Re: Working Draft feedback items

Steve Harris wrote:
> On Sun, Mar 06, 2005 at 04:58:16 +0000, Andy Seaborne wrote:
> 
>>== 4 == Syntactic support for reification
>>
>>Some people use reification and use it a lot - some people use named
>>graph-like approaches and avoid reification - most people just don't use 
>>either.
> 
> 
> The others seem good, and I have no opinion on this...

In progress. I have put this in and have a @@ToDo@@ to explain it in rq23.

Also  ?id << ?s ?p ?o >>

  ?id   rdf:subject  ?s
  ?id   rdf:pedicate ?p
  ?id   rdf:object   ?o

Also  x:z << ?s ?p ?o >>

  x:z   rdf:subject  ?s
  x:z   rdf:pedicate ?p
  x:z   rdf:object   ?o


The underlying concept is that << ?s ?p ?o >> expands:

  [ rdf:subject  ?s ; rdf:predicate ?p ; rdf:object  ?o ]

so it is

  _:a  rdf:subject  ?s
  _:a  rdf:pedicate ?p
  _:a  rdf:object   ?o

for some freshly minted bNode _:a (RDF collections are much the same). There can 
be a predfined subject which is how "?id << ?s ?p ?o >>" works.

If you want to experiment, you can parse/print such queries with ARQ from CVS. 
There will be a collection of cryptic queries to test parsing but I haven't 
though of a way to test the parsing outcome (I print in a canonical debugging 
form and look at the moment).

>  
> 
>>== 5 == Sorting/Grouping

<snip/>

> 
>>== 6 == Optionals and order dependencies.
>>
>>The alternatives I think if or know about are (briefly):
>>
>>  A/ An order rule that states variables must be used in fixed
>>     patterns before optionals if possible.  We can either make
>>     a query that does not do this illegal, execute in this
>>     canonical order or leave to implementations.
> 
> 
> I have a preference for this. Currently I process all the non-OPTIONAL
> blocks, then all the optionals in order, so I have a preference for doing
> that :)

OK - hearing no other input - I'll try for that approach.

> 
> - Steve 
> 

	Andy

Received on Thursday, 10 March 2005 16:29:55 UTC