Simon Schultze-Florey
Simon Schultze-Florey
Frankly said: no idea (yet). But from a bit of digging around it doesn't look like Xtext has some support implemented for this protocol.
This however is not really a dependency on "the system", but in this case clearly on the availability of `DefaultScriptScopeProvider`. I know, code-wise the `ScriptFileWatcher` couldn't care less about them,...
Wow, that's what I'd call a detailed problem description. Thanks for that effort! I'll have a look and try reproducing it at home with a Windows PC.
I remember having seen the same happening in good, old designer (R.I.P.) too. So it indeed might not be related to the language server but to the actual validation... IIRC...
> while there might be good reasons for sticking to GSON, IMHO we should at least make them transparent With the findings above, I happily consider this to be done....
The org.glassfish.json JAR is titled with "JSR 374:Java API for Processing JSON RI". Which in turn only _is_ JSON-P. Sounds like some legacy. Therefore I'd second the assumption that this...
> One can take action in ThingHandler::dispose() No, bindings _must_ do that! When bundles are stopped, their `ThingHandlerRegistry` gets removed. The framework detects that and disposes all its handlers. To...
For the sake of consistency I'd suggest doing it in a different way though: Creating EMF model instances (i.e. these EObject derrived classes we have in the model bundles) created...
We already have it for [Items](https://github.com/eclipse/smarthome/blob/36c9e2717e4f89cf60a0d885e7ecc6069b338f48/bundles/model/org.eclipse.smarthome.model.item/src/org/eclipse/smarthome/model/formatting/ItemsFormatter.xtend). For [Things](https://github.com/eclipse/smarthome/blob/36c9e2717e4f89cf60a0d885e7ecc6069b338f48/bundles/model/org.eclipse.smarthome.model.thing/src/org/eclipse/smarthome/model/thing/formatting/ThingFormatter.xtend) it looks a bit - errr... empty. But it doesn't look too complicated. I agree, Xtext is a heavy tool _just_ for...
> Who tells you that no one might come up with an alternative lightweight parser for those files? Can it do the pretty-printing too? 😉 Okayokay, point taken.