Re: Continued: DMD Source Hot Spot for Expression Type Deduction

[Nordlöw]

I just realized that maybe the best way is to recurse down the 
AST until I find node(s) whose token range cover my query point 
right?