Luigi Auriemma

any ideas ?

I have looked at the file and I have tried to cut the first 0x40 bytes but nothing good in the result.
in short the only requirement in using comtype_scanner is that the compressed data starts exactly at offset 0 of the file.

it's a scanner of algorithms, not a scanner of offsets :)

hmmm, it's not working even with simply files

http://rapidshare.com/files/448317648/e ... rontend.ff

Even if I removed first 12 bytes header and start scanner, it's extract a lot of crap files with nothing...

I have downloaded that file, removed the first 12 bytes of the header and passed to comtype_scan2.bat and it works perfectly as expected: 1.dmp contains the decompressed data.

also 167.dmp and 170.dmp contain the same data because they are zlib too.

do you mean that the file generated on your computer have all a size of 0 bytes?

nope. It's extract about 100 file with different data. Most of them are full with NULL bytes. Some of them containts unreadeble data.

UPDATE. ahh, I'm checked carefully. As u said, 1.dmp, 167.dmp and 170.dmp are the same files and contains true uncompressed data. But what about other files ? I mean why scanner/extractor extract other files too, with NULL bytes.

just because the script tries to extract anything using each algorithm supported by quickbms so the output can be anything or even a crash, it's up to the user to check each file and know what's the good one.
this is a thing for skilled developers who want to know if it's used a known algorithm in less than one minute without reversing the game, NOT final users

Thx for detail explanation !