Re: FYI: Blog on HTTP Link header and host-wide well-known URI-s, and Linked Data

On Wed, 7 Jul 2010 07:03:21 +0200
Ivan Herman <ivan@w3.org> wrote:

> http://www.w3.org/QA/2010/07/new_opportunities_for_linked_d.html

Not sure why my comment yesterday has still not shown up, but for the
benefit of these lists...

I've been supporting some of these technologies in my Perl modules
<http://search.cpan.org/~tobyink/> for a few months now. For example,
my SPARQL Protocol client is able to auto-discover SPARQL endpoints
from resource URIs (including documents, FTP addresses, e-mail
addresses) using the discovery stack.

So you could just do something like:

	$sparql = "SELECT ?name
	           WHERE {
	               [] foaf:name ?name ;
	                  foaf:mbox <mailto:joe@example.com> .
	           }";
	$query = RDF::Query::Client->new($sparql);
	$results = $query->discover_execute('mailto:joe@example.com');

It would use the discovery techniques discussed in that blog article to
find a SPARQL endpoint relevant to the e-mail address, and then execute
the query on that endpoint.

-- 
Toby A Inkster
<mailto:mail@tobyinkster.co.uk>
<http://tobyinkster.co.uk>