FHIR Chat · Element.id has no type code? · conformance

Stream: conformance

Topic: Element.id has no type code?

Michel Rutten (Dec 18 2018 at 15:23):

In R4, the element definition for "Element.id" has slightly changed.
In STU3, the type.code property used to specify "string":

<type>
  <code value="string" />
 </type>

In R4, type.code now contains only special "compiler magic" extensions:

<!-- Note: primitive values do not have an assigned type. e.g. this is compiler magic. XML, JSON and RDF types provided by extension -->
<type>
  <extension url="http://hl7.org/fhir/StructureDefinition/regex">
    <valueString value="[^\s]+(\s[^\s]+)*"/>
  </extension>
  <code>
    <extension url="http://hl7.org/fhir/StructureDefinition/structuredefinition-json-type">
      <valueString value="string"/>
    </extension>
    <extension url="http://hl7.org/fhir/StructureDefinition/structuredefinition-xml-type">
      <valueString value="xsd:token"/>
    </extension>
    <extension url="http://hl7.org/fhir/StructureDefinition/structuredefinition-rdf-type">
      <valueString value="xsd:token"/>
    </extension>
  </code>
</type>

I assume that this change is intentional?

Grahame Grieve (Dec 18 2018 at 20:37):

yes

Grahame Grieve (Dec 18 2018 at 20:37):

it's not a FHIR string - it doesn't allow extensions, for instance

Michel Rutten (Dec 18 2018 at 23:01):

OK, makes sense. This also implies that Element.id no longer inherits constraints (e.g. ele-1) from root Element, as confirmed by the snapshot expansion in official profiles-types.xml - correct?

Michel Rutten (Dec 18 2018 at 23:01):

(just making sure our compiler magic aligns)

Grahame Grieve (Dec 19 2018 at 03:12):

umm I have't thought about that. I guess it doesn't

Brian Postlethwaite (Dec 19 2018 at 19:31):

Is this the same definition in extension.url?

Ewout Kramer (Dec 19 2018 at 21:31):

should be, I guess, same principle...

Grahame Grieve (Dec 19 2018 at 22:03):

yes

Eric Haas (Dec 21 2018 at 01:06):

OK so I am adding fhirprimitive extensions to the Python model... I am excluding them for extension.url and id and what about all the valueXs in the extensions can they have id and extensions?

Eric Haas (Dec 21 2018 at 01:07):

are there any other places where a primitive is not a 'fhirprimitive'?

Grahame Grieve (Dec 21 2018 at 02:32):

the primitive value elements

Eric Haas (Dec 21 2018 at 03:16):

OK and what about for example a valueString element in an extension?

Grahame Grieve (Dec 21 2018 at 03:56):

that's a full string that can have extensions of it's own

Last updated: Apr 12 2022 at 19:14 UTC