SOFA in C++ and Code Optimization Tips

Hi everyone,

I am currently trying to use SOFA with an external library built in C++ for evolving Artificial Neural Networks. The best solution so far seems to be implementing the SOFA simulation in C++ for flexible integration. I have used Bullet Physics in the past with same approach, and I would like to get similar results with SOFA as my simulation environment this time.

A few notes first:
– I am already familiar with the XML scenes and Python scenes files for the runSofa binary
– I also know about the PythonScriptController for modifying/accessing simulation data from .pyscn files during animation
– I have studied most of the C++ examples at application/tutorials folder
– I am using cmake to compile and link my SOFA/ANN cpp files

The problem is that many SOFA examples that I am interested in (including plugin-based ones) are not provided in C++, so I have been spending some time doing reverse engineering from Python wrapped code to C++.

Questions:
1) What are other available resources that I can use to help with my C++ implementation?
2) Also, I have been using the SOFA base modules in my C++ code. Should I switch to using the SimpleAPI library or SceneCreator plugin instead for code optimization (shortening code length as well)?