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                 if( preg_match_all('/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  68                         foreach( $m as $terms ) {
  69                                 if (strlen($terms[1])) {
  70                                         $searchstring .= ' & !';
  71                                 }
  72                                 if (strtolower($terms[2]) === 'and') {
  73                                         $searchstring .= ' & ';
  74                                 }
  75                                 else if (strtolower($terms[2]) === 'or' or $terms[2] === '|') {
  76                                         $searchstring .= ' | ';
  77                                 }
  78                                 else if (strtolower($terms[2]) === 'not') {
  79                                         $searchstring .= ' & !';
  80                                 }
  81                                 else {
  82                                         $searchstring .= " & $terms[2]";
  83                                 }
  84                         }
  85                 }
  86
  87                 ## Strip out leading junk
  88                 $searchstring = preg_replace('/^[\s\&\|]+/', '', $searchstring);
  89
  90                 ## Remove any doubled-up operators
  91                 $searchstring = preg_replace('/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring);
  92
  93                 ## Remove any non-spaced operators (e.g. "Zounds!")
  94                 $searchstring = preg_replace('/([^ ])[\!\&\|]/', "$1", $searchstring);
  95
  96                 ## Remove any trailing whitespace or operators
  97                 $searchstring = preg_replace('/[\s\!\&\|]+$/', '', $searchstring);
  98
  99                 ## Remove unnecessary quotes around everything
 100                 $searchstring = preg_replace('/^[\'"](.*)[\'"]$/', "$1", $searchstring);
 101
 102                 ## Quote the whole thing
 103                 $searchstring = $this->db->addQuotes($searchstring);
 104
 105                 wfDebug( "parseQuery returned: $searchstring" );
 106
 107                 return $searchstring;
 108
 109         }
 110
 111         /**
 112          * Construct the full SQL query to do the search.
 113          * @param string $filteredTerm
 114          * @param string $fulltext
 115          * @private
 116          */
 117         function searchQuery( $term, $fulltext, $colname ) {
 118
 119                 $searchstring = $this->parseQuery( $term );
 120
 121                 ## We need a separate query here so gin does not complain about empty searches
 122                 $SQL = "SELECT to_tsquery('default',$searchstring)";
 123                 $res = $this->db->doQuery($SQL);
 124                 if (!$res) {
 125                         ## TODO: Better output (example to catch: one 'two)
 126                         die ("Sorry, that was not a valid search string. Please go back and try again");
 127                 }
 128                 $top = pg_fetch_result($res,0,0);
 129
 130                 if ($top === "") { ## e.g. if only stopwords are used XXX return something better
 131                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score ".
 132                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 133                                 "AND r.rev_text_id = c.old_id AND 1=0";
 134                 }
 135                 else {
 136                         $m = array();
 137                         if( preg_match_all("/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 138                                 foreach( $m as $terms ) {
 139                                         $this->searchTerms[$terms[1]] = $terms[1];
 140                                 }
 141                         }
 142
 143                         $query = "SELECT page_id, page_namespace, page_title, ".
 144                         "rank($fulltext, to_tsquery('default',$searchstring),5) AS score ".
 145                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 146                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery('default',$searchstring)";
 147                 }
 148
 149                 ## Redirects
 150                 if (! $this->showRedirects)
 151                         $query .= ' AND page_is_redirect = 0'; ## IS FALSE
 152
 153                 ## Namespaces - defaults to 0
 154                 if ( count($this->namespaces) < 1)
 155                         $query .= ' AND page_namespace = 0';
 156                 else {
 157                         $namespaces = implode( ',', $this->namespaces );
 158                         $query .= " AND page_namespace IN ($namespaces)";
 159                 }
 160
 161                 $query .= " ORDER BY score DESC, page_id DESC";
 162
 163                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 164
 165                 wfDebug( "searchQuery returned: $query" );
 166
 167                 return $query;
 168         }
 169
 170         ## Most of the work of these two functions are done automatically via triggers
 171
 172         function update( $pageid, $title, $text ) {
 173                 ## We don't want to index older revisions
 174                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id = ".
 175                                 "(SELECT rev_text_id FROM revision WHERE rev_page = $pageid ".
 176                                 "ORDER BY rev_text_id DESC LIMIT 1 OFFSET 1)";
 177                 $this->db->doQuery($SQL);
 178                 return true;
 179         }
 180
 181         function updateTitle( $id, $title ) {
 182                 return true;
 183         }
 184
 185 } ## end of the SearchPostgres class
 186
 187 /**
 188  * @addtogroup Search
 189  */
 190 class PostgresSearchResult extends SearchResult {
 191         function PostgresSearchResult( $row ) {
 192                 $this->mTitle = Title::makeTitle( $row->page_namespace, $row->page_title );
 193                 $this->score = $row->score;
 194         }
 195         function getScore() {
 196                 return $this->score;
 197         }
 198 }
 199
 200 /**
 201  * @addtogroup Search
 202  */
 203 class PostgresSearchResultSet extends SearchResultSet {
 204         function PostgresSearchResultSet( $resultSet, $terms ) {
 205                 $this->mResultSet = $resultSet;
 206                 $this->mTerms = $terms;
 207         }
 208
 209         function termMatches() {
 210                 return $this->mTerms;
 211         }
 212
 213         function numRows() {
 214                 return $this->mResultSet->numRows();
 215         }
 216
 217         function next() {
 218                 $row = $this->mResultSet->fetchObject();
 219                 if( $row === false ) {
 220                         return false;
 221                 } else {
 222                         return new PostgresSearchResult( $row );
 223                 }
 224         }
 225 }
 226
 227
 228 ?>