FHIR Chat · Illegally constrained XXX.id from II to string

Stream: cda/publish

Topic: Illegally constrained XXX.id from II to string

Sean McIlvenna (Dec 12 2019 at 22:11):

Getting this error in the cda-core-2.0 IG:

org.hl7.fhir.exceptions.DefinitionException: StructureDefinition http://hl7.org/fhir/cda/StructureDefinition/Entity at Entity.id: illegal constrained type http://hl7.org/fhir/cda/StructureDefinition/II from string in http://hl7.org/fhir/StructureDefinition/Element
    at org.hl7.fhir.r5.conformance.ProfileUtilities.updateFromDefinition(ProfileUtilities.java:2030)
    at org.hl7.fhir.r5.conformance.ProfileUtilities.processPaths(ProfileUtilities.java:736)
    at org.hl7.fhir.r5.conformance.ProfileUtilities.generateSnapshot(ProfileUtilities.java:472)
    at org.hl7.fhir.igtools.publisher.Publisher.generateSnapshot(Publisher.java:3773)
    at org.hl7.fhir.igtools.publisher.Publisher.generateSnapshots(Publisher.java:3726)
    at org.hl7.fhir.igtools.publisher.Publisher.loadConformance(Publisher.java:3181)
    at org.hl7.fhir.igtools.publisher.Publisher.createIg(Publisher.java:768)
    at org.hl7.fhir.igtools.publisher.Publisher.execute(Publisher.java:656)
    at org.hl7.fhir.igtools.publisher.Publisher.main(Publisher.java:6622)

Sean McIlvenna (Dec 12 2019 at 22:12):

I believe this is related to other discussions we've had about basing things on Element, and that Element has Element.id

Sean McIlvenna (Dec 12 2019 at 22:12):

@Grahame Grieve , can't remember the solution you proposed... I think maybe you were suggesting creating BaseElement which has BaseElement.id and derives from Element, which is empty.

Sean McIlvenna (Dec 12 2019 at 22:13):

I believe that would solve this

Sean McIlvenna (Dec 12 2019 at 22:13):

But the FHIR spec would have to change baseDefinition references from Element to BaseElement

Sean McIlvenna (Dec 12 2019 at 22:13):

or... maybe instead we could create BaseElement as being the empty profile, and Element derives from BaseElement

Sean McIlvenna (Dec 12 2019 at 22:13):

then no references would need to change

Sean McIlvenna (Dec 12 2019 at 22:13):

and the cda-core-2.0 IG would just need to use a baseDefinition of BaseElement instead of Element

Sean McIlvenna (Dec 12 2019 at 22:16):

Until this is fixed, though, I am going to push a commit that comments out the <type> constraints on all XXX.id elements so that the build can run succesfully

Grahame Grieve (Dec 12 2019 at 22:23):

I don't follow how that fixes it

Grahame Grieve (Dec 12 2019 at 22:23):

@Oliver Egger did we have any other solution for this?

Grahame Grieve (Dec 12 2019 at 22:23):

I have defined Base formally in R5 now

Sean McIlvenna (Dec 12 2019 at 22:36):

right now, Element has "id" which is of type string

Sean McIlvenna (Dec 12 2019 at 22:36):

in CDA, not all elements have "id"

Sean McIlvenna (Dec 12 2019 at 22:36):

and the ones that do are of type "II"

Sean McIlvenna (Dec 12 2019 at 22:36):

so, CDA elements can currently be based on Element because of that conflict

Sean McIlvenna (Dec 12 2019 at 22:37):

we either need to remove "id" from "Element"

Sean McIlvenna (Dec 12 2019 at 22:37):

or we need to create something new that doesn't have "id" in it that CDA elements can be based on

Oliver Egger (Dec 12 2019 at 22:52):

In the fork I created a Base Element for R4 in the Implementation Guide, since the discussion on Zulip Logical Models was not finished I did not create a pull request.

However if a Base is integrated in R5 I see two options:
1) if you accept a Base Definition in the cda-core-2.0 R4 IG I create a pull request based on the commit above
or
2) change the cda-core-2.0 guide to a R5 version. Since it is a logical model the FHIR version does anyhow not depend much on it .

which approach do you prefer? Given the development of R5 I don't see alternatives to above.

Grahame Grieve (Dec 12 2019 at 22:58):

let me investigate

Grahame Grieve (Dec 13 2019 at 00:42):

ok, so this is what I'm going to use as my test case:

Grahame Grieve (Dec 13 2019 at 00:42):

cda.xml example.xml ii.xml any.xml

Grahame Grieve (Dec 13 2019 at 00:43):

that captures the essence the problem, yes?

Grahame Grieve (Dec 13 2019 at 02:06):

well, that will work next release. So just go ahead and change everything to Base

Grahame Grieve (Dec 13 2019 at 02:09):

next release is not that many hours away :-(

Sean McIlvenna (Dec 13 2019 at 03:20):

so there will be a "Base" type that is empty?

Grahame Grieve (Dec 13 2019 at 03:21):

yes

Sean McIlvenna (Dec 13 2019 at 03:25):

awesome!

Grahame Grieve (Dec 13 2019 at 06:05):

ok so I pushed more to the tests. I'm going to get terminlogy validation working

Grahame Grieve (Dec 13 2019 at 06:19):

and the tests are here:

Grahame Grieve (Dec 13 2019 at 06:29):

https://github.com/FHIR/fhir-test-cases/tree/master/validator/cda

Sean McIlvenna (Dec 13 2019 at 06:47):

ok, please let me know when a new build is ready for me to play with. I'll check out the Base stuff

Oliver Egger (Dec 13 2019 at 07:36):

cool what happens while sleeping :-) will check it out also to test

Grahame Grieve (Dec 13 2019 at 10:57):

ok terminology validation works... sort of.

Grahame Grieve (Dec 13 2019 at 10:58):

we will have to work on CDA value sets - they don't seem to exist in UTG?

Grahame Grieve (Dec 13 2019 at 10:58):

we will also have to work on OID 2 URI translations

Grahame Grieve (Dec 13 2019 at 10:58):

otherwise... I think it's solid

Sean McIlvenna (Dec 13 2019 at 16:25):

I assume the "Base" stuff is not ready, yet? I just got latest and tried with http://hl7.org/fhir/StructureDefinition/Base and it failed

Grahame Grieve (Dec 13 2019 at 19:52):

I haven't released it yet. I will sometime in the next few hours

Sean McIlvenna (Dec 14 2019 at 03:48):

Awesome. Given that your last message was eight hours ago, should I download the latest publisher and try it out?

Grahame Grieve (Dec 14 2019 at 04:09):

yes

Sean McIlvenna (Dec 14 2019 at 04:19):

Hmmm... I must be missing something.
I changed <baseDefinition value="http://hl7.org/fhir/StructureDefinition/Element"/> to <baseDefinition value="http://hl7.org/fhir/StructureDefinition/Base"/> and it's giving a NullPointerException

Sean McIlvenna (Dec 14 2019 at 04:25):

and fyi, I removed hl7.fhir.r4.core#4.0.1 from my package cache and had it re-download it

Grahame Grieve (Dec 14 2019 at 04:38):

what's the stack?

Sean McIlvenna (Dec 14 2019 at 05:00):

Stack Dump (for debugging):                                                      (00:30.0910)
java.lang.NullPointerException
        at org.hl7.fhir.igtools.publisher.Publisher.generateSnapshot(Publisher.java:3813)
        at org.hl7.fhir.igtools.publisher.Publisher.generateSnapshots(Publisher.java:3762)
        at org.hl7.fhir.igtools.publisher.Publisher.loadConformance(Publisher.java:3217)
        at org.hl7.fhir.igtools.publisher.Publisher.createIg(Publisher.java:768)
        at org.hl7.fhir.igtools.publisher.Publisher.execute(Publisher.java:655)
        at org.hl7.fhir.igtools.publisher.Publisher.main(Publisher.java:6662)