Add parallelization feature

[christophfroehlich]

The dispersion analysis, basically a monte-carlo simulation, is predestinated for parallelization.

How could one implement this within this python/java framework?
Maybe https://docs.python.org/3/library/multiprocessing.html would do the job.

[christophfroehlich]

I played a little bit with Multiprocessing and concurrent.futures.

I got to the point

TypeError: cannot pickle '_jpype._JPackage' object

which means that I cannot run more than one simulation with the same orh instance. There is a pickle support of jpype, but I've no glue how to start with it.. jpype-project/jpype#533

Trying to create more than one instance of orhelper.OpenRocketInstance() gives

File "....\ortools\lib\site-packages\orhelper_orhelper.py", line 50, in enter
jpype.startJVM(jvm_path, "-ea", f"-Djava.class.path={self.jar_path}")
File "....\ortools\lib\site-packages\jpype_core.py", line 177, in startJVM
raise OSError('JVM cannot be restarted')

It seems to be impossible to launch the JVM (i.e., orhelper.OpenRocketInstance()) more than once within the same python script: jpype-project/jpype#959

@PatrickKa Any ideas how to overcome this?

Maybe we can duplicate the jar file and load two orhelper for different files. But for creating more than one OR-jar file we would have to adapt orhelper, because the jar file is hardcoded:
https://github.com/SilentSys/orhelper/blob/bc240170b8fde9077c7e4ab6cd2aa01ace142a36/orhelper/_orhelper.py#L14