Re: str() and bNodes

On Wed, Nov 09, 2005 at 11:37:56AM +0000, Steve Harris wrote:
> 
> A while back I had some concerns over the change of behviour in the
> current rq23 draft w.r.t. bNodes.
> 
> Suppose you have a load of biblographic entries which have dc:source data
> attched, you wish to return the data which does not come from amazon.com.
> A reasonable query might look like:
> 
> SELECT ?thing ?source
> WHERE { ?thing dc:source ?source .
> 	 FILTER(!regex(str(?source), "amazon.com")) }
> 
> Which works fine for things like
> 
> [ dc:source <http://amazon.com/whatever/> ]
> [ dc:source "amazon.com" ]
> 
> but does something unexpected for expressions like
> 
> [ dc:source [ a :Library ; dc:title "Library of Congress" ] ]
> 
> Its not a huge problem, but its worth noting. Its seems a bit odd that
> str(<http://foo/>) is fine and str("1.0"^^:footSize) is fine, but
> str(_:a) rasies a type error.

This seems connected to persistent names for blank nodes, which is big
and scary and useful and not something I want to tackle in this version.
-- 
-eric

office: +81.466.49.1170 W3C, Keio Research Institute at SFC,
                        Shonan Fujisawa Campus, Keio University,
                        5322 Endo, Fujisawa, Kanagawa 252-8520
                        JAPAN
        +1.617.258.5741 NE43-344, MIT, Cambridge, MA 02144 USA
cell:   +81.90.6533.3882

(eric@w3.org)
Feel free to forward this message to any list for any purpose other than
email address distribution.