1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
|
2026-02-08 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.26 released.
* New option '--append'.
* byte_repair.cc (debug_decompress): Show member-relative positions.
* decoder.cc (check_trailer): With -vv print sizes, not ratios.
* fec_create.cc (write_fec): If input file is larger than RAM / 2,
create all packets concurrently to read the file just once.
(Reported by Phako).
(fec_create): Accept non-regular protected file if --force.
* fec_repair.cc: Test and repair protected files larger than RAM.
(read_fecfile): Return early if no fec_magic found.
(Fec_index): With -i accept a fec file without fec packets.
(fec_df): New function.
(check_prodata): 'size_dif > 0' is not mismatch.
* list.cc (list_files): Print size of tdata below sizes of members.
With -vv print a blank line between files.
* lzip_index.cc (skip_box): New function.
* main.cc (insert_fixed) [!append_lz]: Insert '_fixed' before '.tar'.
(show_help): Print "*Exit status*" for section in man page.
* main_common.cc (getnum): Accept underscore separators ('_').
* Print large numbers with underscore separators ('_').
* reproduce.cc (bursted_pos): Detect more than one burst.
* lziprecover.texi: Document how to use a fec recovery record.
2025-01-08 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.25 released.
* New options '-F, --fec', '-0' to '-9', '-b, --block-size',
'--fec-file', '-r, --recursive', and '-R, --dereference-recursive'.
* Change short name of option '--byte-repair' to '-B'.
* byte_repair.cc: Repair a nonzero first LZMA byte.
* Make '-i' ignore empty members and nonzero first LZMA byte.
* Rename option '--clear-marking' to '--nonzero-repair'.
* Remove options '--empty-error' and '--marking-error'.
* decoder.cc (decode_member): Remove support for Sync Flush marker.
* list.cc (list_files): Detect write error on stdout.
* merge.cc (copy_file): Add name arguments, use 'show_file_error'.
* lziprecover.texi: New chapter 'Syntax of command-line arguments'.
* check.sh: Use 'cp' instead of 'cat'.
* testsuite: Require lzip/clzip. Change several test files.
2024-01-20 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.24 released.
* New options '--empty-error', '--marking-error', '--clear-marking'.
* dump_remove.cc, main.cc: Accept 'empty' in --dump, --remove, --strip.
* main.cc: Rename '--repair' to '--byte-repair'.
Rename '--debug-repair' to '--debug-byte-repair'.
Reformat file diagnostics as 'PROGRAM: FILE: MESSAGE'.
(show_option_error): New function showing argument and option name.
(main): Make -o preserve date/mode/owner if 1 input file.
(open_outstream): Create missing intermediate directories.
* lzip.h: Rename verify_* to check_*.
* repair.cc: Rename to byte_repair.cc.
* unzcrash.cc: Rename '--no-verify' to '--no-check'.
* configure, Makefile.in: New variable 'MAKEINFO'.
* Makefile.in: Remove target 'install-as-lzip'; -e needs a real lzip.
* testsuite: New test files test_3m.txt.lz.md5, fox6_mark.lz.
2022-01-21 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.23 released.
* Decompression time has been reduced by 5-12% depending on the file.
* main_common.cc (getnum): Show option name and valid range if error.
* dump_remove.cc (dump_members): Check tty except for --dump=tdata.
* Option '-U, --unzcrash' now takes an argument ('1' or 'B<size>').
* mtester.cc (duplicate_buffer): Use an external buffer.
* repair.cc (debug_decompress): Continue decoding on CRC mismatch.
* unzcrash.cc: Make zcmp_command a string of unlimited size.
Use execvp instead of popen to avoid invoking /bin/sh.
Print byte or block position in messages.
* New file common.h.
* Improve several descriptions in manual, '--help', and man page.
* lziprecover.texi: Change GNU Texinfo category to 'Compression'.
(Reported by Alfred M. Szmidt).
2021-01-02 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.22 released.
* New options '-e, --reproduce', '--lzip-level', '--lzip-name',
'--reference-file', and '-E, --debug-reproduce'.
* Remove '--dump-tdata', '--remove-tdata', and '--strip-tdata'.
* main.cc (main): Report an error if a file name is empty.
Make '-o' behave like '-c', but writing to file.
Make '-c' and '-o' check whether the output is a terminal only once.
Do not open output if input is a terminal.
* main.cc (decompress): With '-i', ignore data errors, keep files.
* range_dec.cc: '-i -D' now decompresses a truncated last member.
* '-i -D' now returns 0 if only ignored errors are found.
* '-i' now considers any block > 36 with header a member, not a gap.
* Replace 'decompressed', 'compressed' with 'out', 'in' in output.
* Fix several compiler warnings. (Reported by Nissanka Gooneratne).
* lzip_index.cc: Improve messages for corruption in last header.
* New debug options '-M, --md5sum' and '-U, --unzcrash'.
* main.cc: Set a valid invocation_name even if argc == 0.
* Document extraction from tar.lz in manual, '--help', and man page.
* New files lunzcrash.cc, md5.h, md5.cc, nrep_stats.cc, reproduce.cc.
* lziprecover.texi: New chapter 'Reproducing one sector'.
New sections 'Merging with a backup' and 'Reproducing a mailbox'.
Document the debug options for experts.
* check.sh: Lzip 1.16 or newer is required to run the tests.
* testsuite: Add 9 new test files.
2019-01-04 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.21 released.
* Rename File_* to Lzip_*.
* New options '--dump', '--remove', and '--strip'. They
replace '--dump-tdata', '--remove-tdata', and '--strip-tdata',
which are now aliases and will be removed in version 1.22.
* lzip.h (Lzip_trailer): New function 'verify_consistency'.
* lzip_index.cc: Lzip_index now detects gaps between members,
some kinds of corrupt trailers and
some fake trailers embedded in trailing data.
* split.cc: Use Lzip_index to split members, gaps and trailing data.
* split.cc: Check last member before writing anything.
* list.cc (list_files): With '-i', ignore format errors, show gaps.
* range_dec.cc: With '-i', ignore a truncated last member.
* main.cc (main): Check return value of close( infd ).
* Improve and add new diagnostic messages.
* Print '\n' instead of '\r' if !isatty( 1 ) in merge, repair.
* main.cc: Compile on DOS with DJGPP.
* lziprecover.texi: New chapter 'Tarlz'.
* configure: Accept appending to CXXFLAGS; 'CXXFLAGS+=OPTIONS'.
* INSTALL: Document use of CXXFLAGS+='-D __USE_MINGW_ANSI_STDIO'.
* testsuite: New test files fox.lz, fox6_sc[1-6].lz.
2018-02-12 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.20 released.
* split.cc: Fix splitting of files > 64 KiB broken since 1.16.
* New options '--dump-tdata', '--remove-tdata', '--strip-tdata', and
'--loose-trailing'.
* Improve corrupt header detection to HD=3.
* main.cc: Show corrupt or truncated header in multimember file.
* Replace 'bits/byte' with inverse compression ratio in output.
* Show progress of decompression at verbosity level 2 (-vv).
* Show progress of decompression only if stderr is a terminal.
* main.cc: Show final diagnostic when testing multiple files.
* decoder.cc (verify_trailer): Show stored sizes also in hex.
Show dictionary size at verbosity level 4 (-vvvv).
2017-04-10 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.19 released.
* merge.cc: Fix members with thousands of scattered errors.
* Option '-a' now works with '-l' and '-D'.
* The output of option '-l, --list' has been simplified.
* main.cc: Continue testing if any input file is a terminal.
* main.cc: Show trailing data in both hexadecimal and ASCII.
* lzip_index.cc: Improve detection of bad dict and trailing data.
* lzip_index.cc: Skip trailing data more efficiently.
* lzip.h: Unify messages for bad magic, trailing data, etc.
* New struct Bad_byte allows delta and flip modes for bad_value.
* unzcrash.cc: New option '-e, --set-byte'.
2016-05-12 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.18 released.
* New option '-a, --trailing-error'.
* merge.cc (open_input_files): Use CRC to test identical files.
* repair.cc (repair_file): Detect gross damage before repairing.
* repair.cc: Repair a damaged dictionary size in the header.
* repair.cc: Try bytes at offsets 7 to 11 first.
* Decompression time has been reduced by 2%.
* main.cc (main): Delete '--output' file if infd is a terminal.
(main): Don't use stdin more than once.
(decompress): Print 6 bytes of trailing data at verbosity level 4.
* decoder.cc (verify_trailer): Remove test of final code.
* Use 'close_and_set_permissions' and 'set_signals' in all modes.
* range_dec.cc (list_file): Show dictionary size and size of
trailing data (if any) with '-lv'.
* New options '-A, --alone-to-lz', '-W, --debug-decompress', and
'-X, --show-packets'.
* Change short name of option '--debug-delay' to '-Y'.
* Change short name of option '--debug-repair' to '-Z'.
* unzcrash.cc: New options '-B, --block', '-d, --delta',
'-t, --truncate', and '-z, --zcmp'.
* unzcrash.cc: Read files as large as RAM allows.
* unzcrash.cc: Compare output using zcmp if decompressor returns 0.
* unzcrash.cc: Accept negative position and size.
* lziprecover.texi: New chapter 'Trailing data'.
* configure: Avoid warning on some shells when testing for g++.
* Makefile.in: Detect the existence of install-info.
* check.sh: Require a POSIX shell. Don't check error messages.
2015-05-28 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.17 released.
* New block selection algorithm makes merge up to 100 times faster.
* repair.cc: Repair time has been reduced by 15%.
* New options '-y, --debug-delay' and '-z, --debug-repair'.
* Makefile.in: New targets 'install*-compress'.
* testsuite/unzcrash.cc: Move to top directory.
* lziprecover.texi: New chapter 'File names'.
2014-08-29 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.16 released.
* New class LZ_mtester makes repair up to 10 times faster.
* main.cc (close_and_set_permissions): Behave like 'cp -p'.
* lziprecover.texinfo: Rename to lziprecover.texi.
* Change license to GPL version 2 or later.
2013-09-14 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.15 released.
* repair.cc: Repair multimember files with up to one byte error per
member.
* merge.cc: Merge multimember files.
* main.cc (show_header): Don't show header version.
* lziprecover.texinfo: New chapters 'Repairing files',
'Merging files', and 'Unzcrash'.
2013-05-31 Antonio Diaz Diaz <antonio@gnu.org>
* Version 1.14 released.
* New option '-i, --ignore-errors'.
* Option '-l, --list' now accepts more than one file.
* Decompression time has been reduced by 12%.
* split.cc: Use as few digits as possible in file names.
In verbose mode show names of files being created.
* main.cc (show_header): Show header version if verbosity >= 4.
(main): Use 'setmode' instead of '_setmode' on Windows and OS/2.
* configure: Options now accept a separate argument.
* Makefile.in: New targets 'install-as-lzip' and 'install-bin'.
2012-02-24 Antonio Diaz Diaz <ant_diaz@teleline.es>
* Version 1.13 released.
* Lziprecover is now distributed in its own package. Until version
1.12 it was included in the lzip package. Previous entries in this
file are taken from there.
* lziprecover.cc: Rename to main.cc.
* New files merge.cc, repair.cc, split.cc, and range_dec.cc.
* main.cc: Add decompressor options (-c, -d, -k, -t) so that an
external decompressor is not needed for recovery and 'make check'.
* New option '-D, --range-decompress', which extracts a range of
bytes decompressing only the members containing the desired data.
* New option '-l, --list', which prints correct total file sizes
even for multimember files.
* merge.cc, repair.cc: Remove output file if recovery fails.
* Change quote characters in messages as advised by GNU Standards.
* split.cc: Use Boyer-Moore algorithm to search for headers.
* configure: Rename 'datadir' to 'datarootdir'.
2011-04-30 Antonio Diaz Diaz <ant_diaz@teleline.es>
* Version 1.12 released.
* lziprecover.cc: If '-v' is not specified show errors only.
* unzcrash.cc: Use Arg_parser.
New options '-b, --bits', '-p, --position', and '-s, --size'.
2010-09-16 Antonio Diaz Diaz <ant_diaz@teleline.es>
* Version 1.11 released.
* decoder.cc (decode_member): Detect file errors earlier to improve
efficiency of lziprecover's new repair capability.
This change also prevents (harmless) access to uninitialized
memory when decompressing a corrupt file.
* lziprecover.cc: New option '-m, --merge', which tries to produce a
correct file by merging the good parts of two or more damaged copies.
New option '-R, --repair' for repairing a 1-byte error in
single-member files.
New options '-f, --force' and '-o, --output'.
New option '-s, --split' to select the until now only operation of
splitting multimember files.
If no operation is specified, warn the user and do nothing.
2009-06-22 Antonio Diaz Diaz <ant_diaz@teleline.es>
* Version 1.6 released.
* lziprecover.1: New man page.
* check.sh: Test lziprecover.
2009-01-24 Antonio Diaz Diaz <ant_diaz@teleline.es>
* Version 1.4 released.
* Add 'lziprecover', a member recoverer program.
* unzcrash.cc: Test all 1-byte errors.
Copyright (C) 2009-2026 Antonio Diaz Diaz.
This file is a collection of facts, and thus it is not copyrightable, but just
in case, you have unlimited permission to copy, distribute, and modify it.
|
