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  * Search engine hook base class for Postgres
  22  * @addtogroup Search
  23  */
  24
  25 class SearchPostgres extends SearchEngine {
  26
  27         function SearchPostgres( $db ) {
  28                 $this->db = $db;
  29         }
  30
  31         /**
  32          * Perform a full text search query via tsearch2 and return a result set.
  33          * Currently searches a page's current title (page.page_title) and
  34          * latest revision article text (pagecontent.old_text)
  35          *
  36          * @param string $term - Raw search term
  37          * @return PostgresSearchResultSet
  38          * @access public
  39          */
  40         function searchTitle( $term ) {
  41                 $resultSet = $this->db->resultObject( $this->db->query( $this->searchQuery( $term , 'titlevector', 'page_title' )));
  42                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  43         }
  44         function searchText( $term ) {
  45                 $resultSet = $this->db->resultObject( $this->db->query( $this->searchQuery( $term, 'textvector', 'old_text' )));
  46                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  47         }
  48
  49
  50         /*
  51          * Transform the user's search string into a better form for tsearch2
  52         */
  53         function parseQuery( $term ) {
  54
  55                 wfDebug( "parseQuery received: $term" );
  56
  57                 ## No backslashes allowed
  58                 $term = preg_replace('/\\\/', '', $term);
  59
  60                 ## Collapse parens into nearby words:
  61                 $term = preg_replace('/\s*\(\s*/', ' (', $term);
  62                 $term = preg_replace('/\s*\)\s*/', ') ', $term);
  63
  64                 ## Treat colons as word separators:
  65                 $term = preg_replace('/:/', ' ', $term);
  66
  67                 $this->searchTerms = array();
  68                 $m = array();
  69                 $searchstring = '';
  70                 if( preg_match_all('/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  71                         foreach( $m as $terms ) {
  72                                 if (strlen($terms[1])) {
  73                                         $searchstring .= ' & !';
  74                                 }
  75                                 if (strtolower($terms[2]) === 'and') {
  76                                         $searchstring .= ' & ';
  77                                 }
  78                                 else if (strtolower($terms[2]) === 'or' or $terms[2] === '|') {
  79                                         $searchstring .= ' | ';
  80                                 }
  81                                 else if (strtolower($terms[2]) === 'not') {
  82                                         $searchstring .= ' & !';
  83                                 }
  84                                 else {
  85                                         $searchstring .= " & $terms[2]";
  86                                         $safeterm = preg_replace('/\W+/', '', $terms[2]);
  87                                         $this->searchTerms[$safeterm] = $safeterm;
  88                                 }
  89                         }
  90                 }
  91
  92                 ## Strip out leading junk
  93                 $searchstring = preg_replace('/^[\s\&\|]+/', '', $searchstring);
  94
  95                 ## Remove any doubled-up operators
  96                 $searchstring = preg_replace('/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring);
  97
  98                 ## Remove any non-spaced operators (e.g. "Zounds!")
  99                 $searchstring = preg_replace('/([^ ])[\!\&\|]/', "$1", $searchstring);
 100
 101                 ## Remove any trailing operators
 102                 $searchstring = preg_replace('/(?: [\!\&\|])*$/', '', $searchstring);
 103
 104                 ## Quote the whole thing
 105                 $searchstring = $this->db->addQuotes($searchstring);
 106
 107                 wfDebug( "parseQuery returned: $searchstring" );
 108
 109                 return $searchstring;
 110
 111         }
 112
 113         /**
 114          * Construct the full SQL query to do the search.
 115          * @param string $filteredTerm
 116          * @param string $fulltext
 117          * @private
 118          */
 119         function searchQuery( $term, $fulltext, $colname ) {
 120
 121                 $searchstring = $this->parseQuery( $term );
 122
 123                 ## We need a separate query here so gin does not complain about empty searches
 124                 $SQL = "SELECT to_tsquery('default',$searchstring)";
 125                 $res = $this->db->doQuery($SQL);
 126                 if (!$res) {
 127                         ## TODO: Better output (example to catch: one 'two)
 128                         die ("Sorry, that was not a valid search string. Please go back and try again");
 129                 }
 130                 $top = pg_fetch_result($res,0,0);
 131
 132                 if ($top === "") { ## e.g. if only stopwords are used XXX return something better
 133                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score ".
 134                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 135                                 "AND r.rev_text_id = c.old_id AND 1=0";
 136                 }
 137                 else {
 138                         $query = "SELECT page_id, page_namespace, page_title, ".
 139                         "rank($fulltext, to_tsquery('default',$searchstring),5) AS score ".
 140                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 141                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery('default',$searchstring)";
 142                 }
 143
 144                 ## Redirects
 145                 if (! $this->showRedirects)
 146                         $query .= ' AND page_is_redirect = 0'; ## IS FALSE
 147
 148                 ## Namespaces - defaults to 0
 149                 if ( count($this->namespaces) < 1)
 150                         $query .= ' AND page_namespace = 0';
 151                 else {
 152                         $namespaces = implode( ',', $this->namespaces );
 153                         $query .= " AND page_namespace IN ($namespaces)";
 154                 }
 155
 156                 $query .= " ORDER BY score DESC, page_id DESC";
 157
 158                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 159
 160                 wfDebug( "searchQuery returned: $query" );
 161
 162                 return $query;
 163         }
 164
 165         ## Most of the work of these two functions are done automatically via triggers
 166
 167         function update( $pageid, $title, $text ) {
 168                 ## We don't want to index older revisions
 169                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id = ".
 170                                 "(SELECT rev_text_id FROM revision WHERE rev_page = $pageid ".
 171                                 "ORDER BY rev_text_id DESC LIMIT 1 OFFSET 1)";
 172                 $this->db->doQuery($SQL);
 173                 return true;
 174         }
 175
 176         function updateTitle( $id, $title ) {
 177                 return true;
 178         }
 179
 180 } ## end of the SearchPostgres class
 181
 182
 183 class PostgresSearchResult extends SearchResult {
 184         function PostgresSearchResult( $row ) {
 185                 $this->mTitle = Title::makeTitle( $row->page_namespace, $row->page_title );
 186                 $this->score = $row->score;
 187         }
 188         function getScore() {
 189                 return $this->score;
 190         }
 191 }
 192
 193 class PostgresSearchResultSet extends SearchResultSet {
 194         function PostgresSearchResultSet( $resultSet, $terms ) {
 195                 $this->mResultSet = $resultSet;
 196                 $this->mTerms = $terms;
 197         }
 198
 199         function termMatches() {
 200                 return $this->mTerms;
 201         }
 202
 203         function numRows() {
 204                 return $this->mResultSet->numRows();
 205         }
 206
 207         function next() {
 208                 $row = $this->mResultSet->fetchObject();
 209                 if( $row === false ) {
 210                         return false;
 211                 } else {
 212                         return new PostgresSearchResult( $row );
 213                 }
 214         }
 215 }
 216
 217
 218 ?>