Elda basic configuration not working

36 views
Skip to first unread message

Mikel Egaña Aranguren

unread,
May 2, 2017, 10:19:50 AM5/2/17
to linked-data-api-discuss
Hi;

I'm exploring the possibilty of using ELDA as a part of a Government Linked Open Data project. I'm having issues with even the simplest configuration, so any help would be recommended. I'm using the standalone jar with a configuration of my own, pointing to a local triple store:

java -jar elda-standalone-1.3.25-SNAPSHOT-war-exec.jar -Delda.spec=/etc/elda/conf.d/elda-common/hello-mikel.ttl

De hello-mikel.ttl configuration is the same as minimal-named-or-ROOT-split-8080-config.ttl, but the endpoint points to a local blazegraph instance (http://localhost:9999/blazegraph/sparql):

spec:minimal a api:API
; api:base "/"
; rdfs:label "a minimal demonstration API"

When I visit http://localhost:8080/standalone/anything, I get the following pink error:

Internal error

Something broke. Sorry.
Problem running query for SparqlSource{http://localhost:9999/blazegraph/sparql; unauthenticated}: Bad Request
SPARQL-QUERY: queryStr=CONSTRUCT 
  { }
WHERE
  { { VALUES ?item {           } } }

java.util.concurrent.ExecutionException: org.openrdf.query.MalformedQueryException: CONSTRUCT WHERE only permits statement patterns in the WHERE clause.

It seems to me that elda is building an empty SPARQL construct query. Any clues? 

Thanks

Regards


chris dollin

unread,
May 2, 2017, 10:48:47 AM5/2/17
to linked-data-api-discuss
I'll see if I can replicate the problem using Fuseki.
The local data in the blazegraph does have some rdf:type
statements
in it, yes?


On 2 May 2017 at 15:19, Mikel Egaña Aranguren
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to linked-data-api-d...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.



--
Chris "allusive" Dollin

Mikel Egaña Aranguren

unread,
May 3, 2017, 4:30:02 AM5/3/17
to linked-data...@googlegroups.com
thanks;

yes, the local data includes rdf:type predicates. The following query:

SELECT *
WHERE {
  ?s rdf:type ?o
}

Works fine at the blazegraph SPARQL endpoint

regards

2017-05-02 16:48 GMT+02:00 'chris dollin' via linked-data-api-discuss <linked-data...@googlegroups.com>:
I'll see if I can replicate the problem using Fuseki.
The local data in the blazegraph does have some rdf:type
statements
 in it, yes?


On 2 May 2017 at 15:19, Mikel Egaña Aranguren

> For more options, visit https://groups.google.com/d/optout.



--
Chris "allusive" Dollin

--
You received this message because you are subscribed to the Google Groups "linked-data-api-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linked-data-api-discuss+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Mikel Egaña Aranguren, Ph.D.

https://mikel-egana-aranguren.github.io


chris dollin

unread,
May 3, 2017, 5:51:13 AM5/3/17
to linked-data-api-discuss
Hi Mikel

Are the subjects ?s blank nodes or named
resources?

(Elda rejects blank nodes as items in the query,
that is, the bindings of ?item, because you can't
then request properties of those same blank nodes.)

Chris

[Do you have exposable data that I can try here? A
minimal dataset that shows the problem?]



On 3 May 2017 at 09:30, Mikel Egaña Aranguren
<mikel.egan...@gmail.com> wrote:
> thanks;
>
> yes, the local data includes rdf:type predicates. The following query:
>
> SELECT *
> WHERE {
> ?s rdf:type ?o
> }
>
> Works fine at the blazegraph SPARQL endpoint
>
> regards
>
> 2017-05-02 16:48 GMT+02:00 'chris dollin' via linked-data-api-discuss
> <linked-data...@googlegroups.com>:
>>
>> I'll see if I can replicate the problem using Fuseki.
>> The local data in the blazegraph does have some rdf:type
>> statements
>> in it, yes?
>>
>>
>> On 2 May 2017 at 15:19, Mikel Egaña Aranguren
>> > email to linked-data-api-d...@googlegroups.com.
>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to linked-data-api-d...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to linked-data-api-d...@googlegroups.com.

Mikel Egaña Aranguren

unread,
May 3, 2017, 6:00:05 AM5/3/17
to linked-data...@googlegroups.com
2017-05-03 11:51 GMT+02:00 'chris dollin' via linked-data-api-discuss <linked-data...@googlegroups.com>:
Hi Mikel

Are the subjects ?s blank nodes or named
resources?

Hi;

the subjects are resources named by URIs, there are no blank nodes.
 

(Elda rejects blank nodes as items in the query,
that is, the bindings of ?item, because you can't
then request properties of those same blank nodes.)

Chris

[Do you have exposable data that I can try here? A
minimal dataset that shows the problem?]

This is the actual complete dataset (in blazegraph I store it in the default graph):


In fact I have just realised that the rdf:type triples are the first ones on the file. 

Thanks

Regards


 



On 3 May 2017 at 09:30, Mikel Egaña Aranguren
<mikel.egana.aranguren@gmail.com> wrote:
> thanks;
>
> yes, the local data includes rdf:type predicates. The following query:
>
> SELECT *
> WHERE {
>   ?s rdf:type ?o
> }
>
> Works fine at the blazegraph SPARQL endpoint
>
> regards
>
> 2017-05-02 16:48 GMT+02:00 'chris dollin' via linked-data-api-discuss

>>
>> I'll see if I can replicate the problem using Fuseki.
>> The local data in the blazegraph does have some rdf:type
>> statements
>>  in it, yes?
>>
>>
>> On 2 May 2017 at 15:19, Mikel Egaña Aranguren

>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an

>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an

> For more options, visit https://groups.google.com/d/optout.



--
Chris "allusive" Dollin

--
You received this message because you are subscribed to the Google Groups "linked-data-api-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linked-data-api-discuss+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

chris dollin

unread,
May 3, 2017, 7:10:41 AM5/3/17
to linked-data-api-discuss
Thanks for the data.

Could you run the query below against your data? This is the
SELECT query that /anything generates.

SELECT ?item
WHERE
{ ?item <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> ?v1
FILTER ( ! isBlank(?item) )
}
OFFSET 0
LIMIT 10

(This is to help establish where the empty query is coming
from when we don't expect it. But in any case, Elda will
generate empty queries when there are no items -- eg when
the offset gets pushed outside the result set size -- and as far
as I (and the SPARQL validator) can see the empty queries
are legal.)

Chris



On 3 May 2017 at 11:00, Mikel Egaña Aranguren
<mikel.egan...@gmail.com> wrote:
>
>
> 2017-05-03 11:51 GMT+02:00 'chris dollin' via linked-data-api-discuss
> <linked-data...@googlegroups.com>:
>>
>> Hi Mikel
>>
>> Are the subjects ?s blank nodes or named
>> resources?
>
>
> Hi;
>
> the subjects are resources named by URIs, there are no blank nodes.
>
>>
>>
>> (Elda rejects blank nodes as items in the query,
>> that is, the bindings of ?item, because you can't
>> then request properties of those same blank nodes.)
>>
>> Chris
>>
>> [Do you have exposable data that I can try here? A
>> minimal dataset that shows the problem?]
>
>
> This is the actual complete dataset (in blazegraph I store it in the default
> graph):
>
> https://raw.githubusercontent.com/opendata-euskadi/LOD-datasets/URIs-con-ID-envezde-RECURSO/calidad-aire-en-euskadi-2017/rdf/statements.nq
>
> In fact I have just realised that the rdf:type triples are the first ones on
> the file.
>
> Thanks
>
> Regards
>
>
>
>>
>>
>>
>>
>> On 3 May 2017 at 09:30, Mikel Egaña Aranguren
>> <mikel.egan...@gmail.com> wrote:
>> > thanks;
>> >
>> > yes, the local data includes rdf:type predicates. The following query:
>> >
>> > SELECT *
>> > WHERE {
>> > ?s rdf:type ?o
>> > }
>> >
>> > Works fine at the blazegraph SPARQL endpoint
>> >
>> > regards
>> >
>> > 2017-05-02 16:48 GMT+02:00 'chris dollin' via linked-data-api-discuss
>> > <linked-data...@googlegroups.com>:
>> >>
>> >> I'll see if I can replicate the problem using Fuseki.
>> >> The local data in the blazegraph does have some rdf:type
>> >> statements
>> >> in it, yes?
>> >>
>> >>
>> >> On 2 May 2017 at 15:19, Mikel Egaña Aranguren
>> >> > email to linked-data-api-d...@googlegroups.com.
>> >> > For more options, visit https://groups.google.com/d/optout.
>> >>
>> >>
>> >>
>> >> --
>> >> Chris "allusive" Dollin
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google
>> >> Groups
>> >> "linked-data-api-discuss" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send
>> >> an
>> >> email to linked-data-api-d...@googlegroups.com.
>> >> For more options, visit https://groups.google.com/d/optout.
>> >
>> >
>> >
>> >
>> > --
>> > Mikel Egaña Aranguren, Ph.D.
>> >
>> > https://mikel-egana-aranguren.github.io
>> >
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups
>> > "linked-data-api-discuss" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> > an
>> > email to linked-data-api-d...@googlegroups.com.
>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to linked-data-api-d...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to linked-data-api-d...@googlegroups.com.

Mikel Egaña Aranguren

unread,
May 3, 2017, 7:54:58 AM5/3/17
to linked-data...@googlegroups.com
2017-05-03 13:10 GMT+02:00 'chris dollin' via linked-data-api-discuss <linked-data...@googlegroups.com>:
Thanks for the data.

Could you run the query below against your data? This is the
SELECT query that /anything generates.

SELECT  ?item
WHERE
  { ?item <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> ?v1
    FILTER ( ! isBlank(?item) )
  }
OFFSET  0
LIMIT   10

I have executed it against the Triple Store and it does work, these are the entities retrieved:

 

(This is to help establish where the empty query is coming
from when we don't expect it. But in any case, Elda will
generate empty queries when there are no items -- eg when
the offset gets pushed outside the result set size -- and as far
as I (and the SPARQL validator) can see the empty queries
are legal.)

Chris



On 3 May 2017 at 11:00, Mikel Egaña Aranguren
<mikel.egana.aranguren@gmail.com> wrote:
>
>
> 2017-05-03 11:51 GMT+02:00 'chris dollin' via linked-data-api-discuss

>>
>> Hi Mikel
>>
>> Are the subjects ?s blank nodes or named
>> resources?
>
>
> Hi;
>
> the subjects are resources named by URIs, there are no blank nodes.
>
>>
>>
>> (Elda rejects blank nodes as items in the query,
>> that is, the bindings of ?item, because you can't
>> then request properties of those same blank nodes.)
>>
>> Chris
>>
>> [Do you have exposable data that I can try here? A
>> minimal dataset that shows the problem?]
>
>
> This is the actual complete dataset (in blazegraph I store it in the default
> graph):
>
> https://raw.githubusercontent.com/opendata-euskadi/LOD-datasets/URIs-con-ID-envezde-RECURSO/calidad-aire-en-euskadi-2017/rdf/statements.nq
>
> In fact I have just realised that the rdf:type triples are the first ones on
> the file.
>
> Thanks
>
> Regards
>
>
>
>>
>>
>>
>>
>> On 3 May 2017 at 09:30, Mikel Egaña Aranguren
>> <mikel.egana.aranguren@gmail.com> wrote:
>> > thanks;
>> >
>> > yes, the local data includes rdf:type predicates. The following query:
>> >
>> > SELECT *
>> > WHERE {
>> >   ?s rdf:type ?o
>> > }
>> >
>> > Works fine at the blazegraph SPARQL endpoint
>> >
>> > regards
>> >
>> > 2017-05-02 16:48 GMT+02:00 'chris dollin' via linked-data-api-discuss

>> >>
>> >> I'll see if I can replicate the problem using Fuseki.
>> >> The local data in the blazegraph does have some rdf:type
>> >> statements
>> >>  in it, yes?
>> >>
>> >>
>> >> On 2 May 2017 at 15:19, Mikel Egaña Aranguren

>> >> > For more options, visit https://groups.google.com/d/optout.
>> >>
>> >>
>> >>
>> >> --
>> >> Chris "allusive" Dollin
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google
>> >> Groups
>> >> "linked-data-api-discuss" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send
>> >> an

>> >> For more options, visit https://groups.google.com/d/optout.
>> >
>> >
>> >
>> >
>> > --
>> > Mikel Egaña Aranguren, Ph.D.
>> >
>> > https://mikel-egana-aranguren.github.io
>> >
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups
>> > "linked-data-api-discuss" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> > an

>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an

>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an

> For more options, visit https://groups.google.com/d/optout.



--
Chris "allusive" Dollin

--
You received this message because you are subscribed to the Google Groups "linked-data-api-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linked-data-api-discuss+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

chris dollin

unread,
May 3, 2017, 8:27:59 AM5/3/17
to linked-data-api-discuss
I assume that was against the default graph, yes?

I wonder where the "no results" is coming from.

Chris

On 3 May 2017 at 12:54, Mikel Egaña Aranguren
>> <mikel.egan...@gmail.com> wrote:
>> >
>> >
>> > 2017-05-03 11:51 GMT+02:00 'chris dollin' via linked-data-api-discuss
>> > <linked-data...@googlegroups.com>:
>> >>
>> >> Hi Mikel
>> >>
>> >> Are the subjects ?s blank nodes or named
>> >> resources?
>> >
>> >
>> > Hi;
>> >
>> > the subjects are resources named by URIs, there are no blank nodes.
>> >
>> >>
>> >>
>> >> (Elda rejects blank nodes as items in the query,
>> >> that is, the bindings of ?item, because you can't
>> >> then request properties of those same blank nodes.)
>> >>
>> >> Chris
>> >>
>> >> [Do you have exposable data that I can try here? A
>> >> minimal dataset that shows the problem?]
>> >
>> >
>> > This is the actual complete dataset (in blazegraph I store it in the
>> > default
>> > graph):
>> >
>> >
>> > https://raw.githubusercontent.com/opendata-euskadi/LOD-datasets/URIs-con-ID-envezde-RECURSO/calidad-aire-en-euskadi-2017/rdf/statements.nq
>> >
>> > In fact I have just realised that the rdf:type triples are the first
>> > ones on
>> > the file.
>> >
>> > Thanks
>> >
>> > Regards
>> >
>> >
>> >
>> >>
>> >>
>> >>
>> >>
>> >> On 3 May 2017 at 09:30, Mikel Egaña Aranguren
>> >> <mikel.egan...@gmail.com> wrote:
>> >> > thanks;
>> >> >
>> >> > yes, the local data includes rdf:type predicates. The following
>> >> > query:
>> >> >
>> >> > SELECT *
>> >> > WHERE {
>> >> > ?s rdf:type ?o
>> >> > }
>> >> >
>> >> > Works fine at the blazegraph SPARQL endpoint
>> >> >
>> >> > regards
>> >> >
>> >> > 2017-05-02 16:48 GMT+02:00 'chris dollin' via linked-data-api-discuss
>> >> > <linked-data...@googlegroups.com>:
>> >> >>
>> >> >> I'll see if I can replicate the problem using Fuseki.
>> >> >> The local data in the blazegraph does have some rdf:type
>> >> >> statements
>> >> >> in it, yes?
>> >> >>
>> >> >>
>> >> >> On 2 May 2017 at 15:19, Mikel Egaña Aranguren
>> >> >> > email to linked-data-api-d...@googlegroups.com.
>> >> >> > For more options, visit https://groups.google.com/d/optout.
>> >> >>
>> >> >>
>> >> >>
>> >> >> --
>> >> >> Chris "allusive" Dollin
>> >> >>
>> >> >> --
>> >> >> You received this message because you are subscribed to the Google
>> >> >> Groups
>> >> >> "linked-data-api-discuss" group.
>> >> >> To unsubscribe from this group and stop receiving emails from it,
>> >> >> send
>> >> >> an
>> >> >> email to linked-data-api-d...@googlegroups.com.
>> >> >> For more options, visit https://groups.google.com/d/optout.
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > --
>> >> > Mikel Egaña Aranguren, Ph.D.
>> >> >
>> >> > https://mikel-egana-aranguren.github.io
>> >> >
>> >> >
>> >> > --
>> >> > You received this message because you are subscribed to the Google
>> >> > Groups
>> >> > "linked-data-api-discuss" group.
>> >> > To unsubscribe from this group and stop receiving emails from it,
>> >> > send
>> >> > an
>> >> > email to linked-data-api-d...@googlegroups.com.
>> >> > For more options, visit https://groups.google.com/d/optout.
>> >>
>> >>
>> >>
>> >> --
>> >> Chris "allusive" Dollin
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google
>> >> Groups
>> >> "linked-data-api-discuss" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send
>> >> an
>> >> email to linked-data-api-d...@googlegroups.com.
>> >> For more options, visit https://groups.google.com/d/optout.
>> >
>> >
>> >
>> >
>> > --
>> > Mikel Egaña Aranguren, Ph.D.
>> >
>> > https://mikel-egana-aranguren.github.io
>> >
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups
>> > "linked-data-api-discuss" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> > an
>> > email to linked-data-api-d...@googlegroups.com.
>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to linked-data-api-d...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to linked-data-api-d...@googlegroups.com.

Mikel Egaña Aranguren

unread,
May 3, 2017, 8:48:19 AM5/3/17
to linked-data...@googlegroups.com
2017-05-03 14:27 GMT+02:00 'chris dollin' via linked-data-api-discuss <linked-data...@googlegroups.com>:
I assume that was against the default graph, yes?

Yes, against the default graph

Thanks

 

I wonder where the "no results" is coming from.

Chris

On 3 May 2017 at 12:54, Mikel Egaña Aranguren
<mikel.egana.aranguren@gmail.com> wrote:
>
>
> 2017-05-03 13:10 GMT+02:00 'chris dollin' via linked-data-api-discuss
>> <mikel.egana.aranguren@gmail.com> wrote:
>> >
>> >
>> > 2017-05-03 11:51 GMT+02:00 'chris dollin' via linked-data-api-discuss

>> >>
>> >> Hi Mikel
>> >>
>> >> Are the subjects ?s blank nodes or named
>> >> resources?
>> >
>> >
>> > Hi;
>> >
>> > the subjects are resources named by URIs, there are no blank nodes.
>> >
>> >>
>> >>
>> >> (Elda rejects blank nodes as items in the query,
>> >> that is, the bindings of ?item, because you can't
>> >> then request properties of those same blank nodes.)
>> >>
>> >> Chris
>> >>
>> >> [Do you have exposable data that I can try here? A
>> >> minimal dataset that shows the problem?]
>> >
>> >
>> > This is the actual complete dataset (in blazegraph I store it in the
>> > default
>> > graph):
>> >
>> >
>> > https://raw.githubusercontent.com/opendata-euskadi/LOD-datasets/URIs-con-ID-envezde-RECURSO/calidad-aire-en-euskadi-2017/rdf/statements.nq
>> >
>> > In fact I have just realised that the rdf:type triples are the first
>> > ones on
>> > the file.
>> >
>> > Thanks
>> >
>> > Regards
>> >
>> >
>> >
>> >>
>> >>
>> >>
>> >>
>> >> On 3 May 2017 at 09:30, Mikel Egaña Aranguren
>> >> <mikel.egana.aranguren@gmail.com> wrote:
>> >> > thanks;
>> >> >
>> >> > yes, the local data includes rdf:type predicates. The following
>> >> > query:
>> >> >
>> >> > SELECT *
>> >> > WHERE {
>> >> >   ?s rdf:type ?o
>> >> > }
>> >> >
>> >> > Works fine at the blazegraph SPARQL endpoint
>> >> >
>> >> > regards
>> >> >
>> >> > 2017-05-02 16:48 GMT+02:00 'chris dollin' via linked-data-api-discuss

>> >> >>
>> >> >> I'll see if I can replicate the problem using Fuseki.
>> >> >> The local data in the blazegraph does have some rdf:type
>> >> >> statements
>> >> >>  in it, yes?
>> >> >>
>> >> >>
>> >> >> On 2 May 2017 at 15:19, Mikel Egaña Aranguren

>> >> >> > For more options, visit https://groups.google.com/d/optout.
>> >> >>
>> >> >>
>> >> >>
>> >> >> --
>> >> >> Chris "allusive" Dollin
>> >> >>
>> >> >> --
>> >> >> You received this message because you are subscribed to the Google
>> >> >> Groups
>> >> >> "linked-data-api-discuss" group.
>> >> >> To unsubscribe from this group and stop receiving emails from it,
>> >> >> send
>> >> >> an

>> >> >> For more options, visit https://groups.google.com/d/optout.
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > --
>> >> > Mikel Egaña Aranguren, Ph.D.
>> >> >
>> >> > https://mikel-egana-aranguren.github.io
>> >> >
>> >> >
>> >> > --
>> >> > You received this message because you are subscribed to the Google
>> >> > Groups
>> >> > "linked-data-api-discuss" group.
>> >> > To unsubscribe from this group and stop receiving emails from it,
>> >> > send
>> >> > an

>> >> > For more options, visit https://groups.google.com/d/optout.
>> >>
>> >>
>> >>
>> >> --
>> >> Chris "allusive" Dollin
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google
>> >> Groups
>> >> "linked-data-api-discuss" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send
>> >> an

>> >> For more options, visit https://groups.google.com/d/optout.
>> >
>> >
>> >
>> >
>> > --
>> > Mikel Egaña Aranguren, Ph.D.
>> >
>> > https://mikel-egana-aranguren.github.io
>> >
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups
>> > "linked-data-api-discuss" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> > an

>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an

>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an

> For more options, visit https://groups.google.com/d/optout.



--
Chris "allusive" Dollin

--
You received this message because you are subscribed to the Google Groups "linked-data-api-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linked-data-api-discuss+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

chris dollin

unread,
May 4, 2017, 6:12:01 AM5/4/17
to linked-data-api-discuss
OK.

First, a red herring. The message Mikel reported

CONSTRUCT
{ }
WHERE
{ { VALUES ?item { } } }
is actually concealing the ?item values because they
are <URI>s and the browser (or mailer) is discarding
them as bad html.

So where does the empty CONSTRUCT {} come from?
It is because there are no property chains in the view
that is being used, which will be the default viewer
(because Mikel didn't say he'd set one on their
query) and that is api:labelledDescribeViewer (
as seen in the minimal config that they used).

Elda responses are built from 1-3 subresponses
with different view queries:

* a description viewer generates a DESCRIBE

* a labelled describe viewer will have a label-fetching
query which looks for an rdfs:label for each subject
returned by a DESCRIBE or a property chain

* the property chains from the configuration and any
_properties in the URL will have a CONSTRUCT
query which builds a graph from the properties
specified by the chain; if there are no property
chains the CONSTRUCT will be empty.

Normally this (empty property chains) is not a problem,
because an empty CONSTRUCT {} is legal syntax and
it just generates an empty graph. However it looks
like blazegraph doesn't permit this and so, BOOM.

Workaround: use _properties=label or =type in the
query, or make the default viewer one that has
properties defined.

As it happens, Elda's issue #119 requests the
removal of the empty CONSTRUCT clause when
there are no explicit properties. I have pushed a
fix for this -- if you're comfortable building Elda from
source you might want to try it and see if it fixes
your problem. Or you could try the current snapshot
for 1.3.26 (but be warned snapshots are not stable
and may change).

Chris

On 3 May 2017 at 13:48, Mikel Egaña Aranguren
<mikel.egan...@gmail.com> wrote:
>
>
> 2017-05-03 14:27 GMT+02:00 'chris dollin' via linked-data-api-discuss
> <linked-data...@googlegroups.com>:
>>
>> I assume that was against the default graph, yes?
>
>
> Yes, against the default graph
>
> Thanks
>
>
>>
>>
>> I wonder where the "no results" is coming from.
>>
>> Chris
>>
>> On 3 May 2017 at 12:54, Mikel Egaña Aranguren
>> <mikel.egan...@gmail.com> wrote:
>> >
>> >
>> > 2017-05-03 13:10 GMT+02:00 'chris dollin' via linked-data-api-discuss
>> > <linked-data...@googlegroups.com>:
>> >> <mikel.egan...@gmail.com> wrote:
>> >> >
>> >> >
>> >> > 2017-05-03 11:51 GMT+02:00 'chris dollin' via linked-data-api-discuss
>> >> > <linked-data...@googlegroups.com>:
>> >> >>
>> >> >> Hi Mikel
>> >> >>
>> >> >> Are the subjects ?s blank nodes or named
>> >> >> resources?
>> >> >
>> >> >
>> >> > Hi;
>> >> >
>> >> > the subjects are resources named by URIs, there are no blank nodes.
>> >> >
>> >> >>
>> >> >>
>> >> >> (Elda rejects blank nodes as items in the query,
>> >> >> that is, the bindings of ?item, because you can't
>> >> >> then request properties of those same blank nodes.)
>> >> >>
>> >> >> Chris
>> >> >>
>> >> >> [Do you have exposable data that I can try here? A
>> >> >> minimal dataset that shows the problem?]
>> >> >
>> >> >
>> >> > This is the actual complete dataset (in blazegraph I store it in the
>> >> > default
>> >> > graph):
>> >> >
>> >> >
>> >> >
>> >> > https://raw.githubusercontent.com/opendata-euskadi/LOD-datasets/URIs-con-ID-envezde-RECURSO/calidad-aire-en-euskadi-2017/rdf/statements.nq
>> >> >
>> >> > In fact I have just realised that the rdf:type triples are the first
>> >> > ones on
>> >> > the file.
>> >> >
>> >> > Thanks
>> >> >
>> >> > Regards
>> >> >
>> >> >
>> >> >
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >> On 3 May 2017 at 09:30, Mikel Egaña Aranguren
>> >> >> <mikel.egan...@gmail.com> wrote:
>> >> >> > thanks;
>> >> >> >
>> >> >> > yes, the local data includes rdf:type predicates. The following
>> >> >> > query:
>> >> >> >
>> >> >> > SELECT *
>> >> >> > WHERE {
>> >> >> > ?s rdf:type ?o
>> >> >> > }
>> >> >> >
>> >> >> > Works fine at the blazegraph SPARQL endpoint
>> >> >> >
>> >> >> > regards
>> >> >> >
>> >> >> > 2017-05-02 16:48 GMT+02:00 'chris dollin' via
>> >> >> > linked-data-api-discuss
>> >> >> > <linked-data...@googlegroups.com>:
>> >> >> >>
>> >> >> >> I'll see if I can replicate the problem using Fuseki.
>> >> >> >> The local data in the blazegraph does have some rdf:type
>> >> >> >> statements
>> >> >> >> in it, yes?
>> >> >> >>
>> >> >> >>
>> >> >> >> On 2 May 2017 at 15:19, Mikel Egaña Aranguren
>> >> >> >> > email to linked-data-api-d...@googlegroups.com.
>> >> >> >> > For more options, visit https://groups.google.com/d/optout.
>> >> >> >>
>> >> >> >>
>> >> >> >>
>> >> >> >> --
>> >> >> >> Chris "allusive" Dollin
>> >> >> >>
>> >> >> >> --
>> >> >> >> You received this message because you are subscribed to the
>> >> >> >> Google
>> >> >> >> Groups
>> >> >> >> "linked-data-api-discuss" group.
>> >> >> >> To unsubscribe from this group and stop receiving emails from it,
>> >> >> >> send
>> >> >> >> an
>> >> >> >> email to linked-data-api-d...@googlegroups.com.
>> >> >> >> For more options, visit https://groups.google.com/d/optout.
>> >> >> >
>> >> >> >
>> >> >> >
>> >> >> >
>> >> >> > --
>> >> >> > Mikel Egaña Aranguren, Ph.D.
>> >> >> >
>> >> >> > https://mikel-egana-aranguren.github.io
>> >> >> >
>> >> >> >
>> >> >> > --
>> >> >> > You received this message because you are subscribed to the Google
>> >> >> > Groups
>> >> >> > "linked-data-api-discuss" group.
>> >> >> > To unsubscribe from this group and stop receiving emails from it,
>> >> >> > send
>> >> >> > an
>> >> >> > email to linked-data-api-d...@googlegroups.com.
>> >> >> > For more options, visit https://groups.google.com/d/optout.
>> >> >>
>> >> >>
>> >> >>
>> >> >> --
>> >> >> Chris "allusive" Dollin
>> >> >>
>> >> >> --
>> >> >> You received this message because you are subscribed to the Google
>> >> >> Groups
>> >> >> "linked-data-api-discuss" group.
>> >> >> To unsubscribe from this group and stop receiving emails from it,
>> >> >> send
>> >> >> an
>> >> >> email to linked-data-api-d...@googlegroups.com.
>> >> >> For more options, visit https://groups.google.com/d/optout.
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > --
>> >> > Mikel Egaña Aranguren, Ph.D.
>> >> >
>> >> > https://mikel-egana-aranguren.github.io
>> >> >
>> >> >
>> >> > --
>> >> > You received this message because you are subscribed to the Google
>> >> > Groups
>> >> > "linked-data-api-discuss" group.
>> >> > To unsubscribe from this group and stop receiving emails from it,
>> >> > send
>> >> > an
>> >> > email to linked-data-api-d...@googlegroups.com.
>> >> > For more options, visit https://groups.google.com/d/optout.
>> >>
>> >>
>> >>
>> >> --
>> >> Chris "allusive" Dollin
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google
>> >> Groups
>> >> "linked-data-api-discuss" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send
>> >> an
>> >> email to linked-data-api-d...@googlegroups.com.
>> >> For more options, visit https://groups.google.com/d/optout.
>> >
>> >
>> >
>> >
>> > --
>> > Mikel Egaña Aranguren, Ph.D.
>> >
>> > https://mikel-egana-aranguren.github.io
>> >
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups
>> > "linked-data-api-discuss" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> > an
>> > email to linked-data-api-d...@googlegroups.com.
>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to linked-data-api-d...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to linked-data-api-d...@googlegroups.com.

Mikel Egaña Aranguren

unread,
May 4, 2017, 10:31:27 AM5/4/17
to linked-data...@googlegroups.com
2017-05-04 12:11 GMT+02:00 'chris dollin' via linked-data-api-discuss <linked-data...@googlegroups.com>:
OK.

First, a red herring. The message Mikel reported

    CONSTRUCT
    { }
    WHERE
    { { VALUES ?item {   } } }
is actually concealing the ?item values because they
are <URI>s and the browser (or mailer) is discarding
them as bad html.

Yes, the URIs are present in the source of the error webpage. didn't noticed.
Yes, that worked! I have cloned master 1.3.26, built the standalone war, executed it with my configuration, and it does render the results.

Now I will try with the proper production setting, ie building the common and assets wars

Thanks

Regards


 

Chris

On 3 May 2017 at 13:48, Mikel Egaña Aranguren
<mikel.egana.aranguren@gmail.com> wrote:
>
>
> 2017-05-03 14:27 GMT+02:00 'chris dollin' via linked-data-api-discuss

>>
>> I assume that was against the default graph, yes?
>
>
> Yes, against the default graph
>
> Thanks
>
>
>>
>>
>> I wonder where the "no results" is coming from.
>>
>> Chris
>>
>> On 3 May 2017 at 12:54, Mikel Egaña Aranguren
>> <mikel.egana.aranguren@gmail.com> wrote:
>> >
>> >
>> > 2017-05-03 13:10 GMT+02:00 'chris dollin' via linked-data-api-discuss
>> >> <mikel.egana.aranguren@gmail.com> wrote:
>> >> >
>> >> >
>> >> > 2017-05-03 11:51 GMT+02:00 'chris dollin' via linked-data-api-discuss

>> >> >>
>> >> >> Hi Mikel
>> >> >>
>> >> >> Are the subjects ?s blank nodes or named
>> >> >> resources?
>> >> >
>> >> >
>> >> > Hi;
>> >> >
>> >> > the subjects are resources named by URIs, there are no blank nodes.
>> >> >
>> >> >>
>> >> >>
>> >> >> (Elda rejects blank nodes as items in the query,
>> >> >> that is, the bindings of ?item, because you can't
>> >> >> then request properties of those same blank nodes.)
>> >> >>
>> >> >> Chris
>> >> >>
>> >> >> [Do you have exposable data that I can try here? A
>> >> >> minimal dataset that shows the problem?]
>> >> >
>> >> >
>> >> > This is the actual complete dataset (in blazegraph I store it in the
>> >> > default
>> >> > graph):
>> >> >
>> >> >
>> >> >
>> >> > https://raw.githubusercontent.com/opendata-euskadi/LOD-datasets/URIs-con-ID-envezde-RECURSO/calidad-aire-en-euskadi-2017/rdf/statements.nq
>> >> >
>> >> > In fact I have just realised that the rdf:type triples are the first
>> >> > ones on
>> >> > the file.
>> >> >
>> >> > Thanks
>> >> >
>> >> > Regards
>> >> >
>> >> >
>> >> >
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >> On 3 May 2017 at 09:30, Mikel Egaña Aranguren
>> >> >> <mikel.egana.aranguren@gmail.com> wrote:
>> >> >> > thanks;
>> >> >> >
>> >> >> > yes, the local data includes rdf:type predicates. The following
>> >> >> > query:
>> >> >> >
>> >> >> > SELECT *
>> >> >> > WHERE {
>> >> >> >   ?s rdf:type ?o
>> >> >> > }
>> >> >> >
>> >> >> > Works fine at the blazegraph SPARQL endpoint
>> >> >> >
>> >> >> > regards
>> >> >> >
>> >> >> > 2017-05-02 16:48 GMT+02:00 'chris dollin' via
>> >> >> > linked-data-api-discuss

>> >> >> >>
>> >> >> >> I'll see if I can replicate the problem using Fuseki.
>> >> >> >> The local data in the blazegraph does have some rdf:type
>> >> >> >> statements
>> >> >> >>  in it, yes?
>> >> >> >>
>> >> >> >>
>> >> >> >> On 2 May 2017 at 15:19, Mikel Egaña Aranguren

>> >> >> >> > For more options, visit https://groups.google.com/d/optout.
>> >> >> >>
>> >> >> >>
>> >> >> >>
>> >> >> >> --
>> >> >> >> Chris "allusive" Dollin
>> >> >> >>
>> >> >> >> --
>> >> >> >> You received this message because you are subscribed to the
>> >> >> >> Google
>> >> >> >> Groups
>> >> >> >> "linked-data-api-discuss" group.
>> >> >> >> To unsubscribe from this group and stop receiving emails from it,
>> >> >> >> send
>> >> >> >> an

>> >> >> >> For more options, visit https://groups.google.com/d/optout.
>> >> >> >
>> >> >> >
>> >> >> >
>> >> >> >
>> >> >> > --
>> >> >> > Mikel Egaña Aranguren, Ph.D.
>> >> >> >
>> >> >> > https://mikel-egana-aranguren.github.io
>> >> >> >
>> >> >> >
>> >> >> > --
>> >> >> > You received this message because you are subscribed to the Google
>> >> >> > Groups
>> >> >> > "linked-data-api-discuss" group.
>> >> >> > To unsubscribe from this group and stop receiving emails from it,
>> >> >> > send
>> >> >> > an

>> >> >> > For more options, visit https://groups.google.com/d/optout.
>> >> >>
>> >> >>
>> >> >>
>> >> >> --
>> >> >> Chris "allusive" Dollin
>> >> >>
>> >> >> --
>> >> >> You received this message because you are subscribed to the Google
>> >> >> Groups
>> >> >> "linked-data-api-discuss" group.
>> >> >> To unsubscribe from this group and stop receiving emails from it,
>> >> >> send
>> >> >> an

>> >> >> For more options, visit https://groups.google.com/d/optout.
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > --
>> >> > Mikel Egaña Aranguren, Ph.D.
>> >> >
>> >> > https://mikel-egana-aranguren.github.io
>> >> >
>> >> >
>> >> > --
>> >> > You received this message because you are subscribed to the Google
>> >> > Groups
>> >> > "linked-data-api-discuss" group.
>> >> > To unsubscribe from this group and stop receiving emails from it,
>> >> > send
>> >> > an

>> >> > For more options, visit https://groups.google.com/d/optout.
>> >>
>> >>
>> >>
>> >> --
>> >> Chris "allusive" Dollin
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google
>> >> Groups
>> >> "linked-data-api-discuss" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send
>> >> an

>> >> For more options, visit https://groups.google.com/d/optout.
>> >
>> >
>> >
>> >
>> > --
>> > Mikel Egaña Aranguren, Ph.D.
>> >
>> > https://mikel-egana-aranguren.github.io
>> >
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups
>> > "linked-data-api-discuss" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> > an

>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Chris "allusive" Dollin
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "linked-data-api-discuss" group.
>> To unsubscribe from this group and stop receiving emails from it, send an

>> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> Mikel Egaña Aranguren, Ph.D.
>
> https://mikel-egana-aranguren.github.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "linked-data-api-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an

> For more options, visit https://groups.google.com/d/optout.



--
Chris "allusive" Dollin

--
You received this message because you are subscribed to the Google Groups "linked-data-api-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linked-data-api-discuss+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Andy Seaborne

unread,
May 9, 2017, 6:57:22 AM5/9/17
to linked-data-api-discuss
The form

CONSTRUCT 
  { }
WHERE { ... }

is not a CONSTRUCT-WHERE form; that looks like:

CONSTRUCT WHERE {}

and then the patterns in {} are quite restricted as per the error message.

https://www.w3.org/TR/sparql11-query/#constructWhere
Reply all
Reply to author
Forward
0 new messages