Defining the tree:relation in terms of SPARQL algebra

[constraintAutomaton]

I think it is fair to assume that the intention behind the tree:relation is to use them in the context of SPARQL. The current tree:relation(s) (apart from the geospatial one, but there is an extension to SPARQL for geospatial queries , called GeoSPARQL) are strongly linked to the SPARQL algebra. Thus, to improve the understanding of that relation and to help their implementation inside query engines, it might be a good idea to define them in terms of SPARQL algebra. The task should be simple and not have a big impact on how the current relation has been defined.

I will make soon a proposal.