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

Re: Suspicious mismatch between SPARQL grammar rules and examples [OK?]

From: Seaborne, Andy <andy.seaborne@hp.com>
Date: Tue, 10 Jan 2006 10:56:02 +0000
Message-ID: <43C392C2.3030206@hp.com>
To: Ivan Mikhailov <imikhailov@openlinksw.com>
CC: public-rdf-dawg-comments@w3.org



Ivan Mikhailov wrote:
> Hi all,
> 
> It seemes to me that the grammar rule
> 
> [20] GraphPattern ::=
>    	Triples? ( GraphPatternNotTriples '.'? GraphPattern )?
> 
> should be written as
> 
> [20] GraphPattern ::=
>    	( Triples '.'? )? ( GraphPatternNotTriples '.'? GraphPattern )?
> 
> to let examples in the spec match the grammar. Consider examples in 
> sections 5.1, 5.2 etc. and note occurencies of '.' immediately before 
> OPTIONAL keywords.
> 
> If changed, the grammar will remain 'yacc-friendly' LL(1).
> 
> Best Regards,
> Ivan Mikhailov,
> OpenLink Software.
> 

Ivan,

The "Triples" rules can end in a DOT itself so it does cover the case of '.' 
followed by OPTIONAL.

[28] Triples ::= Triples1 ( '.' Triples? )?

This allows for

   { :s :p :o . }

The test suite has
http://www.w3.org/2001/sw/DataAccess/tests/#syntax-struct-03-rq
http://www.w3.org/2001/sw/DataAccess/tests/#syntax-struct-04-rq

which contain
   { :p :q :r OPTIONAL { :a :b :c } }
and
   { :p :q :r . OPTIONAL { :a :b :c } }
respectively to capture the case you highlight.

If this message addresses the comment raised, please let us know.  (If you
respond with [CLOSED] in the subject line it will allow the issue tracking
scripts to close this issue.)

	Andy
Received on Tuesday, 10 January 2006 11:12:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:14:49 GMT