Raw

hCards! And potentially a lot more where they came from.

I grabbed some data using this Java/ Jena-based makeshift scutter. I was going to use Slug, but I had trouble getting it going. I suspect a path issue moving it from MS Win, but I thought it'd be quicker to just hack something than trying to fix today (I'm still keen to use Slug asap, mostly because I'm hopeless with threads...).

PS. I should've been more observant. Leigh's pointed me to the proper notes. for Slug. Worthwhile exercise making (yet another) scutter though, good reminder that programming can be very easy, until you get to concurrency.    

So I used the scutterplan from the FOAFBulletinBoard, going 2 steps, producing 17382 statements, containing I think it was 2035 foaf:Person nodes. I'd estimate about 2/3 of the 260ish linked profiles got picked up, a lot failed all kinds of reasons. Here's that data: all.n3 (check the namespaces!). There were still 3 borked URIs, tipjars! (removed manually from that file), which choked Jena's RDF/XML serialiser, hence the Turtle/N3.

To that data I applied this SPARQL ( foaf.rq):

PREFIX foaf: <http://xmlns.com/foaf/0.1/>



SELECT DISTINCT ?name ?homepage ?mbox ?photo



WHERE {

?x foaf:name ?name.

?x foaf:weblog ?homepage.

?x foaf:depiction ?photo.

OPTIONAL {?x foaf:mbox ?mbox. }

}

LIMIT 10

To the results I applied sparql2hcard.xsl using xsltproc, script was sparql-hcard.sh ( I do hope this was the latest SPARQL XML results format...), producing these hCards . If you mouse-over you'll note Andrew Savory has 2 different homepage URIs, hence the dupe. Funnily enough Leigh showed up in every run I tried, making me feel a little guilty about not persisting with Slug. Hey ho.

The results are a bit light data-wise. I'm not sure what's the best vocab to use for addresses etc - Norm was looking at updating vCard in RDF not long ago, must revisit. The ability to round trip between microformats and RDF should help lubricate things a bit.

I must apologise about still using firewall-unfriendly port 88 for various things, but sometime soon I'm going to try Pound to slap everything on 80 (thanks for the pointer Rich!)

I've not been very productive time the past week or so, mostly because it's been too hot to think straight. Net result, two pieces of work should've been ready for tomorrow not yet ready. One thing was finishing a chapter of microformats, the other exploring alternative ways of presenting SPARQL results. So I think you have to agree the above was fairly controlled procrastination...

Oh yeah, if anyone's got any nice scripts for SPARQL/JSON rendering that might be usable server-side with Rhino (should be all of 'em, I'd have thought) please ping me.

@en