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
  30 /**
  31 * includes and requires
  32 */
  33
  34 define('NO_OUTPUT_BUFFERING', true);
  35
  36 require_once('../config.php');
  37 require_once($CFG->dirroot.'/search/lib.php');
  38
  39 //this'll take some time, set up the environment
  40 @set_time_limit(0);
  41
  42     ini_set('include_path', $CFG->dirroot.DIRECTORY_SEPARATOR.'search'.PATH_SEPARATOR.ini_get('include_path'));
  43
  44 /// only administrators can index the moodle installation, because access to all pages is required
  45
  46     require_login();
  47
  48     if (empty($CFG->enableglobalsearch)) {
  49         print_error('globalsearchdisabled', 'search');
  50     }
  51
  52     if (!has_capability('moodle/site:config', get_context_instance(CONTEXT_SYSTEM))) {
  53         print_error('beadmin', 'search', get_login_url());
  54     }
  55
  56 /// confirmation flag to prevent accidental reindexing (indexersplash.php is the correct entry point)
  57
  58     $sure = strtolower(optional_param('areyousure', '', PARAM_ALPHA));
  59
  60     if ($sure != 'yes') {
  61         mtrace("<pre>Sorry, you need to confirm indexing via <a href='indexersplash.php'>indexersplash.php</a>"
  62               .". (<a href='index.php'>Back to query page</a>).</pre>");
  63
  64         exit(0);
  65     }
  66
  67 /// check for php5 (lib.php)
  68
  69     //php5 found, continue including php5-only files
  70     //require_once("$CFG->dirroot/search/Zend/Search/Lucene.php");
  71     require_once($CFG->dirroot.'/search/indexlib.php');
  72
  73     mtrace('<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8" /></head><body>');
  74     mtrace('<pre>Server Time: '.date('r',time())."\n");
  75
  76     if (isset($CFG->search_indexer_busy) && $CFG->search_indexer_busy == '1') {
  77         //means indexing was not finished previously
  78         mtrace("Warning: Indexing was not successfully completed last time, restarting.\n");
  79     }
  80
  81 /// turn on busy flag
  82
  83     set_config('search_indexer_busy', '1');
  84
  85     //paths
  86     $index_path = SEARCH_INDEX_PATH;
  87     $index_db_file = "{$CFG->dirroot}/search/db/$CFG->dbtype.sql";
  88     $dbcontrol = new IndexDBControl();
  89
  90 /// setup directory in data root
  91
  92     if (!file_exists($index_path)) {
  93         mtrace("Data directory ($index_path) does not exist, attempting to create.");
  94         if (!mkdir($index_path, $CFG->directorypermissions)) {
  95             search_pexit("Error creating data directory at: $index_path. Please correct.");
  96         }
  97         else {
  98             mtrace("Directory successfully created.");
  99         }
 100     }
 101     else {
 102         mtrace("Using {$index_path} as data directory.");
 103     }
 104
 105     Zend_Search_Lucene_Analysis_Analyzer::setDefault(new Zend_Search_Lucene_Analysis_Analyzer_Common_Utf8_CaseInsensitive());
 106     $index = new Zend_Search_Lucene($index_path, true);
 107
 108 /// New regeneration
 109
 110     mtrace('Deleting old index entries.');
 111     $DB->delete_records(SEARCH_DATABASE_TABLE);
 112
 113 /// begin timer
 114
 115     search_stopwatch();
 116     mtrace("Starting activity modules\n");
 117
 118     //the presence of the required search functions -
 119     // * mod_iterator
 120     // * mod_get_content_for_index
 121     //are the sole basis for including a module in the index at the moment.
 122
 123     $searchables = search_collect_searchables();
 124
 125 /// start indexation
 126
 127     if ($searchables){
 128         foreach ($searchables as $mod) {
 129
 130             //mark last update times for mods to now.
 131             $indexdatestring = 'search_indexer_update_date_'.$mod->name;
 132             set_config($indexdatestring, time());
 133             $indexdatestring = 'search_indexer_run_date_'.$mod->name;
 134             set_config($indexdatestring, time());
 135
 136             mtrace("starting indexing {$mod->name}\n");
 137
 138             $key = 'search_in_'.$mod->name;
 139             if (isset($CFG->$key) && !$CFG->$key) {
 140                 mtrace("module $key has been administratively disabled. Skipping...\n");
 141                 continue;
 142             }
 143
 144             if ($mod->location == 'internal'){
 145                 $class_file = $CFG->dirroot.'/search/documents/'.$mod->name.'_document.php';
 146             } else {
 147                 $class_file = $CFG->dirroot.'/'.$mod->location.'/'.$mod->name.'/search_document.php';
 148             }
 149
 150             if (file_exists($class_file)) {
 151                 include_once($class_file);
 152
 153                 //build function names
 154                 $iter_function = $mod->name.'_iterator';
 155                 $index_function = $mod->name.'_get_content_for_index';
 156                 $counter = 0;
 157                 if (function_exists($index_function) && function_exists($iter_function)) {
 158                     mtrace("Processing module function $index_function ...");
 159                     $sources = $iter_function();
 160                     if ($sources){
 161                         foreach ($sources as $i) {
 162                             $documents = $index_function($i);
 163
 164                             //begin transaction
 165                             if ($documents){
 166                                 foreach($documents as $document) {
 167                                     $counter++;
 168
 169                                     // temporary fix until MDL-24822 is resolved
 170                                     if ($document->group_id == -1 and $mod->name ='forum') {
 171                                         $document->group_id = 0;
 172                                     }
 173                                     //object to insert into db
 174                                     $dbid = $dbcontrol->addDocument($document);
 175
 176                                     //synchronise db with index
 177                                     $document->addField(Zend_Search_Lucene_Field::Keyword('dbid', $dbid));
 178
 179                                     //add document to index
 180                                     $index->addDocument($document);
 181
 182                                     //commit every x new documents, and print a status message
 183                                     if (($counter % 2000) == 0) {
 184                                         $index->commit();
 185                                         mtrace(".. $counter");
 186                                     }
 187                                 }
 188                             }
 189                             //end transaction
 190                         }
 191                     }
 192
 193                     //commit left over documents, and finish up
 194                     $index->commit();
 195
 196                     mtrace("-- $counter documents indexed");
 197                     mtrace("done.\n");
 198                 }
 199             } else {
 200                mtrace ("No search document found for plugin {$mod->name}. Ignoring.");
 201             }
 202         }
 203     }
 204
 205 /// finished modules
 206
 207     mtrace('Finished activity modules');
 208     search_stopwatch();
 209
 210     mtrace(".<br/><a href='index.php'>Back to query page</a>.");
 211     mtrace('</pre>');
 212
 213 /// finished, turn busy flag off
 214
 215     set_config('search_indexer_busy', '0');
 216
 217 /// mark the time we last updated
 218
 219     set_config('search_indexer_run_date', time());
 220
 221 /// and the index size
 222
 223     set_config('search_index_size', (int)$index->count());
 224
 225 ?>