Re: [dotNetRDF-Support] Variable passing

[Rob V. <rv...@do...>]

Hi Dennis

I don't know of any obvious reason why you should end up with an extra slash
in one URI, certainly the triple store should not be doing that.

That looks like a data issue however with the examples you have given I
cannot reproduce this, it is possible there is some error in your original
data files that did not translate in your email since I cannot reproduce
this myself.  Can you provide example data files that reproduce the issue?

Rob

From:  Dennis Ludl <Den...@St...>
Reply-To:  Dennis Ludl <Den...@St...>,
dotNetRDF User Help and Support <dot...@li...>
Date:  Friday, February 15, 2013 3:08 PM
To:  dotNetRDF User Help and Support
<dot...@li...>
Subject:  Re: [dotNetRDF-Support] Variable passing

> I've implemeted it with the VALUES clause now and it works perfectly fine
> (performance is way better now). But I came across another question while
> comparing nodes for the VALUES clause.
> I have two RDF files (turtle syntax) which I load both of them to a triple
> store with LoadFromFile.
> Here are two examples:
>  
> File 1:
> <http://example.org/example.ttl#1-part0> a bla:Part;
>                 bla:hasFile <http://localhost/uploads/xyz.xml#Bla1>;
>  
> File 2: 
> <http://localhost/uploads/xyz.xml#Bla1> a bla:File.
>  
> So I would expact that 1 node "http://localhost/uploads/xyz.xml#Bla1" is
> created and used for the triples from both files. My problem right now is that
> it reads:
>  
> File1:
> http://localhost/uploads/xyz.xml#Bla1
>  
> File 2:
> http:///localhost/uploads/xyz.xml#Bla1
>  
> I'm not sure if the additional / is an issue with the triplestore, but I can't
> find any issues in the RDF files. Could it be the case, that the triplestore
> adds the additional / to create 2 seperate nodes?
> 
>  
> -------------------
> Hi Rob, 
> thanks for your answer. Populating over the results value by value (1 query
> per result) leads to very long processing times for me (A few k values per
> result -> 2-3minutes). Thats way to long for my application. I'm not able to
> reduce the number of results and checked some benchmarks which shows that
> current TripleStores can process ~200-700 (simple) queries per second. So
> right now I'm trying to reduce my used queries to a minimum.
> Thanks for your suggestion with the VALUES clause, this looks like very
> promising. I'll implement it that way and hope for a better performance.
>  
> Thank you,
> Dennis
> -------------------
> Hi Dennis
> 
> The code you suggest should work fine, make your query to get the values for
> ?var and then loop over the results populating and making a new query. The
> downside to this is that you have to make a query per result from your first
> query.
> 
> The other option if you are trying to get to minimize the number of queries
> would be to use the results from the first query to build up a VALUES clause
> for the second query and then execute that getting back all the final values
> in one go, see http://www.w3.org/TR/sparql11-query/#inline-data
> <http://www.w3.org/TR/sparql11-query/#inline-data>  for syntax of the VALUES
> clause.
> 
> Hope this helps,
> 
> Rob
> ------------------------------------------------------------------------------
> The Go Parallel Website, sponsored by Intel - in partnership with Geeknet, is
> your hub for all things parallel software development, from weekly thought
> leadership blogs to news, videos, case studies, tutorials, tech docs,
> whitepapers, evaluation guides, and opinion stories. Check out the most recent
> posts - join the conversation now.
> http://goparallel.sourceforge.net/____________________________________________
> ___ dotNetRDF-Support mailing list dot...@li...
> https://lists.sourceforge.net/lists/listinfo/dotnetrdf-support