Extraction and unpacking of game archives and compression, encryption, obfuscation, decoding of unknown files
Golokopitenko
Posts: 5Joined: Fri Mar 09, 2018 9:27 pm
Post
by Golokopitenko Fri Mar 09, 2018 9:37 pm
Hi! I made a thread
(link) at the ZeNTaX forums, which so kindly helped me to extract the files contained in Techland's 2001 RTS Mission Humanity.
The code used to extract the files contained in the .rsr using the .nam (which acts like an index?) is this:
Code: Select all
# script for QuickBMS http://aluigi.altervista.org/quickbms.htmThe problem is that it seems to stop extracting files very early, and only extracts a very small fraction of all the resources. At ZeNTaX I was told that it seemed that the problem was related to the compression used by the .rsr file, and that someone called Aluigi could help.
Link with the files:
https://www.mediafire.com/folder/tuaws9 ... 82w/shared EDIT: I just realised aluigi is the site's admin. Heh.
EDIT II: fixed the code "quote"
Last edited by
Golokopitenko on Thu Mar 15, 2018 7:44 pm, edited 1 time in total.
aluigi
Site Admin
Posts: 12984Joined: Wed Jul 30, 2014 9:32 pm
Post
by aluigi Fri Mar 09, 2018 11:19 pm
There are at least 4 algorithms compatible with that one, but all of them can't handle the compressed data of some files:
unlzwx, SCUMMVM20, KB_LZW and ultima6.
No idea what's the customization used in the game.
I leave the work-in-progress code mainly because it's more readable than the one you posted
Code: Select all
comtype unlzwx # SCUMMVM20, KB_LZW, ultima6
Acewell
Posts: 706Joined: Fri Aug 08, 2014 1:06 am
Post
by Acewell Sat Mar 10, 2018 2:47 am
aluigi wrote: There are at least 4 algorithms compatible with that one, but all of them can't handle the compressed data of some files:
and milestone_lzw
aluigi wrote: I leave the work-in-progress code mainly because it's more readable than the one you posted
nah, that was my draft script and is perfectly readable, Golokopitenko is just not using code tags.
this is what i have now and is currently set to extract compressed files until a solution is discovered.
Code: Select all
# script for QuickBMS http://aluigi.altervista.org/quickbms.htm
Golokopitenko
Posts: 5Joined: Fri Mar 09, 2018 9:27 pm
Post
by Golokopitenko Sat Mar 10, 2018 10:12 am
Acewell wrote: nah, that was my draft script and is perfectly readable, Golokopitenko is just not using code tags.
Oops! Sorry about that
aluigi wrote: There are at least 4 algorithms compatible with that one, but all of them can't handle the compressed data of some files:
So if I understand, there are at least 4 compression algorithms that seem to be compatible with the .nam/.rsr files, but some of the resources they contain are incompatible with those same algorithms?
And what do you mean by customization used in game?
Also, thanks for your effort!
EDIT: I tried Acewell's code again, and it seemed to work, it seemed to extract all the files
Code: Select all
00000478 1172 mh\fnt\PFMENU0.FNTHowever, it seems that the extracted files are corrupted.
aluigi
Site Admin
Posts: 12984Joined: Wed Jul 30, 2014 9:32 pm
Post
by aluigi Sat Mar 10, 2018 1:50 pm
Golokopitenko wrote: However, it seems that the extracted files are corrupted.
Because it performs no decompression.
Golokopitenko
Posts: 5Joined: Fri Mar 09, 2018 9:27 pm
Post
by Golokopitenko Sat Mar 10, 2018 2:57 pm
aluigi wrote: Golokopitenko wrote: However, it seems that the extracted files are corrupted.
Because it performs no decompression.
It's weird because I tried using the first code Acewell made, which only extracted a few files, but they were not corrupted. Using the same code now extracts all of them but corrupted. Why is that?
And how could we fix the compression issue?
BCGhost
Posts: 35Joined: Fri Dec 15, 2017 1:42 pm
Post
by BCGhost Sat Mar 10, 2018 4:30 pm
Golokopitenko wrote: It's weird because I tried using the first code Acewell made, which only extracted a few files, but they were not corrupted. Using the same code now extracts all of them but corrupted. Why is that?
Because the first script performs a decompression on the data whereas it doesn't work for all files with that algorithm, while the second one simply splits the raw compressed data from the .rsr archive into a bunch of small files. Before you solve the compressions you can't do too much with either the script. So just be patient with that.
aluigi
Site Admin
Posts: 12984Joined: Wed Jul 30, 2014 9:32 pm
Post
by aluigi Sat Mar 10, 2018 9:54 pm
BCGhost wrote: So just be patient with that.
To be honest I don't think someone will ever work on that.
For sure it's not something related to the lzw bits (starting amount of bits for codewords and maximum bits) and not related to the starting size of the buffer. u6decode.cc is quite flexible for these tests, and all of them failed.
Probably it's something related to the size of the compressed data, like resetting the compressor every X bits/bytes... just guessing.
BCGhost
Posts: 35Joined: Fri Dec 15, 2017 1:42 pm
Post
by BCGhost Sun Mar 11, 2018 4:32 am
aluigi wrote: To be honest I don't think someone will ever work on that.
I know. Not everyone interested in that game happens to be an expert in compression after all.
aluigi wrote: For sure it's not something related to the lzw bits (starting amount of bits for codewords and maximum bits) and not related to the starting size of the buffer. u6decode.cc is quite flexible for these tests, and all of them failed.
I literally have no idea what you're talking about.
But if even for you it cannot be solved, then I guess we'll have to give up.
Golokopitenko
Posts: 5Joined: Fri Mar 09, 2018 9:27 pm
Post
by Golokopitenko Sun Mar 11, 2018 12:00 pm
aluigi wrote: To be honest I don't think someone will ever work on that.
Please don't say that! D: I have hoped to see those files ever since I got this game back in 2001. I have a deep emotional link with this little gem which little know and fewer appreciate...
If it's of some use, one of the properly extracted files weights 39,104 bytes, while the same file, when improperly extracted (when the script extracts everything but does not decompress) weights 23,872 bytes. The compressed file is 38.95% smaller. Does this help...?
