Looking into how the semantic annotations are currently translated, we need to rethink due to the following points:
- Semantic annotations should not be converted like key-value pairs. Currently, something like
"eclass:hasQuantity": "eclass14-0:0173-1-Z4-BAJ272-003" becomes {"idShort": "unit", "valueType": "xs:string", "value": "eclass14-0:0173-1-05-AAA153-004", "modelType": "Property"}. It should be done via semanticId, like how it is done for the WoT terms.
- IDTA prefers the usage of
conceptDescriptions to store internal and even external semantic references. That means that all our semantic references need to point to a concept description in the root of the AAS. Implementing that requires rethinking since we cannot do that if we generate only the submodel.
I would propose starting with the first point already. The second one can be done after a thorough discussion.
Looking into how the semantic annotations are currently translated, we need to rethink due to the following points:
"eclass:hasQuantity": "eclass14-0:0173-1-Z4-BAJ272-003"becomes{"idShort": "unit", "valueType": "xs:string", "value": "eclass14-0:0173-1-05-AAA153-004", "modelType": "Property"}. It should be done via semanticId, like how it is done for the WoT terms.conceptDescriptionsto store internal and even external semantic references. That means that all our semantic references need to point to a concept description in the root of the AAS. Implementing that requires rethinking since we cannot do that if we generate only the submodel.I would propose starting with the first point already. The second one can be done after a thorough discussion.