Remove the old parser's AST from StatementRewrite

[sgrif]

It is no longer used, so we can get rid of it. I opted to change the signature of maybe_rewrite to take a mutable reference to the statement rather than copy it and return a new one. This resulted in a significant indentation change, but no actual code changes (beyond replacing a few as_ref and as_muts with stmt and stmt_mut.

[levkk]

This is because we couldn't mutate ASTs before, right? Nice. Bye bye pg_query, say hello to sage_query. No qualms with you naming the library after yourself.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[sgrif]

This is because we couldn't mutate ASTs before, right?

No, we could. It's purely a question of where make::owned is called. The reason I opted to move it up a level instead of returning the copied/mutated node from this function is multi-statement ASTs. This function only operates on a single node, but we still need to keep the unmanipulated nodes around, meaning it's not just a question of creating a new single statement list of the transformed node. We'd need to assign it to index 0 in the old list, which requires a mutable reference to it, which requires a make::owned block at that level.