Cleanup aggregation planning

[2010YOUY01]

Is your feature request related to a problem or challenge?

part of #22710

Now the aggregation planning is implemented with a list of rules:

• datafusion/datafusion/physical-plan/src/aggregates/mod.rs

  Line 1189 in bbb552f

  if self.should_use_ordered_partial_aggregate_stream(context) {

After the refactor in the EPIC is done, we could probably clean them up into a simpler form.

A related idea is to add a better top-level specification for aggregation planning, to better describe this code. Currently, the comment at the top of the file explains the planning logic, but only at the level of intuition and high-level ideas.

Describe the solution you'd like

No response

Describe alternatives you've considered

No response

Additional context

No response

[Pranav2612000]

Hey @2010YOUY01 ! Is this safe to pick up now? I'm looking to start contributing to datafusion and think this could be a good issue to start with. Planning to do a change on the lines of

        if context
            .session_config()
            .options()
            .execution
            .enable_migration_aggregate
        {
            let mode = self.mode;
            let input_order_mode = &self.input_order_mode;
            let is_true_no_grouping = self.group_by.is_true_no_grouping();
            let limit_options = self.limit_options_supported_by_hash_stream();

            match (mode, input_order_mode, is_true_no_grouping, limit_options) {
                (AggregateMode::Partial, InputOrderMode::Linear, false, true) => {
                    return Ok(StreamType::PartialHash(PartialHashAggregateStream::new(
                        self, context, partition,
                    )?));
                }
                /// arms for other stream types
                _ => {}
            }

        }

Let me know if this is what you had in mind and I can submit a PR.