Issue 534: Support for property paths in SHACL UI

[smessie]

The document currently lacks screenshots but provides descriptions of the UI. We can add screenshots of potential UIs later, e.g., once we have a first implementation of a SHACL UI library.

Contributes to #534

• See this document rendered online here

[bergos]

I propose removing the targets from the examples. Only the section about target finding should contain them. The other sections should not need them, which makes the examples shorter and easier to grasp.

The rest looks good to me.

[edmondchuc]

I’m wondering if we should mention the restrictions on alternative paths in edit mode as described in #534 (comment), specifically on which predicate paths are allowed.

Should we also add a paragraph on the challenges of supporting complex paths in edit mode?

Otherwise, everything else looks good to me.

[smessie]

@edmondchuc I’ve pushed a commit that addresses your comment. Is it what you had in mind, or are there specific things you’re missing, or is it too verbose? I’m open to suggestions.

[bergos]

@smessie I think your last additions to alternative path and other complex paths are very useful.

One small thing: Can you add a shape to the Other Complex Paths example? I think it would be useful to show sh:path in that case.

[smessie]

@bergos I wasn't sure yet if a sequence path (sh:path ( ex:hasPart ex:hasPart ) ;) would maybe make more sense than the zero-or-more-path I've used now, but if you think it does, I can still change it.

[bergos]

@smessie not required to change anything. The path is fine like it is.

[TallTed]

Screenshots may still be overly ambitious, but wireframes may be helpful to comprehension.