W3C

– DRAFT –
FHIR RDF

23 February 2023

Attendees

Present
Daniel Stone, David Booth, EricP, James Champion, Jim Balhoff, Rob Hausam
Regrets
-
Chair
David Booth
Scribe
dbooth

Meeting minutes

PR status

eric: PR for shex is not ready yet. Needs testing against examples.
… The schema for Patient needs the import statements.

dbooth: Daniel, any issues on your PR?

daniel: Not sure if integer64 change to xsd:integer will cause a problem.

dbooth: for fhir:date definition, xsd:dateTime is wrong.

ACTION: jim to give daniel the correct OWL disjunction (for the union of types) instead of xsd:dateTime for defining fhir:date

looking at Patient.birthDate

jim: _:x rdf:type rdfs:Datatype .

_: x owl:unionOf T(SEQ DR1 ... DRn) .

http://build.fhir.org/datatypes.html#code

jim: some of the owl:someValuesFrom have an extra blank node that causes weird display in protege:

fhir:dateTime ....
                                          owl:someValuesFrom  [ a               rdfs:Datatype ;
                                                                owl:onDatatype  xsd:dateTime
                                                              ]

eric: For fhir:dateTime and a few others, we'll want a union. But for most of the primitive types, we don't want the extra bnode with owl:someValuesFrom

eric: Specifically, fhir:date, fhir:dateType and fhir:decimal

fhir:date rdf:type owl:Class ;
          rdfs:subClassOf fhir:Primitive ,
                          [ rdf:type owl:Restriction ;
                            owl:onProperty fhir:v ;
                            owl:someValuesFrom [ rdf:type rdfs:Datatype ;
                                                 owl:unionOf ( xsd:date
                                                               xsd:gYear
                                                               xsd:gYearMonth
                                                             )
                                               ]
                          ] ;

rob: One PR needs a rebase

ACTION: Jim to rebase his PR

Playground

james: Which version of R5 should I use?

rob: Snapshot 3 might be a better choice.

rob: For v4, you should use 4.3.0 -- R4B

james: Patient._birthDate example fails.

eric: I didn't work hard to make sure extensions work.

eric: You could just take out that example.

james: Eventually we'll add in a validator.

dbooth: Change "Show JSON-LD Editor" to "Show JSON-LD Viewer" if it is not editable by the user.

james: I fixed the bug of re-rendering many times.

dbooth: Please change "Click here to use R4" to a drop-down list of versions.

james: Where in the code does it load the default version?

eric: Try moving the files and let it do a stack dump, and see where it failed.
… Use the first entry in the manifest, instead of hard-coded default.

james: I'll make these changes, then update my branch.
… Then update the main.

eric: Maybe needs to be gh-pages branch, to show up on github.io
… I'll check which branch

Summary of action items

  1. jim to give daniel the correct OWL disjunction (for the union of types) instead of xsd:dateTime for defining fhir:date
  2. Jim to rebase his PR
Minutes manually created (not a transcript), formatted by scribe.perl version 210 (Wed Jan 11 19:21:32 2023 UTC).

Diagnostics

Succeeded: s/modifier/

Succeeded: s/Needs/Maybe needs/

No scribenick or scribe found. Guessed: dbooth

Maybe present: _, daniel, dbooth, eric, james, jim, rob

All speakers: _, daniel, dbooth, eric, james, jim, rob

Active on IRC: dbooth