Help About LZMA Header Format

[Inatna Tsurt]

I apologize for bothering you about this question and I thank you anyway for your time.
I have an old (at least 2-3 years old I think) file compressed with .lzma (maybe created by lzma_alone?).
I have no knowledge about what tool made it but I need to know if it's
a real .lzma format and how to extract it.
I have not found very detailed  info about lzma format header at least on the
header created by lmza_alone tool. The header processing seems to pass the decode tool inizial check..
but after that the decoding fails..with: Decode Error

I really appreciate if someone can give me some url/info/doc about all lzma format (old and new..)
if exist or/and if you have some spare minutes to give me some answers about
the header of my nasty file..:

Here are first 32bytes:

0000000   5d 00 00 40 00 00 08 00  2f 3e 61 3f fb 7d ac 43  |]..@..../>a?.}.C|
00000010  a5 74 50 73 17 c9 55 1f  e6 2a 58 0c 1f d9 6d 4f  |.tPs..U..*X...mO|

The file has lenght of 44708 byte.

I dont know about the byte at offset 0, but i think the byte 1-4 was the dictionary size...

I have some other files that i can cleanly decode that have similar header but with
0x6d as the first byte and 8byte decompressed file size after the dictionary size bytes..

Tnx a lot
- Inatna

[Igor Pavlov]

read about 13-bytes header format ln lzma docs.

First byte  in .LZMA compressed stream must be 0.
So there are two versions about your file:
1) replace 00 08 by FF FF FF FF FF FF FF FF (8 bytes).
2) insert FF FF FF FF FF FF FF FF (8 bytes) before 00 08.

Just check both these versions.

[Kallisto]

Look here: http://tukaani.org/

Quote: 'please read and give feedback about new .lzma file format specification draft'

Seems this is a special format these utilities use. Just download the utilities and try them.