Can you share the technology & methods sourcegraph use for syntax/semantic analysis?
If there is a brief comparison of tools you have tried and abandoned it would be really really helpful.
The accuracy of whatever working behind the scene is simply amazing.