> > Strange thing is, most functions are properly demangled but 2 > > aren't. > > Is this a (known) bug? > > Yes, core.demangle can't do some symbols because DMD applies > a one-way hash to them once they reach a certain length because > such long symbols tend to break linkers.
Ah I see, but what about the short one: _D4arsd3web3runFC4arsd3cgi3CgiPS4arsd3web14ReflectionInfoZv