search/indexer.php

   1 <?php
   2 /**
   3 * Global Search Engine for Moodle
   4 *
   5 * @package search
   6 * @category core
   7 * @subpackage search_engine
   8 * @author Michael Champanis (mchampan) [cynnical@gmail.com], Valery Fremaux [valery.fremaux@club-internet.fr] > 1.8
   9 * @date 2008/03/31
  10 * @version prepared for Moodle 2.0
  11 * @license http://www.gnu.org/copyleft/gpl.html GNU Public License
  12 *
  13 * The indexer logic -
  14 *
  15 * Look through each installed module's or block's search document class file (/search/documents)
  16 * for necessary search functions, and if they're present add the content to the index.
  17 * Repeat this for blocks.
  18 *
  19 * Because the iterator/retrieval functions are now stored in /search/documents/<mod>_document.php,
  20 * /mod/mod/lib.php doesn't have to be modified - and thus the search module becomes quite
  21 * self-sufficient. URL's are now stored in the index, stopping us from needing to require
  22 * the class files to generate a results page.
  23 *
  24 * Along with the index data, each document's summary gets stored in the database
  25 * and synchronised to the index (flat file) via the primary key ('id') which is mapped
  26 * to the 'dbid' field in the index
  27 * */
  28
  29 //this'll take some time, set up the environment
  30 @set_time_limit(0);
  31 @ob_implicit_flush(true);
  32 @ob_end_flush();
  33
  34 /**
  35 * includes and requires
  36 */
  37 require_once('../config.php');
  38 require_once($CFG->dirroot.'/search/lib.php');
  39
  40     ini_set('include_path', $CFG->dirroot.DIRECTORY_SEPARATOR.'search'.PATH_SEPARATOR.ini_get('include_path'));
  41
  42 /// only administrators can index the moodle installation, because access to all pages is required
  43
  44     require_login();
  45
  46     if (empty($CFG->enableglobalsearch)) {
  47         print_error('globalsearchdisabled', 'search');
  48     }
  49
  50     if (!has_capability('moodle/site:config', get_context_instance(CONTEXT_SYSTEM))) {
  51         print_error('beadmin', 'search', get_login_url());
  52     }
  53
  54 /// confirmation flag to prevent accidental reindexing (indexersplash.php is the correct entry point)
  55
  56     $sure = strtolower(optional_param('areyousure', '', PARAM_ALPHA));
  57
  58     if ($sure != 'yes') {
  59         mtrace("<pre>Sorry, you need to confirm indexing via <a href='indexersplash.php'>indexersplash.php</a>"
  60               .". (<a href='index.php'>Back to query page</a>).</pre>");
  61
  62         exit(0);
  63     }
  64
  65 /// check for php5 (lib.php)
  66
  67     //php5 found, continue including php5-only files
  68     //require_once("$CFG->dirroot/search/Zend/Search/Lucene.php");
  69     require_once($CFG->dirroot.'/search/indexlib.php');
  70
  71     mtrace('<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8" /></head><body>');
  72     mtrace('<pre>Server Time: '.date('r',time())."\n");
  73
  74     if (isset($CFG->search_indexer_busy) && $CFG->search_indexer_busy == '1') {
  75         //means indexing was not finished previously
  76         mtrace("Warning: Indexing was not successfully completed last time, restarting.\n");
  77     }
  78
  79 /// turn on busy flag
  80
  81     set_config('search_indexer_busy', '1');
  82
  83     //paths
  84     $index_path = SEARCH_INDEX_PATH;
  85     $index_db_file = "{$CFG->dirroot}/search/db/$CFG->dbtype.sql";
  86     $dbcontrol = new IndexDBControl();
  87
  88 /// setup directory in data root
  89
  90     if (!file_exists($index_path)) {
  91         mtrace("Data directory ($index_path) does not exist, attempting to create.");
  92         if (!mkdir($index_path, $CFG->directorypermissions)) {
  93             search_pexit("Error creating data directory at: $index_path. Please correct.");
  94         }
  95         else {
  96             mtrace("Directory successfully created.");
  97         }
  98     }
  99     else {
 100         mtrace("Using {$index_path} as data directory.");
 101     }
 102
 103     Zend_Search_Lucene_Analysis_Analyzer::setDefault(new Zend_Search_Lucene_Analysis_Analyzer_Common_Utf8_CaseInsensitive());
 104     $index = new Zend_Search_Lucene($index_path, true);
 105
 106 /// New regeneration
 107
 108     mtrace('Deleting old index entries.');
 109     $DB->delete_records(SEARCH_DATABASE_TABLE);
 110
 111 /// begin timer
 112
 113     search_stopwatch();
 114     mtrace("Starting activity modules\n");
 115
 116     //the presence of the required search functions -
 117     // * mod_iterator
 118     // * mod_get_content_for_index
 119     //are the sole basis for including a module in the index at the moment.
 120
 121     $searchables = search_collect_searchables();
 122
 123 /// start indexation
 124
 125     if ($searchables){
 126         foreach ($searchables as $mod) {
 127
 128             mtrace("starting indexing {$mod->name}\n");
 129
 130             $key = 'search_in_'.$mod->name;
 131             if (isset($CFG->$key) && !$CFG->$key) {
 132                 mtrace("module $key has been administratively disabled. Skipping...\n");
 133                 continue;
 134             }
 135
 136             if ($mod->location == 'internal'){
 137                 $class_file = $CFG->dirroot.'/search/documents/'.$mod->name.'_document.php';
 138             } else {
 139                 $class_file = $CFG->dirroot.'/'.$mod->location.'/'.$mod->name.'/search_document.php';
 140             }
 141
 142             if (file_exists($class_file)) {
 143                 include_once($class_file);
 144
 145                 //build function names
 146                 $iter_function = $mod->name.'_iterator';
 147                 $index_function = $mod->name.'_get_content_for_index';
 148                 $counter = 0;
 149                 if (function_exists($index_function) && function_exists($iter_function)) {
 150                     mtrace("Processing module function $index_function ...");
 151                     $sources = $iter_function();
 152                     if ($sources){
 153                         foreach ($sources as $i) {
 154                             $documents = $index_function($i);
 155
 156                             //begin transaction
 157                             if ($documents){
 158                                 foreach($documents as $document) {
 159                                     $counter++;
 160
 161                                     //object to insert into db
 162                                     $dbid = $dbcontrol->addDocument($document);
 163
 164                                     //synchronise db with index
 165                                     $document->addField(Zend_Search_Lucene_Field::Keyword('dbid', $dbid));
 166
 167                                     //add document to index
 168                                     $index->addDocument($document);
 169
 170                                     //commit every x new documents, and print a status message
 171                                     if (($counter % 2000) == 0) {
 172                                         $index->commit();
 173                                         mtrace(".. $counter");
 174                                     }
 175                                 }
 176                             }
 177                             //end transaction
 178                         }
 179                     }
 180
 181                     //commit left over documents, and finish up
 182                     $index->commit();
 183
 184                     mtrace("-- $counter documents indexed");
 185                     mtrace("done.\n");
 186                 }
 187             } else {
 188                mtrace ("No search document found for plugin {$mod->name}. Ignoring.");
 189             }
 190         }
 191     }
 192
 193 /// finished modules
 194
 195     mtrace('Finished activity modules');
 196     search_stopwatch();
 197
 198     mtrace(".<br/><a href='index.php'>Back to query page</a>.");
 199     mtrace('</pre>');
 200
 201 /// finished, turn busy flag off
 202
 203     set_config('search_indexer_busy', '0');
 204
 205 /// mark the time we last updated
 206
 207     set_config('search_indexer_run_date', time());
 208
 209 /// and the index size
 210
 211     set_config('search_index_size', (int)$index->count());
 212
 213 ?>