Re: IRIs for Direct Mapping

* Juan Sequeda <juanfederico@gmail.com> [2011-02-21 11:42-0600]
> Eric,
> 
> but have we decided on
> 
> base_IRI/table_name/column_name=value#_
> 
> or
> 
> base_IRI/table_name/column_name.value#_
> 
> note the differente between (=) and (.)

I recall my original desideratum, "=" is a lot less common and table
or attribute names than is ".". The spec says "=" and I've updated
  http://this-db-really.does-not-exist.org/
to use "=". I'm not sure we need to visit this unless we have a good
reason to make colums like foo.bar harder to read in the DM.

I also updated the minimal escaping. See <http://tinyurl.com/6jpqh9u>.


> Juan Sequeda
> +1-575-SEQ-UEDA
> www.juansequeda.com
> 
> 
> On Mon, Feb 21, 2011 at 11:37 AM, Eric Prud'hommeaux <eric@w3.org> wrote:
> 
> > * Boris Villazón Terrazas <bvillazon@fi.upm.es> [2011-02-21 17:35+0100]
> > > Hi Juan
> > > On 18/02/2011 9:09, Juan Sequeda wrote:
> > > >I just realized that in [1], we generate a IRI the following way:
> > > >
> > > >base_IRI/table_name/column_name=value#_
> > > >
> > > >but in the test cases and in the direct mapping demo, it is in the form:
> > > >
> > > >base_IRI/table_name/column_name.value#_
> > > I rely on the direct mapping demo for generating direct mapping test
> > cases.
> > > >
> > > >The difference is between = and .
> > > >
> > > >Is this a general mixup. Or did we decide to use a dot (.) instead
> > > >of an equal sign?
> > > Honestly I don't remember, Eric? Alexandre?
> >
> > '.' is allowed in localnames, though i'm not confident that this is of
> > material value to us.
> >
> > > Boris
> > >
> > > >
> > > >Thanks for the clarification
> > > >
> > > >[1] http://www.w3.org/2001/sw/rdb2rdf/directMapping/#gen_iri
> > > >
> > > >Juan Sequeda
> > > >+1-575-SEQ-UEDA
> > > >www.juansequeda.com <http://www.juansequeda.com>
> > >
> >
> > --
> > -ericP
> >

-- 
-ericP

Received on Monday, 21 February 2011 19:14:24 UTC