john-users - unicode conversion in NT

[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Date: Mon, 27 Sep 2010 20:44:37 +0200
From: "Magnum, P.I." <rawsmooth@...dband.net>
To: john-users@...ts.openwall.com
Subject: unicode conversion in NT_fmt ?

Hi folks

I'm experimenting with alternate charset conversion for the various MS 
hashes. All of NetLMv2, NetNTLM, NETNTLMv2 and MSCHAPv2 use the same 
function (or an identical copy of it) for conversion to UTF-16, namely 
mbstowcs(), which blindly assumes the input is ISO-8859-1 or ASCII. 
These are isolated functions that can easily be replaced with something 
else. I think I have something working for UTF-8 wordlists (let's ignore 
the problems with rules applied to UTF-8 for now) but have yet to 
produce test hashes to verify (but they do pass the self tests). I'll 
post a patch if/when I have something that works.

But the NET_fmt.c I just can't understand. I get the feeling it is 
optimized to the point we can't just replace a function like that. Is it 
even possible to point your finger at a line in that code where the 
"conversion" happens? Is it just a bit shift? The more I look at that 
code, the more lost I get.

I guess the same goes for mscash_fmt.c, that code looks very similar.

Any help appreciated.

magnum