W3C

– DRAFT –
FHIR RDF

18 November 2021

Attendees

Present
Brad Simons, David Booth, EricP, Gopi, James Champion, Jim Balhoff
Regrets
-
Chair
David Booth
Scribe
dbooth

Meeting minutes

Examples of most proposed R5 changes

eric: I wrote up all the variations: https://www.w3.org/2021/12/fhir-egp/

eric: Also want to change fhir:valueX (where X is a datatype) to fhir:value

david: That means that for round tripping, the valueX mapping to RDF data types must be 1:1.

eric: These examples illustrate multiple R5 changes I'd like to see.

eric: I changed fhir:valueQuantity to fhir:value

david: That raises the OWL question of whether fhir:value is a datatype property or an object property

eric: We could run a check to make sure it isn't declared as both.

brad: To go from FHIR RDF back to FHIR JSON, do you anticipate HAPI being updated with these R5 changes?

eric: Yes, I'm responsible for that HAPI code.

brad: HAPI can convert FHIR JSON to FHIR RDF?

eric: Yes.

eric: Look here: https://github.com/hapifhir/hapi-fhir/blob/master/hapi-fhir-structures-r4/src/test/java/ca/uhn/fhir/parser/RDFParserTest.java#L91-L115

eric: That tests round-tripping using HAPI.
… Runs for about an hour testing all of the FHIR examples.

eric: This example shows all of the R5 changes I hope for: https://www.w3.org/2021/12/fhir-egp/short-resource-and-datatype-names-no-datatype-currying-short-scalars-rdf-collection.ttl

david: In shortening property names, what is the consequent if there is a conflict when a name is shortened?

eric: Minimal. I suspect nobody will complain.

https://researchsoftwareinstitute.github.io/data-translator/apps/camp-fhir

david: You manually edited these examples. (Thank you!) I think we need to be able to auto-generate them, in order to test any desired example, before we try to finalize these changes in the HL7 balloting.

eric: Biggest risks we face are places where the FHIR folks have not yet though much about it, such as new resources.

eric: James Champion is working on code to turn FHIR JSON to FHIR RDF, with proposed R5 changes. Deepak is working on synthesis. I'm working on the shex code.

david: Running out of time for Jan R5 balloting.

james: working on campFHIR project (URL above). Connects to a DB, ETL queries the DB, hands data to HAPI, and outputs FHIR JSON. But I can choose what format I output.
… JSON, RDF or XML
… This could be a way to test this.
… This could output 100k examples as FHIR R4, then another tool could convert R4 to R5.

jim: Using framing for RDF lists. But when I round trip back to FHIR JSON, ordered lists have an @list property.

eric: I think using Framing to go back to FHIR JSON is nice-to-have but not essential, because it can be done using other techniques instead of framing.

ADJOURNED

Minutes manually created (not a transcript), formatted by scribe.perl version 159 (Fri Nov 5 17:37:14 2021 UTC).

Diagnostics

Succeeded: s/Issues 77 and 93 (extensions and modifier extensions)/Examples of most proposed R5 changes

Succeeded: s|https://github.com/w3c/hcls-fhir-rdf/issues/77||

Succeeded: s|https://github.com/w3c/hcls-fhir-rdf/issues/93||

No scribenick or scribe found. Guessed: dbooth

Maybe present: brad, david, eric, james, jim