AnyInsteadOfEmpty rule uses wrong node position

[jwoertink]

In Lucky, we have a query method .any?. This actually runs SQL, so Performance/AnyInsteadOfEmpty rule here doesn't apply.

This code passed on 0.14.4, but fails on 1.0.0

TransactionQuery.new
  .receiver(user)
  .any? # ameba:disable Performance/AnyInsteadOfEmpty

src/serializers/post_serializer.cr:80:24 [Correctable]
[C] Performance/AnyInsteadOfEmpty: Use `!{...}.empty?` instead of `{...}.any?`
> TransactionQuery.new
  ^

Maybe solved by #265 ?

[Sija]

That's a regression, and shouldn't depend on #265 in any way.