storage/sql: Refactor subsource purification (prep for source-fed table purification) [ENG-TASK-19]

[rjobanp]

Motivation

Inspired by & replaces Sean's draft PR #27320 to shuffle when subsource statements are generated.

This PR is prep work for the implementation described in #27907

This PR shouldn't introduce any functional changes -- it is a refactor of subsource purification such that we now generate CreateSubsourceStatements after purification and using the same method in both 'create source' and 'alter source' statements.

The intent is to tease apart the purification logic (obtaining state from external systems) from the statement generation logic, such that I can re-use as much of the subsource purification logic as possible to purify & plan CREATE TABLE .. FROM SOURCE statements in subsequent PRs.

In the spirit of yesterday's cluster-team whiteboard discussions I also cleaned up the naming in much of the purification code.

Tips for reviewer

The first commit is the actual refactor

The 2nd and 3rd commits are a rename of many overloaded structs/variables using the word 'subsource', to clarify whether they were referring to an 'external reference' (e.g. upstream postgres table) or a 'source export' - the object (e.g. either a subsource or table) being exported by a source. This ended up more LOC than expected but clarifying these terms will likely make later PRs easier to grok.

Checklist

• This PR has adequate test coverage / QA involvement has been duly considered. (trigger-ci for additional test/nightly runs)
• This PR has an associated up-to-date design doc, is a design doc (template), or is sufficiently small to not require a design.
• If this PR evolves an existing $T ⇔ Proto$T mapping (possibly in a backwards-incompatible way), then it is tagged with a T-proto label.
• If this PR will require changes to cloud orchestration or tests, there is a companion cloud PR to account for those changes that is tagged with the release-blocker label (example).
• This PR includes the following user-facing behavior changes:

[petrosagg]

Looks good! I think a nice future step would be to replace some of the enums with a generic parameter that describes the source, which will get rid of the unreachable! statements that we must have now.

Also thanks for separating the commit of the renames, made reviewing so much easier!

[notion-workspace]

Refactor subsource purification (prep for source-fed table purification)