W3C home > Mailing lists > Public > public-data-shapes-wg@w3.org > January 2016

Re: when is a node a class, when is a node a shape, when is a node an instance of a class

From: Peter F. Patel-Schneider <pfpschneider@gmail.com>
Date: Fri, 15 Jan 2016 16:21:02 -0800
To: Holger Knublauch <holger@topquadrant.com>, public-data-shapes-wg@w3.org
Message-ID: <56998CEE.40809@gmail.com>
My guess is that every place more than one triple might be involved needs this
determination, given that it is no longer the case that all the triples needs
to be in the data graph except for the determination whether a node is a shape
(which is an exception to a more general notion, I guess).

peter


On 01/14/2016 08:16 PM, Holger Knublauch wrote:
> On 15/01/2016 12:14 PM, Peter F. Patel-Schneider wrote:
>> That is not sufficient because there are several sources of information input
>> into SHACL.  For example, is a node an instance of a class for scoping if the
>> rdf:type triple is in the shapes graph, or do only triples in the data graph
>> count for this?
> 
> Ok, these questions need to be answered on a case-to-case basis. For your
> specific issue above (if I understand it correctly), I have added a clarifying
> sentence to state that the triples must be in the data graph.
> 
> https://github.com/w3c/data-shapes/commit/de401e2c411b616bb821fd7d313cebad0835d4c5
> 
> 
> Which other places would require such clarifications?
> 
> Thanks,
> Holger
> 
> 
>>
>> peter
>>
>>
>> On 01/14/2016 03:14 PM, Holger Knublauch wrote:
>>> The spec currently has section 1.1 where we state that certain terms from RDFS
>>> are used with slightly different meaning in the SHACL spec. Maybe, to satisfy
>>> the issue you raise below, we could expand this section a little bit to
>>> clarify that "being in instance" means something like
>>>
>>> ASK {
>>>      $type rdfs:subClassOf* ?class .
>>>      $instance a ?class .
>>> }
>>>
>>> Then, "being a class" means $type=rdfs:Class and "being a shape" means
>>> $type=sh:Shape.
>>>
>>> Would this help?
>>>
>>> Holger
>>>
>>>
>>> On 15/01/2016 2:09 AM, Peter F. Patel-Schneider wrote:
>>>> SHACL currently depends on the answers to several class-based questions,
>>>> including
>>>> - when is a node a class
>>>> - when is a node a shape
>>>> - when is a node an instance of a class
>>>> but how these are determined is not completely spelled out in the SHACL
>>>> documment.
>>>>
>>>> I think that there needs to be a complete definition of these
>>>> relationships in
>>>> the SHACL document and test cases to back up the definition.
>>>>
>>>> peter
>>>>
>>>
> 
Received on Saturday, 16 January 2016 00:21:34 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:30:29 UTC