Stream: conformance
Topic: extension: structuredefinition-wg
Simone Heckmann (Jun 19 2017 at 12:19):
When creating a new profile with forge, the resulting XML maintains the core profile's extension "structuredefinition-wg" and it's value. e.g.:
<StructureDefinition xmlns="http://hl7.org/fhir"> <extension url="http://hl7.org/fhir/StructureDefinition/structuredefinition-wg"> <valueCode value="pa"/> </extension> <url value="http://fhir.de/StructureDefinition/patient-de-basis"/> [...] <type value="Patient"/> <baseDefinition value="http://hl7.org/fhir/StructureDefinition/Patient"/> <derivation value="constraint"/>
I don't think that's correct, as derived profile are usually not maintained by the HL7 workgroup.
Is it legitimate to remove the extension in the profiles?
Mirjam Baltus (Jun 19 2017 at 12:56):
I think you have a point there, so I will raise an issue for Forge to leave this out in the future.
Lloyd McKenzie (Jun 19 2017 at 14:48):
That's actually an interesting question. I don't know about the root of the StructureDefinition, but there will certainly be extensions on elements inside the StructureDefinition that you do want to inherit to child profiles. (E.g. min length, regex, etc.) Do we inherit everything on element? If not, how do we distinguish what gets inherited and what doesn't?
Chris Grenz (Jun 21 2017 at 15:00):
IMO, None of the StructureDefinition elements should inherit. And all extensions on elements should inherit/aggregate in order (most "base" first) in with no capability to overwrite/replace.
Lloyd McKenzie (Jun 21 2017 at 16:57):
@Chris Grenz I'm not sure what you mean by "none of the StructureDefinition elements should inherit". They absolutely inherit. If you create StructureDefinition B that is a constraint on StructureDefinition A, B will inherit the bindings, comments, definitions, etc. from A unless it overrides them. There are rules for what sort of overrides can happen, but overrides can definitely happen too. The challenge with extensions is that it's not necessarily clear what the rules should be.
Chris Grenz (Jun 21 2017 at 16:59):
None of the top-level elements of StructureDefinition (extensions, name, title, publisher, etc...) should inherit.
Lloyd McKenzie (Jun 21 2017 at 17:00):
Oh, agree with that. I'm talking about extensions on element.
Chris Grenz (Jun 21 2017 at 17:01):
Only elementdefinition elements inherit. And, I'd say (although not authoritatively) that elementdefinition.extension should inherit as an aggregable element (https://github.com/chrisgrenz/FHIR-Primer/wiki/Aggregating-Profile-Differentials#aggregating-elements)
Lloyd McKenzie (Jun 21 2017 at 17:51):
That sort of works, though for things like minLength, etc. it's pretty inefficient. And if any of them are documentation-type elements, you wouldn't necessarily want them to repeat
Last updated: Apr 12 2022 at 19:14 UTC
