registered namespaces ignored in sparql construct query results?

[joern...@gmail.com]

Hi,

is there any way to make stardog use the registered namespaces in the returned results sparql construct queries (e.g., TTL or JSON-LD)?

Currently it seems to me that the URIs are always returned in expanded form, which obviously is correct, but sometimes (e.g., in a GUI showing the resulting triples) it would be nice to be able to just use the namespaces of the input TTL files without having to hard-code them in the client logics as well.

Best,

Jörn

[Evren Sirin]

HTTP requests will return results using only the prefixes defined in
the query. At some point we can change this behavior but since the
server returns results in a streaming fashion we don't know which
namespaces are used in the construct output and including all
namespaces every time is not great. If you are using the API then you
can retrieve the registered namespaces from the database (so no
hard-coding on the client-side, just an extra call) and use them to
serialize/display your results.

Best,
Evren

> --
> -- --
> You received this message because you are subscribed to the C&P "Stardog"
> group.
> To post to this group, send email to sta...@clarkparsia.com
> To unsubscribe from this group, send email to
> stardog+u...@clarkparsia.com
> For more options, visit this group at
> http://groups.google.com/a/clarkparsia.com/group/stardog?hl=en

[Ray Rashif]

Could you point me to any SPARQL implementation that returns (not
displays) results using prefixes? I recall most of the triple stores
I've tried show me the prefixes in their endpoint interface (if any),
but give me the fully expanded URIs when I fetch them
programmatically.

On 17 July 2016 at 19:08, <ronal...@gmail.com> wrote:
> I thought the SPARQL standard included the possibility of including
> namespace prefixes in the query itself. In that light, there is absolutely
> no reason why the server couldn't examine those and perform the
> substitutions. This is a poor implementation of the standard if we are now
> prevented from using all the convenience IO methods for outputting results
> since those are rendered almost useless for most practical situations.

> ---
> You received this message because you are subscribed to the Google Groups
> "Stardog" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to stardog+u...@clarkparsia.com.



--
GPG/PGP ID: C0711BF1

[Ron Ayoub]

I'm new to semantic web technology but at this point I'm skeptical since I haven't found anything that matches the literature and every product out there has flaws. Star dog gave me the least trouble but even then has missing entries in the API docs and mangled maven dependencies. Would like to see a product like Postgres in this space. If I recall Virtuoso had expected behavior but it's docs are perhaps the most incoherent I have ever scene for something billed as an enterprise product. I did figure out that this problem can be worked around by using rdfs:label effectively. That is, resource names themselves shouldn't be the part of a query.

> You received this message because you are subscribed to a topic in the Google Groups "Stardog" group.
> To unsubscribe from this topic, visit https://groups.google.com/a/clarkparsia.com/d/topic/stardog/H3RMnqITD_I/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to stardog+u...@clarkparsia.com.
>

[Evren Sirin]

Hi Ron,

Sorry but I'm not following what you are saying. What exactly is the
problem you are seeing?

Best,
Evren