Re: [Myexperiment-discuss] SPARQL endpoint problems

[David R Newman]

Hi Jerzy,

Apologies, I should have learnt my lesson from last time that even a small 
change can effect a lot of queries.

Recently the myExperiment data model itself changed so that admins could 
define content types and give them their own human-readable labels.  This 
slightly changed the RDF model so that instead of having the dcterms:type 
property, a workflow now points at a ContentType via 
mebase:has-content-type.  This means that to run the same query as you 
posted you now need to use the query below:

PREFIX dcterms: <http://purl.org/dc/terms/>
PREFIX mebase: <http://rdf.myexperiment.org/ontologies/base/>
PREFIX snarm: <http://rdf.myexperiment.org/ontologies/snarm/>
PREFIX sioc: <http://rdfs.org/sioc/ns#>
PREFIX mevd: <http://rdf.myexperiment.org/ontologies/viewings_downloads/>
SELECT distinct ?url ?title ?description ?author ?viewed ?downloaded
FROM <http://rdf.myexperiment.org>
WHERE {
 ?workflow mebase:has-content-type 
<http://rdf.myexperiment.org/ContentType/1> .
 ?workflow dcterms:title ?title .
 ?workflow dcterms:description ?description .
 ?workflow mebase:has-policy ?policy .
 ?workflow mebase:content-url ?url .
 ?workflow sioc:has_owner ?author .
 ?workflow mevd:viewed ?viewed .
 ?workflow mevd:downloaded ?downloaded .
 ?policy snarm:has-access 
<http://rdf.myexperiment.org/ontologies/specific/PublicDownload>
}

Getting rid of the filter statement that using ContentType objects allows 
should mean that as the number of workflows increases the query should 
still stay fairly quick.

To try to avoid problems in future I have written a CHANGELOG page 
(http://rdf.myexperiment.org/ontologies/CHANGELOG) for the myExperiment 
ontologies / RDF data model.  At least this will serve to reassure people 
that the error is not their fault.  Again, apologies, I don't and didn't 
foresee there being and major changes to the RDF data model but as I can 
now see, even small changes can produce large ripples.

Regards

David Newman

--On Monday, June 08, 2009 17:31:25 +0200 Jerzy Orlowski 
<address@hidden> wrote:

> Hi
>
> Some time ago I made a SPARQL query for getting all the T1 workflows.
> It worked, but now it returns an empty list. Did You change anything in
> the endpoint?
>
> PREFIX dcterms: <http://purl.org/dc/terms/> PREFIX mebase:
> <http://rdf.myexperiment.org/ontologies/base/> PREFIX snarm:
> <http://rdf.myexperiment.org/ontologies/snarm/> PREFIX sioc:
> <http://rdfs.org/sioc/ns#> PREFIX mevd:
> <http://rdf.myexperiment.org/ontologies/viewings_downloads/> select
> distinct ?url ?title ?description ?author ?viewed ?downloaded FROM
> <http://rdf.myexperiment.org> where {   ?workflow dcterms:type ?thetype
> .   ?workflow dcterms:title ?title .   ?workflow dcterms:description
> ?description .   ?workflow mebase:has-policy ?policy .  ?workflow
> mebase:content-url ?url .   ?workflow sioc:has_owner ?author .
> ?workflow mevd:viewed ?viewed .   ?workflow mevd:downloaded ?downloaded
> .   ?policy snarm:has-access
> <http://rdf.myexperiment.org/ontologies/specific/PublicDownload> .
> FILTER (REGEX(?thetype,'application/vnd.taverna.scufl\\\+xml')) }
>
>
> Jerzy Orlowski
>
>
> _______________________________________________
> Myexperiment-discuss mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/myexperiment-discuss