Smarter fuzzywuzzy ratio #1

Parsing two similar expecting files where one is a substring of another causes problems when using fuzzywuzzy's partial_ratio.

This is particularly problematic in the _load method of Project when selecting the best matching expecting.

Example

expecting = ['functions.py', 'test_functions.py']
file = 'student123456_test_functions.py'
ratios = [unicity._check_fuzzy_ratio(file, expect) for expect in expecting]

>>> ratios = [100, 100]

Request

Use a different method for fuzzy compare that deals with substring collisions in expecting