Thinking about why the PRQL syntax is as it is, does it have something to do with chaining CTEs together? I roughly see how given a pipeline with discrete compo

Thinking about why the PRQL syntax is as it is, does it have something to do with chaining CTEs together? I roughly see how given a pipeline with discrete components you could translate it into a collection of CTEs which lead into each other.