Stream: IG creation
Topic: Specifying US edition
Lloyd McKenzie (Dec 04 2020 at 14:17):
The newest validator raises a new error: "The IG does not specify the US edition of SNOMED CT...". When I go looking for what parameter to set in the Confluence documentation (https://confluence.hl7.org/display/FHIR/Implementation+Guide+Parameters), it's not clear what the parameter name should be. Guidance @Rob Hausam @Grahame Grieve
Rob Hausam (Dec 04 2020 at 14:57):
Hmm. That's a good question. It's not clear to me, either. We used to specify that explicitly (in the ini file?) before we moved to the new template infrastructure, but, unless I'm overlooking it somehow, I don't see that I have anything that specifies it in any of my current IGs. And I'm also not locating guidance on where and how to specify it.
Eric Haas (Dec 04 2020 at 15:14):
See US core. Parameter expansion soutce or something like that and then stick the parameter in your directory
Eric Haas (Dec 04 2020 at 15:15):
Ours is in _resources
Eric Haas (Dec 04 2020 at 15:15):
Look at the sushi config file
Eric Haas (Dec 04 2020 at 16:58):
Eric Haas (Dec 04 2020 at 16:59):
Eric Haas (Dec 04 2020 at 16:59):
{
"resourceType" : "Parameters",
"id" : "exp-params",
"parameter" : [{
"name" : "system-version",
"valueUri" : "http://snomed.info/sct|http://snomed.info/sct/731000124108"
}]
}
Eric Haas (Dec 04 2020 at 17:01):
...
parameters: # see https://confluence.hl7.org/display/FHIR/Implementation+Guide+Parameters
path-resource:
- input/resources
- fsh-generated/resources
path-pages:
- input/pagecontent
- input/intro-notes
- fsh-generated/includes
#- input/summaries
path-expansion-params: '../../input/_resources/exp-params.json' # for using US ed of SNOMED
active-tables: 'false'
apply-contact: 'true'
...
Lloyd McKenzie (Dec 04 2020 at 19:33):
Thanks Eric. @Grahame Grieve, I presume there's some specific convention for use of this that should be added to the documentation on the confluence page?
Rob Hausam (Dec 04 2020 at 19:36):
Yeah. I don't know how Eric found or figured this out. And it doesn't seem to be the most intuitive way that I could probably think of - but as long as it works that's the main thing.
Lloyd McKenzie (Dec 04 2020 at 19:41):
I added that parameter to my IG and it didn't help with the error :(
Grahame Grieve (Dec 04 2020 at 20:12):
which documentation on which page? The actual technical documentation is
https://confluence.hl7.org/display/FHIR/Implementation+Guide+Parameters + http://hl7.org/fhir//valueset-operation-expand.html but it's one of those cases where the documentation only makes sense when you already know the answer
Lloyd McKenzie (Dec 05 2020 at 02:27):
The documentation I was looking at was https://confluence.hl7.org/display/FHIR/Implementation+Guide+Parameters. I didn't see it there. So what is the answer?
Grahame Grieve (Dec 05 2020 at 18:46):
Eric already answered that
Lloyd McKenzie (Dec 06 2020 at 04:00):
I tried. It didn't work...
Grahame Grieve (Dec 06 2020 at 09:33):
well, that's what I did for US core. So double check, and I guess I'll have to look
Lloyd McKenzie (Dec 06 2020 at 20:10):
I think I've done it the same, though obviously I'm not using FSH. IG is davinci-crd.
Grahame Grieve (Dec 06 2020 at 20:27):
I don't see any entry in parameters for it
Lloyd McKenzie (Dec 06 2020 at 21:06):
Sorry. I guess committing would be helpful
Grahame Grieve (Dec 06 2020 at 22:34):
so you haven't copied what Eric had above. In you IG:
parameter
name = path-expansion-params
path = {{some path}}
at {{some path}}
"resourceType" : "Parameters",
"id" : "exp-params",
"parameter" : [{
"name" : "system-version",
"valueUri" : "http://snomed.info/sct|http://snomed.info/sct/731000124108"
}]
}
Lloyd McKenzie (Dec 06 2020 at 22:48):
Why is it not just an IG parameter? How do we keep this Parameters instance from publishing?
Grahame Grieve (Dec 06 2020 at 22:54):
there might be any number of parameters you want to add, and it's possible that there would be name clashes between IG parameters and terminology parameters. You might even want to have different sets of expansion parameters (e.g. different versions, in the future). So there's several good reasons to keep the parameter lists separate
It's up to the author to decide whether to publish it. It won't be automatically published unless you put in a path where resources are automatically picked up from
Lloyd McKenzie (Dec 06 2020 at 23:08):
How does it get picked up if you put it in a different path?
Lloyd McKenzie (Dec 06 2020 at 23:08):
never mind
Lloyd McKenzie (Dec 06 2020 at 23:13):
@Sean McIlvenna Is there a way to do this in Trifolia?
Lloyd McKenzie (Dec 06 2020 at 23:16):
@Grahame Grieve Committed, and it seems to be finding the file (because it yells if I get the path wrong). However, I'm still seeing the QA issue...
Grahame Grieve (Dec 06 2020 at 23:40):
well, you committed it wrong...
Grahame Grieve (Dec 07 2020 at 01:14):
but also so did I. Fixed in the release happening now
Lloyd McKenzie (Dec 07 2020 at 03:39):
What did I commit wrong?
Grahame Grieve (Dec 07 2020 at 09:26):
you had the path wrong - was pointing to \input\input
Lloyd McKenzie (Dec 07 2020 at 14:25):
Argh. Committed the one I'd tested and known was wrong. Sorry. And thanks for the pointer in the right direction.
Sarah Gaunt (Dec 07 2020 at 19:52):
Little late, but @Lloyd McKenzie I've done this in Trifolia. Just created the Parameter resource and then set parameters on the "other" tab for the IG. BFDR and eCR both do this (and maybe Vital Records Common Profiles Library, can't remember)
Lloyd McKenzie (Dec 07 2020 at 22:47):
But doesn't that put the Parameters resource in the list of resources that get published?
Sarah Gaunt (Dec 07 2020 at 23:28):
Yes. Didn't really realize that was a problem.
Lloyd McKenzie (Dec 07 2020 at 23:31):
It's not something that's relevant to the reader, so we really shouldn't be publishing it.
Mark Kramer (Dec 09 2020 at 11:46):
This does not work for me, either. I'm following @Eric Haas's recipe, and I am positive that the publisher is picking up the exp-params.json file (if I move it, the publishing process fails). The exp-params.json file itself is an exact copy of Eric's. But I'm still getting the message that "The IG does not specify the US edition of SNOMED CT version in the parameters (http://snomed.info/sct/731000124108)".
Grahame Grieve (Dec 09 2020 at 11:47):
what version the IG publisher are you using?
Mark Kramer (Dec 09 2020 at 11:56):
IG Publisher Version: v1.1.48
Grahame Grieve (Dec 09 2020 at 12:03):
thought so. That's like so yesterday...
Grahame Grieve (Dec 09 2020 at 12:03):
actually, I fixed this for Lloyd yesterday
Mark Kramer (Dec 09 2020 at 12:13):
Updated to 1.1.51 but that gave:
onGenerate:
Exception in thread "main" java.lang.NoClassDefFoundError: org/hl7/fhir/igtools/renderers/CodeSystemRenderer
at org.hl7.fhir.igtools.publisher.Publisher.generateOutputsCodeSystem(Publisher.java:7508)
at org.hl7.fhir.igtools.publisher.Publisher.generateResourceHtml(Publisher.java:6686)
at org.hl7.fhir.igtools.publisher.Publisher.generateHtmlOutputs(Publisher.java:6647)
at org.hl7.fhir.igtools.publisher.Publisher.generate(Publisher.java:4901)
at org.hl7.fhir.igtools.publisher.Publisher.createIg(Publisher.java:885)
at org.hl7.fhir.igtools.publisher.Publisher.execute(Publisher.java:729)
at org.hl7.fhir.igtools.publisher.Publisher.main(Publisher.java:8414)
Caused by: java.lang.ClassNotFoundException: org.hl7.fhir.igtools.renderers.CodeSystemRenderer
at java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:581)
at java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:178)
at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:522)
... 7 more
Press any key to continue . . .
Mark Kramer (Dec 09 2020 at 12:13):
Guess I'll try 1.1.50
Grahame Grieve (Dec 09 2020 at 12:13):
wooah. that's serious. @Mark Iantorno can you look into this? that's certainly not a new class
Mark Iantorno (Dec 09 2020 at 12:36):
I'll take a look this morning
Mark Kramer (Dec 09 2020 at 12:51):
1.1.50 doesn't have the above error, so it crept in lately.
Grahame Grieve (Dec 09 2020 at 13:01):
java infrastructure errors like that tend to be picked up pretty quickly and fixed pronto. Mark I, I have nothing to commit - just release a new version when you figure this out
Mark Iantorno (Dec 09 2020 at 14:18):
@Mark Kramer what exact steps are you following to get this error?
Mark Iantorno (Dec 09 2020 at 14:24):
I am running 1.1.51 without that error, but please send me the steps once you are online and I will reproduce and fix
Mark Kramer (Dec 10 2020 at 16:03):
I can't reproduce it.
And with 1.1.50 I can get the US Version of SNOMED recognized.
So all is calm, all is bright. :santa:
Last updated: Apr 12 2022 at 19:14 UTC
