special character problem with standarad hash generators

[BringInTheCat]

Hopefully this is the correct area ^^. I have searched the forum but couldn’t find any help for my problem and google wasn’t successful with that either, so it was time to make a post.

I am trying to crack some self-created md5 hashes (the hash type doesn’t really matter for me, it’s just an example/test) from words with special characters like könig, lözt, €uro, wutbürger, Ömmaken. The hashes have been created with different websites and checked with Linux’s (mint 17) md5sum command (echo –en | könig md5sum) to "verify" the correctness.

Linux Mint, http://www.md5hashgenerator.com/ and many other websites show the same hash for the words listed above while http://md5-generator.de/ creats a different result AND this is only one that actually works with oclHashcat.

I have also read about problems with special character sets reported at http://hashcat.net/forum/archive/index.p...-2763.html and that md5 hashes from the md5hashgenerator website are wrong, most likely due encoding problems but then this would be true for all the other website (with one exception) and Linux Mint/Ubuntu.

While using the ?b option for the charset, some of the hashes are cracked correctly but others have strange characters (the encoding seems to be wrong (and yes I set my hex editor (wxHexEditor) to Windows CP1252 Ansi) ;-))

My Question is rather simple. How do I get the md5 hashes from the other sources solved with oclHashcat?






========================================================================

Here’s the configuration of my System, but that’s not the problem from my point of view.

Win7x64 PC with AMD HD 7800 (Catalyst 14.9) and oclHashcat-1.32

And that's the code I ran with a batch script (removed some gpu options (temperature stuff, accel, loops) for a better overview) just in case someone asks.

"C:\Users\%username%\Downloads\oclHashcat-1.32\oclHashcat64.exe" --potfile-disable -m 0 -a 3 -i --increment-min=1 --increment-max=8 --custom-charset1="C:\Users\%username%\Downloads\oclHashcat-1.32\charsets\standard\German\de_cp1252.hcchr" --custom-charset2=?1?a "C:\Users\%username%\Desktop\ocl\md5\md5.txt" ?2?2?2?2?2?2?2?2

[atom]

If you know for example that the hash has been generated using utf8 you could run your wordlist through iconv. hashcat does not know about encoding, it reads the data bytewise. If you feed it with utf8 data, it will crack utf8, if you feed it with iso-5589-1 it will crack iso-8859-1. There's also a possibility to crack utf-8 with masks, i think rurapenthe wrote about this on his blog some time ago

[BringInTheCat]

Thank you for your answer, I appreciate your help – problem solved. There was also an entry in the wiki, which I must have overlooked (shame on me).

The crucial information was that oclHashcat does not know anything about encoding and the bytewise reading. The approach from rurapenthe link is an interesting way and I understand what he’s doing but couldn’t figure out why it works.


I will make here a short summary for other people with the same problem:
Just put your character set (everything form the ?a option with the special stuff like ä, ö, ü) into a file and save it with the proper encoding (e.g. in linux: leafpad -> save as and choose the encoding you want, windows: notepad++ just change the encoding and save).

Always more fun under Linux is to use the command line ^^

To convert a wordlist with iconv just use:
iconv -f UTF-8 -t ISO-8859-15 wordlistUTF8 > wordlistISO885915

to calculate the hashes (e.g. md5) use:
while read line; do echo -n $line | md5sum | cut -c 1-32; done < wordlistISO885915 >> md5wordlistISO885915