Re: Question on "Test of path sequence with duplicate 001" test

See https://www.w3.org/TR/shacl/#value-nodes:

Forproperty shapes <https://www.w3.org/TR/shacl/#dfn-property-shape>with 
avalue <https://www.w3.org/TR/shacl/#dfn-value>for|sh:path||p|thevalue 
nodes <https://www.w3.org/TR/shacl/#dfn-value-nodes>are the set ofnodes 
<https://www.w3.org/TR/shacl/#dfn-node>in thedata graph 
<https://www.w3.org/TR/shacl/#dfn-data-graph>that can be reached from 
thefocus node <https://www.w3.org/TR/shacl/#dfn-focus-node>with thepath 
mapping <https://www.w3.org/TR/shacl/#dfn-path-mapping>of|p|.

In other words, the maxCount applies to a *set* of nodes, excluding 
duplicates. No matter if the value can be reached from two paths, the 
set of value nodes is still just one value.

Hope this clarifies it
Holger

PS: glad to hear about a SHACL implementation for AllegroGraph.


On 31/01/2019 12:31 am, gwking - metabang wrote:
> Hi Holger,
>
> Thanks. I was talking about that test. The issue is occurring in the context of the SHACL implementation I’m building for AllegroGraph. Also, I mistaken when I said:
>
>>> However, the validation report only contains the single result for maxCount. What does it not also contains results for the nodeKind constraint?
> I meant that it only had the nodeKind constraint violation.
>
> thanks again,
>
>
>> On Jan 29, 2019, at 6:02 PM, Holger Knublauch <holger@topquadrant.com> wrote:
>>
>> Are you referring to this test
>>
>> https://github.com/w3c/data-shapes/blob/gh-pages/data-shapes-test-suite/tests/core/path/path-sequence-duplicate-001.ttl
>>
>> - it only contains the sh:nodeKind violation.
>>
>> What implementation have you tried that delivers the maxCount violation?
>>
>> Holger
>>
>>
>> On 30/01/2019 2:21 am, gwking - metabang wrote:
>>> The focus node :A is
>>>
>>> ex:A
>>>    ex:p1 [
>>>        ex:p2 "value" ;
>>>      ] ;
>>>    ex:p1 [
>>>        ex:p2 "value" ;
>>>      ] ;
>>>
>>> This has two paths from :p1/:p2 to the literal “value”.
>>>
>>> The shape :SP is
>>>
>>>
>>> ex:SP
>>>    rdf:type sh:PropertyShape ;
>>>    sh:path (
>>>        ex:p1
>>>        ex:p2
>>>      ) ;
>>>    sh:maxCount 1 ;
>>>    sh:nodeKind sh:IRI ;
>>>
>>> which says that :A should:
>>>
>>> 1. have at most one path :p1/:p2, and
>>> 2. that the value at the end of the path should be an IRI.
>>>
>>> However, the validation report only contains the single result for maxCount. What does it not also contains results for the nodeKind constraint?
>>>
>>> thanks,