> On Sep 18, 2017, at 8:46 PM, Steven Folsom <sf...@cornell.edu> wrote:
>
> My understanding is that the semantics of the following shape is that every
> individual related to a ex:Work through ex:subject must be a member of ALL of
> the classes listed. Am I wrong to think this?
>
> sh:property [
> sh:path ex:subject ;
> sh:target ex:Work ;
> sh:class ex:Instance ;
> sh:class ex:Item ;
> sh:class ex:Temporal ;
> sh:class ex:Work ;
> sh:class ex:Event ;
> sh:class ex:Language ;
> sh:class ex:Concept ;
> sh:class ex:Location ;
> sh:class ex:Agent ;
> ] .
>
You need to use sh:targetClass
If you say
ex:MyNodeShape sh:targetClass ex:Work;
sh:property [
sh:path ex:subject ;
sh:class ex:Instance ;
sh:class ex:Item ;
sh:class ex:Temporal;
sh:class ex:Work ;
sh:class ex:Event ;
sh:class ex:Language ;
sh:class ex:Concept ;
sh:class ex:Location ;
sh:class ex:Agent ;
] .
You will indeed be saying that every resource that is an object of a triple
where a subject is a member (transitively) of ex:Work and a predicate is
ex:subject must be an instance (transitively) of all these classes.
> Based on the examples I've seen, using sh:in seems to be reserved for listing
> individuals. Am I wrong to think this? For example is it permitted to
> construct the following shape using sh:in for classes?
>
> sh:property [
> sh:path ex:subject ;
> sh:target ex:Work ;
> sh:in ( ex:Work ex:Instance ex:Item ex:Temporal ex:Event ex:Language
> ex:Concept ex:Location ex:Agent ) ;
> ] .
>
In your example, the property shape is contained within some node shape that
you are not listing here. I am concluding this because you are using
sh:property. In these cases, the target statement for the property shape will
be ignored. See my example above as to how to use targets.
You can use anything in sh:in, but what you listing above says that the object
of a triple with ex:subject predicate must be either ex:Work, ex:Instance, etc.
Not that the objects must be of certain types.
You can do:
ex:MyPropertyShape
a sh:PropertyShape ;
sh:targetClass a ex:Works ;
sh:path (ex:subject rdf:type) ;
sh:in ( ex:Work ex:Instance ex:Item ex:Temporal ex:Event ex:Language ex:Concept
ex:Location ex:Agent ) .
or
ex:MyNodeShape
a sh:NodeShape ;
sh:targetClass ex:Work;
sh:property [
sh:path (ex:subject rdf:type) ;
sh:in ( ex:Work ex:Instance ex:Item ex:Temporal ex:Event ex:Language ex:Concept
ex:Location ex:Agent )
].
The advantage of the second option is that you can then group together any
other constraints that members of the ex:Work class may have e.g.,
ex:MyNodeShape
a sh:NodeShape ;
sh:targetClass ex:Work;
sh:property [
sh:path (ex:subject rdf:type) ;
sh:in ( ex:Work ex:Instance ex:Item ex:Temporal ex:Event ex:Language ex:Concept
ex:Location ex:Agent );
];
sh:property [
sh:path ex:author ;
sh:minCount 1;
].
etc.
If you want to also allow subclasses of ex:Work, ex:Instance, etc. and you do
not want to list them explicitly, then you need to go with sh:class and use
sh:or
sh:property [
sh:path ex:subject ;
sh:or (
[
sh:class ex:Work ;
]
[
sh:class ex:Instance ;
]
[ …]# your other classes
)
] .
> My use cases are to be able to say in a dataset only these types of things
> can be the subject of a work AND to build a form that narrows the types of
> things that can be the subject of a ex:Work.
>
> Any help to understand how to say these types of constraints using SHACL
> would be greatly appreciated.
>
> Thanks,
> Steven
>
> ------------
> Steven Folsom
> Metadata Specialist
> Cornell University Library
> http://orcid.org/0000-0003-3427-5769
> http://vivo.cornell.edu/individual/sf433
> @sf433
>
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "TopBraid Suite Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to topbraid-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups
"TopBraid Suite Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to topbraid-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
