Sorry, of course you are right – both.
When I looked at the output for the last time, I had a SELECT query
running – thus completely foreseeable getting
application/sparql-results+xml.
What I was originally looking for is to format the result of a CONSTRUCT
query, which as of now comes out quite linear like:
<rdf:Description
rdf:about="http://resources.semaworx.eu/inventory/ActivityStreams/current/">
<rdf:type
rdf:resource="http://namespaces.semaworx.org/inventoryControl/ActivityStreamGroup"/>
(… some ?p ?o here …)
</rdf:Description>
Though, what I have been looking for, is supposed to come out more like:
<rdf:RDF
xmlns:ive="http://namespaces.semaworx.org/inventoryControl/">
<ive:ActivityStreamGroup
rdf:about="http://resources.semaworx.eu/inventory/ActivityStreams/current/">
(… some ?p ?o here …)
</ive:ActivityStreamGroup>
 </rdf:RDF>
I’m quite aware I may have missed a change of defaults here.
All I need to know is learning how to do this from now on.
— Bardo
On 22.04.21 11.09 Uhr, Andy Seaborne wrote:
On 22/04/2021 09:55, Rob Vesse wrote:
What is the query?
RDF/XML is a graph format and so only makes sense for
CONSTRUCT/DESCRIBE queries
ASK/SELECT queries are going to produce tabular results which uses
the SPARQL XML Results format by default.
Historically there was an ability to transcribe the tabular format
into a graph but this was a very niche use case (primarily only used
in the SPARQL test suite itself). There were a bunch of changes in
Jena 4.x to simplify Content-Type handling to only use canonical
formats by default.
Rob
Using
curl --header 'Accept: application/rdf+xml'
works for me.
Use of "pretty", to become the more basic form of RDF/XML changed
several versions ago at 3.15.0. But that is not
application/sparql-results+xml.
https://issues.apache.org/jira/browse/JENA-1826
Please can we have an example with query, curl call and versions numbers?
Andy
On 22/04/2021, 08:10, "Bardo Nelgen"
<mailing.list.in...@bnnperformances.de> wrote:
Hi all,
�
Maybe there is something about Jena 4 with outputting
„pretty“
RDF/XML via Fuseki ?
�
Whenever asking Fuseki (i.e. via HTTP…
– be it from the
GUI or via
curl) for RDF/XML results, it now ALWAYS defaults to SRX
(application/SPARQL-results+xml).
�
The behaviour could be changed in older versions by sending an
Accept-Header for application/rdf+xml. With version 4.0,
this option
appears to be gone.
For now, I am through with the release notes,
existing bug
reports and
the list archive already but find nothing coming similar to
what we
currently experience.�
What is the contemporary approach to achieve the desired
result ?
�
Thanks in advance for any input.
�
Best,
Bardo
On 22.04.21 11.09 Uhr, Andy Seaborne wrote:
On 22/04/2021 09:55, Rob Vesse wrote:
What is the query?
RDF/XML is a graph format and so only makes sense for
CONSTRUCT/DESCRIBE queries
ASK/SELECT queries are going to produce tabular results which uses
the SPARQL XML Results format by default.
Historically there was an ability to transcribe the tabular format
into a graph but this was a very niche use case (primarily only used
in the SPARQL test suite itself). There were a bunch of changes in
Jena 4.x to simplify Content-Type handling to only use canonical
formats by default.
Rob
Using
curl --header 'Accept: application/rdf+xml'
works for me.
Use of "pretty", to become the more basic form of RDF/XML changed
several versions ago at 3.15.0. But that is not
application/sparql-results+xml.
https://issues.apache.org/jira/browse/JENA-1826
Please can we have an example with query, curl call and versions numbers?
Andy
On 22/04/2021, 08:10, "Bardo Nelgen"
<mailing.list.in...@bnnperformances.de> wrote:
Hi all,
�
Maybe there is something about Jena 4 with outputting
„pretty“
RDF/XML via Fuseki ?
�
Whenever asking Fuseki (i.e. via HTTP…
– be it from the
GUI or via
curl) for RDF/XML results, it now ALWAYS defaults to SRX
(application/SPARQL-results+xml).
�
The behaviour could be changed in older versions by sending an
Accept-Header for application/rdf+xml. With version 4.0,
this option
appears to be gone.
For now, I am through with the release notes,
existing bug
reports and
the list archive already but find nothing coming similar to
what we
currently experience.�
What is the contemporary approach to achieve the desired
result ?
�
Thanks in advance for any input.
�
Best,
Bardo