FHIR Chat · Language codes · terminology

Stream: terminology

Topic: Language codes

Grahame Grieve (Feb 16 2017 at 21:46):

Proposal to add the following properties for the BCP-47 code system:
- language
- ext-lang
- script
- region
- variant
- extension
- private-use

Grahame Grieve (Feb 16 2017 at 21:46):

all are able to be used for with $lookup, and also for using in value sets filters, for extensional value sets

Grahame Grieve (Feb 16 2017 at 21:47):

this creates the language for existing and proposed US conformance rules, and is probably relevant for other countries too

Rob Hausam (Feb 16 2017 at 23:52):

looks good to me

Grahame Grieve (Feb 28 2017 at 05:19):

ok, so, then, given this valueset:

<?xml version="1.0" encoding="UTF-8"?>
<ValueSet xmlns="http://hl7.org/fhir">
    <id value="simple-languages"/>
    <url value="http://hl7.org/fhir/ValueSet/simple-languages"/>
    <version value="1.9.0"/>
    <name value="All Language codes with language and optionally a region modifier"/>
    <status value="draft"/>
    <experimental value="true"/>
    <publisher value="HL7 International - FHIR-Infrastructure"/>
    <contact>
        <telecom>
            <system value="url"/>
            <value value="http://hl7.org/fhir"/>
        </telecom>
    </contact>
    <description value="This value set includes all possible codes from BCP-47 (http://tools.ietf.org/html/bcp47)"/>
    <compose>
        <include>
            <system value="urn:ietf:bcp:47"/>
            <filter>
              <property value="ext-lang"/>
              <op value="exists"/>
              <value value="false"/>
            </filter>
            <filter>
              <property value="script"/>
              <op value="exists"/>
              <value value="false"/>
            </filter>
            <filter>
              <property value="variant"/>
              <op value="exists"/>
              <value value="false"/>
            </filter>
            <filter>
              <property value="extension"/>
              <op value="exists"/>
              <value value="false"/>
            </filter>
            <filter>
              <property value="private-use"/>
              <op value="exists"/>
              <value value="false"/>
            </filter>
        </include>
    </compose>
</ValueSet>

Grahame Grieve (Feb 28 2017 at 05:21):

then the following is valid:

GET http://fhir3.healthintersections.com.au/open/ValueSet/simple-languages/$validate-code?system=urn:ietf:bcp:47&code=en-AU

Grahame Grieve (Feb 28 2017 at 05:21):

and this is not:

Grahame Grieve (Feb 28 2017 at 05:21):

GET http://fhir3.healthintersections.com.au/open/ValueSet/simple-languages/$validate-code?system=urn:ietf:bcp:47&code=zh-Hans

Grahame Grieve (Feb 28 2017 at 05:21):

Hans is a script, not a region

Grahame Grieve (Feb 28 2017 at 05:21):

those won't work right now, but they will when I next upgrade my server - in the next 24 hours

Grahame Grieve (Feb 28 2017 at 05:22):

response:

<?xml version="1.0" encoding="UTF-8"?>
<Parameters xmlns="http://hl7.org/fhir">
    <parameter>
        <name value="result"/>
        <valueBoolean value="false"/>
    </parameter>
    <parameter>
        <name value="message"/>
        <valueString value="The language code zh-Hans contains a script, and it is not allowed to"/>
    </parameter>
    <parameter>
        <name value="message"/>
        <valueString value="The code provided (urn:ietf:bcp:47#zh-Hans) is not  valid in the value set All Language codes with language and optionally a region modifier"/>
    </parameter>
</Parameters>

Grahame Grieve (Feb 28 2017 at 05:40):

@Robert McClure this is for you

Robert McClure (Mar 02 2017 at 04:13):

@Grahame Grieve Many thanks. I suppose I need to figure out how to test this. Can you notify Brett when it's done? And will there be a "all codes" value set - I assume not?

Eric Haas (Mar 02 2017 at 05:27):

Brett has been notified and it is applied in us core. Still a mystery how this aligns with "use two letters code first".

Grahame Grieve (Mar 02 2017 at 06:50):

that's not stated explicitly, but the terminology server knows it

Grahame Grieve (Mar 09 2017 at 20:25):

@Robert McClure getting back to this:
- no there won't be an 'all codes' value set, but we could do what we do for UCUM and generate a common values expansion - but, in fact, we already have that in that all the common language values are valid in this value set - so US usage restricts the max value set
- you can test this using these URLS:
http://fhir3.healthintersections.com.au/open/ValueSet/simple-languages/$validate-code?system=urn:ietf:bcp:47&code=en-AU
http://fhir3.healthintersections.com.au/open/ValueSet/simple-languages/$validate-code?system=urn:ietf:bcp:47&code=zh-Hans

Robert McClure (Mar 13 2017 at 23:24):

Thanks. Remind me, is "simple-language" the code system name you guys decided for this? I think we need to clarify how these validation operations deliver on a subset of the entire grammar. That is correct right?

And we need to figure out if this represents the same thing the C-CDA referenced content means. I'm assuming it should. Agree @Brett Marquard ?

Grahame Grieve (Mar 14 2017 at 00:51):

'simple-language' is the arbitrary id for the value set on my server.

Grahame Grieve (Mar 14 2017 at 00:51):

you can see the definition of the value set here;

Grahame Grieve (Mar 14 2017 at 00:52):

http://fhir3.healthintersections.com.au/open/ValueSet/simple-languages

Grahame Grieve (Mar 14 2017 at 00:53):

it explicitly defines the subset of BCP 47 that applies. The fact that BCP says 2 letter codes then 3 is implicit in the code system, and therefore not explicit in the value set

Grahame Grieve (Mar 14 2017 at 00:53):

my server knows this fact

Last updated: Apr 12 2022 at 19:14 UTC