Support Hydra pagination

[ddeboer]

As seen for instance on https://opendata.picturae.com/catalog.ttl.

May also be relevant for Nationaal Archief.

Or support N-Quad Streaming?

Each page must be complete in itself, for example have an Organization and its properties for the publisher predicate.

[ddeboer]

Comunica already seems to support Hydra:

comunica-sparql 'https://opendata.picturae.com/catalog.ttl' 'select * { ?s ?p ?o}'

^ Does not work because <https://opendata.picturae.com/catalog.ttl?page=1> a hydra:PagedCollection.

comunica-sparql 'https://opendata.picturae.com/catalog.ttl?page=1' 'select * { ?s ?p ?o}'
[
...
]

returns all datasets, including those from following pages. However, we want to dereference per page to reduce memory usage (#384).

[coret]

Picturae (or rather CKAN) has implemented hydra.
Hub3 (NA+CollectieNederland) will implement hydra.

[ddeboer]

@coret Do you have the URL for one or more paginated Hub3 collections? That way I can test at least two providers.

I found https://data.collectienederland.nl/doc/datacatalog but that isn’t using Hydra as of now.

[ddeboer]

This is currently blocked by comunica/comunica#1180.