Check recovered password last bytes to be in expected character set

[kimci86]

During password recovery, the user-provided character set is used to bruteforce the n-6 first characters. The 6 last bytes are derived from there and they can have any value not necessarily in the character set.

Example originally posted by @mindked in #69 (comment):

I tried recovery password of bkcrack with -r 14 ?a
It found the password after 12 hours of running but there is a strange character. The password didn't work

as bytes: **32 69 93 69
as text: *****iôiU

Combined with tricky encoding issues, the recovered password might be difficult to use.

An improvement would be to check if every byte of the recovered password is in the provided character set.
If they are not, the password would still be printed because it should technically be a valid password anyways, but the search would continue.

This is related to #34, #35, #55.