SPARQL Endpoint?

[Tim McIver]

I can't seem to find out whether or not Trellis has a built-in SPARQL endpoint.  Does it have one and, if so, where is it accessed?  The LDP spec mentions SPARQL several times but it's not clear to me if it's actually a part of the spec.

Tim

[Aaron Coburn]

Hello Tim,

SPARQL is actually two specs: SPARQL-Query and SPARQL-Update. When LDP refers to SPARQL, it is referring to SPARQL-Update, which Trellis supports. A SPARQL endpoint, however, falls under SPARQL-Query, which is not part of LDP and not part of Trellis.

If you would like to add a SPARQL endpoint to a Trellis server, there is a mechanism for doing so via the notification engine. If you are familiar with Camel, there are some pre-built integrations at https://github.com/trellis-ldp/camel-ldp-recipes, but pretty much any integration platform should be workable. If you do build a SPARQL endpoint, you will just want to consider the security and performance implications of making that public; typically, you'd want to keep that internal or else populate it only with resources that are public.

Cheers, Aaron

On Thu, 6 Feb 2020 at 10:39, Tim McIver <t...@timmciver.com> wrote:

I can't seem to find out whether or not Trellis has a built-in SPARQL endpoint.  Does it have one and, if so, where is it accessed?  The LDP spec mentions SPARQL several times but it's not clear to me if it's actually a part of the spec.

Tim

--
You received this message because you are subscribed to the Google Groups "Trellis LDP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trellis-ldp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/trellis-ldp/2795be59-853f-4607-9817-fea582245af1%40googlegroups.com.

[Tim McIver]

Aaron,

Thanks for the quick response as usual.  I'll check out the camel integrations.  Are you aware of any blogs/articles that walk through the process of adding a SPARQL endpoint to Trellis?

To view this discussion on the web visit https://groups.google.com/d/msgid/trellis-ldp/CAD4uyLdBVuKg7JOUP8HByMizTktxjadE18dHHZg4RUvSTSc3yw%40mail.gmail.com.

[Soroka, Adam]

In addition to the approach Aaron describes using Camel to provide asynchronous indexing via notifications, which is certainly a powerful and scalable model, you can use a datastore that supports SPARQL Update and Query (and almost any modern triplestore will) as the backend to Trellis for RDF storage. This would allow you to use the backend directly to support SPARQL with no intervening Camel, but it may limit throughput or scale compared to other backends, and it would mean that users doing queries compete with users doing updates for resources.

I would emphasize Aaron's point about what you expose to the world this way-- would you open a SQL endpoint that supports an important application to the public Internet? At the very least, it's wise to ensure that only Query is available to the public, and that Update is kept safely tucked away.

---
A. Soroka
Research Computing : Office of the CIO : the Smithsonian Institution

> On Feb 6, 2020, at 10:52 AM, Aaron Coburn <aco...@apache.org> wrote:
>
> External Email - Exercise Caution

> To view this discussion on the web visit https://groups.google.com/d/msgid/trellis-ldp/CAD4uyLdBVuKg7JOUP8HByMizTktxjadE18dHHZg4RUvSTSc3yw%40mail.gmail.com.