raise NumberParseException on invalid number with unicode

[houqp]

HI,

We just got an UnicodeDecodeError for an invalid user input: '\xe2\x80\x8b1-123-456-789'. I suggest we catch that in the library and throw out a NumberParseException.

[daviddrysdale]

That input is the byte string for the UTF-8 encoding of u'\u200b1-123-456-789' (where \u200b is ZERO WIDTH SPACE), but the library is currently expecting (on Python 2.x) to receive a valid string -- either a unicode string or a plain str string with ASCII contents.

I'll think about whether it makes sense to unilaterally assume any str input is actually UTF-8, but in the meanwhile you can probably work around the problem by ensuring you feed in unicode objects.

[houqp]

Good call, I will decode all the input before feeding to the library. thank you :)