Feature/finalize fhir rdf #3
Conversation
…feature/finalize-fhir-rdf # Conflicts: # input/pagecontent/dynamic-data.md
rolandgroen
left a comment
rolandgroen
left a comment
There was a problem hiding this comment.
PR Review
Summary
This PR updates the RDF observation documentation in dynamic-data.md and extends the WellDataObservation profile to support additional value types (dateTime, string).
What's Good
- Profile and documentation are now aligned - the FSH profile supports all value types shown in the RDF example
- Version bump and changelog are properly updated
- Whitespace cleanup improves consistency
Issues
1. Invalid RDF/Turtle Syntax in Mock Example
The mock observation shows 4 separate fhir:value statements, each ending with . (period):
fhir:value [ ... ] . # ends the subject
fhir:value [ ... ] . # orphaned - no subject
fhir:value [ ... ] . # orphaned
fhir:value [ ... ] . # orphanedIn Turtle, . terminates a statement about a subject. Only the first fhir:value is actually attached to the observation - the other three are syntactically invalid/orphaned.
Suggestion: Show each value type as a separate commented-out alternative, or create 4 mini-examples with distinct subjects.
2. CodeableConcept Example Uses Wrong Structure
The profile supports CodeableConcept, but the RDF example shows:
fhir:value [
fhir:coding ( [ ... ] )
]This should be wrapped as:
fhir:value [
a fhir:CodeableConcept ;
fhir:coding ( [ ... ] )
]3. Practical Examples Removed
The previous documentation had individual RDF examples for specific observations (cholesterol ratio, HDL, height, weight, BMI, waist circumference, blood pressure). These practical examples are now gone.
Suggestion: Either restore simplified versions, or add a note pointing readers to the FSH example instances in input/fsh/instances/example-welldata-observation.fsh.
|
@rolandgroen Also added a comment on the importance of writing individual observations with uuid's in the observation folder. |
Align with FHIR ShEx