rev_text_id FROM revision WHERE rev_page = " . intval( $pageid ) .
 206                                 " ORDER BY rev_text_id DESC OFFSET 1)";
 207                 $this->db->doQuery($SQL);
 208                 return true;
 209         }
 210
 211         function updateTitle( $id, $title ) {
 212                 return true;
 213         }
 214
 215 } ## end of the SearchPostgres class
 216
 217 /**
 218  * @ingroup Search
 219  */
 220 class PostgresSearchResult extends SearchResult {
 221         function __construct( $row ) {
 222                 parent::__construct($row);
 223                 $this->score = $row->score;
 224         }
 225         function getScore() {
 226                 return $this->score;
 227         }
 228 }
 229
 230 /**
 231  * @ingroup Search
 232  */
 233 class PostgresSearchResultSet extends SearchResultSet {
 234         function __construct( $resultSet, $terms ) {
 235                 $this->mResultSet = $resultSet;
 236                 $this->mTerms = $terms;
 237         }
 238
 239         function termMatches() {
 240                 return $this->mTerms;
 241         }
 242
 243         function numRows() {
 244                 return $this->mResultSet->numRows();
 245         }
 246
 247         function next() {
 248                 $row = $this->mResultSet->fetchObject();
 249                 if( $row === false ) {
 250                         return false;
 251                 } else {
 252                         return new PostgresSearchResult( $row );
 253                 }
 254         }
 255 }