Context action service one function housing every action? or one function per action?

Lets say you were to bind 'e', 'f', 'v'

would it be better to have one function, and in that function it checks was 'e', 'f', or 'v' key pressed and if so do a different action

or would it be better to have a unique function for the 'e' key and bind that one the 'f' key and bind that one the 'v' key and bind that one