Why does expand_column_types not respect on_schema_change config?

[nik-hinge]

Hi!

I'm trying to understand why for the incremental materialization, the adapter calls expand_column_types on every run.
This is something I expected would be influenced by the on_schema_change config, but it is called before the process_schema_changes macro, which is influenced by on_schema_change.

See the code here: https://github.com/dbt-labs/dbt-core/blob/main/core/dbt/include/global_project/macros/materializations/models/incremental/incremental.sql#L42-L44

What is the purpose of the expand_column_types macro, and why do we handle it differently from other types of schema changes?