[redland-dev] Questions about example1.c

[Dave Beckett]

On Tue, 3 Jun 2003 10:45:12 +0100
Paul Keogh <paul.keogh at anam.com> wrote:

> > 
> > You might benefit from the use of Redland contexts, which I've just
> > discovered aren't in a full example yet.  There is, however, some
> > test code at the end of librdf/rdf_model.c that might help.
> 
> Thanks for the pointer. Where can I get a good understanding of what
> a context is ?

I've failed to explain that properly in the redland documentation.
But I did discuss this earlier in the year:
  http://lists.usefulinc.com/pipermail/redland-dev/2003-February/000307.html

contexts are a redland-artifact, not part of the RDF model and so although
they allow you to potentially, for example, give URIs for RDF subgraphs,
they can be used for other things such as provenance - tracking who said
what or some other feature from the RDF family such as n3 formulae.

> Also, I've now got my code loading UAProf documents and creating models,
> parsers, uris etc. So far so good. I now want to query one of the documents
> but I'm having a hard time finding good examples of query strings. I need
> some for-dummy examples that say here's a document, here's the query string
> and here's the results. Again, any pointers ?

The query *language* support in Redland has yet to be written and I
aim to do that over the summer, at least for RDQL.  The current triple
query is at the API level where you have software calls that let you
return matches to (s, p, o) triples where any of the s,p,o can be blank
to match triples.

The current query language stuff is a very simple test language that I
used to flesh out the initial API parts that the graph (librdf_model)
and storage (librdf_storage) classes would need for such support.  It is
incomplete.

Dave