Closed vinipsmaker2 closed 7 years ago
You are right. I have removed the pre-condition in e1d8cef3dfb8da206a6d4b310bb5953567aaf14a, and added some test cases instead.
Good idea about the advertising. I will add that to the readme.md file when I have added a library synopsis there.
Thanks. I predict I heavier usage of the parser (the dynamic::variable
) on our code base once we start to rewrite one other of our projects in C++ too. If we use more features of the parser there, I may come back with more feedback.
Thanks for the excellent library so far.
I mean,
reader::value<T>()
will already do aswitch
onsymbol
to possibly throw a runtime error. This precondition makes the API easier to use wrong and I fail to see a benefit on this.Btw, we're using trial.protocol "heavily" (all JSON messages) on our new project at BlinkTrade, so if you need to advertise "is there a real usage case?", you can mention it (only boring thing til now is the increased number of
assert(false)
and code complexity thanks to #3).