W3C home > Mailing lists > Public > public-rdf-tap@w3.org > August 2003

Re: panic.stanford.edu/data

From: Rob McCool <robm@robm.com>
Date: Thu, 7 Aug 2003 09:32:45 -0700
To: "Richard H. McCullough" <rhm@cdepot.net>
Cc: "Turner, Aaron" <Aaron.Turner@goodrich.com>, KR-language <KR-language@YahooGroups.com>, "Latimer, Richard S." <rslatimer@comcast.net>, "Cerra, James F." <jimbobbs@hotmail.com>, "Bray, Tim" <tbray@textuality.com>, "Apollo, Adrian" <Thinkonaut@aol.com>, public-rdf-tap@w3.org
Message-ID: <20030807093245.A10780@flapjack.stanford.edu>

> 1. It looks like a URI problem.
> I'm beginning to make some progress by using GetResource
> to find the right URI.  It's a pain trying to keep straight
> whether it's rdf or rdfs or dc, for example.
> 2. Do you have a Qname capability, e.g. rdfs:subClassOf,
> in your new system?  I'm 100% in favor of Qnames for a 
> user interface instead of long URIs.  Even better, when
> possible (which is most of the time), would be unqualified
> names, with the URI being looked up in a local dictionary.

Yes, if you ask for rdfs:subClassOf or tap:AthleteJordan,_Michael, then 
a system called the Abbreviator expands "rdfs:" to the actual namespace.

The abbreviator also takes care of remembering which are rdfs: and which
are rdf:, and other such hassles by managing a set of "standard tags". You'd
gotten away earlier by putting the TAP namespace on everything, and frankly
I did that for a long time too, but we have to be more compliant now. 

The way to tell if a tag doesn't need a namespace is to call the Abbreivator's
IsStandardTag static method. So in GetData.java, you'd do something like this:

  if(!Abbrev.IsStandardTag(tag)) {
      tag = "http://tap.stanford.edu/data/" + tag;

Off the cuff, the tags I use the most that require no namespace are 
rdfs:Class, rdfs:subClassOf, rdfs:label, rdf:type, and rdf:Resource.
For these tags, if you just say the tag, such as "Class", the system will 
automaticall put the "rdfs:" on for you.
Received on Thursday, 7 August 2003 12:32:56 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:43:14 UTC