Scientific float numbers are not supported correctly

[christophepersoz]

Hi there,

Thanks for this fast and simple use library. I noticed that scientific notation for float values are not well supported by the library.

For example if you got the following JSON value {"val":-2E-1} the parse of the value will return a float value of -2.0 instead of -0.2... which can lead to some problems.

I will try to go a bit deeper on how to get a good extraction of exponent part of a float number. In the meantime if you have a good idea to solve this, I'm in!

Thanks a lot!
Christophe