I very much doubt there would be any libraries specifically for converting voice to cat.
You are probably best off experimenting with audio filters and pitch shifters in an audio program like Fruitloop to discover what effects you need to apply in order to transform man into cat, research how those filters can be implemented digitally and try to write youwn version of them.
You'll need to know quite a bit about signal processing to do that, which, in turn, needs a lot of math.
The solution would probably be something like either filtering out low frequencies, or shifting low frequencies to higher frequencies. I doubt that would be enough, though, you'll probably need something else to make it all the way.