Reserved words conflict: The jak2stat5 example fails with mrgsolve about heta-compiler HOT 3 OPEN

Hi @martinmodrak ,
Thank you for your feedback. I have the same issue.
The model included a compartment with "default" id.
This is a reserved word in mrgsolve, as I understand.

I've updated the repository with jak2stat5 model changing "default" => "def".
Now everything is fine. Compilation and simulations are OK.
Pull the latest commit from hetalang/qsp-benchmark and recompile the model.

This is a workaround. But the unexpected issues with reserved words are still a problem.
We need a better solution.

from heta-compiler.

Wouldn't a systematic solution be that the XXExport classes expose a list of keywords and the compiler checks the model against this list. It could then either:

• Raise an error/warning when an id conflicts with the keywords, or
• Automatically rename all ids to avoid the conflict (e.g. by adding underscorses or something)

(I don't insist on this being done for the review, it just seems like a relatively cheap but useful solution)

from heta-compiler.

@martinmodrak, agree, this would be good.
As a fast temporal solution, I've added "default" into a general keyword list.
Now the usage of such id throws an error. A user should change the id of such a component by himself.
But I still doubt if it's the best solution.

1. It seems that each "export" type should have its own stop list. We need time to collect and analyze them all.
2. We should also think of automatic renaming rules.

So I guess we suggest a better solution later.

from heta-compiler.