Igor Dejanović

We have module/class level mutable objects in multiple places which should be converted to thread local storage for multi-threading to work correctly. The aim of this project was never focused...

> Can you explain in a little bit more detail? Scheduling 100 threads or processes and doing parsing with textx seems safe enough to be included in a test suite....

> It might be practical to exclude these tests from the main test suite for now until there is a good confidence in the overall thread safety (to avoid fighting...

> Maybe it is useful to provide my use case for the parallelization of textX. @stanislaw BTW, very interesting use case. Seems that your project will be a great test-bed...

> Doing a quick scan over the shared variables, I have tried to un-globalize some of them starting from `_tx_obj_attrs`. It is indeed a non-trivial effort and requires doing them...

That rule would not be used during parsing so you actually have to consume non-interesting stuff until you find something of interest, then you try to parse it and repeat...

Usually when dealing with your AST/model you need to know structure and semantics to do anything useful. But, as you say, there are times when some generic processing is needed...

> ( I need some time to read ) Sure. This is an important stuff so we should take our time to think thoroughly before jumping into implementation.

@goto40 Thanks a lot for a detailed analysis. Sorry that I'm not so agile in response these days, been quite busy. I've started to work out the details of the...

I have put together an [initial version of TEP](https://github.com/igordejanovic/textX/wiki/Reference-resolving-expression-language-(RREL)) (textX ehnacement proposal :) ) for this in the form of Wiki page as I think that issue is more for...