Re: propose to make syntax-qname-08-rq and syntax-qname-14-rq OBSOLETE

that proposal seems really fine to me + thanks for good catch
with that proposal it is fine to query triples like

:postSurgicalState a :PatientState.
pc:Jane a rim:Patient.
pc:trial rim:Patient.confidentiality_cd rim:Confidentiality.ByAccessKind.

-- 
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/




"Seaborne, Andy" <andy.seaborne@hp.com>
Sent by: public-rdf-dawg-request@w3.org
28/06/2005 12:04
Please respond to andy.seaborne

 
        To:     Jos De_Roo/AMDUS/MOR/Agfa-NV/BE/BAYER@AGFA
        cc:     public-rdf-dawg@w3.org
        Subject:        Re: propose to make syntax-qname-08-rq and syntax-qname-14-rq OBSOLETE




jos.deroo@agfa.com wrote:
> I see that
> 
> http://www.w3.org/2001/sw/DataAccess/tests/#syntax-qname-08-rq
> http://www.w3.org/2001/sw/DataAccess/tests/#syntax-qname-14-rq
> 
> are approved..
> (must have missed that vote..)
> 
> 
> I propose to make them OBSOLETE and change rq23 to either
> 
> [84]   NCCHAR   ::=   NCCHAR1 | '_' | '-' | [0-9] | #x00B7 |
>                       [#x0300-#x036F] | [#x203F-#x2040]
> 
> i.e. disallow '.' in qname
> and instead write absoluteized iris like
>   <http://protege.stanford.edu/kb#Patient.confidentiality_cd>

Since web pages often have ".html" in them, I think that it would be 
inconvenient not allow DOTs inside qnames.

> 
> or
> 
> [86]   NCNAME   ::=   ( "_" | NCCHAR1 ) NCCHAR* ( "_" | NCCHAR1 )

Err - that is at least 2 chars long.

> 
> i.e. allow '.' in but not at-end-of qname
> e.g.
>   rim:Patient.confidentiality_cd

Full productions for no trailing DOTs, but allowing DOTs inside prefixed 
names 
below.

                 Andy


## SPARQL prefixed names
## (Just for reference - unchanged)
QNAME ::= NCNAME_PREFIX? ":" NCNAME


## NCCHAR1 is XML 1.1's NCNameStartChar without the "_"
## (No change)

NCHAR ::= ...

## This is XML NCCHAR without the "." and without "_"

NCCHAR      ::=    NCCHAR1 | '_' | '-' | | [0-9]
                  | #x00B7 | [#x0300-#x036F] | [#x203F-#x2040]

## No leading "_" (blank nodes)
## Could allow a triling DOT here but that would be very confusing.

NCNAME_PREFIX ::= NCCHAR1 (( NCCHAR |".")* NCCHAR )?

## No trailing DOT
NCNAME       ::= ( "_" | NCCHAR1 ) (( NCCHAR |".")* NCCHAR )?

Received on Tuesday, 28 June 2005 10:32:04 UTC