/home/palakars/workspace-inprogress/code-in-progress/aitools3/aitools3-aq-invertedindex2-java/src/de/aitools/aq/invertedindex/usage/UsingIndexer.java

Go to the documentation of this file.

// Copyright (C) 2011 webis.de. All rights reserved.
package de.aitools.aq.invertedindex.usage;

import java.io.File;

import de.aitools.aq.invertedindex.core.Configuration;
import de.aitools.aq.invertedindex.core.Configuration.KeySorting;
import de.aitools.aq.invertedindex.core.Configuration.ValueSorting;
import de.aitools.aq.invertedindex.core.Indexer;
import de.aitools.aq.invertedindex.core.Properties;
import de.aitools.aq.util.Memory;
import de.aitools.aq.util.Record;
import de.aitools.aq.value.single.SingleLong;

public class UsingIndexer {

        public static void main(String[] args) {
                
                // ---------------------------------------------------------------------
                // Setup the Indexer
                // ---------------------------------------------------------------------
                Configuration config = new Configuration();
                config.setIndexDirectory(new File("/path/to/your/index"));
                config.setKeySorting(KeySorting.UNSORTED);          // optional
                config.setValueSorting(ValueSorting.ASCENDING);     // optional
                config.setMaxMemoryUsage(Memory.MB1024);            // optional
                // If you know the exact or approximated number of records you
                // are going to put into the indexer, you can set this information
                // in the configuration to increase the indexing performance.
                config.setExpectedNumberOfRecords(Short.MAX_VALUE); // optional
                
                // ---------------------------------------------------------------------
                // Instantiate a new Indexer
                // ---------------------------------------------------------------------
                Class<SingleLong> valueClass = SingleLong.class;
                Indexer<SingleLong> indexer = Indexer.open(valueClass, config);
                
                // ---------------------------------------------------------------------
                // Fill the index with a number of records
                // ---------------------------------------------------------------------
                for (int i = 0; i != Short.MAX_VALUE; ++i) {
                        // Create some key
                        String key = "key" + (i % Byte.MAX_VALUE);
                        // Create some value
                        SingleLong value = new SingleLong(i);
                        // Create and put the record
                        indexer.put(new Record<SingleLong>(key, value));
                        
                        // -----------------------------------------------------------------
                        // In some situations, you might get an approximation of the
                        // total number of records expected to be inserted during the
                        // indexing process. It is always a good idea to pass this
                        // information to the indexer to optimize the running procedure.
                        // The overall aim of this estimation is to avoid "rehashing".
                        // Please take a look at your logging output for something like:
                        // "2011-03-17 15:26:25 INFO Rehash current inserted records : 123"
                        // If you know the (exact or approximated) total number of records
                        // before indexing, you can set this value in the configuration.
                        // -----------------------------------------------------------------
                        if (i == 1000) {
                                indexer.setExpectedNumberOfRecords(Short.MAX_VALUE);
                        }
                }
                
                Properties props = indexer.index(); // be patient ...
                
                props.printTo(System.out).println();
        }
}

---