Preface

The first commit milestone from refactoring of NamedExpressions.
Essentially there are only 3 types of NamedExpressions:

1. Alias - user define (implicit or explicit) name
2. FieldAttribute - field from Elasticsearch
3. ReferenceAttribute - a reference to another source acting as an Attribute.

To recap, Attributes form the properties of a derived table an each LogicalPlan has Attributes as output since each one can be part of a query and its result sent to the user.

This change essentially removes the name id comparison so any changes applied to existing functions should work as long as the functions are semantically equivalent.
By removing any property from an Attribute, the various components need to look the original source for comparison which, while annoying, should prevent nodes from getting out of sync due to optimizations.
For the most part, I expect by eliminating secondary aliases, this issue will be isolated and further more, by tweaking the rules of the Analyzer, we could opt for copying the source instead of a ReferenceAttribute (though the latter makes things more robust as working against Attribute is better since it means working against the official output/abstraction).

Actual commit

Restrict named expression hierarchy
Remove different types of Attributes and allow only FieldAttribute, UnresolvedAttribute and ReferenceAttribute.
Updated Analyzer, Verifier and VerifierErrorMessageTests are compiling and passing.
There are still compilation errors down the stream however I'm taking a staged approach to introduce the concept in stages before ending up with final, big PR.