FHIR Chat · fixed · IG creation

Stream: IG creation

Topic: fixed


view this post on Zulip David Hay (Jul 13 2020 at 00:36):

I'm getting this error:

Unsupported fixed pattern type for discriminator(system) for slice Location.identifier:HpifacId: org.hl7.fhir.r5.model.UriType

Here's a snippet from the SD: (I've tried both pattern and value discriminators)

{
        "id": "Location.identifier",
        "path": "Location.identifier",
        "slicing": {
          "discriminator": [
            {
              "type": "value",
              "path": "use"
            },
            {
              "type": "value",
              "path": "system"
            }
          ],
          "rules": "openAtEnd"
        }
      },
      {
        "id": "Location.identifier:HpifacId",
        "path": "Location.identifier",
        "sliceName": "HpifacId",
        "short": "The current Facility id for this location",
        "min": 0,
        "max": "1",
        "mustSupport": true
      },
      {
        "id": "Location.identifier:HpifacId.use",
        "path": "Location.identifier.use",
        "min": 1,
        "fixedCode": "official"
      },
      {
        "id": "Location.identifier:HpifacId.system",
        "path": "Location.identifier.system",
        "min": 1,
        "patternUri": "https://standards.digital.health.nz/ns/hpi-facility-id"
      },
      {

view this post on Zulip David Hay (Jul 13 2020 at 00:39):

FWIW: Here's the FSH I used: (and tried with and without the (exactly) keyword

* identifier ^slicing.discriminator.type = #value
* identifier ^slicing.discriminator.path = "use"

* identifier ^slicing.discriminator[1].type = #value
* identifier ^slicing.discriminator[1].path = "system"

* identifier ^slicing.rules = #openAtEnd

* identifier contains
    HpifacId 0..1 MS and
    dormantHpiFacId 0..* MS and
    MOHFacId 0..1 MS

* identifier[HpifacId].system = "https://standards.digital.health.nz/ns/hpi-facility-id"
* identifier[HpifacId].use = #official (exactly)
* identifier[HpifacId] ^short = "The current Facility id for this location"

view this post on Zulip Grahame Grieve (Jul 13 2020 at 00:44):

it's unusual to have a pattern value for a primitive

view this post on Zulip Grahame Grieve (Jul 13 2020 at 00:53):

but, sure, I'll add support for it in the next release

view this post on Zulip David Hay (Jul 13 2020 at 01:04):

Is that not value?

view this post on Zulip David Hay (Jul 13 2020 at 01:05):

ie where's the pattern coming from?

view this post on Zulip Grahame Grieve (Jul 13 2020 at 01:09):

the profile, presumably

view this post on Zulip David Hay (Jul 13 2020 at 01:19):

I did set the discriminator type to 'value' - not 'pattern' - (see snippet above) so a bit confused TBH. Never mind, as long as it's not a show stopper error...

view this post on Zulip Keith Boone (Jul 16 2020 at 00:11):

Did you try: * identifier[HpifacId].system = "https://standards.digital.health.nz/ns/hpi-facility-id" (exactly)

view this post on Zulip David Hay (Jul 16 2020 at 19:06):

pretty sure I did - but will confirm...

view this post on Zulip David Hay (Jul 16 2020 at 19:19):

Actually - that fixed it! Thanks @Keith Boone It actually needed (exactly) on both system and use:

* identifier[HpifacId].system = "https://standards.digital.health.nz/ns/hpi-facility-id" (exactly)
* identifier[HpifacId].use = #official (exactly)
* identifier[HpifacId] ^short = "The current Facility id for this location"

view this post on Zulip Chris Moesel (Jul 16 2020 at 20:11):

Grahame Grieve said:

it's unusual to have a pattern value for a primitive

Actually, I think @Lloyd McKenzie convinced me this was the best approach, otherwise you are denying implementors the opportunity to apply an extension or id to the primitive value.

view this post on Zulip Grahame Grieve (Jul 16 2020 at 23:59):

I think that you generally do want to do that

view this post on Zulip Grahame Grieve (Jul 16 2020 at 23:59):

but not always, I agree

view this post on Zulip Keith Boone (Jul 17 2020 at 12:13):

Good to hear. I was having a VERY similar problem, and thanks to you, was able to track it back to the fixed/pattern discussions eslewhere, so in a way, you helped me too!


Last updated: Apr 12 2022 at 19:14 UTC