minor bugs with empty_count

[BluMist]

example 1: if I just have a variable named count not associated with an array/object, the rule will trigger. maybe restrict the rule to only check for .count?
var count = 0
return count == 0 (trigger error warning)

as an extension, if I do if num + count > 0, the rule will trigger as well

example 2: (probably undetectable) - classes (ex. NSArray) can have .count but not .isEmpty

[jpsim]

These issues are known, which is why the rule isn't on by default. See #206 for a discussion around that.

[BluMist]

ah, got it, didn't surface in the search.

[marcelofabri]

I don't get why the rule shouldn't check for .count. It wouldn't catch a violation if self is a collection object (and count is used without self.), but I think that's a valid tradeoff.

[jpsim]

Anecdotally I can say that enabling this rule helped find some inefficient zero comparisons in some Array extensions. More false positives but also more true positives.