Add support for published dataset with transform on read

[gra-moore]

We have several use cases where standards such as ICAR and OGC are adopting UDA like protocols for data sharing. While they support the core semantics of the push / pull protocol the entity representation is more specialised than the EGDM instances. To support this and other datasets like publishing endpoints there is a need for a dataset or endpoint that allows more flexible behaviour.

Use cases are such that:

1. Create an alternate syntactic view over a dataset but support UDA semantics
2. Arbitrary data transforms and query parameters that can be used to build richer API experiences hosted and secured by the datahub.