All Downloads are FREE. Search and download functionalities are using the official Maven repository.

src.it.unimi.dsi.big.mg4j.index.cluster.ContiguousDocumentalStrategy Maven / Gradle / Ivy

Go to download

MG4J (Managing Gigabytes for Java) is a free full-text search engine for large document collections written in Java. The big version is a fork of the original MG4J that can handle more than 2^31 terms and documents.

The newest version!
package it.unimi.dsi.big.mg4j.index.cluster;

/*		 
 * MG4J: Managing Gigabytes for Java (big)
 *
 * Copyright (C) 2006-2011 Sebastiano Vigna 
 *
 *  This library is free software; you can redistribute it and/or modify it
 *  under the terms of the GNU Lesser General Public License as published by the Free
 *  Software Foundation; either version 3 of the License, or (at your option)
 *  any later version.
 *
 *  This library is distributed in the hope that it will be useful, but
 *  WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
 *  or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public License
 *  for more details.
 *
 *  You should have received a copy of the GNU Lesser General Public License
 *  along with this program; if not, see .
 *
 */

import it.unimi.dsi.util.Properties;

import java.io.Serializable;
import java.util.Arrays;

/** A documental partitioning and clustering strategy that partitions an index into contiguous segments.
 * 
 * 

To partition an index in contiguous parts, you must provide an array * of cutpoints, which define each part. More precisely, given * cutpoints c0,c2,…,ck, * the global index will be partitioned into k local indices containing the documents * from c0 (included) to c1 (excluded), from * c1 (included) to c2 and so on. Note that * necessarily c0=0 and ck=N, * where N is the number of globally indexed documents. * *

The {@link #properties()} method provides two properties, pointerfrom and pointerto, * that contain the first (included) and last (excluded) pointer in the local index. * * @author Alessandro Arrabito * @author Sebastiano Vigna */ public class ContiguousDocumentalStrategy implements DocumentalPartitioningStrategy, DocumentalClusteringStrategy,Serializable{ private static final long serialVersionUID = 0L; /** The cutpoints.*/ private final long[] cutPoint; /** The (cached) number of segments. */ private final int k; /** Creates a new contiguous strategy with the given cutpoints. * *

Note that {@link DocumentalStrategies} has ready-made factory * methods for the common cases. * * @param cutPoint an array of cutpoints (see the class description}. */ public ContiguousDocumentalStrategy( final long... cutPoint ) { if ( cutPoint.length == 0 ) throw new IllegalArgumentException( "Empty cutpoint array" ); if ( cutPoint[ 0 ] != 0 ) throw new IllegalArgumentException( "The first cutpoint must be 0" ); this.cutPoint = cutPoint; this.k = cutPoint.length - 1; } public int numberOfLocalIndices() { return k; } public int localIndex( final long globalPointer ) { if ( globalPointer >= cutPoint[ k ] ) throw new IndexOutOfBoundsException( Long.toString( globalPointer ) ); for ( int i = k; i-- != 0; ) if ( globalPointer >= cutPoint[ i ] ) return i; throw new IndexOutOfBoundsException( Long.toString( globalPointer ) ); } public long localPointer( final long globalPointer ) { return globalPointer - cutPoint[ localIndex( globalPointer ) ]; } public long globalPointer( final int localIndex, final long localPointer ) { return localPointer + cutPoint[ localIndex ]; } public long numberOfDocuments( final int localIndex ) { return cutPoint[ localIndex + 1 ] - cutPoint[ localIndex ]; } public Properties[] properties() { Properties[] properties = new Properties[ k ]; for( int i = 0; i < k; i++ ) { properties[ i ] = new Properties(); properties[ i ].addProperty( "pointerfrom", cutPoint[ i ] ); properties[ i ].addProperty( "pointerto", cutPoint[ i + 1 ] ); } return properties; } public String toString() { return Arrays.toString( cutPoint ); } }





© 2015 - 2025 Weber Informatics LLC | Privacy Policy