Stream: IG creation
Topic: language valueset expansion error
Eric Haas (Feb 24 2021 at 02:35):
In US Core this error has been around a long time and has yet to be resolved:
http://hl7.org/fhir/us/core/2021JAN/ValueSet-simple-language.html
valueset-simple-language.xml "Error from server: Language valuesets cannot be expanded as they are based on a grammar"
Is there any roadmap to fix or should we redefine the CLD or capture the Value set from elsewhere (@Gay Dolin suggested using http://www.loc.gov/standards/iso639-2/php/code_list.php to the value set description page http://hl7.org/fhir/us/core/2021JAN/ValueSet-simple-language.html ?
@Grahame Grieve ? @Robert McClure ?
Grahame Grieve (Feb 24 2021 at 03:22):
do you need to resolve it?
Eric Haas (Feb 24 2021 at 16:35):
Or would be nice for the reader to have a peek. I could just go with gay suggestion and show a bunch of codes. But then that would risk it becoming the defacing standard. Monkey see monkey do
Eric Haas (Feb 24 2021 at 16:35):
Gay’s Suggestion
Eric Haas (Feb 24 2021 at 16:35):
Amy advice is welcome.
Grahame Grieve (Feb 24 2021 at 20:10):
well, the value set is not a fixed list of codes. Should it actually be a fixed list?
Eric Haas (Feb 24 2021 at 20:36):
no not technically but i think it would be handy to have a starter set. I think ill add an intro that lists some as an example and make clear is not only representative and incomplete
Eric Haas (Feb 24 2021 at 20:37):
We get ballot comments re this every time.
Grahame Grieve (Feb 24 2021 at 21:59):
ok. so don't change the binding. In the medium term, this will be resolved by my next round of binding work, but that's probably a few months away
Robert McClure (Feb 25 2021 at 03:43):
@Grahame Grieve What we need to do I think is allow that the bcp47 code system is this set of atoms and the initial include is all atoms with type="language" or type="region" so we have the basic allowed atoms.
Then what I would like to do is add into the include for this value set a new value set that is an enumerated set of defined known expressions that meet the requirement noted. Lloyd's initial list based on members is a meager first step and we need to put in a lot more. @Gay Dolin @Eric Haas I can help work on this if you want. If GG can render the code system as I've suggested (is it already?) then perhaps we can make a value set that works now.
Grahame Grieve (Feb 25 2021 at 03:51):
that aligns with my intent for the next round of binding work, but, as I said, it's few months away
Eric Haas (Feb 25 2021 at 15:26):
@Lloyd McKenzie can you share your "meager first step" (see above)?
Lloyd McKenzie (Feb 25 2021 at 15:37):
I don't think it's Lloyd's. I think Rob's referring to the Common Languages value set, which is driven by the languages typically used by HL7 affiliates. Grahame created it.
Eric Haas (Jun 08 2021 at 23:53):
@Grahame Grieve where are we at with this one.
Grahame Grieve (Jun 14 2021 at 07:06):
no where. I need to do more work on binding
Eric Haas (Nov 16 2021 at 13:48):
@Grahame Grieve , I am circling back on this error for US Core which was granted a variance in June:
input/resources/valueset-simple-language.xml
Path Severity Message
ValueSet.where(id = 'simple-language') error Error from server: Language valuesets cannot be expanded as they are based on a grammar
Grahame Grieve (Nov 18 2021 at 01:11):
well, I admit that I do not have time for this yet but it does keep coming up. So as soon as the ballot crunch eases.
Last updated: Apr 12 2022 at 19:14 UTC
