Mailing-List: contact lucene-dev-help@jakarta.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Lucene Developers List" <lucene-dev@jakarta.apache.org>
Message-ID: <EAFF6534E57BD44B9C51F9EDCAAB58750E76F0@exchange.visator.local>
From: Anders Nielsen <anders@visator.com>
Reply-To: Lucene Developers List <lucene-dev@jakarta.apache.org>
To: "'lucene-dev@jakarta.apache.org '" <lucene-dev@jakarta.apache.org>
Subject: RE: PhrasePrefixQuery.java and MultipleTermPositions.java
Date: Wed, 15 May 2002 23:54:59 +0200
MIME-Version: 1.0
Content-Type: multipart/mixed;
	boundary="----_=_NextPart_000_01C1FC5B.289DD3A0"

------_=_NextPart_000_01C1FC5B.289DD3A0
Content-Type: text/plain;
	charset="iso-8859-1"


Here's an updated version of MultipleTermPositions.java.

In this version next() uses adjustTop() instead of pop()-push() as per
Doug's suggestion.

And instead of a SortedSet for keeping positions, it keeps them in a class
that reuses an int[].


regards,
Anders Nielsen


------_=_NextPart_000_01C1FC5B.289DD3A0
Content-Type: application/octet-stream;
	name="MultipleTermPositions.java"
Content-Disposition: attachment;
	filename="MultipleTermPositions.java"

package org.apache.lucene.index;

import java.io.IOException;
import java.util.Arrays;
import java.util.Iterator;
import java.util.LinkedList;
import java.util.List;

import org.apache.lucene.util.PriorityQueue;

public 
class MultipleTermPositions 
implements TermPositions 
{
	private final static
	class TermPositionsQueue
	extends PriorityQueue
	{
		TermPositionsQueue(List termPositions)
		throws IOException
		{
			initialize(termPositions.size());

			Iterator i = termPositions.iterator();
			while (i.hasNext()) 
			{
				TermPositions tp = (TermPositions)i.next();				
				if (tp.next())
					put(tp);
			}			
		}

		final TermPositions peek()
		{
			return (TermPositions)top();
		}		
		
		public final boolean lessThan(Object a, Object b)
		{
			return ((TermPositions)a).doc() < ((TermPositions)b).doc();
		}
	}
	
	private final static
	class IntQueue
	{
		private int _arraySize = 16;

		private int _index = 0;
		private int _lastIndex = 0;
		
		private int[] _array = new int[_arraySize];

		final void add(int i)
		{
			if (_lastIndex == _arraySize)
				growArray();
			
			_array[_lastIndex++] = i;
		}
		
		final int next()
		{
			return _array[_index++];
		}

		final void sort()
		{
			Arrays.sort(_array, _index, _lastIndex);
		}
		
		final void clear()
		{
			_index = 0;
			_lastIndex = 0;
		}
		
		final int size()
		{
			return (_lastIndex-_index);
		}	
		
		private void growArray()
		{
			int[] newArray = new int[_arraySize*2];
			System.arraycopy(_array, 0, newArray, 0, _arraySize);
			_array = newArray;
			_arraySize *= 2;
		}			 		
	}
	
	private int _doc;
	private int _freq;

	private TermPositionsQueue _termPositionsQueue;
	private IntQueue _posList;

	public MultipleTermPositions(IndexReader indexReader, Term[] terms)
	throws IOException
	{
		List termPositions = new LinkedList();
		
		for (int i=0;i<terms.length;i++) 
			termPositions.add(indexReader.termPositions(terms[i]));
		
		_termPositionsQueue = new TermPositionsQueue(termPositions);
		_posList = new IntQueue();
	}			
	
	/**
	 * @see TermDocs#next()
	 */
	public final boolean next() 
	throws IOException 
	{
		if (_termPositionsQueue.size() == 0)
			return false;
			
		_posList.clear();
		_doc = _termPositionsQueue.peek().doc();

		TermPositions tp;		
		do
		{
			tp = _termPositionsQueue.peek();
			
			for (int i=0;i<tp.freq();i++) 
				_posList.add(tp.nextPosition());
	
			if (tp.next())
				_termPositionsQueue.adjustTop();
			else {
				_termPositionsQueue.pop();
				tp.close();
			}
		}
		while (_termPositionsQueue.size() > 0 && _termPositionsQueue.peek().doc() == _doc);
	
		_posList.sort();
		_freq = _posList.size();

		return true;
	}

	/**
	 * @see TermPositions#nextPosition()
	 */
	public final int nextPosition() 
	throws IOException 
	{
		return _posList.next();
	}

	/** 
	 * @see TermDocs#skipTo(int)
	 */
	public final boolean skipTo(int target) 
	throws IOException 
	{
		while (target > _termPositionsQueue.peek().doc())
	    {
			TermPositions tp = (TermPositions)_termPositionsQueue.pop();

			if (tp.skipTo(target)) 
				_termPositionsQueue.put(tp);
			else
				tp.close();
		} 

		return next();
	}		

	/**
	 * @see TermDocs#doc()
	 */
	public final int doc() 
	{
		return _doc;
	}

	/**
	 * @see TermDocs#freq()
	 */
	public final int freq() 
	{
		return _freq;
	}

	/**
	 * @see TermDocs#close()
	 */
	public final void close() 
	throws IOException 
	{
		while (_termPositionsQueue.size() > 0)
			((TermPositions)_termPositionsQueue.pop()).close();
	}

	/**
	 * @see TermDocs#seek(Term)
	 */
	public void seek(Term arg0) 
	throws IOException 
	{
		throw new UnsupportedOperationException();
	}

	/**
	 * @see TermDocs#read(int[], int[])
	 */
	public int read(int[] arg0, int[] arg1) 
	throws IOException 
	{
		throw new UnsupportedOperationException();
	}
}

------_=_NextPart_000_01C1FC5B.289DD3A0
Content-Type: text/plain; charset=us-ascii

--
To unsubscribe, e-mail:   <mailto:lucene-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:lucene-dev-help@jakarta.apache.org>
------_=_NextPart_000_01C1FC5B.289DD3A0--