feat: Compress HANA expand queries by reducing duplicated statements

[BobdenOs]

Concept

When constructing an expand query it is required to reference the parent query in the child query. This was originally prevented by storing the results into a table value, but when switching away from the DO BEGIN syntax the queries where being duplicated into each child query. This change attempts to remove the query duplication by defining them inside the WITH clause which allows the child queries to directly reference the parent query. This SHOULD allow for the SQL parser to parse less and SHOULD allow for the execution plan optimizer to easier recognize that the query is doing the same thing. The big danger is that the execution plan recognizes that the query is doing the same thing and results in the queries executing in sequence. Which might result in reverting the behavior back to before #342. So the hope is that this makes everything much better and faster. 🤞

[johannes-vogel]

@BobdenOs I've tried to resolve the conflicts but I don't know whether the failing tests are a result of a bad conflict resolution. Could you double check? At least the commit history tells me that this was anyway not yet working?!