kitaev@213: /*
kitaev@213:  * Copyright (c) 2010-2011 TMate Software Ltd
kitaev@213:  *  
kitaev@213:  * This program is free software; you can redistribute it and/or modify
kitaev@213:  * it under the terms of the GNU General Public License as published by
kitaev@213:  * the Free Software Foundation; version 2 of the License.
kitaev@213:  *
kitaev@213:  * This program is distributed in the hope that it will be useful,
kitaev@213:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
kitaev@213:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
kitaev@213:  * GNU General Public License for more details.
kitaev@213:  *
kitaev@213:  * For information on how to redistribute this software under
kitaev@213:  * the terms of a license other than GNU General Public License
kitaev@213:  * contact TMate Software at support@hg4j.com
kitaev@213:  */
kitaev@213: package org.tmatesoft.hg.repo;
kitaev@213: 
kitaev@213: import static org.tmatesoft.hg.repo.HgRepository.BAD_REVISION;
kitaev@213: import static org.tmatesoft.hg.repo.HgRepository.TIP;
kitaev@213: 
kitaev@213: import java.io.IOException;
kitaev@213: import java.nio.ByteBuffer;
kitaev@213: import java.util.Arrays;
kitaev@213: import java.util.Collection;
kitaev@213: import java.util.HashSet;
kitaev@213: import java.util.LinkedList;
kitaev@213: import java.util.List;
kitaev@213: 
kitaev@213: import org.tmatesoft.hg.core.HgBadStateException;
kitaev@213: import org.tmatesoft.hg.core.HgException;
kitaev@213: import org.tmatesoft.hg.core.Nodeid;
kitaev@213: import org.tmatesoft.hg.internal.DataAccess;
kitaev@213: import org.tmatesoft.hg.internal.RevlogStream;
kitaev@213: import org.tmatesoft.hg.util.ByteChannel;
kitaev@213: import org.tmatesoft.hg.util.CancelSupport;
kitaev@213: import org.tmatesoft.hg.util.CancelledException;
kitaev@213: import org.tmatesoft.hg.util.ProgressSupport;
kitaev@213: 
kitaev@213: 
kitaev@213: /**
kitaev@213:  * Base class for all Mercurial entities that are serialized in a so called revlog format (changelog, manifest, data files).
kitaev@213:  * 
kitaev@213:  * Implementation note:
kitaev@213:  * Hides actual actual revlog stream implementation and its access methods (i.e. RevlogStream.Inspector), iow shall not expose anything internal
kitaev@213:  * in public methods.
kitaev@213:  *   
kitaev@213:  * @author Artem Tikhomirov
kitaev@213:  * @author TMate Software Ltd.
kitaev@213:  */
kitaev@213: abstract class Revlog {
kitaev@213: 
kitaev@213: 	private final HgRepository repo;
kitaev@213: 	protected final RevlogStream content;
kitaev@213: 
kitaev@213: 	protected Revlog(HgRepository hgRepo, RevlogStream contentStream) {
kitaev@213: 		if (hgRepo == null) {
kitaev@213: 			throw new IllegalArgumentException();
kitaev@213: 		}
kitaev@213: 		if (contentStream == null) {
kitaev@213: 			throw new IllegalArgumentException();
kitaev@213: 		}
kitaev@213: 		repo = hgRepo;
kitaev@213: 		content = contentStream;
kitaev@213: 	}
kitaev@213: 	
kitaev@213: 	// invalid Revlog
kitaev@213: 	protected Revlog(HgRepository hgRepo) {
kitaev@213: 		repo = hgRepo;
kitaev@213: 		content = null;
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	public final HgRepository getRepo() {
kitaev@213: 		return repo;
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	public final int getRevisionCount() {
kitaev@213: 		return content.revisionCount();
kitaev@213: 	}
kitaev@213: 	
kitaev@213: 	public final int getLastRevision() {
kitaev@213: 		return content.revisionCount() - 1;
kitaev@213: 	}
kitaev@213: 	
kitaev@213: 	public final Nodeid getRevision(int revision) {
kitaev@213: 		// XXX cache nodeids?
kitaev@213: 		return Nodeid.fromBinary(content.nodeid(revision), 0);
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	public final int getLocalRevision(Nodeid nid) {
kitaev@213: 		int revision = content.findLocalRevisionNumber(nid);
kitaev@213: 		if (revision == BAD_REVISION) {
kitaev@213: 			throw new IllegalArgumentException(String.format("%s doesn't represent a revision of %s", nid.toString(), this /*XXX HgDataFile.getPath might be more suitable here*/));
kitaev@213: 		}
kitaev@213: 		return revision;
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	// Till now, i follow approach that NULL nodeid is never part of revlog
kitaev@213: 	public final boolean isKnown(Nodeid nodeid) {
kitaev@213: 		final int rn = content.findLocalRevisionNumber(nodeid);
kitaev@213: 		if (Integer.MIN_VALUE == rn) {
kitaev@213: 			return false;
kitaev@213: 		}
kitaev@213: 		if (rn < 0 || rn >= content.revisionCount()) {
kitaev@213: 			// Sanity check
kitaev@213: 			throw new IllegalStateException();
kitaev@213: 		}
kitaev@213: 		return true;
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	/**
kitaev@213: 	 * Access to revision data as is (decompressed, but otherwise unprocessed, i.e. not parsed for e.g. changeset or manifest entries) 
kitaev@213: 	 * @param nodeid
kitaev@213: 	 */
kitaev@213: 	protected void rawContent(Nodeid nodeid, ByteChannel sink) throws HgException, IOException, CancelledException {
kitaev@213: 		rawContent(getLocalRevision(nodeid), sink);
kitaev@213: 	}
kitaev@213: 	
kitaev@213: 	/**
kitaev@213: 	 * @param revision - repo-local index of this file change (not a changelog revision number!)
kitaev@213: 	 */
kitaev@213: 	protected void rawContent(int revision, ByteChannel sink) throws HgException, IOException, CancelledException {
kitaev@213: 		if (sink == null) {
kitaev@213: 			throw new IllegalArgumentException();
kitaev@213: 		}
kitaev@213: 		ContentPipe insp = new ContentPipe(sink, 0);
kitaev@213: 		insp.checkCancelled();
kitaev@213: 		content.iterate(revision, revision, true, insp);
kitaev@213: 		insp.checkFailed();
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	/**
kitaev@213: 	 * XXX perhaps, return value Nodeid[2] and boolean needNodeids is better (and higher level) API for this query?
kitaev@213: 	 * 
kitaev@213: 	 * @param revision - revision to query parents, or {@link HgRepository#TIP}
kitaev@213: 	 * @param parentRevisions - int[2] to get local revision numbers of parents (e.g. {6, -1})
kitaev@213: 	 * @param parent1 - byte[20] or null, if parent's nodeid is not needed
kitaev@213: 	 * @param parent2 - byte[20] or null, if second parent's nodeid is not needed
kitaev@213: 	 * @return
kitaev@213: 	 */
kitaev@213: 	public void parents(int revision, int[] parentRevisions, byte[] parent1, byte[] parent2) {
kitaev@213: 		if (revision != TIP && !(revision >= 0 && revision < content.revisionCount())) {
kitaev@213: 			throw new IllegalArgumentException(String.valueOf(revision));
kitaev@213: 		}
kitaev@213: 		if (parentRevisions == null || parentRevisions.length < 2) {
kitaev@213: 			throw new IllegalArgumentException(String.valueOf(parentRevisions));
kitaev@213: 		}
kitaev@213: 		if (parent1 != null && parent1.length < 20) {
kitaev@213: 			throw new IllegalArgumentException(parent1.toString());
kitaev@213: 		}
kitaev@213: 		if (parent2 != null && parent2.length < 20) {
kitaev@213: 			throw new IllegalArgumentException(parent2.toString());
kitaev@213: 		}
kitaev@213: 		class ParentCollector implements RevlogStream.Inspector {
kitaev@213: 			public int p1 = -1;
kitaev@213: 			public int p2 = -1;
kitaev@213: 			public byte[] nodeid;
kitaev@213: 			
kitaev@213: 			public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, DataAccess da) {
kitaev@213: 				p1 = parent1Revision;
kitaev@213: 				p2 = parent2Revision;
kitaev@213: 				this.nodeid = new byte[20];
kitaev@213: 				// nodeid arg now comes in 32 byte from (as in file format description), however upper 12 bytes are zeros.
kitaev@213: 				System.arraycopy(nodeid, nodeid.length > 20 ? nodeid.length - 20 : 0, this.nodeid, 0, 20);
kitaev@213: 			}
kitaev@213: 		};
kitaev@213: 		ParentCollector pc = new ParentCollector();
kitaev@213: 		content.iterate(revision, revision, false, pc);
kitaev@213: 		parentRevisions[0] = pc.p1;
kitaev@213: 		parentRevisions[1] = pc.p2;
kitaev@213: 		if (parent1 != null) {
kitaev@213: 			if (parentRevisions[0] == -1) {
kitaev@213: 				Arrays.fill(parent1, 0, 20, (byte) 0);
kitaev@213: 			} else {
kitaev@213: 				content.iterate(parentRevisions[0], parentRevisions[0], false, pc);
kitaev@213: 				System.arraycopy(pc.nodeid, 0, parent1, 0, 20);
kitaev@213: 			}
kitaev@213: 		}
kitaev@213: 		if (parent2 != null) {
kitaev@213: 			if (parentRevisions[1] == -1) {
kitaev@213: 				Arrays.fill(parent2, 0, 20, (byte) 0);
kitaev@213: 			} else {
kitaev@213: 				content.iterate(parentRevisions[1], parentRevisions[1], false, pc);
kitaev@213: 				System.arraycopy(pc.nodeid, 0, parent2, 0, 20);
kitaev@213: 			}
kitaev@213: 		}
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	/*
kitaev@213: 	 * XXX think over if it's better to do either:
kitaev@213: 	 * pw = getChangelog().new ParentWalker(); pw.init() and pass pw instance around as needed
kitaev@213: 	 * or
kitaev@213: 	 * add Revlog#getParentWalker(), static class, make cons() and #init package-local, and keep SoftReference to allow walker reuse.
kitaev@213: 	 * 
kitaev@213: 	 *  and yes, walker is not a proper name
kitaev@213: 	 */
kitaev@213: 	public final class ParentWalker {
kitaev@213: 
kitaev@213: 		
kitaev@213: 		private Nodeid[] sequential; // natural repository order, childrenOf rely on ordering
kitaev@213: 		private Nodeid[] sorted; // for binary search
kitaev@213: 		private int[] sorted2natural;
kitaev@213: 		private Nodeid[] firstParent;
kitaev@213: 		private Nodeid[] secondParent;
kitaev@213: 
kitaev@213: 		// Nodeid instances shall be shared between all arrays
kitaev@213: 
kitaev@213: 		public ParentWalker() {
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		public HgRepository getRepo() {
kitaev@213: 			return Revlog.this.getRepo();
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		public void init() {
kitaev@213: 			final RevlogStream stream = Revlog.this.content;
kitaev@213: 			final int revisionCount = stream.revisionCount();
kitaev@213: 			firstParent = new Nodeid[revisionCount];
kitaev@213: 			// although branches/merges are less frequent, and most of secondParent would be -1/null, some sort of 
kitaev@213: 			// SparseOrderedList might be handy, provided its inner structures do not overweight simplicity of an array
kitaev@213: 			secondParent = new Nodeid[revisionCount];
kitaev@213: 			//
kitaev@213: 			sequential = new Nodeid[revisionCount];
kitaev@213: 			sorted = new Nodeid[revisionCount];
kitaev@213: 		
kitaev@213: 			RevlogStream.Inspector insp = new RevlogStream.Inspector() {
kitaev@213: 				
kitaev@213: 				int ix = 0;
kitaev@213: 				public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, DataAccess da) {
kitaev@213: 					if (ix != revisionNumber) {
kitaev@213: 						// XXX temp code, just to make sure I understand what's going on here
kitaev@213: 						throw new IllegalStateException();
kitaev@213: 					}
kitaev@213: 					if (parent1Revision >= revisionNumber || parent2Revision >= revisionNumber) {
kitaev@213: 						throw new IllegalStateException(); // sanity, revisions are sequential
kitaev@213: 					}
kitaev@213: 					final Nodeid nid = new Nodeid(nodeid, true);
kitaev@213: 					sequential[ix] = sorted[ix] = nid;
kitaev@213: 					if (parent1Revision != -1) {
kitaev@213: 						assert parent1Revision < ix;
kitaev@213: 						firstParent[ix] = sequential[parent1Revision];
kitaev@213: 					}
kitaev@213: 					if (parent2Revision != -1) { // revlog of DataAccess.java has p2 set when p1 is -1
kitaev@213: 						assert parent2Revision < ix;
kitaev@213: 						secondParent[ix] = sequential[parent2Revision];
kitaev@213: 					}
kitaev@213: 					ix++;
kitaev@213: 				}
kitaev@213: 			};
kitaev@213: 			stream.iterate(0, TIP, false, insp);
kitaev@213: 			Arrays.sort(sorted);
kitaev@213: 			sorted2natural = new int[revisionCount];
kitaev@213: 			for (int i = 0; i < revisionCount; i++) {
kitaev@213: 				Nodeid n = sequential[i];
kitaev@213: 				int x = Arrays.binarySearch(sorted, n);
kitaev@213: 				assertSortedIndex(x);
kitaev@213: 				sorted2natural[x] = i;
kitaev@213: 			}
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		private void assertSortedIndex(int x) {
kitaev@213: 			if (x < 0) {
kitaev@213: 				throw new HgBadStateException();
kitaev@213: 			}
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		// FIXME need to decide whether Nodeid(00 * 20) is always known or not
kitaev@213: 		// right now Nodeid.NULL is not recognized as known if passed to this method, 
kitaev@213: 		// caller is supposed to make explicit check 
kitaev@213: 		public boolean knownNode(Nodeid nid) {
kitaev@213: 			return Arrays.binarySearch(sorted, nid) >= 0;
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		/**
kitaev@213: 		 * null if none. only known nodes (as per #knownNode) are accepted as arguments
kitaev@213: 		 */
kitaev@213: 		public Nodeid firstParent(Nodeid nid) {
kitaev@213: 			int x = Arrays.binarySearch(sorted, nid);
kitaev@213: 			assertSortedIndex(x);
kitaev@213: 			int i = sorted2natural[x];
kitaev@213: 			return firstParent[i];
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		// never null, Nodeid.NULL if none known
kitaev@213: 		public Nodeid safeFirstParent(Nodeid nid) {
kitaev@213: 			Nodeid rv = firstParent(nid);
kitaev@213: 			return rv == null ? Nodeid.NULL : rv;
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		public Nodeid secondParent(Nodeid nid) {
kitaev@213: 			int x = Arrays.binarySearch(sorted, nid);
kitaev@213: 			assertSortedIndex(x);
kitaev@213: 			int i = sorted2natural[x];
kitaev@213: 			return secondParent[i];
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		public Nodeid safeSecondParent(Nodeid nid) {
kitaev@213: 			Nodeid rv = secondParent(nid);
kitaev@213: 			return rv == null ? Nodeid.NULL : rv;
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		public boolean appendParentsOf(Nodeid nid, Collection<Nodeid> c) {
kitaev@213: 			int x = Arrays.binarySearch(sorted, nid);
kitaev@213: 			assertSortedIndex(x);
kitaev@213: 			int i = sorted2natural[x];
kitaev@213: 			Nodeid p1 = firstParent[i];
kitaev@213: 			boolean modified = false;
kitaev@213: 			if (p1 != null) {
kitaev@213: 				modified = c.add(p1);
kitaev@213: 			}
kitaev@213: 			Nodeid p2 = secondParent[i];
kitaev@213: 			if (p2 != null) {
kitaev@213: 				modified = c.add(p2) || modified;
kitaev@213: 			}
kitaev@213: 			return modified;
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		// XXX alternative (and perhaps more reliable) approach would be to make a copy of allNodes and remove 
kitaev@213: 		// nodes, their parents and so on.
kitaev@213: 		
kitaev@213: 		// @return ordered collection of all children rooted at supplied nodes. Nodes shall not be descendants of each other!
kitaev@213: 		// Nodeids shall belong to this revlog
kitaev@213: 		public List<Nodeid> childrenOf(List<Nodeid> roots) {
kitaev@213: 			HashSet<Nodeid> parents = new HashSet<Nodeid>();
kitaev@213: 			LinkedList<Nodeid> result = new LinkedList<Nodeid>();
kitaev@213: 			int earliestRevision = Integer.MAX_VALUE;
kitaev@213: 			assert sequential.length == firstParent.length && firstParent.length == secondParent.length;
kitaev@213: 			// first, find earliest index of roots in question, as there's  no sense 
kitaev@213: 			// to check children among nodes prior to branch's root node
kitaev@213: 			for (Nodeid r : roots) {
kitaev@213: 				int x = Arrays.binarySearch(sorted, r);
kitaev@213: 				assertSortedIndex(x);
kitaev@213: 				int i = sorted2natural[x];
kitaev@213: 				if (i < earliestRevision) {
kitaev@213: 					earliestRevision = i;
kitaev@213: 				}
kitaev@213: 				parents.add(sequential[i]); // add canonical instance in hope equals() is bit faster when can do a ==
kitaev@213: 			}
kitaev@213: 			for (int i = earliestRevision + 1; i < sequential.length; i++) {
kitaev@213: 				if (parents.contains(firstParent[i]) || parents.contains(secondParent[i])) {
kitaev@213: 					parents.add(sequential[i]); // to find next child
kitaev@213: 					result.add(sequential[i]);
kitaev@213: 				}
kitaev@213: 			}
kitaev@213: 			return result;
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		/**
kitaev@213: 		 * @param nid possibly parent node, shall be {@link #knownNode(Nodeid) known} in this revlog.
kitaev@213: 		 * @return <code>true</code> if there's any node in this revlog that has specified node as one of its parents. 
kitaev@213: 		 */
kitaev@213: 		public boolean hasChildren(Nodeid nid) {
kitaev@213: 			int x = Arrays.binarySearch(sorted, nid);
kitaev@213: 			assertSortedIndex(x);
kitaev@213: 			int i = sorted2natural[x];
kitaev@213: 			assert firstParent.length == secondParent.length; // just in case later I implement sparse array for secondParent
kitaev@213: 			assert firstParent.length == sequential.length;
kitaev@213: 			// to use == instead of equals, take the same Nodeid instance we used to fill all the arrays.
kitaev@213: 			final Nodeid canonicalNode = sequential[i];
kitaev@213: 			i++; // no need to check node itself. child nodes may appear in sequential only after revision in question
kitaev@213: 			for (; i < sequential.length; i++) {
kitaev@213: 				// FIXME likely, not very effective. 
kitaev@213: 				// May want to optimize it with another (Tree|Hash)Set, created on demand on first use, 
kitaev@213: 				// however, need to be careful with memory usage
kitaev@213: 				if (firstParent[i] == canonicalNode || secondParent[i] == canonicalNode) {
kitaev@213: 					return true;
kitaev@213: 				}
kitaev@213: 			}
kitaev@213: 			return false;
kitaev@213: 		}
kitaev@213: 	}
kitaev@213: 
kitaev@213: 	protected static class ContentPipe implements RevlogStream.Inspector, CancelSupport {
kitaev@213: 		private final ByteChannel sink;
kitaev@213: 		private final CancelSupport cancelSupport;
kitaev@213: 		private Exception failure;
kitaev@213: 		private final int offset;
kitaev@213: 
kitaev@213: 		/**
kitaev@213: 		 * @param _sink - cannot be <code>null</code>
kitaev@213: 		 * @param seekOffset - when positive, orders to pipe bytes to the sink starting from specified offset, not from the first byte available in DataAccess
kitaev@213: 		 */
kitaev@213: 		public ContentPipe(ByteChannel _sink, int seekOffset) {
kitaev@213: 			assert _sink != null;
kitaev@213: 			sink = _sink;
kitaev@213: 			cancelSupport = CancelSupport.Factory.get(_sink);
kitaev@213: 			offset = seekOffset;
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		protected void prepare(int revisionNumber, DataAccess da) throws HgException, IOException {
kitaev@213: 			if (offset > 0) { // save few useless reset/rewind operations
kitaev@213: 				da.seek(offset);
kitaev@213: 			}
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, DataAccess da) {
kitaev@213: 			try {
kitaev@213: 				prepare(revisionNumber, da); // XXX perhaps, prepare shall return DA (sliced, if needed)
kitaev@213: 				final ProgressSupport progressSupport = ProgressSupport.Factory.get(sink);
kitaev@213: 				ByteBuffer buf = ByteBuffer.allocate(512);
kitaev@213: 				progressSupport.start(da.length());
kitaev@213: 				while (!da.isEmpty()) {
kitaev@213: 					cancelSupport.checkCancelled();
kitaev@213: 					da.readBytes(buf);
kitaev@213: 					buf.flip();
kitaev@213: 					// XXX I may not rely on returned number of bytes but track change in buf position instead.
kitaev@213: 					int consumed = sink.write(buf); 
kitaev@213: 					// FIXME in fact, bad sink implementation (that consumes no bytes) would result in endless loop. Need to account for this 
kitaev@213: 					buf.compact();
kitaev@213: 					progressSupport.worked(consumed);
kitaev@213: 				}
kitaev@213: 				progressSupport.done(); // XXX shall specify whether #done() is invoked always or only if completed successfully.
kitaev@213: 			} catch (IOException ex) {
kitaev@213: 				recordFailure(ex);
kitaev@213: 			} catch (CancelledException ex) {
kitaev@213: 				recordFailure(ex);
kitaev@213: 			} catch (HgException ex) {
kitaev@213: 				recordFailure(ex);
kitaev@213: 			}
kitaev@213: 		}
kitaev@213: 		
kitaev@213: 		public void checkCancelled() throws CancelledException {
kitaev@213: 			cancelSupport.checkCancelled();
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		protected void recordFailure(Exception ex) {
kitaev@213: 			assert failure == null;
kitaev@213: 			failure = ex;
kitaev@213: 		}
kitaev@213: 
kitaev@213: 		public void checkFailed() throws HgException, IOException, CancelledException {
kitaev@213: 			if (failure == null) {
kitaev@213: 				return;
kitaev@213: 			}
kitaev@213: 			if (failure instanceof IOException) {
kitaev@213: 				throw (IOException) failure;
kitaev@213: 			}
kitaev@213: 			if (failure instanceof CancelledException) {
kitaev@213: 				throw (CancelledException) failure;
kitaev@213: 			}
kitaev@213: 			if (failure instanceof HgException) {
kitaev@213: 				throw (HgException) failure;
kitaev@213: 			}
kitaev@213: 			throw new HgBadStateException(failure);
kitaev@213: 		}
kitaev@213: 	}
kitaev@213: }