W3C home > Mailing lists > Public > public-rif-comments@w3.org > September 2009

Re: EBNF grammar of presentation syntax

From: Axel Polleres <axel.polleres@deri.org>
Date: Sat, 19 Sep 2009 13:50:54 +0100
Cc: <public-rif-comments@w3.org>
Message-Id: <5DECACA8-B4B1-415B-AA2B-1B34380AF403@deri.org>
To: Carlos Damásio <cd@di.fct.unl.pt>
Dear Carlos,

Thanks for spotting this. We have updated the EBNF productions in RIF  
FLD and BLD to only
allow NCNames (instead of arbitrary unicode strings) for localnames,  
Prefix names,
Variable names and Profiles names, cf. [1,2,3]

Please acknowledge receipt of this email to <mailto:public-rif-comments@w3.org 
 >
(replying to this email should suffice). In your acknowledgment please  
let us
know whether or not you are satisfied with the working group's  
response to your
comment.

with best regards,
Axel

1. http://www.w3.org/2005/rules/wiki/FLD#EBNF_Grammar_for_the_Presentation_Syntax_of_RIF-FLD
2. http://www.w3.org/2005/rules/wiki/BLD#EBNF_Grammar_for_the_Presentation_Syntax_of_RIF-BLD_.28Informative.29
3. http://www.w3.org/2005/rules/wiki/DTB#Shortcuts_for_Constants_in_RIF.27s_Presentation_Syntax



On 9 Sep 2009, at 16:54, Carlos Damásio wrote:

> Dear all,
>
>
> In first place, my congratulations to this nice and necessary work.  
> I have just started  playing around with Prolog parsers for RIF and  
> realized that theEBNF grammar is ambiguous, as noticed before by Axel
> (see http://lists.w3.org/Archives/Public/public-rif-wg/2009Aug/0003.html 
> ).
>
> I found very problematic the Name and Var productions which use  
> Unicode Strings which are not delimited. For instance the fragment,
>
> Document(
> Dialect(  anasty ) dialect ))Base( <http://www.dot.com> )
> )
>
> What is the appropriate parsing of this Document? Where does Dialect  
> end?
> Even for this
>
> Document(
> Dialect(FOL)Base( <http://www.dot.com> )
> )
>
> a possible parsing according to the grammar is to understand this as  
> a single dialect declaration "FOL)Base( <http://www.dot.com> ".
>
> I even think that Vars are more problematic because they can appear  
> in sequences like
>
> atom(?x?y ?z)
>
> How many vars are there? 1(a var named "x?y ?z), 2( a var named "x? 
> y" and other named "z") or 3?
>
> The same applies to UNITERMs like this
>
> atom( ?x->y->z)
>
> A possible way out is to say that some special characters are not  
> allowed in Names and use escaped chars for spaces, question marks, "- 
> >", etc. The exact set of non allowed sequences is to be carefully
> defined because of possible interactions with other grammar  
> productions.
>
> Best regards,
>
> Carlos
>
>
>
>

-- 
Dr. Axel Polleres
Digital Enterprise Research Institute, National University of Ireland,  
Galway
email: axel.polleres@deri.org  url: http://www.polleres.net/
Received on Saturday, 19 September 2009 12:51:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 19 September 2009 12:51:36 GMT