Possible problem in interpolate_angle3

[Ameir Al-Zoubi]

I am working on a Swift library that implements the algorithms in Jean Meeus' Astronomical Algorithms. I ran into the issue of interpolating right ascension values as they crossed the angle boundary of 360. Your implementation of interpolate_angle3 helped me to see how a solution can be implemented, I think though that you may want to look at that again. You use diff_angle for determining the difference of a and b in the value c, but c is meant to be the difference in the two intervals, it is not meant to be an angle itself. Also you should be able to just re-use the modpi2 function on the result of y1 - y0 and y2 - y1 instead of needing the diff_angle function.