Comment on 11.2 "filter evaluation" (CR 6 Apr 2006)

• From: Fred Zemke <fred.zemke@oracle.com>
• Date: Fri, 09 Jun 2006 00:14:23 +0000
• To: public-rdf-dawg@w3.org
• Message-ID: <4488BD17.9090608@oracle.com>

11.2 Filter evaluation
The rules do not say how to handle an unbound variable
in an expression.  Clearly it must be possible
for the argument of BOUND to be unbound.  I believe the desired
semantics are that it is an error to have an unbound argument
for any other function.  This should be stated explicitly.
Also, note that the first bullet says
"SPARQL functions do not process node sequences.  When interpreting
the semantics of XPath functions assume that each argument is
a sequence of a single node".  Given that an argument might be
unbound, it is not true that arguments are always a sequence
of length 1; I think the truth is that the argument may be a
sequence of length 0 or 1.

Fred

Received on Friday, 9 June 2006 05:40:12 UTC