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 class SearchPostgres extends SearchEngine {
  25
  26         function SearchPostgres( $db ) {
  27                 $this->db = $db;
  28         }
  29
  30         /**
  31          * Perform a full text search query via tsearch2 and return a result set.
  32          * Currently searches a page's current title (page.page_title) and
  33          * latest revision article text (pagecontent.old_text)
  34          *
  35          * @param string $term - Raw search term
  36          * @return PostgresSearchResultSet
  37          * @access public
  38          */
  39         function searchTitle( $term ) {
  40                 $resultSet = $this->db->resultObject( $this->db->query( $this->searchQuery( $term , 'titlevector', 'page_title' )));
  41                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  42         }
  43         function searchText( $term ) {
  44                 $resultSet = $this->db->resultObject( $this->db->query( $this->searchQuery( $term, 'textvector', 'old_text' )));
  45                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  46         }
  47
  48
  49         /*
  50          * Transform the user's search string into a better form for tsearch2
  51         */
  52         function parseQuery( $term ) {
  53
  54                 wfDebug( "parseQuery received: $term" );
  55
  56                 ## No backslashes allowed
  57                 $term = preg_replace('/\\\/', '', $term);
  58
  59                 ## Collapse parens into nearby words:
  60                 $term = preg_replace('/\s*\(\s*/', ' (', $term);
  61                 $term = preg_replace('/\s*\)\s*/', ') ', $term);
  62
  63                 ## Treat colons as word separators:
  64                 $term = preg_replace('/:/', ' ', $term);
  65
  66                 $searchstring = '';
  67                 $m = array();
  68                 if( preg_match_all('/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  69                         foreach( $m as $terms ) {
  70                                 if (strlen($terms[1])) {
  71                                         $searchstring .= ' & !';
  72                                 }
  73                                 if (strtolower($terms[2]) === 'and') {
  74                                         $searchstring .= ' & ';
  75                                 }
  76                                 else if (strtolower($terms[2]) === 'or' or $terms[2] === '|') {
  77                                         $searchstring .= ' | ';
  78                                 }
  79                                 else if (strtolower($terms[2]) === 'not') {
  80                                         $searchstring .= ' & !';
  81                                 }
  82                                 else {
  83                                         $searchstring .= " & $terms[2]";
  84                                 }
  85                         }
  86                 }
  87
  88                 ## Strip out leading junk
  89                 $searchstring = preg_replace('/^[\s\&\|]+/', '', $searchstring);
  90
  91                 ## Remove any doubled-up operators
  92                 $searchstring = preg_replace('/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring);
  93
  94                 ## Remove any non-spaced operators (e.g. "Zounds!")
  95                 $searchstring = preg_replace('/([^ ])[\!\&\|]/', "$1", $searchstring);
  96
  97                 ## Remove any trailing whitespace or operators
  98                 $searchstring = preg_replace('/[\s\!\&\|]+$/', '', $searchstring);
  99
 100                 ## Remove unnecessary quotes around everything
 101                 $searchstring = preg_replace('/^[\'"](.*)[\'"]$/', "$1", $searchstring);
 102
 103                 ## Quote the whole thing
 104                 $searchstring = $this->db->addQuotes($searchstring);
 105
 106                 wfDebug( "parseQuery returned: $searchstring" );
 107
 108                 return $searchstring;
 109
 110         }
 111
 112         /**
 113          * Construct the full SQL query to do the search.
 114          * @param string $filteredTerm
 115          * @param string $fulltext
 116          * @private
 117          */
 118         function searchQuery( $term, $fulltext, $colname ) {
 119
 120                 $searchstring = $this->parseQuery( $term );
 121
 122                 ## We need a separate query here so gin does not complain about empty searches
 123                 $SQL = "SELECT to_tsquery('default',$searchstring)";
 124                 $res = $this->db->doQuery($SQL);
 125                 if (!$res) {
 126                         ## TODO: Better output (example to catch: one 'two)
 127                         die ("Sorry, that was not a valid search string. Please go back and try again");
 128                 }
 129                 $top = pg_fetch_result($res,0,0);
 130
 131                 if ($top === "") { ## e.g. if only stopwords are used XXX return something better
 132                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score ".
 133                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 134                                 "AND r.rev_text_id = c.old_id AND 1=0";
 135                 }
 136                 else {
 137                         $m = array();
 138                         if( preg_match_all("/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 139                                 foreach( $m as $terms ) {
 140                                         $this->searchTerms[$terms[1]] = $terms[1];
 141                                 }
 142                         }
 143
 144                         $query = "SELECT page_id, page_namespace, page_title, ".
 145                         "rank($fulltext, to_tsquery('default',$searchstring),5) AS score ".
 146                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 147                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery('default',$searchstring)";
 148                 }
 149
 150                 ## Redirects
 151                 if (! $this->showRedirects)
 152                         $query .= ' AND page_is_redirect = 0'; ## IS FALSE
 153
 154                 ## Namespaces - defaults to 0
 155                 if ( count($this->namespaces) < 1)
 156                         $query .= ' AND page_namespace = 0';
 157                 else {
 158                         $namespaces = implode( ',', $this->namespaces );
 159                         $query .= " AND page_namespace IN ($namespaces)";
 160                 }
 161
 162                 $query .= " ORDER BY score DESC, page_id DESC";
 163
 164                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 165
 166                 wfDebug( "searchQuery returned: $query" );
 167
 168                 return $query;
 169         }
 170
 171         ## Most of the work of these two functions are done automatically via triggers
 172
 173         function update( $pageid, $title, $text ) {
 174                 ## We don't want to index older revisions
 175                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id = ".
 176                                 "(SELECT rev_text_id FROM revision WHERE rev_page = $pageid ".
 177                                 "ORDER BY rev_text_id DESC LIMIT 1 OFFSET 1)";
 178                 $this->db->doQuery($SQL);
 179                 return true;
 180         }
 181
 182         function updateTitle( $id, $title ) {
 183                 return true;
 184         }
 185
 186 } ## end of the SearchPostgres class
 187
 188 /**
 189  * @addtogroup Search
 190  */
 191 class PostgresSearchResult extends SearchResult {
 192         function PostgresSearchResult( $row ) {
 193                 $this->mTitle = Title::makeTitle( $row->page_namespace, $row->page_title );
 194                 $this->score = $row->score;
 195         }
 196         function getScore() {
 197                 return $this->score;
 198         }
 199 }
 200
 201 /**
 202  * @addtogroup Search
 203  */
 204 class PostgresSearchResultSet extends SearchResultSet {
 205         function PostgresSearchResultSet( $resultSet, $terms ) {
 206                 $this->mResultSet = $resultSet;
 207                 $this->mTerms = $terms;
 208         }
 209
 210         function termMatches() {
 211                 return $this->mTerms;
 212         }
 213
 214         function numRows() {
 215                 return $this->mResultSet->numRows();
 216         }
 217
 218         function next() {
 219                 $row = $this->mResultSet->fetchObject();
 220                 if( $row === false ) {
 221                         return false;
 222                 } else {
 223                         return new PostgresSearchResult( $row );
 224                 }
 225         }
 226 }
 227
 228
 229 ?>