Stream: IG creation
Topic: Feelings about conversion of YAML files
Eric Haas (Jul 21 2020 at 19:56):
I use a bash script and a python one liner to convert my ig.yml to ig.json and I make a lot of edits in YAML and convert to json in my text editor with a couple of keystrokes. What is the feeling about having the publisher autoconverting YAML resources to JSON on the fly?
Chris Moesel (Jul 21 2020 at 20:11):
I'm sure there must be some really great, easy-to-use Java libraries that could be leveraged...
Grahame Grieve (Jul 21 2020 at 21:04):
there's no ambiguity about the conversion?
Lloyd McKenzie (Jul 22 2020 at 04:44):
I'm not a fan of divergence in authoring. When you say ig.json, are you referring to the ImplementationGuide resource instance? If so, I'm even more opposed. We shouldn't have something in source control for a 'standard' resource that HL7 hasn't defined. If you don't like authoring in JSON, why not use FSH?
Eric Haas (Jul 22 2020 at 16:26):
It is just a convenience...YAML is good for quick and dirty hand edits. like adding an example into an ig resource, (and fsh could probably easily generate in YAML as JSON since technically YAML is a superset of JSON. This means that, in theory at least, a YAML parser can understand JSON, but not necessarily the other way around).
Lloyd McKenzie (Jul 22 2020 at 16:30):
We don't want to introduce more syntaxes than we need to have. Every syntax increases the learning curve. And every variation in how different HL7 guides are maintained reduces the ability to maintain them and increases publishing complexity (and the chances of things going wrong)
Eric Haas (Jul 22 2020 at 16:40):
Its not a learning curve thing , its a choice/preference thing but I am ok either way.
Lloyd McKenzie (Jul 22 2020 at 16:45):
If it's choice, then that's variability. If it's not a choice, then it's learning curve.
Last updated: Apr 12 2022 at 19:14 UTC
