LHX
Jump to navigation
Jump to search
NAME
lhx - A file test/recovery tool for LHARC archives
SYNOPSIS
lhx [-command] [@recovery_info] arcname [filename ...]
DESCRIPTION
The newest archive tool out of Japan, LHARC, does a very good job of file
compression but it does not handle damaged archives well at all. LHX is a
tool designed to assist in the recovery of data from damaged LHARC
archives.
LHX can be used to list the contents of an LHARC archive, test the
integrity of an archive, scan a damaged archive to look for recoverable
files, and to extract files from an archive.
The commands are: "l" to list archive contents, "t" to test the integrity
of an archive, "e" to extract files from an archive, and "s" to scan an
archive for possible file headers. The default command is "l".
'arcname' is the name of the archive files to be used. This is the only
required argument. MS-DOS wildcard characters '*' and '?' are allowed. If
no extension is given, .LZH is assumed.
'filename' arguments are files to list/test/extract from the archive. If
no 'filename's are given, all files will be selected. MS-DOS wildcard
characters '*' and '?' are allowed. LHX ignores any pathnames stored in
the archive when selecting files.
LIST COMMAND
The list command generates an 'ls -l' like listing of the selected
files in the archive. Any file headers in the archive with checksum
errors will be marked with a '?'.
> lhx -l test
Listing archive TEST.LZH
--rw- 651 / 331 11 Apr 89 22:01:34 TEST\DIR
? --rw- 1161 / 715 5 Apr 89 09:07:38 REaDME
--rw- 8435 / 8435 12 Apr 89 19:55:38 TEST2.LZH
The file README has a bad header. In this case, the error is due to
the change in the filename.
TEST COMMAND
The test command checks the file headers and file contents of the
selected files in the archive.
> lhx -t test
Testing archive TEST.LZH
TEST\DIR -- data CRC error
REaDME -- bad file header, data OK
TEST2.LZH -- data OK
2 errors detected
The file TEST\DIR has an error in the data portion of the file. LHX
will be able to extract the file, but the data will be bad. The file
README has an error in the file header, but its data is intact. LHX
will be able to extract the file with no errors.
EXTRACT COMMAND
This command extracts the selected files from the archive. If one of
the files exists, the user will be prompted whether or not he wants to
overwrite the file. LHX ignores the path and attribute information in
the archive; the files will be created in the current working directory
with standard attributes.
> lhx -e test
Extracting archive TEST.LZH
TEST\DIR -- extracted with CRC error
REaDME --
file exists; overwrite (y/n) ? y
exacted OK
TEST2.LZH --
file exists; overwrite (y/n) ? n
1 error detected
File DIR did not exist in the current working directory so it was
extracted without query; it is probably bad due to the CRC error. The
files README and TEST2.LZH already exist so LHX asks if they should be
overwritten.
SCAN COMMAND
This command scans damaged archives looking for anything that looks
like it might be a file header. Information about these possible file
headers is reported. This information can be given to LHX test/list/
extract commands via the '@' option so that it can attempt to recover
files from the damaged archive.
> lhx -s test
Scanning file TEST.LZH
@0,{32 | 32},651,1 TEST\DIR
@363,{393 | 393},1161,1 REaDME
@1108,{1141 | 1141},8435,0 TEST2.LZH
@1141,{1175 | 1175},955,1 ADDBFCRC.C
@1758,{1791 | 1791},2560,1 CRCDEFS.C
@2851,{2880 | 2880},5058,1 LHX.C
@4721,{4752 | 4752},15337,1 LZHUF.C
The numbers in each of these output lines are:
the starting offset of the file header,
the starting offset of the file data--calculated from two different
values in the header--the first value is the one that LHX
normally uses. If the values are different and the filename
looks correct, the second number is more likely to be correct.
the output file length, and
the compression flag.
These numbers are used in the same order in the '@' parameter. Any
values that are not supplied to the '@' parameter are read from the
header. If a filename is supplied in the '@' parameter, only one file
will be processed by LHX.
In this example, LHX found the headers for all the files in TEST.LZH
and also found all the headers in TEST2.LZH that was stored
uncompressed in the archive. TEST2 could be directly listed/tested/
extracted by using '@1141'.
SELF-EXTRACTING ARCHIVES
Due to the growing number of "virus" programs that are circulating in
the PC community, people are leery about running self-extracting
archives because they may not be as advertised. LHX can be used to
extract the files from a self-extracting archive.
> lhx -s lharc10e.com
Scanning file LHARC10E.COM
@1290,{1323 | 1323},27498,1 LHARC.EXE
@18852,{18885 | 18885},21984,1 LHARC.MAN
@27919,{27949 | 27949},1161,1 README
This file can be manipulated by LHX like any other archive by using the
"@1290" option:
> lhx -l @1290 lharc10e.com
Listing archive LHARC10E.COM
--rw- 27498 / 17529 4 Mar 89 18:07:22 LHARC.EXE
--rw- 21984 / 9034 5 Apr 89 09:07:28 LHARC.MAN
--rw- 1161 / 715 5 Apr 89 09:07:38 README
> lhx -e @1290 lharc10e.com
Extracting archive ..\LHARC10E.COM
LHARC.EXE -- extracted OK
LHARC.MAN -- extracted OK
README -- extracted OK
A JOURNEY INTO THE ABSURD
Just to show what can be done with LHX, let's play with LHARC10E.COM a
little bit more; first we'll extract LHARC.MAN, still compressed, into
LHARCMAN.Z, then we'll uncompress LHARCMAN.Z into LHARC.MAN.
> lhx -e @18852,,9034,0,lharcman.z lharc10e.com
Extracting archive LHARC10E.COM
lharcman.z -- extracted with CRC error
1 error detected
The CRC error is expected since the CRC in the header is that of the
uncompressed file. LHARCMAN.Z now contains the raw compressed data for
LHARC.MAN. This file can be decompressed by:
> lhx -e @0,0,21984,1,lharc.man lharcman.z
Extracting archive LHARC.Z
lharc.man --
file exists; overwrite (y/n) ? y
extracted with CRC error
1 error detected
The CRC error is expected since there was no header on the file.
(Actually, we told LHX to read the header from the beginning of the raw
data file, but it didn't use any of the data it read.)
ACKNOWLEDGEMENTS
This program was written using code from the Usenet distribution of LZHUF.C
written by Haruyasu Yoshizaki, Haruhiko Okumura, and Kenji Rikitake. This
is copyrighted code with permission granted for non-commercial use.
I also used ADDBFCRC.C and CRCDEFS.C from the Usenet ZOO 2.01 distribution.
Rahul Dhesi claims no copyright on them but I acknowledge him anyway--he
saved me a bunch of typing.
AUTHOR
Mark Armbrust
maa@nbires.nbi.com
Back to Packer/Depacker