Fork me on GitHub
#rdf
<
2017-02-13
>
ricroberts15:02:41

Sesame is now called rdf4j btw

rickmoynihan23:02:58

@nblumoe: Nice to hear from you again 🙂 as ricroberts said http://grafter.org/ is a clojure layer over sesame (will update at some point to RDF4j) — half of grafter is concerned with RDF… with a few bits and pieces in grafter.tabular - to integrate it with incanter - though this side is less well developed — and will soon likely be split out into a separate dependency. The sesame layer is pretty well developed, and mainly provides low level coercion between clojure/rdf datatypes, and access to sesame’s rdf serialisers/deserialisers… I keep meaning to find some time to write some better docs on how to use this side of it. Main focus isn’t on ontologies/semantics though… We also have a yes-sparql style sparql api in latest builds - which supports binding style query generation and coerces RDF types into clojure types. Happy to explain more if you’re interested.

rickmoynihan23:02:00

It started life as an ETL tool to get data from tabular data into RDF… and we still mainly use it for that but probably most of the code which is under grafter.rdf is about raw RDF processing.

rickmoynihan23:02:53

@nblumoe: I know you have been an incanter user/contributor too — basically in the early days of grafter we adopted incanter Datasets so we can retain both the order of columns and access them by name/id… however incanter has a habit of eagerly loading everything — which makes it problematic for large ETL — so the functions in grafter.tabular tend towards lazily loading datasets… which is not without its issues too. Anyway curious what you think the future of incanter is; 1.9/2.0 and the promised implementation in terms of core.matrix protocols seems increasingly less likely to happen.... Is that a fair assessment?