speech.py

class Voice:
  
    def __init__(self, **config_options):
        self.config(config_options)
        
    def say(phonemes):
        # This returns a generator (or C implemented iterator) of audio frames
        ...
            
    def sing(phonemes):
        # This returns a generator (or C implemented iterator) of audio frames
        ...
        
#Don't know what the best name for this is.
def recite(text)->phonemes:
    ...
    
def say(phonemes, **config_options):
    Voice(**config_options).say(phonemes)
    
def sing(phonemes, **config_options):
    Voice(**config_options).sing(phoneme)
  
def read(text, **config_options):
    Voice(**config_options).say(recite(text))