Wrong Line length for lines with Unicode characters.

[Adi Roiban]

Here is an example for an unicode line, that when split it is encoded

>>> initial_line = u'mâț mițișor:x:2000:2010:Mâț Mițișor,,,,:/home/mâț mițișor:/bin/bash\n'
>>> line_encoded = "u'm\xc3\xa2\xc8\x9b mi\xc8\x9bi\xc8\x99or:x:2000:2010:M\xc3\xa2\xc8\x9b Mi\xc8\x9bi\xc8\x99or,,,,:/home/m\xc3\xa2\xc8\x9b mi\xc8\x9bi\xc8\x99or:/bin/bash\\n'"

>>> len(line_encoded)
84
>>> len(line_encoded.decode('utf-8'))
72

----

I am not sure if always converting long lines to Unicode will solve all problems.

Maybe we can change pocket-lint to be smart and convert only if the Python headers contains an unicode declaration.

I will attach a branch with a naive fix and the required tests.

Please let me know how do you think this problem should be solved.

Cheers,
Adi