Can somebody check these errornous files for a possible compression? samples: http://www2.zippyshare.com/v/HarM8toj/file.html The folder "normal" shows how the xp files should look, the folder "error" shows the variant. Extraction script:
endian big get UNK long 0 get UNK long 0 get FILES long 0 get NAMESTART long 0 set NAMEOFF FILES math NAMEOFF *= 0x1c math NAMEOFF += 0x10 math NAMEOFF += NAMESTART # only in some files get FOLDER basename get FSIZE asize 1 for i = 1 <= FILES get CRC long 0 get OFFSET long 0 get OFFSET2 long 0 get SIZE long 0 get FFFF long 0 get UNK long 0 get ZERO long 0 savepos MYOFF 0 goto NAMEOFF 0 get NAMEL long 0 getDstring NAME NAMEL 0 string NAME p= "%s~%s.swa" FOLDER NAME savepos NAMEOFF 0 math OFFSET2 -= 0x800 goto OFFSET2 1 getDstring DUMMY 0x806 1 get TEST byte 1 if TEST != 0xfc string NAME p= "%s_0x%08x" NAME OFFSET2 endif xmath TEST "OFFSET2 + SIZE" if TEST <= FSIZE log NAME OFFSET2 SIZE 1 endif goto MYOFF 0 next i
The xp files are compressed with xbcompress.exe. quickbms can handle them automatically when you specify comtype xmemdecompress.
For some unknown reasons cp_kaz.re.nv.xp doesn't decompress and quickbms remains freeded. I will chek it but that feature has been tested very well some months ago so I don't know why that file gives problems.
As I already said, you are supposed to use xbdecompress.exe to decompress them.
Consider that supporting those files is a feature of quickbms and not part of the xmemcompress algorithm, it's like uclpack for UCL. I will review and try to fix the problem in quickbms 0.7.2
