ZenHAX

Hello guys !

I'm trying to extract Attack on Titan 2 PC game models, I have successfully unpacked "LINKDATA_A.Bin" with http://aluigi.org/bms/attack_on_titan.bms which is for Attack on Titan : Wings of Freedom.

Then ,seems like it doesn't work for "LINKDATA_B.Bin", I've got this error from quickbms :

Error: the compressed zlib/deflate input is wrong or incomplete (-3)
Info: algorithm 1
offset 000000001978f608
input size 0x00000000000008a2 2210
output size 0x0000000000008000 32768
result 0xffffffffffffffff -1

Error: the uncompressed data (-1) is bigger than the allocated buffer (68157588)

Last script line before the error or that produced the error:
29 clog MEMORY_FILE OFFSET CHUNK_ZSIZE CHUNK_SIZE

I don't know if the file is encrypted .. Here is the filecutter in case someone wants to check, tell me if I did it wrong
Thank you !

Does anyone has found something ? Thank you

up

Use this script on that linkdata file and upload the file it generates:

log "uploadme.dat" 0x1978f608 0x8a2

Thank you Aluigi !

Here is the file

It's not an error of quickbms or the script.
Are you sure that archive is original and it's not corrupted?
The algorithm is correct (zlib) and there is clearly zlib data starting from offset 0 of the uploadme.dat file BUT it's invalid since offzip finds nothing and there is another zlib stream at offset 0x800.

Maybe it's possible that the format has some old entries that must be skipped while the script read them... don't know.
Anyway currently I have no other ideas.

Yes the archive is original and I don't think it's corrupted since it works in game.
I did retry to extract it by redownloading the Attack on Titan script.bms, the archive is about 7gb so I use quickbms_4gb_file,
the thing it extracts a bunch of "000000000000000.dat, 000000000000001.dat" till 00000000000199.dat, then I got the error. When extracting the .dat file, there is some like 000000000000189.dat then 00000000000018a.dat, 00000000000018b etc..

Hello,
I am trying to extract the same LINKDATA_B.bin and running into the same issue.
The data is not corrupted, because I tried with the PC version, and then the PS Vita version (twice), and the script stops after generating the same amount of files.

Error: the compressed zlib/deflate input is wrong or incomplete (-3)
Info: algorithm 1
offset 0291a508
input size 0x0000425e 16990
output size 0x00008000 32768
result 0xffffffff -1

The hexa screenshot if it's any help :


So I am interested in your idea to "skip old entries" but I don't know how to modify the .bms to make a trial.
And sorry I was not able to generate uploadme.dat

The only "workaround" is adding "comtype zlib_noerror" at the beginning of the script.

It will just decompress the invalid data without reporting any error... a sort of skip, yeah.

Okay, I tried and it just ignored all the data after the faulty offset.
So it didn't help me ! Must be an unknown format.
Don't worry, in the 1st place I will just edit what I am able to, and it will already keep me occupied for some time (^,^)_b