FHIR Chat · Invalid Consent StructureDefinition Identifier Example

Stream: conformance

Topic: Invalid Consent StructureDefinition Identifier Example

Brian Postlethwaite (Sep 02 2020 at 22:57):

I have a profile on Consent in my IG, and now I always get this error message that I can't suppress, and I can't work out how to make it go away, as the snapshot generation always puts it back in, even though I've profiled the element out...

<element id="Consent.identifier">
      <path value="Consent.identifier"/>
     ...
      <min value="0"/>
      <max value="0"/>
     ...
      <example>     (not in my differential, it just has max=0)
        <label value="General"/>
        <valueIdentifier>
          <system value="urn:ietf:rfc:3986"/>
          <value value="Local eCMS identifier"/>
        </valueIdentifier>
      </example>

The IG Publisher generates this error message:

StructureDefinition.snapshot.element[9].example[0].value.ofType(Identifier)
error
if identifier.system is ''urn:ietf:rfc:3986'', then the identifier.value must be a full URI (e.g. start with a scheme)

Anyway around this, or I have to live with the error :(

Brian Postlethwaite (Sep 02 2020 at 23:01):

Can see the above example is being snapshot re-generated from here:
http://hl7.org/fhir/R4/consent.profile.xml.html
A suggestion for the snapshot generator is to exclude the example if the element is profiled out (would fix my case, but doesn't fix the general issue of a dud base definition)
(and if I was to suppress other things in the snapshot when max=0, I'd consider mapping and invariants too)

Lloyd McKenzie (Sep 02 2020 at 23:17):

So the issue is that the base spec is broken? Should at minimum submit a technical correction to get it fixed

Brian Postlethwaite (Sep 02 2020 at 23:33):

Yes, yes I should do that.

Brian Postlethwaite (Sep 02 2020 at 23:39):

Done: https://jira.hl7.org/browse/FHIR-28426

Grahame Grieve (Sep 03 2020 at 00:39):

I'm positive I've already fixed this. What version does your IG depend on?

Shovan Roy (Sep 03 2020 at 00:43):

I've also raised this issue before : https://jira.hl7.org/browse/FHIR-27940. Created 10th July 2020. Showing at unresolved

John Moehrke (Sep 03 2020 at 11:59):

in what way is the IG publisher broken? Because the system given is 3986, yet your value is NOT a valid URL. Put in a URL from example.com domain?

John Moehrke (Sep 03 2020 at 12:08):

ah, so it is in the definition of consent? I don't see it in R5. Where did that come from?

David Pyke (Sep 03 2020 at 12:13):

It was in R4 (and STU3) but was removed from the build. A second bug was opened for R4 in case of a technical update

John Moehrke (Sep 03 2020 at 12:15):

the good news is that there is some work going on profiling Consent... Yeah!. Would be good to get visibility into these.

John Moehrke (Sep 03 2020 at 13:55):

so @Lloyd McKenzie and @Grahame Grieve -- should this result in a technical correction to previous releases? It certainly is preventing profiling of current release

Lloyd McKenzie (Sep 03 2020 at 14:09):

Yes, it should get treated as a technical correction to R4 - when/if we ever do one of those.

John Moehrke (Sep 03 2020 at 14:12):

@David Pyke You mention a CR for R4 technical correction. what number?
@Lloyd McKenzie how do we assign that CR to the backlog for technical correction to R4?