Stream: implementers
Topic: cardinality for backbone in schema and data elements in d...
Marvin Greenberg (Feb 28 2020 at 15:36):
Hi, I'm new to this chat so apologies if I'm asking in the wrong place. We have some code that we create validations schemas. Previously (dstu2, stu3) we used the dataelements.json to do this. But the structure of this changed significantly in r4 (https://www.hl7.org/fhir/definitions.json.zip) AND it seems like the new structure has lost some information, specifically the cardinality for the backbone elements. I also looked at the fhir.schema.json and it also seems not to have the cardinality. Am I just missing something? Thanks!
Lloyd McKenzie (Feb 28 2020 at 15:37):
@Grahame Grieve
Grahame Grieve (Feb 28 2020 at 19:00):
the json schema looks like it has the cardinality to me but it's subtle in json schema
Marvin Greenberg (Feb 28 2020 at 19:04):
So there is something that says array, 1..1 for 'source' in AuditEvent and 1..* for agent in AuditEvent?
Grahame Grieve (Feb 28 2020 at 19:06):
not in json schema, no. The type is array or not, and the element is in the required list or not
Marvin Greenberg (Feb 28 2020 at 19:08):
required list... Ahah
Marvin Greenberg (Feb 28 2020 at 19:08):
I get it. So a required elt of type array is 1..*
Grahame Grieve (Feb 28 2020 at 19:09):
as for the definitions file... that seems pretty screwed up on backbone elements. E.g. Account.coverage somehow magically morphs into http://hl7.org/fhir/StructureDefinition/de-Account.coverage.coverage and the cardinalities are wrong
Marvin Greenberg (Feb 28 2020 at 19:10):
Thanks. And I guess, the data elements StructureDefinitions don't have that info for backbone elements because they aren't intended to be a schema
Grahame Grieve (Feb 28 2020 at 19:10):
I think they are intended to be there
Marvin Greenberg (Feb 28 2020 at 19:10):
Um, that is what I guess I concluded. OK.
Marvin Greenberg (Feb 28 2020 at 19:11):
It isn't obvious where to put them the way dataelement is currently structured.
Grahame Grieve (Feb 28 2020 at 19:11):
Added to the list of R4 known issues: https://confluence.hl7.org/display/FHIR/Known+Issues+with+the+published+FHIR+Specifications
Marvin Greenberg (Feb 28 2020 at 19:12):
But thanks. I'll see if it is worth right now for me to switch to using the schema.
Grahame Grieve (Feb 28 2020 at 19:14):
if you're going to switch, just use the structure definitions... they are the formal authority
Marvin Greenberg (Feb 28 2020 at 19:17):
I am not sure I understand. The structure definitions in dataelements.json is what we were using - that is missing the backbone cardinality.
Marvin Greenberg (Feb 28 2020 at 19:19):
Maybe you are saying that the schema has certain elements that ARE StructureDefns? I'm not very familiar with json schema
Grahame Grieve (Feb 28 2020 at 19:22):
you should use the structure definitions in resources.json and types.json. Or even better, switch to the core package (hl7.fhir.r4.core), which is the same definitions, but the format we'll be maintaining going forward
Marvin Greenberg (Feb 28 2020 at 19:29):
OK, thanks
Last updated: Apr 12 2022 at 19:14 UTC
