Reading the wavefile one by one from the database and finding the feature vectors for each wave

Respected sir, Am doing project in Emotion recognition from speech.In my project i extracted the features of energy, mfcc for each frame segmented from each wave. But i did manually.This feature extraction should model it with GMM. Here the problem is i have to read the wave one by one from the database and performing the function of energy extraction and then save the energy feature vectors for each frame in one file.so that i could use all those vectors for to model with GMM. i dont have idea to automate this(ie.reading the speech wave one by one and performing the energy extraction function and collect all feature vectors in one file) with code sir.so help me please..