W3C home > Mailing lists > Public > public-rdf-dawg-comments@w3.org > August 2012

Re: BIND semantics (2)

From: Andy Seaborne <andy.seaborne@epimorphics.com>
Date: Tue, 14 Aug 2012 12:07:18 +0100
Message-ID: <502A3166.4030703@epimorphics.com>
To: public-rdf-dawg-comments@w3.org
(not an official reply)

Hi Jeremy,

On 14/08/12 06:24, Jeremy Carroll wrote:
> This is a formal last call comment on
> http://www.w3.org/TR/2012/WD-sparql11-query-20120724/
> following on from my colleague's comment
> http://lists.w3.org/Archives/Public/public-rdf-dawg-comments/2012Aug/0014.html
> Do the following sections of the specification contradict each other?
> A) http://www.w3.org/TR/2012/WD-sparql11-query-20120724/#selectExpressions
> "The rules of assignment in SELECT expression are the same as for
> assignment in BIND."

Rules that matter are:

1/ scope - the variable must not be already in scope

What the reach of the BIND operator actually is at that point is related 
to the substance of your comment but the rule is still that the variable 
must be fresh.

2/ evaluation - e.g. error in evaluation leaves assigned the variable 
unbound.  For example, BIND(?undef AS ?x) leaves ?x unbound.

3/ SELECT expression also work left-to-right

SELECT (1 AS ?one) (?one+1 AS ?two) {}

Received on Tuesday, 14 August 2012 11:07:50 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:52:13 UTC