Update code in similarity.py

[dschult]

The code in similarity.py uses some out of date Python constructions and would be IMO clearer and easier to maintain if it was updated.

Looking through the code for match_edge while reviewing #5516 I saw some rewrites that could take advantage of current Python features. For example, there are many list functions calls which should be made into list comprehensions. Similarly the zip(range(m), g_ind) and similar structures should either use enumerate or avoid indexing completely. e.g. for g in g_ind instead of for i, g in zip(range(m), g_ind). Furthermore, the match_edge function (and likely others) would be easier to read IMO if we handled the case of an empty matched_uv up front. That case falls through the logic, executing various else clauses when we know what will happen to it. And there should be no default value on the keyword argument matched_uv.

[kpetridis24]

I will start working on this as well! @dschult @MridulS