Mercurial > jhg
comparison src/org/tmatesoft/hg/internal/PatchGenerator.java @ 533:e6f72c9829a6
Generate patches using diff algorithm
author | Artem Tikhomirov <tikhomirov.artem@gmail.com> |
---|---|
date | Wed, 30 Jan 2013 15:48:36 +0100 |
parents | |
children | dd4f6311af52 |
comparison
equal
deleted
inserted
replaced
532:688c1ab113bb | 533:e6f72c9829a6 |
---|---|
1 /* | |
2 * Copyright (c) 2013 TMate Software Ltd | |
3 * | |
4 * This program is free software; you can redistribute it and/or modify | |
5 * it under the terms of the GNU General Public License as published by | |
6 * the Free Software Foundation; version 2 of the License. | |
7 * | |
8 * This program is distributed in the hope that it will be useful, | |
9 * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
11 * GNU General Public License for more details. | |
12 * | |
13 * For information on how to redistribute this software under | |
14 * the terms of a license other than GNU General Public License | |
15 * contact TMate Software at support@hg4j.com | |
16 */ | |
17 package org.tmatesoft.hg.internal; | |
18 | |
19 import java.util.ArrayList; | |
20 import java.util.HashMap; | |
21 import java.util.Map; | |
22 | |
23 import org.tmatesoft.hg.repo.HgDataFile; | |
24 import org.tmatesoft.hg.repo.HgLookup; | |
25 import org.tmatesoft.hg.repo.HgRepository; | |
26 | |
27 /** | |
28 * Mercurial cares about changes only up to the line level, e.g. a simple file version bump in manifest looks like (RevlogDump output): | |
29 * | |
30 * 522: 233748 0 103 17438 433 522 521 -1 756073cf2321df44d3ed0585f2a5754bc8a1b2f6 | |
31 * <PATCH>: | |
32 * 3487..3578, 91:src/org/tmatesoft/hg/core/HgIterateDirection.java\00add61a8a665c5d8f092210767f812fe0d335ac8 | |
33 * | |
34 * I.e. for the {fname}{revision} entry format of manifest, not only {revision} is changed, but the whole line, with unchanged {fname} is recorded | |
35 * in the patch. | |
36 * | |
37 * Mercurial paper describes reasons for choosing this approach to delta generation, too. | |
38 * | |
39 * | |
40 * @author Artem Tikhomirov | |
41 * @author TMate Software Ltd. | |
42 */ | |
43 public class PatchGenerator { | |
44 | |
45 private Map<ChunkSequence.ByteChain, IntVector> chunk2UseIndex; | |
46 private ChunkSequence seq1, seq2; | |
47 | |
48 // get filled by #longestMatch, track start of common sequence in seq1 and seq2, respectively | |
49 private int matchStartS1, matchStartS2; | |
50 // get filled by #findMatchingBlocks, track start of changed/unknown sequence in seq1 and seq2 | |
51 private int changeStartS1, changeStartS2; | |
52 | |
53 public void init(byte[] data1, byte[] data2) { | |
54 seq1 = new ChunkSequence(data1); | |
55 seq1.splitByNewlines(); | |
56 seq2 = new ChunkSequence(data2); | |
57 seq2.splitByNewlines(); | |
58 prepare(seq2); | |
59 } | |
60 | |
61 private void prepare(ChunkSequence s2) { | |
62 chunk2UseIndex = new HashMap<ChunkSequence.ByteChain, IntVector>(); | |
63 for (int i = 0, len = s2.chunkCount(); i < len; i++) { | |
64 ChunkSequence.ByteChain bc = s2.chunk(i); | |
65 IntVector loc = chunk2UseIndex.get(bc); | |
66 if (loc == null) { | |
67 chunk2UseIndex.put(bc, loc = new IntVector()); | |
68 } | |
69 loc.add(i); | |
70 // bc.registerUseIn(i) - BEWARE, use of bc here is incorrect | |
71 // in this case need to find the only ByteChain to keep indexes | |
72 // i.e. when there are few equal ByteChain instances, notion of "usedIn" shall be either shared (reference same vector) | |
73 // or kept within only one of them | |
74 } | |
75 // for (ChunkSequence.ByteChain bc : chunk2UseIndex.keySet()) { | |
76 // System.out.printf("%s: {", new String(bc.data())); | |
77 // for (int x : chunk2UseIndex.get(bc).toArray()) { | |
78 // System.out.printf(" %d,", x); | |
79 // } | |
80 // System.out.println("}"); | |
81 // } | |
82 } | |
83 | |
84 public void findMatchingBlocks() { | |
85 changeStartS1 = changeStartS2 = 0; | |
86 findMatchingBlocks(0, seq1.chunkCount(), 0, seq2.chunkCount()); | |
87 if (changeStartS1 < seq1.chunkCount() || changeStartS2 < seq2.chunkCount()) { | |
88 reportDeltaElement(seq1.chunkCount(), seq2.chunkCount()); | |
89 } | |
90 } | |
91 | |
92 /** | |
93 * implementation based on Python's difflib.py and SequenceMatcher | |
94 */ | |
95 public int longestMatch(int startS1, int endS1, int startS2, int endS2) { | |
96 matchStartS1 = matchStartS2 = 0; | |
97 int maxLength = 0; | |
98 IntMap<Integer> chunkIndex2MatchCount = new IntMap<Integer>(8); | |
99 for (int i = startS1; i < endS1; i++) { | |
100 ChunkSequence.ByteChain bc = seq1.chunk(i); | |
101 IntMap<Integer> newChunkIndex2MatchCount = new IntMap<Integer>(8); | |
102 IntVector occurencesInS2 = chunk2UseIndex.get(bc); | |
103 if (occurencesInS2 == null) { | |
104 // chunkIndex2MatchCount.clear(); // TODO need clear instead of new instance | |
105 chunkIndex2MatchCount = newChunkIndex2MatchCount; | |
106 continue; | |
107 } | |
108 for (int j : occurencesInS2.toArray()) { | |
109 // s1[i] == s2[j] | |
110 if (j < startS2) { | |
111 continue; | |
112 } | |
113 if (j >= endS2) { | |
114 break; | |
115 } | |
116 int prevChunkMatches = chunkIndex2MatchCount.containsKey(j-1) ? chunkIndex2MatchCount.get(j-1) : 0; | |
117 int k = prevChunkMatches + 1; | |
118 newChunkIndex2MatchCount.put(j, k); | |
119 if (k > maxLength) { | |
120 matchStartS1 = i-k+1; | |
121 matchStartS2 = j-k+1; | |
122 maxLength = k; | |
123 } | |
124 } | |
125 chunkIndex2MatchCount = newChunkIndex2MatchCount; | |
126 } | |
127 return maxLength; | |
128 } | |
129 | |
130 public void findMatchingBlocks(int startS1, int endS1, int startS2, int endS2) { | |
131 int matchLength = longestMatch(startS1, endS1, startS2, endS2); | |
132 if (matchLength > 0) { | |
133 final int saveStartS1 = matchStartS1, saveStartS2 = matchStartS2; | |
134 if (startS1 < matchStartS1 && startS2 < matchStartS2) { | |
135 findMatchingBlocks(startS1, matchStartS1, startS2, matchStartS2); | |
136 } | |
137 reportDeltaElement(saveStartS1, saveStartS2); | |
138 changeStartS1 = saveStartS1 + matchLength; | |
139 changeStartS2 = saveStartS2 + matchLength; | |
140 // System.out.printf("match: from line #%d and line #%d of length %d\n", saveStartS1, saveStartS2, matchLength); | |
141 if (saveStartS1+matchLength < endS1 && saveStartS2+matchLength < endS2) { | |
142 findMatchingBlocks(saveStartS1 + matchLength, endS1, saveStartS2 + matchLength, endS2); | |
143 } | |
144 } | |
145 } | |
146 | |
147 private Patch deltaCollector; | |
148 | |
149 private void reportDeltaElement(int i, int j) { | |
150 if (changeStartS1 < i) { | |
151 if (changeStartS2 < j) { | |
152 System.out.printf("changed [%d..%d) with [%d..%d)\n", changeStartS1, i, changeStartS2, j); | |
153 } else { | |
154 assert changeStartS2 == j; | |
155 System.out.printf("deleted [%d..%d)\n", changeStartS1, i); | |
156 } | |
157 if (deltaCollector != null) { | |
158 int from = seq1.chunk(changeStartS1).getOffset(); | |
159 int to = seq1.chunk(i).getOffset(); | |
160 byte[] data = seq2.data(changeStartS2, j); | |
161 deltaCollector.add(from, to, data); | |
162 } | |
163 } else { | |
164 assert changeStartS1 == i; | |
165 if(changeStartS2 < j) { | |
166 System.out.printf("added [%d..%d)\n", changeStartS2, j); | |
167 } else { | |
168 assert changeStartS2 == j; | |
169 System.out.printf("adjustent equal blocks %d, %d and %d,%d\n", changeStartS1, i, changeStartS2, j); | |
170 } | |
171 if (deltaCollector != null) { | |
172 int insPoint = seq1.chunk(changeStartS1).getOffset(); | |
173 byte[] data = seq2.data(changeStartS2, j); | |
174 deltaCollector.add(insPoint, insPoint, data); | |
175 } | |
176 } | |
177 } | |
178 | |
179 public static void main(String[] args) throws Exception { | |
180 HgRepository repo = new HgLookup().detectFromWorkingDir(); | |
181 HgDataFile df = repo.getFileNode("cmdline/org/tmatesoft/hg/console/Main.java"); | |
182 ByteArrayChannel bac1, bac2; | |
183 df.content(80, bac1 = new ByteArrayChannel()); | |
184 df.content(81, bac2 = new ByteArrayChannel()); | |
185 // String s1 = "line 1\nline 2\r\nline 3\n\nline 1\nline 2"; | |
186 // String s2 = "abc\ncdef\r\nline 2\r\nline 3\nline 2"; | |
187 PatchGenerator pg = new PatchGenerator(); | |
188 pg.init(bac1.toArray(), bac2.toArray()); | |
189 pg.findMatchingBlocks(); | |
190 } | |
191 | |
192 public Patch delta(byte[] prev, byte[] content) { | |
193 deltaCollector = new Patch(); | |
194 init(prev, content); | |
195 findMatchingBlocks(); | |
196 return deltaCollector; | |
197 } | |
198 | |
199 private static class ChunkSequence { | |
200 | |
201 private final byte[] input; | |
202 private ArrayList<ByteChain> lines; | |
203 | |
204 public ChunkSequence(byte[] data) { | |
205 input = data; | |
206 } | |
207 | |
208 public void splitByNewlines() { | |
209 lines = new ArrayList<ByteChain>(); | |
210 int lastStart = 0; | |
211 for (int i = 0; i < input.length; i++) { | |
212 if (input[i] == '\n') { | |
213 lines.add(new ByteChain(lastStart, i+1)); | |
214 lastStart = i+1; | |
215 } else if (input[i] == '\r') { | |
216 if (i+1 < input.length && input[i+1] == '\n') { | |
217 i++; | |
218 } | |
219 lines.add(new ByteChain(lastStart, i+1)); | |
220 lastStart = i+1; | |
221 } | |
222 } | |
223 if (lastStart < input.length) { | |
224 lines.add(new ByteChain(lastStart, input.length)); | |
225 } | |
226 } | |
227 | |
228 public ByteChain chunk(int index) { | |
229 return lines.get(index); | |
230 } | |
231 | |
232 public int chunkCount() { | |
233 return lines.size(); | |
234 } | |
235 | |
236 public byte[] data(int chunkFrom, int chunkTo) { | |
237 if (chunkFrom == chunkTo) { | |
238 return new byte[0]; | |
239 } | |
240 int from = chunk(chunkFrom).getOffset(), to = chunk(chunkTo).getOffset(); | |
241 byte[] rv = new byte[to - from]; | |
242 System.arraycopy(input, from, rv, 0, rv.length); | |
243 return rv; | |
244 } | |
245 | |
246 | |
247 final class ByteChain { | |
248 private final int start, end; | |
249 private final int hash; | |
250 | |
251 ByteChain(int s, int e) { | |
252 start = s; | |
253 end = e; | |
254 hash = calcHash(input, s, e); | |
255 } | |
256 | |
257 /** | |
258 * byte offset of the this ByteChain inside ChainSequence | |
259 */ | |
260 public int getOffset() { | |
261 return start; | |
262 } | |
263 | |
264 public byte[] data() { | |
265 byte[] rv = new byte[end - start]; | |
266 System.arraycopy(input, start, rv, 0, rv.length); | |
267 return rv; | |
268 } | |
269 | |
270 @Override | |
271 public boolean equals(Object obj) { | |
272 if (obj == null || obj.getClass() != ByteChain.class) { | |
273 return false; | |
274 } | |
275 ByteChain other = (ByteChain) obj; | |
276 if (other.hash != hash || other.end - other.start != end - start) { | |
277 return false; | |
278 } | |
279 return other.match(input, start); | |
280 } | |
281 | |
282 private boolean match(byte[] oi, int from) { | |
283 for (int i = start, j = from; i < end; i++, j++) { | |
284 if (ChunkSequence.this.input[i] != oi[j]) { | |
285 return false; | |
286 } | |
287 } | |
288 return true; | |
289 } | |
290 | |
291 @Override | |
292 public int hashCode() { | |
293 return hash; | |
294 } | |
295 | |
296 @Override | |
297 public String toString() { | |
298 return String.format("[@%d\"%s\"]", start, new String(data())); | |
299 } | |
300 } | |
301 | |
302 // same as Arrays.hashCode(byte[]), just for a slice of a bigger array | |
303 static int calcHash(byte[] data, int from, int to) { | |
304 int result = 1; | |
305 for (int i = from; i < to; i++) { | |
306 result = 31 * result + data[i]; | |
307 } | |
308 return result; | |
309 } | |
310 } | |
311 } |