Text to Speech -- unit selection / diphone concatenation using Festival TTS

I want to build a text to speech system using unit selection or diphone concatenation in festival framework. One of the two methods can be used (unit selection or diphone concatenation). And all the steps taken to build the system should be documented and shared as well.
Attached is the documentation to use for unit selection and below is the link for Festival framework
https://www.cstr.ed.ac.uk/projects/festival/