ENH: Consider (!) dedicated link to stratigraphic column

[perolavsvendsen]

Currently, each data object which represents a stratigraphic entity (horizon, zone) is linked to the stratigraphic column via data.name. Since the "name" of an object within FMU can be completely arbitrary, we use a lookup function which matches "FMU-name" with "Official name", and replaces. In the stratigraphic column, then name is the identifier.

Example:

In global_variables:

stratigraphy:
  Viking_Top:
    name: VIKING Gp. Top

When exporting e.g. a surface representing "Viking Top", the name will initially be "Viking_Top". We do a lookup in stratigraphy, and if "Viking_Top" is found, we replace the name with the official one. Resulting metadata looks like this:

data:
  name: VIKING Gp. Top
  stratigraphic: true

Similarly, if we export "Some Other Name" which is NOT in the stratigraphy dictionary, resulting metadata becomes:

data:
  name:  Some Other Name
  stratigraphic: false

Meanwhile, other references to master data sits in the masterdata block, e.g.

masterdata:
  smda:
    field: ...
    stratigraphic_column: ...

Suggestion for discussion

Instead of using data.name as the identifier, we use an explicit one. I.e. we create something like:

masterdata:
  smda:
    stratigraphic_column: <uuid4> # id of the strat column in SMDA
    stratigraphic_entity:
      identifier: <Name of the Horizon> # 🆕

This may be easier to both understand and use further. It is not a great pattern than we have a bool ("stratigraphic") which determines if data.name is an actual identifier, or just a string.

[tnatt]

related #401