Stream: shorthand
Topic: Question about jurisdiction
Bart Decuypere (Mar 18 2022 at 13:43):
@Chris Moesel Your solution above does not work for me:
I get this error: The resource should declare its jurisdiction to match the package id (hl7.fhir.be.core, jurisdiction = urn:iso:std:iso:3166#BE) (for sushi-config.yaml: 'jurisdiction: urn:iso:std:iso:3166#BE "null"')
Notification Bot (Mar 18 2022 at 13:43):
Bart Decuypere has marked this topic as unresolved.
Julian Carter (Mar 18 2022 at 15:42):
Hi @Bart Decuypere,
I would suggest adding the following to your sushi-config.yaml file:
jurisdiction: urn:iso:std:iso:3166#BE "Belgium"
I'm not entirely sure why the IG Publisher suggested using null, but I'm not seeing anything beyond a warning of an incorrect text value when running the IG Publisher with the suggested jurisdiction value. What error were you seeing when you included the suggested value in your sushi-config.yaml file?
Elliot Silver (Mar 18 2022 at 16:18):
@Bart Decuypere , what version of sushi and what version of IG publisher are you using? There have been a number of changes, particularly in IG publisher, in this area in the last couple of months.
John Moehrke (Mar 18 2022 at 16:34):
jurisdiction also seems to prefer using m49 codesystem. It does not complain when using 3166, but it handles that a bit more bluntly.
https://unstats.un.org/unsd/methodology/m49/
John Moehrke (Mar 18 2022 at 16:37):
I have never bothered putting the display name on jurisdiction
John Moehrke (Mar 18 2022 at 16:47):
specifically the numbers from m49 work best.
John Moehrke (Mar 18 2022 at 21:29):
just found out from @Lloyd McKenzie that the m49 codes are not recommended, except for "World"... so you should be using the ISO 3166 codes.... BUT, my experience still does not display a flag graphic.
Bart Decuypere (Mar 21 2022 at 09:13):
@Elliot Silver I am using SUSHI v2.2.5 (implements FHIR Shorthand specification v1.2.0) and IG Publisher 1.1.108
using jurisdiction: urn:iso:std:iso:3166#BE "Belgium" does not help, and gives the same (identical) error messages.
John Moehrke (Mar 21 2022 at 11:45):
There is a IG build parameter that will apply the IG jurisdiction to all resources
John Moehrke (Mar 21 2022 at 11:46):
apply-jurisdiction: 'true'
John Moehrke (Mar 21 2022 at 11:46):
You likely want to turn that on.
John Moehrke (Mar 21 2022 at 11:47):
This is my default set of parameters
parameters: # see https://confluence.hl7.org/display/FHIR/Implementation+Guide+Parameters
shownav: 'true'
apply-contact: 'true'
apply-jurisdiction: 'true'
apply-publisher: 'true'
apply-version: 'true'
Bart Decuypere (Mar 22 2022 at 14:57):
@John Moehrke Sorry, but this does not work. I keep having the same error over and over again. My sushi-config.yaml looks like this:
canonical: https://www.ehealth.fgov.be/standards/fhir/core
fhirVersion: 4.0.1
FSHOnly: true
applyExtensionMetadataToRoot: false
id: hl7.fhir.be.core
name: EhealthFgovStandardsFhir
status: active
version: 2.0.0
jurisdiction: http://unstats.un.org/unsd/methods/m49/m49.htm#056 "Belgium"
parameters:
apply-jurisdiction: 'true'
As stated in https://chat.fhir.org/#narrow/stream/215610-shorthand/topic/Question.20about.20jurisdiction/near/276022521 , the other option doesn't work either.
John Moehrke (Mar 22 2022 at 15:22):
so you are using FSHOnly... thus the jurisdiction does not come from sushi-config, but your hand crafted IG xml/json.
When you use FSHOnly, the sushi processing is just to create the json artifacts, and not to setup the IG xml/json.
Mint Thompson (Mar 22 2022 at 15:51):
Hello @Bart Decuypere . I'm investigating this right now. In addition to what was mentioned about setting FSHOnly to false, I found that errors no longer occurred in the publisher when setting jurisdiction to urn:iso:std:iso:3166#BE in sushi-config.yaml. Is this jurisdiction setting acceptable for your IG, or do you need to be using the m49 code?
Mint Thompson (Mar 22 2022 at 16:03):
Based on what I'm seeing, a jurisdiction of http://unstats.un.org/unsd/methods/m49/m49.htm#056 "Belgium (BEL)" seems like it ought to work, but the publisher still emits the error with the unusual suggestion to have a display string of "null", which leads me to believe something unusual may be happening with the publisher.
John Moehrke (Mar 22 2022 at 16:06):
are you specifying a local terminology service?
John Moehrke (Mar 22 2022 at 16:06):
@Jose Costa Teixeira ?
John Moehrke (Mar 22 2022 at 16:07):
@Lloyd McKenzie tells me that m49 should only be used for universal/international... the others should all use 3166.
Jean Duteau (Mar 22 2022 at 16:09):
You should not be using the unstats.un.org codes except for the use of World (and perhaps Europe). All other codes need to be from iso3166. Here is the definition of the value set:
Include all codes defined in urn:iso:std:iso:3166
Include all codes defined in urn:iso:std:iso:3166:-2
Include codes from http://unstats.un.org/unsd/methods/m49/m49.htm where class = region
Belgium isn't a region so 056 is not a valid code in this value set.
When I use urn:iso:std:iso:3166#BE "Belgium", it works in my guide.
Jose Costa Teixeira (Mar 22 2022 at 16:28):
This seems some old stuff that was there in the first IGs.
@Bart Decuypere sorry i didn't see that we're still pointing to the ig.xml
Chris Moesel (Mar 23 2022 at 12:57):
So it looks like at least some of the confusion was because Giorgio was setting parameters in sushi-config.yaml (based on our advice), BUT his config was FSHOnly -- so the parameters were ignored. Ideally, SUSHI would have detected that he was using configuration parameters that are ignored in FSHOnly mode and warned him about that. I'll file a ticket against SUSHI for that.
Chris Moesel (Mar 23 2022 at 13:01):
Last updated: Apr 12 2022 at 19:14 UTC
