Fix unicode on RHS in ini scan

[KevinHock]

Changes

• 🐛 Handle unicode in ini_file_parser.py

This bug was trigged only when Unicode was on the RHS e.g.

foo=bår

&

• merged in master to get coverage back up to the minimum (98%)

Why this works

To clarify why unicode_literals fixes this, we format things into a string literal when doing re.compile, so this makes sure it's unicode and not bytes.

[calvinli]

To clarify, the problem this fixes is that a Python 2.7 str cannot be .format()-ed with unicode:

>>> x = u'helló'
>>> r'foo{}'.format(x)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
UnicodeEncodeError: 'ascii' codec can't encode character u'\xf3' in position 4: ordinal not in range(128)
>>>

The pattern must be unicode in order to do this:

>>> x = u'helló'
>>> ur'foo{}'.format(x)
u'foohell\xf3'
>>>

[calvinli]

🚢