splitlen shrinks dictionary size?

[eXPeri3nc3]

Hi all!

I'm using splitlen on my dictionaries as mentioned in "Essential performance note" here http://hashcat.net/forum/thread-2543.html

However, I noticed that the size of the dictionary after split and merge decreases. I compared the lines before and after, and the count after sorting decreases as well

Is this supposed to happen?

Thanks!

[Kgx Pnqvhm]

Splitlen stops at 15 characters.
Try ULM.

[Mangix]

You need to recompile splitlen and change these two lines accordingly:

#define LEN_MIN 1
#define LEN_MAX 15

You could also try using a one-liner in bash to accomplish the same thing:

awk '{print length, $0}' < wordlist.dict | sort -n | cut -d ' ' -f 2- > wordlist-sorted.dict

[skalderis]

Just changing #define LEN_MAX 15 doesn't do the trick. Dictionary goes from 16gig -> 4mb.

[Kuci]

You have to edit value after LEN_MAX to 55 or something like that...

[atom]

Just changing #define LEN_MAX 15 doesn't do the trick. Dictionary goes from 16gig -> 4mb.

After changing, you need to recompile the binary

[skalderis]

I changed and recompiled with Dev-C++. It drops me an exe. But then it doesn't do what it suppose to. Maybe I should use different compiler.

[Mangix]

I made you a 64-bit binary using MinGW64. Should work.

https://dl.dropboxusercontent.com/u/1020...litlen.exe

[skalderis]

I made you a 64-bit binary using MinGW64. Should work.

https://dl.dropboxusercontent.com/u/1020...litlen.exe

I appreciate that. But I don't open random exe's from internets Thanks anyway. Found that Dev-C++ still have some 500 known bugs. Canceled Visual Studio 2012 install and booted linux. One liner with gcc and done, working correctly.

[KT819GM]

There is good alternative for dev-c++ like codeblocks which by default use MinGW64 as compiller.