Because of HiGHmed there are a lot of new templates out there in German, which includes all sorts of strange characters in template ids and terminology ids. While doing some tests for EHRBASE, I have found this template Clinical Knowledge Manager which uses “§21 KHEntgG” as a terminology_id for cod…

Because of this: terminology_id = name-str, [ '(', name-str, ')' ] ; name-str = letter, { letter | digit | '_' | '-' | '/' | '+' } ; Terminology IDs mentioned in the same spec are invalid: ICD9(1999) ICD10AM(3rd_ed) ICD10AM(4th_ed) Note name-str requires the first character to be a letter, …

This might be related: So, in the Archetype designer , you are allowed to set a “query” URI as the terminology for coded text elements: [image] @Athul_K_Nair figured that the templates don’t get saved with some query strings, however, some others work: Doesn’t work SNOMED-CT::<123456 SNOMED-CT:…

I don’t see an issue with the uri, but in the terminology_id field value.

[image] pablo: §21 KHEntgG This isn’t a valid terminology id in openEHR. It might be the human name of some real terminology, but an id will need to be constructed for it for computational use, just like the string ‘snomed_ct’ of ‘snomed-ct’ etc is used to mean Snomed, even though Snomed’s p…

[image] pablo: §21 KHEntgG That’s actually a paragraph from German hospital financing law. A bit of a stretch to call it a terminology. This is from the oet: <includedValues>§21 KHEntgG::01::Krankenhausbehandlung, vollstationär</includedValues> <includedV…

[image] sebastian.garde: That’s actually a paragraph from German hospital financing law. A bit of a stretch to call it a terminology. :rofl: :rofl: :rofl:

[image] thomas.beale: an id will need to be constructed for it for computational use, just like the string ‘snomed_ct’ of ‘snomed-ct’ etc is used to mean Snomed, even though Snomed’s proper name is ‘SNOMED CT’ (containing a space). Thanks @thomas.beale did you see the current grammar for TERM…

The (Antlr) grammar rules we are actually using are: // e.g. [ICD10AM(1998)::F23]; [ISO_639-1::en] TERM_CODE_REF : '[' TERM_CODE ']' ; TERM_CODE : TERM_CODE_CHAR+ ( '(' TERM_CODE_CHAR+ ')' )? '::' TERM_CODE_CHAR+ ('|' ~[|\]]+ '|')?; fragment TERM_CODE_CHAR: NAME_CHAR | '.' ; fragment NAME_CHAR …

As I mentioned, this is in the spec: Base Types Check TERMINOLOGY_ID currently doesn’t accept “ICD10AM(1998)” Not sure what you mean with the antlr grammar, we are referring to different things. (* ------------------------- UID, OID, UUID -------------------------- *) uid = iso_oid | uuid | i…

[image] pablo: Check TERMINOLOGY_ID currently doesn’t accept “ICD10AM(1998)” Ah sorry, I see - because the ‘1998’ is not matched by ‘name-str’, which has to start with a letter. That’s an error :wink: So we need to fix that. I believe our aim should be to replace these grammars in the spec …

(in)Valid TERMINOLOGY_IDs in templates

Specifications

thomas.beale (Thomas Beale) 20 September 2021 21:03 15

We’re thinking about this

Topic		Replies	Views
SNOMEDCT - correct representation Implementers (archive)	49	79	20 May 2017
constraint binding error Technical (archive)	35	47	24 February 2011
SNOMED-CT postcoordination support in ADL Implementers (archive)	11	66	28 April 2015
Archetype IDs starting with a numeric? Specifications	11	776	3 March 2021
Storing codes of two external terminologies Terminology	37	1833	19 July 2024
one-to-many term bindings in archetypes Clinical (archive)	12	20	3 April 2015
Terminology_id name list Specifications	52	416	18 November 2025
Inconsistency in REST API Create EHR implementations -needs spec clarification? ITS	27	888	9 October 2024
Illegal character in authority at index 14... Archetype Designer	13	228	4 June 2024
Is unbinded constraint code valid in archetypes / templates? ADL archetype , template	4	485	26 October 2021

(in)Valid TERMINOLOGY_IDs in templates

Related topics