Minor inconsistencies in ADL2 spec specialisation section


9.2.5. Redefinition Concept, to ADD a new node
the criteria reads: 'Differential node has a specialised node identifier, and corresponding node in flat parent has max occurrences > 1.'
When adding a node, I would think there is no corresponding node in the flat parent, so there can be no condition on occurrences.

  • 9.3.1 redefinition for refinement:
    The text mentions node id id2 and id2.1, but it is not present in ADL in the example

  • 9.3.2 specialization with cloning
    The node identifiers in the ADL2 source and the workbench screenshot are not consistent, the ADL lists id2.3 as a subnode of id3.2, the workbench screenshot consistently does id3.1 -> id2.1, id3.2 -> id2.2, etc.




Thomas Beale
October 4, 2017, 1:45 PM

If ADDing a new object node is adding a new child either of a container or of a single-valued attribute, where a child already exists, then it will be treated as a specialisation of the existing child, in which case the rule is correct. But it doesn't mention the case when there is a pure ADD, it is the first constraint on a child member of the parent attribute. This needs to be fixed.

Error in 9.3.1 - yes, this is a typo and needs to be corrected.

The screen shot looks correct to me - I assumed this is the one in question: http://www.openehr.org/releases/AM/latest/docs/ADL2/images/specialisation_with_cloning.png . The corresponding ADL text also looks correct to me.

Pieter Bos
October 6, 2017, 8:46 AM

That is the screenshot indeed. It's a very minor inconsistency: the screen shot is the correct structure, but it has different node ids for the id2.x nodes. For example, in the example id2.3 is HDL cholesterol and in the screenshot it is id2.2. I would think the screenshot is better

Thomas Beale
October 6, 2017, 8:58 AM

Ah, got it. I've synchronised the text example to what is in the screenshot.


Pieter Bos




Affects versions