Stream: terminology
Topic: ClaML to fhir
Jim Steel (Mar 17 2020 at 01:43):
If anyone works with terminologies that are distributed as ClaML files, we have a fairly simple program for converting ClaML XML files into FHIR CodeSystems. Its available open-source at https://github.com/aehrc/fhir-claml. Feedback very welcome.
Julian Sass (Mar 19 2020 at 13:50):
Hi @Jim Steel , thanks for posting this! When I run it, I get illegal reflective access warning as it fails to read the external DTD. I tried setting the dependencies to their newest version as suggested on stackoverflow, but this didn't help. Any ideas? Thank you!
2020-03-19-14_41_14-fhir-WSL-Visual-Studio-Code.png
Jim Steel (Mar 19 2020 at 22:13):
Ah yes, that's a current bug! As a workaround, you can delete the DTD line in the ClaML file.
Jim Steel (Mar 20 2020 at 00:21):
Fixed, hopefully (i.e. it should hopefully now just disregard the DTD reference, meaning you don't need to modify the file)
Julian Sass (Mar 20 2020 at 08:43):
Yes, thank you. It's working now :slight_smile:
Patrick Werner (Mar 21 2020 at 18:10):
Hi @Jim Steel thanks for your great tool.
What do you think about adding the functionality of having a defined optional fallback for the -d option? icd10-gm contains for some concepts preferredLong rubrics. It would be nice to create the CS always using the preferredLong rubrics, but defaulting to preferred if there is none. I can make a PR probably next week if you like the idea.
Jim Steel (Mar 22 2020 at 01:03):
Happy to
Patrick Werner (Mar 23 2020 at 14:11):
First (very minor) PR which auto-creates output folders if non-existing instead of throwing an Exception
https://github.com/aehrc/fhir-claml/pull/1
Patrick Werner (Mar 23 2020 at 23:09):
and the second PR @Jim Steel : https://github.com/aehrc/fhir-claml/pull/2 including the fallBack options for -d if -d was not found.
Jim Steel (Mar 24 2020 at 00:29):
Good idea. Another option would be to have a list of rubrics, and the tool takes the first one that's populated as the display, and all of the others are designations?
Jim Steel (Mar 24 2020 at 00:30):
So we change -d to be a list instead of adding -dFallback
Jim Steel (Mar 24 2020 at 00:30):
What do you think?
Patrick Werner (Mar 24 2020 at 07:10):
even better, and more intuitive to use.
Jim Steel (Mar 24 2020 at 07:14):
OK, I'll do that
Patrick Werner (Mar 24 2020 at 07:24):
is was so focused on getting a fallback option, that i didn't realize how much more elegant a ordered -d list would be
Jim Steel (Mar 24 2020 at 07:43):
All good. I'm most of the way done. I'll get it pushed either tonight or tomorrow morning (AU time)
Jim Steel (Mar 25 2020 at 00:23):
OK, done. However, although its implemented now, I really think 'preferred' is a better rubric for display than 'preferredLong' (which is better as a definition and/or designation)
Patrick Werner (Mar 25 2020 at 20:42):
Hmm. The Problem with preferred is that the display isnt always explaining the concept completely. Then you have to lookup the parents and have a logic to build a display value for a GUI. PreferredLong adds this context for these concepts, and is therefore better for GUI consumption.
Last updated: Apr 12 2022 at 19:14 UTC
