Shuffle Writer code is a huge messy file #3354
Description
What is the problem the feature request solves?
While trying to look into the shuffle code to check some encoding inefficiency I got very frustrated with the needlessly huge file, I think it's high time this was split into smaller files, sorted by categories.
It will make it easier to identify where we can reuse code, and when we can extend existing functionality(e.g., by creating a trait and impling it for the existing structs, similar to what we have with the current ShufflePartitioner)
Describe the potential solution
A simple refactor, mostly moving code around and splitting it into file, I can't point at any required logic changes other than that at the moment
Additional context
No response