W3C home > Mailing lists > Public > public-rdf-dawg-comments@w3.org > January 2006

Re: minor technical: are periods separators or terminators?

From: Dan Connolly <connolly@w3.org>
Date: Thu, 12 Jan 2006 18:18:16 -0600
To: Fred Zemke <fred.zemke@oracle.com>
Cc: public-rdf-dawg-comments@w3.org
Message-Id: <1137111497.19546.484.camel@dirk.w3.org>

On Thu, 2006-01-12 at 09:43 -0800, Fred Zemke wrote:
> Appendix A.7 grammar
> In rule [27] "ConstructTriples", and rule [28] "Triples",
> '.' is a separator between triples.
> I suggest that the period should be used as a terminator for all
> triples instead.

The WG did consider this explicitly in consideration of
the punctuationSyntax issue
  http://www.w3.org/2001/sw/DataAccess/issues#punctuationSyntax

in particular
 separator/terminator ; 05 Apr 2005
http://lists.w3.org/Archives/Public/public-rdf-dawg/2005AprJun/0022.html

We now have an LL(1) grammar and a set of 70+ test cases to match
it, and a growing number of increasingly interoperable implementations.

I am reluctant to re-open the punctuationSyntax issue at this
point in the process.

I hope you find this justification satisfactory. Please let us
know whether you do.


>    The RDF metaphor is that each triple is
> like a sentence in a natural language.  When we write English, we put
> a period at the end of each sentence.  The common
> representation of RDF used in examples throughout this specification
> shows data triples ending in a period.  This will be a source of
> user confusion if periods are a terminator in RDF but a separator in
> SPARQL.
> 
> Note that your examples are not consistent about this.
> Compare the query in section 2.1 "Writing a simple query" in which
> the triple within the curly braces ends in a period with the example in
> section 2.1.5 "Examples of query syntax" in which the triple pattern
> does not end in a period.  (The example in section 2.1 is
> illegal according to the current grammar.)
> See also the first example of 2.3 "Triple
> patterns", which shows a triple pattern ending in a period. 
>  From the standpoint of using cut-and-paste to build queries, it will
> be better for users if triple patterns always end in a period, rather
> than having to examine the context to decide whether they need a
> period or not.
> 
> Fred Zemke
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E
Received on Friday, 13 January 2006 00:18:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:14:50 GMT