Mercurial > jhg

annotate src/org/tmatesoft/hg/internal/RevlogDump.java @ 576:3c4db86e8c1f

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Issue 43: poor performance with InflaterDataAccess. Phase 2: inflate into buffer, effective skip and readByte/readBytes()
author Artem Tikhomirov <tikhomirov.artem@gmail.com>
date Tue, 16 Apr 2013 19:31:57 +0200
parents e6f72c9829a6
children 47dfa0ec7e35
rev   line source
22
603806cd2dc6 Status of local working dir against non-tip base revision
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 17
diff changeset
1 /*
533
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
2 * Copyright (c) 2010-2013 TMate Software Ltd
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
3 *
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
4 * This program is free software; you can redistribute it and/or modify
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
5 * it under the terms of the GNU General Public License as published by
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
6 * the Free Software Foundation; version 2 of the License.
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
7 *
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
8 * This program is distributed in the hope that it will be useful,
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
9 * but WITHOUT ANY WARRANTY; without even the implied warranty of
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
11 * GNU General Public License for more details.
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
12 *
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
13 * For information on how to redistribute this software under
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
14 * the terms of a license other than GNU General Public License
102
a3a2e5deb320 Updated contact address to support@hg4j.com
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 88
diff changeset
15 * contact TMate Software at support@hg4j.com
22
603806cd2dc6 Status of local working dir against non-tip base revision
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 17
diff changeset
16 */
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
17 package org.tmatesoft.hg.internal;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
18
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
19 import java.io.BufferedInputStream;
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
20 import java.io.ByteArrayInputStream;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
21 import java.io.DataInput;
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
22 import java.io.DataInputStream;
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
23 import java.io.File;
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
24 import java.io.FileInputStream;
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
25 import java.io.IOException;
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
26 import java.io.UnsupportedEncodingException;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
27 import java.math.BigInteger;
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
28 import java.nio.ByteBuffer;
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
29 import java.nio.channels.FileChannel;
533
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
30 import java.util.regex.Matcher;
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
31 import java.util.regex.Pattern;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
32 import java.util.zip.Inflater;
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
33
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
34 /**
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
35 * Utility to test/debug/troubleshoot
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
36 *
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
37 * @author Artem Tikhomirov
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
38 * @author TMate Software Ltd.
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
39 */
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
40 public class RevlogDump {
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
41
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
42 /**
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
43 * Takes 3 command line arguments -
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
44 * repository path,
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
45 * path to index file (i.e. store/data/hello.c.i) in the repository (relative)
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
46 * and "dumpData" whether to print actual content or just revlog headers
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
47 */
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
48 public static void main(String[] args) throws Exception {
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
49 String repo = "/temp/hg/hello/.hg/";
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
50 String filename = "store/00changelog.i";
22
603806cd2dc6 Status of local working dir against non-tip base revision
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 17
diff changeset
51 // String filename = "store/data/hello.c.i";
4
aa1912c70b36 Fix offset issue for inline revlogs. Commandline processing.
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 3
diff changeset
52 // String filename = "store/data/docs/readme.i";
533
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
53 System.out.println(escape("abc\0def\nzxc\tmnb"));
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
54 boolean dumpDataFull = true;
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
55 boolean dumpDataStats = false;
88
61eedab3eb3e Status between two revisions to recognize copy/rename
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 73
diff changeset
56 if (args.length > 1) {
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
57 repo = args[0];
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
58 filename = args[1];
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
59 dumpDataFull = args.length > 2 ? "dumpData".equals(args[2]) : false;
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
60 dumpDataStats = args.length > 2 ? "dumpDataStats".equals(args[2]) : false;
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
61 }
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
62 final boolean needRevData = dumpDataFull || dumpDataStats;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
63 //
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
64 DataInputStream dis = new DataInputStream(new BufferedInputStream(new FileInputStream(new File(repo, filename))));
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
65 DataInput di = dis;
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
66 dis.mark(10);
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
67 int versionField = di.readInt();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
68 dis.reset();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
69 final int INLINEDATA = 1 << 16;
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
70
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
71 final boolean inlineData = (versionField & INLINEDATA) != 0;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
72 System.out.printf("%#8x, inline: %b\n", versionField, inlineData);
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
73 FileChannel dataStream = null;
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
74 if (!inlineData && needRevData) {
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
75 dataStream = new FileInputStream(new File(repo, filename.substring(0, filename.length()-2) + ".d")).getChannel();
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
76 }
73
0d279bcc4442 Utility for future troubleshooting
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 47
diff changeset
77 System.out.println("Index Offset Flags Packed Actual Base Rev Link Rev Parent1 Parent2 nodeid");
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
78 int entryIndex = 0;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
79 while (dis.available() > 0) {
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
80 long l = di.readLong();
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
81 long offset = entryIndex == 0 ? 0 : (l >>> 16);
420
6c22bdc0bdfd Respect long offsets in revlogs
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 394
diff changeset
82 int flags = (int) (l & 0x0FFFF);
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
83 int compressedLen = di.readInt();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
84 int actualLen = di.readInt();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
85 int baseRevision = di.readInt();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
86 int linkRevision = di.readInt();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
87 int parent1Revision = di.readInt();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
88 int parent2Revision = di.readInt();
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
89 byte[] buf = new byte[32];
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
90 di.readFully(buf, 12, 20);
168
dd525ca65de8 Stupid defect that cost me couple of hours. What a coincidence for merge revision of interest (157) to be right on the BIS buffer boundary.
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 157
diff changeset
91 dis.skipBytes(12);
dd525ca65de8 Stupid defect that cost me couple of hours. What a coincidence for merge revision of interest (157) to be right on the BIS buffer boundary.
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 157
diff changeset
92 // CAN'T USE skip() here without extra precautions. E.g. I ran into situation when
dd525ca65de8 Stupid defect that cost me couple of hours. What a coincidence for merge revision of interest (157) to be right on the BIS buffer boundary.
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 157
diff changeset
93 // buffer was 8192 and BufferedInputStream was at position 8182 before attempt to skip(12).
dd525ca65de8 Stupid defect that cost me couple of hours. What a coincidence for merge revision of interest (157) to be right on the BIS buffer boundary.
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 157
diff changeset
94 // BIS silently skips available bytes and leaves me two extra bytes that ruin the rest of the code.
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
95 System.out.printf("%4d:%14d %6X %10d %10d %10d %10d %8d %8d %040x\n", entryIndex, offset, flags, compressedLen, actualLen, baseRevision, linkRevision, parent1Revision, parent2Revision, new BigInteger(buf));
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
96 String resultString;
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
97 byte[] data = new byte[compressedLen];
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
98 if (inlineData) {
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
99 di.readFully(data);
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
100 } else if (needRevData) {
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
101 dataStream.position(offset);
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
102 dataStream.read(ByteBuffer.wrap(data));
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
103 }
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
104 if (needRevData) {
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
105 if (compressedLen == 0) {
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
106 resultString = "<NO DATA>";
3
24bb4f365164 Rudimentary log functionality with basic infrastructure is in place
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 2
diff changeset
107 } else {
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
108 if (data[0] == 0x78 /* 'x' */) {
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
109 Inflater zlib = new Inflater();
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
110 zlib.setInput(data, 0, compressedLen);
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
111 byte[] result = new byte[actualLen*2];
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
112 int resultLen = zlib.inflate(result);
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
113 zlib.end();
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
114 resultString = buildString(result, 0, resultLen, baseRevision != entryIndex, dumpDataFull);
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
115 } else if (data[0] == 0x75 /* 'u' */) {
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
116 resultString = buildString(data, 1, data.length - 1, baseRevision != entryIndex, dumpDataFull);
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
117 } else {
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
118 resultString = buildString(data, 0, data.length, baseRevision != entryIndex, dumpDataFull);
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
119 }
3
24bb4f365164 Rudimentary log functionality with basic infrastructure is in place
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 2
diff changeset
120 }
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
121 System.out.println(resultString);
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
122 }
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
123 entryIndex++;
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
124 }
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
125 dis.close();
376
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
126 if (dataStream != null) {
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
127 dataStream.close();
d45ad07dc94c Allow content dump for separate .i and .d files
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 189
diff changeset
128 }
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
129 //
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
130 }
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
131
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
132 private static String buildString(byte[] data, int offset, int len, boolean isPatch, boolean completeDataDump) throws IOException, UnsupportedEncodingException {
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
133 if (isPatch) {
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
134 DataInputStream dis = new DataInputStream(new ByteArrayInputStream(data, offset, len));
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
135 StringBuilder sb = new StringBuilder();
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
136 sb.append("<PATCH>:\n");
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
137 while (dis.available() > 0) {
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
138 int s = dis.readInt();
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
139 int e = dis.readInt();
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
140 int l = dis.readInt();
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
141 sb.append(String.format("%d..%d, %d", s, e, l));
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
142 if (completeDataDump) {
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
143 byte[] src = new byte[l];
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
144 dis.read(src, 0, l);
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
145 sb.append(":");
533
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
146 sb.append(escape(new String(src, 0, l, "UTF-8")));
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
147 } else {
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
148 dis.skipBytes(l);
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
149 }
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
150 sb.append('\n');
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
151 }
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
152 return sb.toString();
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
153 } else {
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
154 if (completeDataDump) {
533
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
155 return escape(new String(data, offset, len, "UTF-8"));
392
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
156 }
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
157 return String.format("<DATA>:%d bytes", len-offset);
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
158 }
656a6c1346ff Extra debug option (dumpDataStats) and patch structure decoding for RevlogDump debug utility
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 376
diff changeset
159 }
533
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
160
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
161 private static Pattern controlCharPattern = Pattern.compile("\\p{Cntrl}");
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
162 // \p{Cntrl} A control character: [\x00-\x1F\x7F]
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
163 private static String[] replacements = new String[33];
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
164 static {
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
165 for (int i = 0; i < 32; i++) {
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
166 // no idea why need FOUR backslashes to get only one in printout
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
167 replacements[i] = String.format("\\\\%X", i);
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
168 }
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
169 replacements[32] = String.format("\\\\%X", 127);
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
170 }
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
171 // handy to get newline-separated data printed on newlines.
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
172 // set to false for non-printable data (e.g. binaries, where \n doesn't make sense)
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
173 private static boolean leaveNewlineInData = true;
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
174
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
175 private static String escape(CharSequence possiblyWithBinary) {
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
176 Matcher m = controlCharPattern.matcher(possiblyWithBinary);
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
177 StringBuffer rv = new StringBuffer();
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
178 while (m.find()) {
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
179 char c = m.group().charAt(0);
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
180 if (leaveNewlineInData && c == '\n') {
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
181 continue;
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
182 }
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
183 int x = (int) c;
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
184 m.appendReplacement(rv, replacements[x == 127 ? 32 : x]);
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
185 }
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
186 m.appendTail(rv);
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
187 return rv.toString();
e6f72c9829a6 Generate patches using diff algorithm
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 420
diff changeset
188 }
0
dbd663faec1f Basic changelog parsing
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
189 }