You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jan 30, 2024. It is now read-only.
In either context of searching for a node, displaying it in a tree, an autocomplete, or wherever other case. We should be able to define the logic in which the label of a given node is determined or computed.
Simple property at the sh:PropertyShape level
At its simplest, a property shape could be defined as being used for defining the label of the node. Not advising a particular nomenclature just yet, but it could be defined as such:
A boolean might be enough, but I do like the idea of defining purposes/use-cases for shapes in general (not only linked to labelling, I might created a dedicated issue for that). We could imagine other labelling purposes like shui:AutocompleteLabelPurpose, shui:SearchLabelPurpose, shui:TreeLabelPurpose and so on, depending on what we feel might make sense.
"Templating" at the sh:NodeShape level
The previous proposition would likely work for most cases, but what about label composition? For example, we might want to concatenate, for a given skos:Concept, its skos:prefLabel and the skos:prefLabel of all its "broader chain". For such more complex cases, being able to make use of a sparql query to build the label triples would be a good option:
Defining labelling logic at 2 possible different levels might be both counter-intuitive and hard to follow, so maybe have everything at the sh:NodeShape level instead?
In my experience, the expectation for labels is that they also work in the other direction, e.g. for auto-complete purposes. Do you have thoughts on how this could work if labels are dynamically created?
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
In either context of searching for a node, displaying it in a tree, an autocomplete, or wherever other case. We should be able to define the logic in which the label of a given node is determined or computed.
Simple property at the
sh:PropertyShape
levelAt its simplest, a property shape could be defined as being used for defining the label of the node. Not advising a particular nomenclature just yet, but it could be defined as such:
A boolean might be enough, but I do like the idea of defining purposes/use-cases for shapes in general (not only linked to labelling, I might created a dedicated issue for that). We could imagine other labelling purposes like
shui:AutocompleteLabelPurpose
,shui:SearchLabelPurpose
,shui:TreeLabelPurpose
and so on, depending on what we feel might make sense."Templating" at the
sh:NodeShape
levelThe previous proposition would likely work for most cases, but what about label composition? For example, we might want to concatenate, for a given
skos:Concept
, itsskos:prefLabel
and theskos:prefLabel
of all its "broader chain". For such more complex cases, being able to make use of a sparql query to build the label triples would be a good option:Wrapping things all together
Defining labelling logic at 2 possible different levels might be both counter-intuitive and hard to follow, so maybe have everything at the
sh:NodeShape
level instead?But I'll stop there for now, as we first need to decide it this feature makes sense in shacl-ui or not 😅
The text was updated successfully, but these errors were encountered: