lzip/lziprecover
1
0
Fork 0
Code Issues Activity

Adding upstream version 1.19.

Browse source 
Signed-off-by: Daniel Baumann <daniel@debian.org>
This commit is contained in:
Daniel Baumann 2025-02-21 11:27:50 +01:00
parent f46d8ce0c8
commit d7ceba2005
Signed by: daniel
GPG key ID: FBB4F0E80A80222F
31 changed files with 1468 additions and 963 deletions
Download patch file Download diff file Expand all files Collapse all files

78
NEWS
View file

@ -1,71 +1,17 @@
Changes in version 1.18:
Changes in version 1.19:
The option "-a, --trailing-error", which makes lziprecover exit with
error status 2 if any remaining input is detected after decompressing
the last member, has been added.
'--merge' is now able to fix files with thousands of scattered errors
per member by grouping the errors into clusters and then merging the
files as if each cluster were a single error.
"--merge" now detects identical files by their CRC.
The option '-a, --trailing-error' now works with '-l, --list' and
'-D, --range-decompress'.
"--repair" now tries to detect gross damage in the file before
attempting to repair it.
The output of option '-l, --list' has been simplified to make it easier
to read.
"--repair" now can repair a damaged dictionary size in the header.
In test mode, lziprecover now continues checking the rest of the files
if any input file is a terminal.
"--repair" now tries bytes at member offsets 7 to 11 first because
errors in these bytes sometimes can't be detected until the end of the
member.
Decompression time has been reduced by 2%.
When decompressing or testing, up to 6 bytes of trailing data are
printed if "-vvvv" is specified.
The test of the value remaining in the range decoder has been removed.
(After extensive testing it has been found useless to detect corruption
in the decompressed data. Eliminating it reduces the number of false
positives for corruption and makes error detection more accurate).
When decompressing, the file specified with the '--output' option is now
deleted if the input is a terminal.
"--merge", "--range-decompress", "--repair" and "--split" now preserve
dates, permissions, and, when possible, ownership of the files created
just as "--decompress" does.
Dictionary size and size of trailing data (if any) are now printed when
"-lv" is specified.
The new option "-A, --alone-to-lz", which converts lzma-alone files to
lzip format without recompressing, just adding a lzip header and
trailer, has been added. Only streamed files with default LZMA
properties can be converted; non-streamed lzma-alone files lack the end
of stream marker required in lzip files.
The new option "-W, --debug-decompress=<pos>,<val>", which sets the byte
<pos> to the value <val> and then decompresses to stdout the resulting
corrupt data, has been added.
The new option "-X, --show-packets", which shows the LZMA packets
(coding sequences) coded in a given file, has been added.
The short name of option "--debug-delay" has been changed to "-Y".
The short name of option "--debug-repair" has been changed to "-Z".
The new options "-B, --block", "-d, --delta", "-t, --truncate" and "-z,
--zcmp" have been added to unzcrash.
Unzcrash now can read files as large as RAM allows.
Unzcrash now compares the output of the decompressor for the original
and corrupt files when the decompressor returns with zero status. For
this unzcrash needs a 'zcmp' program able to understand the format being
tested. For example the one provided by 'zutils'.
Unzcrash now accepts negative position (relative to the end of file) and
negative size (relative to the rest of the file).
The new chapter "Trailing data" has been added to the manual.
A harmless check failure on Windows, caused by the failed comparison of
a message in text mode, has been fixed.
Trailing data are now shown both in hexadecimal and as a string of
printable ASCII characters.