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                 $q = $this->searchQuery( $term , 'titlevector', 'page_title' );
  41                 $olderror = error_reporting(E_ERROR);
  42                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  43                 error_reporting($olderror);
  44                 if (!$resultSet) {
  45                         // Needed for "Query requires full scan, GIN doesn't support it"
  46                         return new SearchResultTooMany();
  47                 }
  48                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  49         }
  50         function searchText( $term ) {
  51                 $q = $this->searchQuery( $term, 'textvector', 'old_text' );
  52                 $olderror = error_reporting(E_ERROR);
  53                 $resultSet = $this->db->resultObject( $this->db->query( $q, 'SearchPostgres', true ) );
  54                 error_reporting($olderror);
  55                 if (!$resultSet) {
  56                         return new SearchResultTooMany();
  57                 }
  58                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  59         }
  60
  61
  62         /*
  63          * Transform the user's search string into a better form for tsearch2
  64         */
  65         function parseQuery( $term ) {
  66
  67                 wfDebug( "parseQuery received: $term" );
  68
  69                 ## No backslashes allowed
  70                 $term = preg_replace('/\\\/', '', $term);
  71
  72                 ## Collapse parens into nearby words:
  73                 $term = preg_replace('/\s*\(\s*/', ' (', $term);
  74                 $term = preg_replace('/\s*\)\s*/', ') ', $term);
  75
  76                 ## Treat colons as word separators:
  77                 $term = preg_replace('/:/', ' ', $term);
  78
  79                 $searchstring = '';
  80                 $m = array();
  81                 if( preg_match_all('/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  82                         foreach( $m as $terms ) {
  83                                 if (strlen($terms[1])) {
  84                                         $searchstring .= ' & !';
  85                                 }
  86                                 if (strtolower($terms[2]) === 'and') {
  87                                         $searchstring .= ' & ';
  88                                 }
  89                                 else if (strtolower($terms[2]) === 'or' or $terms[2] === '|') {
  90                                         $searchstring .= ' | ';
  91                                 }
  92                                 else if (strtolower($terms[2]) === 'not') {
  93                                         $searchstring .= ' & !';
  94                                 }
  95                                 else {
  96                                         $searchstring .= " & $terms[2]";
  97                                 }
  98                         }
  99                 }
 100
 101                 ## Strip out leading junk
 102                 $searchstring = preg_replace('/^[\s\&\|]+/', '', $searchstring);
 103
 104                 ## Remove any doubled-up operators
 105                 $searchstring = preg_replace('/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring);
 106
 107                 ## Remove any non-spaced operators (e.g. "Zounds!")
 108                 $searchstring = preg_replace('/([^ ])[\!\&\|]/', "$1", $searchstring);
 109
 110                 ## Remove any trailing whitespace or operators
 111                 $searchstring = preg_replace('/[\s\!\&\|]+$/', '', $searchstring);
 112
 113                 ## Remove unnecessary quotes around everything
 114                 $searchstring = preg_replace('/^[\'"](.*)[\'"]$/', "$1", $searchstring);
 115
 116                 ## Quote the whole thing
 117                 $searchstring = $this->db->addQuotes($searchstring);
 118
 119                 wfDebug( "parseQuery returned: $searchstring" );
 120
 121                 return $searchstring;
 122
 123         }
 124
 125         /**
 126          * Construct the full SQL query to do the search.
 127          * @param string $filteredTerm
 128          * @param string $fulltext
 129          * @private
 130          */
 131         function searchQuery( $term, $fulltext, $colname ) {
 132                 global $wgDBversion;
 133
 134                 if ( !isset( $wgDBversion ) ) {
 135                         $this->db->getServerVersion();
 136                         $wgDBversion = $this->db->numeric_version;
 137                 }
 138                 $prefix = $wgDBversion < 8.3 ? "'default'," : '';
 139
 140                 $searchstring = $this->parseQuery( $term );
 141
 142                 ## We need a separate query here so gin does not complain about empty searches
 143                 $SQL = "SELECT to_tsquery($prefix $searchstring)";
 144                 $res = $this->db->doQuery($SQL);
 145                 if (!$res) {
 146                         ## TODO: Better output (example to catch: one 'two)
 147                         die ("Sorry, that was not a valid search string. Please go back and try again");
 148                 }
 149                 $top = pg_fetch_result($res,0,0);
 150
 151                 if ($top === "") { ## e.g. if only stopwords are used XXX return something better
 152                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score ".
 153                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 154                                 "AND r.rev_text_id = c.old_id AND 1=0";
 155                 }
 156                 else {
 157                         $m = array();
 158                         if( preg_match_all("/'([^']+)'/", $top, $m, PREG_SET_ORDER ) ) {
 159                                 foreach( $m as $terms ) {
 160                                         $this->searchTerms[$terms[1]] = $terms[1];
 161                                 }
 162                         }
 163
 164                         $rankscore = $wgDBversion > 8.2 ? 5 : 1;
 165                         $rank = $wgDBversion < 8.3 ? 'rank' : 'ts_rank';
 166                         $query = "SELECT page_id, page_namespace, page_title, ".
 167                         "$rank($fulltext, to_tsquery($prefix $searchstring), $rankscore) AS score ".
 168                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 169                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery($prefix $searchstring)";
 170                 }
 171
 172                 ## Redirects
 173                 if (! $this->showRedirects)
 174                         $query .= ' AND page_is_redirect = 0';
 175
 176                 ## Namespaces - defaults to 0
 177                 if ( count($this->namespaces) < 1)
 178                         $query .= ' AND page_namespace = 0';
 179                 else {
 180                         $namespaces = implode( ',', $this->namespaces );
 181                         $query .= " AND page_namespace IN ($namespaces)";
 182                 }
 183
 184                 $query .= " ORDER BY score DESC, page_id DESC";
 185
 186                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 187
 188                 wfDebug( "searchQuery returned: $query" );
 189
 190                 return $query;
 191         }
 192
 193         ## Most of the work of these two functions are done automatically via triggers
 194
 195         function update( $pageid, $title, $text ) {
 196                 ## We don't want to index older revisions
 197                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id = ".
 198                                 "(SELECT rev_text_id FROM revision WHERE rev_page = $pageid ".
 199                                 "ORDER BY rev_text_id DESC LIMIT 1 OFFSET 1)";
 200                 $this->db->doQuery($SQL);
 201                 return true;
 202         }
 203
 204         function updateTitle( $id, $title ) {
 205                 return true;
 206         }
 207
 208 } ## end of the SearchPostgres class
 209
 210 /**
 211  * @addtogroup Search
 212  */
 213 class PostgresSearchResult extends SearchResult {
 214         function PostgresSearchResult( $row ) {
 215                 $this->mTitle = Title::makeTitle( $row->page_namespace, $row->page_title );
 216                 $this->score = $row->score;
 217         }
 218         function getScore() {
 219                 return $this->score;
 220         }
 221 }
 222
 223 /**
 224  * @addtogroup Search
 225  */
 226 class PostgresSearchResultSet extends SearchResultSet {
 227         function PostgresSearchResultSet( $resultSet, $terms ) {
 228                 $this->mResultSet = $resultSet;
 229                 $this->mTerms = $terms;
 230         }
 231
 232         function termMatches() {
 233                 return $this->mTerms;
 234         }
 235
 236         function numRows() {
 237                 return $this->mResultSet->numRows();
 238         }
 239
 240         function next() {
 241                 $row = $this->mResultSet->fetchObject();
 242                 if( $row === false ) {
 243                         return false;
 244                 } else {
 245                         return new PostgresSearchResult( $row );
 246                 }
 247         }
 248 }
 249
 250
 251