RE: [RIF] homework for 10/17 telecon (RIF core syntax)

Hi Dan and Dave,

> >    Con      ::= entity
> >
> > which I don't understand. What does "entity" mean?
>
> I agree this isn't very clear, but my understanding is that <Con> is
> just a constant (i.e. #PCDATA).

I also agree: since "entity" has nothing to do with "XML entities" here,
this designation should be avoided, as detailed below (*).

However, we can clarify the notion of "XML entities" there:

> I'm confused by "The non-terminals in all-upercase such as CONDIT
become
> XML entities, which act like macros and will not be visible in
instance
> markups."

These "XML entities" are meant as "parameter entities used within a
DTD".

E.g., the BNF production 

CONDIT   ::= LITFORM | QUANTIF | CONJ | DISJ

becomes the following parameter entity declaration:

<!ENTITY % CONDIT   "(%LITFORM; | %QUANTIF; | %CONJ; | %DISJ;)" >

Best,
Harold

------------------------------------------------------------------

(*) Regarding
http://www.w3.org/2005/rules/wg/wiki/A.1_Basis%3A_Positive_Conditions

the syntax

  Con      ::= entity
  Var      ::= '?' name
  . . .

can be clarified as follows, insulating DTD-style #PCDATA:

  constant ::= #PCDATA
  name     ::= #PCDATA
  Con      ::= constant
  Var      ::= '?' name
  . . .

( The initial group of #PCDATA could instead use XSD-style xs:string:

  constant ::= xs:string
  name     ::= xs:string
  Con      ::= constant
  Var      ::= '?' name )

Considering
http://www.w3.org/2005/rules/wg/wiki/A.1.1_Basis%3A_Positive_Conditions_
over_Bipartitioned_Constants

the syntax

  Data     ::= value
  Ind      ::= object
  Var      ::= '?' name

can be analogously clarified as follows:

  datavalue  ::= #PCDATA  (or xs:string)
  individual ::= #PCDATA  (or xs:string)
  name       ::= #PCDATA  (or xs:string)
  Data       ::= datavalue  (with optional xsi:type attribute)
  Ind        ::= individual
  Var        ::= '?' name


-----Original Message-----
From: public-rif-wg-request@w3.org [mailto:public-rif-wg-request@w3.org]
On Behalf Of David Z. Hirtle
Sent: Friday, October 13, 2006 4:02 PM
To: Dan Connolly
Cc: Public-Rif-Wg (E-mail)
Subject: Re: [RIF] homework for 10/17 telecon (RIF core syntax)


Hi Dan,

Regarding integers, I think you were on the right track with your
previous message: we can use the xsi:type attribute
(http://www.w3.org/TR/xmlschema-1/#xsi_type) on <Data> to point to any
of the XSD datatypes
(http://www.w3.org/TR/xmlschema-2/#built-in-datatypes), e.g.

<Data xsi:type="xs:string">test</Data>
<Data xsi:type="xs:nonNegativeInteger">12</Data>
<Data xsi:type="xs:dateTime">2002-10-10T17:00:00Z</Data>

(XSV properly enforces these, too -- I've tested it.)

The problem with special short-cut syntax for integers (e.g. <Int>) is
what about other datatypes? We could end up with a lot of shortcut
syntax...

>    Con      ::= entity
>
> which I don't understand. What does "entity" mean?

I agree this isn't very clear, but my understanding is that <Con> is
just a constant (i.e. #PCDATA).

David

On 10/13/06, Dan Connolly <connolly@w3.org> wrote:
>
> On Fri, 2006-10-13 at 13:14 -0400, Christopher Welty wrote:
> > All,
> >
> > We expect to spend the bulk of the next telecon discussing the
technical
> > proposal [http://www.w3.org/2005/rules/wg/wiki/CORE], especially the
> > syntax.
>
> I'm not sure which syntax you mean. I see
>   "A human-oriented syntax, an XML syntax, and the semantics of the
> condition language and of the rule language are given."
>
> That's 4 syntaxes, I guess. And the details of them seem to
> be spread around several wiki topics.
>
> The first grammar rule I see on
>
http://www.w3.org/2005/rules/wg/wiki/A.1_Basis%253A_Positive_Conditions
>
> is
>    Con      ::= entity
>
> which I don't understand. What does "entity" mean?
>
> What's the alphabet of this grammar?
>
> I could perhaps disregard the "human-oriented syntax"
> (which I would call programmer-oriented syntax, since most
> humans can't read it) but I don't see how the XML
> syntax for "entity" is specified.
>
> I'm confused by "The non-terminals in all-upercase such as CONDIT
become
> XML entities, which act like macros and will not be visible in
instance
> markups."
>
> The standard definition of "XML entity" is...
>
> [Definition: An XML document may consist of one or many storage units.
> These are called entities; they all have content and are all (except
for
> the document entity and the external DTD subset) identified by entity
> name.]
> http://www.w3.org/TR/xml/#dt-entity
>
> I don't think that's what is meant, so I suggest using a different
term,
> but I'm not sure what term to suggest, since I don't understand what
> is intended.
>
>
> > To prepare, please try to use the XML syntax to encode some rules.
If you
> > find you need something that is not in the syntax, please note it
and be
> > prepared to discuss it.
>
> Encoding of integers.
> http://lists.w3.org/Archives/Public/public-rif-wg/2006Sep/0071.html
>
> >   If the syntax is acceptable (it has been there
> > for some time), we will decide at this telecon to accept it as our
syntax.
>
> I find the XML syntax nearly acceptable, modulo issues about
> encoding integers, datatyped literals, and language information.
>
> If the WG is going to adopt a "human-readable" syntax, I'll ask
> to consider something closer to SPARQL.
>
> --
> Dan Connolly, W3C http://www.w3.org/People/Connolly/
> D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E
>
>
>

Received on Saturday, 14 October 2006 03:30:03 UTC