Architecture Information?

[matanox]

@yunabe Thanks a lot for this project. I went through the original blog announcement, and was wondering whether any documentation currently discusses the way that this project is made possible, meaning, in terms of how you could take a compiled language, and make it act much like it is interpretable in Jupyter. If I followed correctly, you've not built your own custom interpreter mirroring the language spec / official go compiler, so I assume this project may be utilizing compiler features maybe initially meant for IDE and lately called "language server" development.

Even if I happen to be so far correct, I am very happy to learn a bit more.

I'm happy to use this for data science together with gonum, if I can only understand a bit more about the architecture choices upfront, to avoid implied limitations an so forth ....

Thanks again for this project, and thanks in advance for your comments!!