Fix encoding errors

[NeffIsBack]

Better error handling when reading a file with some non utf-8 characters.

Now you can add --ignore-decoding-errors to just skip non utf-8 characters in your password file.

Issues:
byt3bl33d3r/CrackMapExec#806
byt3bl33d3r/CrackMapExec#808
byt3bl33d3r/CrackMapExec#356

Fixing coding error when password uses €

Issue: byt3bl33d3r/CrackMapExec#786

[Marshall-Hallenbeck]

It's interesting that rockyou isn't utf-8...

The question comes down to "should we check the file encoding and fail out if its not utf-8, or silently remove non-utf-8 chars?"

[XiaoliChan]

Think we can ignore the decode error by default, don't need a new argument, just tell the users in self.logger.error(f"{password} decode error")

[NeffIsBack]

Think we can ignore the decode error by default, don't need a new argument, just tell the users in self.logger.error(f"{password} decode error")

That would be possible, but imo telling the user in advance would be better. Otherwise it is easy to miss, especially if you take something like rockyou (where you anyway have to wait >30h). I think that would have more or less the same effect as ignoring bad characters

[XiaoliChan]

That would be possible, but imo telling the user in advance would be better. Otherwise it is easy to miss, especially if you take something like rockyou (where you anyway have to wait >30h). I think that would have more or less the same effect as ignoring bad characters

Yeah yeah, if loading a huge passwords list and users wait for a long time then it break with decode error, that mind be annoy XD

[XiaoliChan]

I don't know which codec windows(maybe I should said the six lib in impacket ) will accept it except utf-8

Maybe --password-codec args is better?

[NeffIsBack]

I don't know which codec windows(maybe I should said the six lib in impacket ) will accept it except utf-8

Maybe --password-codec args is better?

Hmm i dont think the error occures because user want to use another codec. I think this is more likely happening when user are not aware that their files contain non utf-8 characters like here. But the six lib is a valid point, not sure if it would even work if we use other codecs.

[Marshall-Hallenbeck]

LGTM