Open hxnoyd opened 9 years ago
It's the other way around, totalhash published it first. And yes, I suspect this can be done. I can probably do it after I finish my current batch of changes.
Thanks for the correction Wesley. Looking forward to try pehash Yara rules.
I've got this mostly done. Just need to put it through some more tests.
Thanks for the feedback!
I have to hold off on publishing this work, because it turns out that both the Viper and totalhash implementations are wrong, with respect to the paper, which means I have nothing to test against. I'm working with the author of the paper to ensure my implementation is accurate before I commit it.
If you want details of why they are wrong please see https://gist.github.com/wxsBSD/07a5709fdcb59d346e9e.
Interesting. The downside is that we'll end up having different implementations, and that will be very confusing for people already familiar with totalhash/Viper's one. It's a shame they didn't get right from the beginning.
Agreed. Once I have an implementation that is accurate to the paper and produces "correct" results according to the author of the paper I will commit it and also fix both Python implementations. I'm not sure when that will be though.
This is an enhancement request. Is it possible to compute pehash via pe module in the same way imphash is? Viper's implementation seems pretty solid and is already used by totalhash: https://github.com/botherder/viper/blob/master/modules/pehash/pehasher.py Thanks in advance.