Add ability to somehow map field names to coordinate axis names that vector expects

Describe the potential feature

Vector expects coordinates to have very strict names. For example to have an array with a Momentum4D behavior, the fields can be named pt, eta, phi, and mass but if pt is named "pT" for example, it's broken.
We want to have the ability to add a mapping between the fields present in the array and the coordinates that vector actually needs

Motivation

Often arrays are created directly from root files for example in coffea's schemas. And the root files may call things Pt, Eta, Phi, and Mass for example. What we now do in coffea is inside the schema to alias these branches to pt, eta, phi, and mass that vector expects. It's not a full copy as the memory is shared but it's annoting that the resulting array now has Pt and pt because vector cannot work with the former. A way to tell vector what the mapping is from field names to the names it expects would be cool.

Possible Implementation

No response

Start by reviewing how vector currently interprets coordinate field names and how coffea schemas alias ROOT branches such as Pt, Eta, Phi, and Mass. Define the expected mapping behavior and its scope; the work is done when arrays using alternate field names can provide the same coordinate behavior without requiring those aliases.