FHIR Chat · Content-Type for Mapping Language files

Stream: implementers

Topic: Content-Type for Mapping Language files

Alexander Zautke (Mar 11 2019 at 11:03):

At Firely, we are currently planning to implement a component which can transform a mapping file containing FHIR mapping language rules to a StructureMap. The initial idea would be to use $convert for this task. Any suggestions on what the Content-Type would look like for the input parameter, i.e. a mapping language file?

Alexander Zautke (Mar 12 2019 at 19:34):

Or would the first sentence under "General Syntax Notes" indicate that the content-type should be "text/plain" and you would need try to parse the content to see if it actually contains mapping language rules?

Grahame Grieve (Mar 14 2019 at 03:44):

interesting question. I've not considered about a mime type for that language. I'd guess at something like text/x-fhir-mapping

Lloyd McKenzie (Mar 14 2019 at 04:40):

Not sure it meets the requirements of "text"...

Grahame Grieve (Mar 14 2019 at 04:59):

it certainly does. It's UTF-8 text

Lloyd McKenzie (Mar 14 2019 at 05:08):

I thought text had to be something intended to be read by a human

Grahame Grieve (Mar 14 2019 at 05:39):

that's exactly what it is

Lloyd McKenzie (Mar 14 2019 at 05:39):

You have a different idea of 'human' than I do I guess...

Grahame Grieve (Mar 14 2019 at 05:40):

we're talking about this format:

Grahame Grieve (Mar 14 2019 at 05:41):

http://hl7.org/fhir/account-version-maps.html

Alexander Zautke (Mar 18 2019 at 07:28):

In RFC 2046, defining the characteristics of a text/* subtype, there is no requirement that the format needs to be human readable. It only needs to be "principally textual in form". If we have a look at all registered IANA media types, there is no other format having a "*-" prefix. Because of this, I would prefer "text/fhir-mapping".