tikhomirov@13: /*
tikhomirov@74:  * Copyright (c) 2010-2011 TMate Software Ltd
tikhomirov@74:  *  
tikhomirov@74:  * This program is free software; you can redistribute it and/or modify
tikhomirov@74:  * it under the terms of the GNU General Public License as published by
tikhomirov@74:  * the Free Software Foundation; version 2 of the License.
tikhomirov@74:  *
tikhomirov@74:  * This program is distributed in the hope that it will be useful,
tikhomirov@74:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
tikhomirov@74:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
tikhomirov@74:  * GNU General Public License for more details.
tikhomirov@74:  *
tikhomirov@74:  * For information on how to redistribute this software under
tikhomirov@74:  * the terms of a license other than GNU General Public License
tikhomirov@102:  * contact TMate Software at support@hg4j.com
tikhomirov@2:  */
tikhomirov@74: package org.tmatesoft.hg.repo;
tikhomirov@74: 
tikhomirov@218: import static org.tmatesoft.hg.repo.HgRepository.TIP;
tikhomirov@218: 
tikhomirov@232: import java.io.ByteArrayOutputStream;
tikhomirov@157: import java.io.IOException;
tikhomirov@218: import java.util.ArrayList;
tikhomirov@218: import java.util.Arrays;
tikhomirov@157: 
tikhomirov@157: import org.tmatesoft.hg.core.HgBadStateException;
tikhomirov@74: import org.tmatesoft.hg.core.Nodeid;
tikhomirov@157: import org.tmatesoft.hg.internal.DataAccess;
tikhomirov@266: import org.tmatesoft.hg.internal.DigestHelper;
tikhomirov@232: import org.tmatesoft.hg.internal.Experimental;
tikhomirov@218: import org.tmatesoft.hg.internal.Lifecycle;
tikhomirov@196: import org.tmatesoft.hg.internal.Pool;
tikhomirov@77: import org.tmatesoft.hg.internal.RevlogStream;
tikhomirov@232: import org.tmatesoft.hg.util.Path;
tikhomirov@74: 
tikhomirov@2: 
tikhomirov@2: /**
tikhomirov@2:  *
tikhomirov@74:  * @author Artem Tikhomirov
tikhomirov@74:  * @author TMate Software Ltd.
tikhomirov@2:  */
tikhomirov@2: public class HgManifest extends Revlog {
tikhomirov@218: 	private RevisionMapper revisionMap;
tikhomirov@2: 
tikhomirov@13: 	/*package-local*/ HgManifest(HgRepository hgRepo, RevlogStream content) {
tikhomirov@21: 		super(hgRepo, content);
tikhomirov@13: 	}
tikhomirov@13: 
tikhomirov@218: 	/**
tikhomirov@218: 	 * 
tikhomirov@218: 	 * @param start changelog (not manifest!) revision to begin with
tikhomirov@218: 	 * @param end changelog (not manifest!) revision to stop, inclusive.
tikhomirov@218: 	 * @param inspector can't be <code>null</code>
tikhomirov@218: 	 */
tikhomirov@19: 	public void walk(int start, int end, final Inspector inspector) {
tikhomirov@196: 		if (inspector == null) {
tikhomirov@196: 			throw new IllegalArgumentException();
tikhomirov@196: 		}
tikhomirov@218: 		int start0 = fromChangelog(start);
tikhomirov@218: 		int end0 = fromChangelog(end);
tikhomirov@218: 		content.iterate(start0, end0, true, new ManifestParser(inspector));
tikhomirov@19: 	}
tikhomirov@218: 	
tikhomirov@254: 	/**
tikhomirov@254: 	 * "Sparse" iteration of the manifest
tikhomirov@254: 	 * 
tikhomirov@254: 	 * @param inspector
tikhomirov@254: 	 * @param localRevisions local changeset revisions to visit
tikhomirov@254: 	 */
tikhomirov@254: 	public void walk(final Inspector inspector, int... localRevisions) {
tikhomirov@254: 		if (inspector == null || localRevisions == null) {
tikhomirov@254: 			throw new IllegalArgumentException();
tikhomirov@254: 		}
tikhomirov@254: 		int[] manifestLocalRevs = new int[localRevisions.length];
tikhomirov@254: 		boolean needsSort = false;
tikhomirov@254: 		for (int i = 0; i < localRevisions.length; i++) {
tikhomirov@254: 			final int manifestLocalRev = fromChangelog(localRevisions[i]);
tikhomirov@254: 			manifestLocalRevs[i] = manifestLocalRev;
tikhomirov@254: 			if (i > 0 && manifestLocalRevs[i-1] > manifestLocalRev) {
tikhomirov@254: 				needsSort = true;
tikhomirov@254: 			}
tikhomirov@254: 		}
tikhomirov@254: 		if (needsSort) {
tikhomirov@254: 			Arrays.sort(manifestLocalRevs);
tikhomirov@254: 		}
tikhomirov@254: 		content.iterate(manifestLocalRevs, true, new ManifestParser(inspector));
tikhomirov@254: 	}
tikhomirov@254: 	
tikhomirov@232: 	// manifest revision number that corresponds to the given changeset
tikhomirov@218: 	/*package-local*/ int fromChangelog(int revisionNumber) {
tikhomirov@218: 		if (HgInternals.wrongLocalRevision(revisionNumber)) {
tikhomirov@218: 			throw new IllegalArgumentException(String.valueOf(revisionNumber));
tikhomirov@218: 		}
tikhomirov@257: 		if (revisionNumber == HgRepository.WORKING_COPY || revisionNumber == HgRepository.BAD_REVISION) {
tikhomirov@257: 			throw new IllegalArgumentException("Can't use constants like WORKING_COPY or BAD_REVISION");
tikhomirov@257: 		}
tikhomirov@257: 		// revisionNumber == TIP is processed by RevisionMapper 
tikhomirov@218: 		if (revisionMap == null) {
tikhomirov@218: 			revisionMap = new RevisionMapper(getRepo());
tikhomirov@218: 			content.iterate(0, TIP, false, revisionMap);
tikhomirov@218: 		}
tikhomirov@218: 		return revisionMap.at(revisionNumber);
tikhomirov@218: 	}
tikhomirov@218: 	
tikhomirov@232: 	/**
tikhomirov@232: 	 * Extracts file revision as it was known at the time of given changeset.
tikhomirov@232: 	 * 
tikhomirov@232: 	 * @param revisionNumber local changeset index 
tikhomirov@232: 	 * @param file path to file in question
tikhomirov@232: 	 * @return file revision or <code>null</code> if manifest at specified revision doesn't list such file
tikhomirov@232: 	 */
tikhomirov@232: 	@Experimental(reason="Perhaps, HgDataFile shall own this method")
tikhomirov@232: 	public Nodeid getFileRevision(int revisionNumber, final Path file) {
tikhomirov@232: 		int rev = fromChangelog(revisionNumber);
tikhomirov@232: 		final Nodeid[] rv = new Nodeid[] { null };
tikhomirov@232: 		content.iterate(rev, rev, true, new RevlogStream.Inspector() {
tikhomirov@232: 			
tikhomirov@232: 			public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, DataAccess data) {
tikhomirov@232: 				ByteArrayOutputStream bos = new ByteArrayOutputStream();
tikhomirov@232: 				try {
tikhomirov@232: 					byte b;
tikhomirov@232: 					while (!data.isEmpty() && (b = data.readByte()) != '\n') {
tikhomirov@232: 						if (b != 0) {
tikhomirov@232: 							bos.write(b);
tikhomirov@232: 						} else {
tikhomirov@232: 							String fname = new String(bos.toByteArray());
tikhomirov@232: 							bos.reset();
tikhomirov@232: 							if (file.toString().equals(fname)) {
tikhomirov@232: 								byte[] nid = new byte[40];  
tikhomirov@232: 								data.readBytes(nid, 0, 40);
tikhomirov@232: 								rv[0] = Nodeid.fromAscii(nid, 0, 40);
tikhomirov@232: 								break;
tikhomirov@232: 							}
tikhomirov@232: 							// else skip to the end of line
tikhomirov@232: 							while (!data.isEmpty() && (b = data.readByte()) != '\n')
tikhomirov@232: 								;
tikhomirov@232: 						}
tikhomirov@232: 					}
tikhomirov@232: 				} catch (IOException ex) {
tikhomirov@232: 					throw new HgBadStateException(ex);
tikhomirov@232: 				}
tikhomirov@232: 			}
tikhomirov@232: 		});
tikhomirov@232: 		return rv[0];
tikhomirov@232: 	}
tikhomirov@232: 			
tikhomirov@19: 	public interface Inspector {
tikhomirov@222: 		boolean begin(int mainfestRevision, Nodeid nid, int changelogRevision);
tikhomirov@19: 		boolean next(Nodeid nid, String fname, String flags);
tikhomirov@222: 		boolean end(int manifestRevision);
tikhomirov@2: 	}
tikhomirov@267: 	
tikhomirov@267: 	public interface ElementProxy<T> {
tikhomirov@267: 		T get();
tikhomirov@267: 	}
tikhomirov@267: 	
tikhomirov@268: 	/**
tikhomirov@268: 	 * When Pool uses Strings directly,
tikhomirov@268: 	 * ManifestParser creates new String instance with new char[] value, and does byte->char conversion.
tikhomirov@268: 	 * For cpython repo, walk(0..10k), there are over 16 million filenames, of them only 3020 unique.
tikhomirov@268: 	 * This means there are 15.9 million useless char[] instances and byte->char conversions  
tikhomirov@268: 	 * 
tikhomirov@268: 	 * When String is wrapped into {@link StringProxy}, there's extra overhead of byte[] representation
tikhomirov@268: 	 * of the String, but these are only for unique Strings (3020 in the example above). Besides, I save
tikhomirov@268: 	 * useless char[] and byte->char conversions. 
tikhomirov@268: 	 */
tikhomirov@268: 	private static class StringProxy {
tikhomirov@267: 		private byte[] data;
tikhomirov@268: 		private int start; 
tikhomirov@268: 		private final int hash, length;
tikhomirov@267: 		private String result;
tikhomirov@267: 
tikhomirov@268: 		public StringProxy(byte[] data, int start, int length) {
tikhomirov@267: 			this.data = data;
tikhomirov@267: 			this.start = start;
tikhomirov@267: 			this.length = length;
tikhomirov@267: 
tikhomirov@268: 			// copy from String.hashCode(). In fact, not necessarily match result of String(data).hashCode
tikhomirov@268: 			// just need some nice algorithm here
tikhomirov@267: 			int h = 0;
tikhomirov@267: 			byte[] d = data;
tikhomirov@267: 			for (int i = 0, off = start, len = length; i < len; i++) {
tikhomirov@267: 				h = 31 * h + d[off++];
tikhomirov@267: 			}
tikhomirov@267: 			hash = h;
tikhomirov@267: 		}
tikhomirov@267: 
tikhomirov@267: 		@Override
tikhomirov@267: 		public boolean equals(Object obj) {
tikhomirov@268: 			if (false == obj instanceof StringProxy) {
tikhomirov@267: 				return false;
tikhomirov@267: 			}
tikhomirov@268: 			StringProxy o = (StringProxy) obj;
tikhomirov@267: 			if (o.result != null && result != null) {
tikhomirov@267: 				return result.equals(o.result);
tikhomirov@267: 			}
tikhomirov@268: 			if (o.length != length || o.hash != hash) {
tikhomirov@267: 				return false;
tikhomirov@267: 			}
tikhomirov@267: 			for (int i = 0, x = o.start, y = start; i < length; i++) {
tikhomirov@267: 				if (o.data[x++] != data[y++]) {
tikhomirov@267: 					return false;
tikhomirov@267: 				}
tikhomirov@267: 			}
tikhomirov@267: 			return true;
tikhomirov@267: 		}
tikhomirov@267: 		@Override
tikhomirov@267: 		public int hashCode() {
tikhomirov@267: 			return hash;
tikhomirov@267: 		}
tikhomirov@267: 		
tikhomirov@267: 		public String freeze() {
tikhomirov@267: 			if (result == null) {
tikhomirov@267: 				result = new String(data, start, length);
tikhomirov@268: 				// release reference to bigger data array, make a copy of relevant part only
tikhomirov@268: 				byte[] d = new byte[length];
tikhomirov@268: 				System.arraycopy(data, start, d, 0, length);
tikhomirov@268: 				data = d;
tikhomirov@268: 				start = 0;
tikhomirov@267: 			}
tikhomirov@267: 			return result;
tikhomirov@267: 		}
tikhomirov@267: 	}
tikhomirov@196: 
tikhomirov@268: 	private static class ManifestParser implements RevlogStream.Inspector {
tikhomirov@196: 		private boolean gtg = true; // good to go
tikhomirov@196: 		private final Inspector inspector;
tikhomirov@267: 		private Pool<Nodeid> nodeidPool, thisRevPool;
tikhomirov@268: 		private final Pool<StringProxy> fnamePool;
tikhomirov@268: 		private final Pool<StringProxy> flagsPool;
tikhomirov@266: 		private byte[] nodeidLookupBuffer = new byte[20]; // get reassigned each time new Nodeid is added to pool
tikhomirov@262: 		
tikhomirov@196: 		public ManifestParser(Inspector delegate) {
tikhomirov@196: 			assert delegate != null;
tikhomirov@196: 			inspector = delegate;
tikhomirov@267: 			nodeidPool = new Pool<Nodeid>();
tikhomirov@268: 			fnamePool = new Pool<StringProxy>();
tikhomirov@268: 			flagsPool = new Pool<StringProxy>();
tikhomirov@267: 			thisRevPool = new Pool<Nodeid>();
tikhomirov@196: 		}
tikhomirov@268: 		
tikhomirov@196: 		public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, DataAccess da) {
tikhomirov@196: 			if (!gtg) {
tikhomirov@196: 				return;
tikhomirov@196: 			}
tikhomirov@196: 			try {
tikhomirov@222: 				gtg = gtg && inspector.begin(revisionNumber, new Nodeid(nodeid, true), linkRevision);
tikhomirov@196: 				String fname = null;
tikhomirov@196: 				String flags = null;
tikhomirov@196: 				Nodeid nid = null;
tikhomirov@267: 				int i;
tikhomirov@267: 				byte[] data = da.byteArray();
tikhomirov@267: 				for (i = 0; gtg && i < actualLen; i++) {
tikhomirov@267: 					int x = i;
tikhomirov@267: 					for( ; data[i] != '\n' && i < actualLen; i++) {
tikhomirov@267: 						if (fname == null && data[i] == 0) {
tikhomirov@268: 							StringProxy px = fnamePool.unify(new StringProxy(data, x, i - x));
tikhomirov@268: 							// if (cached = fnamePool.unify(px))== px then cacheMiss, else cacheHit
tikhomirov@268: 							// cpython 0..10k: hits: 15 989 152, misses: 3020
tikhomirov@268: 							fname = px.freeze();
tikhomirov@267: 							x = i+1;
tikhomirov@196: 						}
tikhomirov@196: 					}
tikhomirov@267: 					if (i < actualLen) {
tikhomirov@267: 						assert data[i] == '\n'; 
tikhomirov@267: 						int nodeidLen = i - x < 40 ? i-x : 40; // if > 40, there are flags
tikhomirov@267: 						DigestHelper.ascii2bin(data, x, nodeidLen, nodeidLookupBuffer); // ignore return value as it's unlikely to have NULL in manifest
tikhomirov@267: 						nid = new Nodeid(nodeidLookupBuffer, false); // this Nodeid is for pool lookup only, mock object
tikhomirov@267: 						Nodeid cached = nodeidPool.unify(nid);
tikhomirov@267: 						if (cached == nid) {
tikhomirov@267: 							// buffer now belongs to the cached nodeid
tikhomirov@267: 							nodeidLookupBuffer = new byte[20];
tikhomirov@267: 						} else {
tikhomirov@267: 							nid = cached; // use existing version, discard the lookup object
tikhomirov@267: 						} // for cpython 0..10k, cache hits are 15 973 301, vs 18871 misses.
tikhomirov@267: 						thisRevPool.record(nid); // memorize revision for the next iteration. 
tikhomirov@267: 						if (nodeidLen + x < i) {
tikhomirov@267: 							// 'x' and 'l' for executable bits and symlinks?
tikhomirov@267: 							// hg --debug manifest shows 644 for each regular file in my repo
tikhomirov@268: 							// for cpython 0..10k, there are 4361062 flagPool checks, and there's only 1 unique flag
tikhomirov@268: 							flags = flagsPool.unify(new StringProxy(data, x + nodeidLen, i-x-nodeidLen)).freeze();
tikhomirov@267: 						}
tikhomirov@267: 						gtg = gtg && inspector.next(nid, fname, flags);
tikhomirov@267: 					}
tikhomirov@196: 					nid = null;
tikhomirov@196: 					fname = flags = null;
tikhomirov@196: 				}
tikhomirov@196: 				gtg = gtg && inspector.end(revisionNumber);
tikhomirov@262: 				//
tikhomirov@262: 				// keep only actual file revisions, found at this version 
tikhomirov@262: 				// (next manifest is likely to refer to most of them, although in specific cases 
tikhomirov@262: 				// like commit in another branch a lot may be useless)
tikhomirov@262: 				nodeidPool.clear();
tikhomirov@267: 				Pool<Nodeid> t = nodeidPool;
tikhomirov@262: 				nodeidPool = thisRevPool;
tikhomirov@264: 				thisRevPool = t;
tikhomirov@196: 			} catch (IOException ex) {
tikhomirov@196: 				throw new HgBadStateException(ex);
tikhomirov@196: 			}
tikhomirov@196: 		}
tikhomirov@196: 	}
tikhomirov@218: 	
tikhomirov@218: 	private static class RevisionMapper implements RevlogStream.Inspector, Lifecycle {
tikhomirov@218: 		
tikhomirov@218: 		private final int changelogRevisions;
tikhomirov@218: 		private int[] changelog2manifest;
tikhomirov@218: 		private final HgRepository repo;
tikhomirov@218: 
tikhomirov@218: 		public RevisionMapper(HgRepository hgRepo) {
tikhomirov@218: 			repo = hgRepo;
tikhomirov@218: 			changelogRevisions = repo.getChangelog().getRevisionCount();
tikhomirov@218: 		}
tikhomirov@218: 
tikhomirov@257: 		// respects TIP
tikhomirov@218: 		public int at(int revisionNumber) {
tikhomirov@257: 			if (revisionNumber == TIP) {
tikhomirov@257: 				revisionNumber = changelogRevisions - 1;
tikhomirov@257: 			}
tikhomirov@218: 			if (changelog2manifest != null) {
tikhomirov@218: 				return changelog2manifest[revisionNumber];
tikhomirov@218: 			}
tikhomirov@218: 			return revisionNumber;
tikhomirov@218: 		}
tikhomirov@218: 
tikhomirov@218: 		public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, DataAccess data) {
tikhomirov@218: 			if (changelog2manifest != null) {
tikhomirov@220: 				// next assertion is not an error, rather assumption check, which is too development-related to be explicit exception - 
tikhomirov@220: 				// I just wonder if there are manifests that have two entries pointing to single changeset. It seems unrealistic, though -
tikhomirov@220: 				// changeset records one and only one manifest nodeid
tikhomirov@220: 				assert changelog2manifest[linkRevision] == -1 : String.format("revision:%d, link:%d, already linked to revision:%d", revisionNumber, linkRevision, changelog2manifest[linkRevision]);
tikhomirov@218: 				changelog2manifest[linkRevision] = revisionNumber;
tikhomirov@218: 			} else {
tikhomirov@218: 				if (revisionNumber != linkRevision) {
tikhomirov@218: 					changelog2manifest = new int[changelogRevisions];
tikhomirov@218: 					Arrays.fill(changelog2manifest, -1);
tikhomirov@218: 					for (int i = 0; i < revisionNumber; changelog2manifest[i] = i, i++)
tikhomirov@218: 						;
tikhomirov@218: 					changelog2manifest[linkRevision] = revisionNumber;
tikhomirov@218: 				}
tikhomirov@218: 			}
tikhomirov@218: 		}
tikhomirov@218: 		
tikhomirov@218: 		public void start(int count, Callback callback, Object token) {
tikhomirov@218: 			if (count != changelogRevisions) {
tikhomirov@218: 				assert count < changelogRevisions; // no idea what to do if manifest has more revisions than changelog
tikhomirov@218: 				// the way how manifest may contain more revisions than changelog, as I can imagine, is a result of  
tikhomirov@218: 				// some kind of an import tool (e.g. from SVN or CVS), that creates manifest and changelog independently.
tikhomirov@218: 				// Note, it's pure guess, I didn't see such repository yet (although the way manifest revisions
tikhomirov@218: 				// in cpython repo are numbered makes me think aforementioned way) 
tikhomirov@218: 				changelog2manifest = new int[changelogRevisions];
tikhomirov@218: 				Arrays.fill(changelog2manifest, -1);
tikhomirov@218: 			}
tikhomirov@218: 		}
tikhomirov@218: 
tikhomirov@218: 		public void finish(Object token) {
tikhomirov@218: 			if (changelog2manifest == null) {
tikhomirov@218: 				return;
tikhomirov@218: 			}
tikhomirov@218: 			// I assume there'd be not too many revisions we don't know manifest of
tikhomirov@218: 			ArrayList<Integer> undefinedChangelogRevision = new ArrayList<Integer>();
tikhomirov@218: 			for (int i = 0; i < changelog2manifest.length; i++) {
tikhomirov@218: 				if (changelog2manifest[i] == -1) {
tikhomirov@218: 					undefinedChangelogRevision.add(i);
tikhomirov@218: 				}
tikhomirov@218: 			}
tikhomirov@218: 			for (int u : undefinedChangelogRevision) {
tikhomirov@218: 				Nodeid manifest = repo.getChangelog().range(u, u).get(0).manifest();
tikhomirov@220: 				// FIXME calculate those missing effectively (e.g. cache and sort nodeids to speed lookup
tikhomirov@218: 				// right away in the #next (may refactor ParentWalker's sequential and sorted into dedicated helper and reuse here)
tikhomirov@218: 				changelog2manifest[u] = repo.getManifest().getLocalRevision(manifest);
tikhomirov@218: 			}
tikhomirov@218: 		}
tikhomirov@218: 	}
tikhomirov@2: }