view src/org/tmatesoft/hg/internal/RevisionLookup.java @ 709:497e697636fc

Report merged lines as changed block if possible, not as a sequence of added/deleted blocks. To facilitate access to merge parent lines AddBlock got mergeLineAt() method that reports index of the line in the second parent (if any), while insertedAt() has been changed to report index in the first parent always
author Artem Tikhomirov <tikhomirov.artem@gmail.com>
date Wed, 21 Aug 2013 16:23:27 +0200
parents 6526d8adbc0f
children
line wrap: on
line source
/*
 * Copyright (c) 2013 TMate Software Ltd
 *  
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; version 2 of the License.
 *
 * This program 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 General Public License for more details.
 *
 * For information on how to redistribute this software under
 * the terms of a license other than GNU General Public License
 * contact TMate Software at support@hg4j.com
 */
package org.tmatesoft.hg.internal;

import static org.tmatesoft.hg.repo.HgRepository.BAD_REVISION;

import java.util.Arrays;

import org.tmatesoft.hg.core.Nodeid;
import org.tmatesoft.hg.repo.HgInvalidControlFileException;
import org.tmatesoft.hg.repo.HgInvalidRevisionException;
import org.tmatesoft.hg.repo.HgRevisionMap;
import org.tmatesoft.hg.repo.HgRuntimeException;

/**
 * Lite alternative to {@link HgRevisionMap}, to speed up nodeid to index conversion without consuming too much memory.
 * E.g. for a 100k revisions, {@link HgRevisionMap} consumes 3 * (N * sizeof(int)) for indexes plus 48 bytes per 
 * Nodeid instance, total (12+48)*N = 6 mb of memory. {RevisionLookup} instead keeps only Nodeid hashes, (N * sizeof(int) = 400 kb),
 * but is slower in lookup, O(N/2) to find potential match plus disk read operatin (or few, in an unlikely case of hash collisions).
 * 
 * @author Artem Tikhomirov
 * @author TMate Software Ltd.
 */
public class RevisionLookup implements RevlogStream.Inspector {
	
	private final RevlogStream content;
	private int[] nodeidHashes;

	public RevisionLookup(RevlogStream stream) {
		assert stream != null;
		content = stream;
	}
	
	public static RevisionLookup createFor(RevlogStream stream) throws HgRuntimeException {
		RevisionLookup rv = new RevisionLookup(stream);
		int revCount = stream.revisionCount();
		rv.prepare(revCount);
		if (revCount > 0) {
			stream.iterate(0, revCount - 1, false, rv);
		}
		return rv;
	}

	public void prepare(int count) {
		nodeidHashes = new int[count];
		Arrays.fill(nodeidHashes, BAD_REVISION);
	}
	public void next(int index, byte[] nodeid) {
		nodeidHashes[index] = Nodeid.hashCode(nodeid);
	}
	public void next(int index, Nodeid nodeid) {
		nodeidHashes[index] = nodeid.hashCode();
	}
	public int findIndex(Nodeid nodeid) throws HgInvalidControlFileException, HgInvalidRevisionException {
		final int hash = nodeid.hashCode();
		for (int i = 0; i < nodeidHashes.length; i++) {
			if (nodeidHashes[i] == hash) {
				byte[] nodeidAtI = content.nodeid(i);
				if (nodeid.equalsTo(nodeidAtI)) {
					return i;
				}
			}
			// else: false match (only 4 head bytes matched, continue loop
		}
		return BAD_REVISION;
	}

	public void next(int revisionIndex, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, DataAccess data) {
		next(revisionIndex, nodeid);
	}
}