you can use the reimport feature of quickbms if your new file has the same size or smaller than the original one (section 3 of quickbms.txt)

are you sure I don't need to remove also 0x200 of the header to get the correct offset?
I mean: offset = pak_size - (files * 0x150) - 0x200

because if I do only "pak_size - (files * 0x150)" then we have an overlapping of the last 2 entries

crypto++ uses the common algorithms like the others so signsrch will get them.

the only problem is with those that don't use static tables and constants like rc4

seriously I don't understand. in the specifications about the file table you wrote: Offset: (Total Files * 0x150 + 0x3D0) from end of file and it also encrypted. and indeed my script does: get OFFSET asize # full file size math OFFSET -= SIZE # less the size of the table math OFFSET -= 0x3d0 # less ...

script fixed, let me know the other problems

I noticed the EXE parsing code and so I remembered I did something for another project.
I guess you may find the following script interesting and maybe even useful in some occasions:
http://aluigi.org/papers/bms/others/parse_exe.bms

tested with Minecraft skins1.pck of the DLC: # Minecraft xbox360 PCK # tested with skins1.pck # script for QuickBMS http://quickbms.aluigi.org endian big get DUMMY long get NAMES long for i = 0 < NAMES get DUMMY long get NAMESZ long math NAMESZ *= 2 getdstring NAME NAMESZ set NAME unicode NAME get D...

why not starting to put some bms code on the thread?
I don't have files to test so I have simply followed what I have understood, correct everything you see wrong:

*edit*
new scripts in the next posts

thanx, I have updated the script.
it was just a RPMC file out from the archive so an easy job

why you would do it?

that sequence of bytes contains info about the compression like the used algorithms and the compression level.
note that it doesn't affect the decompression algorithms.

you did nothing wrong, but ttgames.bms is not valid for the reimport operations because the lz2k algorithm is not supported in recompression mode and the archives use chunks

can someone upload one of these language files?

thanx for the info, I have updated my script in case someone is interested

with these files the situation is a bit boring because the uncompressed archives (no AsuraCmp and no AsuraZlb) are valid for the rebuilding without problems. instead for the AsuraCmp/Zlb archive you should first unpack them and then you can replace them with the unpacked archive. luckily doing that ...

it uses lzma compression but the dat files are encrypted probably with aes or another block cipher: comtype lzma get EXT extension if EXT == "dat" print "encryption not supported" elif EXT == "ccdat" print "encryption not supported" else get SIZE long savepos ...