Pyprocessing is a (partial) port of processing to python. You can and probably should use a python ide to develop your code, I've been quite impressed with  Eric4, which supports code completion, debugging and much more. 

Whilst I'm not a huge python fanboy (I prefer ruby) there is a lot to be said for this processing implementation. It is based on pyglet a python games engine (which in turn uses ctypes) so the performance is pretty impressive (uses opengl as standard), and as far as I can tell it matches or even exceeds the the current performance of opengl2 in

vanilla processing. Don't take my word for it download pyprocessing here, check out my experiments with pyprocessing here, also my blog here. A vanilla processing sketch "Rod Hilbert" that I posted on open processing and my blog provides an interesting benchmark, the equivalent pyprocessing sketch (and module) meets or exceeds the vanilla processing version which is much optimised. The alternative python implementation processing.py unfortunately lags behind in the performance stakes, although it can access the java libraries (unlike pyprocessing), however there are a lot of python libraries already out there for math/science etc so I think there exists a huge opportunity for pyprocessing.