Implement a new, CONSTRUCT-based DataCubeObservations query instead of using rdf-cube-view-query library

[sosiology]

Currently, Visualize uses rdf-cube-view-query library to fetch observations.

We would like to not use the library for this, but rather implement a custom CONSTRUCT query.

It should be noted that we would still need to keep the library to be able to generate SELECT-based SPARQL query that could be previewed in LINDAS YASGUI or remove the “Run SPARQL query” button completely.

edit: the main advantage using a CONSTRUCT query in this context: the query can fire directly to get the observations w/o relying on the cube shape being fetched first. (this pattern that is already implemented in the Preview query)

User story

I would like the DataCubeObservations query to be CONSTRUCT-based, instead of being SELECT-based.

Definition of Done

DataCubeObservations query is written from scratch and uses CONSTRUCT approach, which should result in improved performance of fetching observations.