[FEA] support collect aggregations in reduction

[sperlingxx]

Is your feature request related to a problem? Please describe.
Currently, cuDF supports collect aggregations in rolling windows and groupBy context (though the support is not complete). But, collect aggregations in reduction context is still missing.

Additional context
I believe we can not support this feature until #5887 got solved, since scalar of ListType is essential in reduction to collect.

[jrhemstad]

A collect reduction on a whole column doesn't make much sense to me. Turn an entire column into a list column with a single list?

[sperlingxx]

@jrhemstad Yes, it basically turns an entire column into a list column with a single list. We need this feature because we want to provide GPU support of spark built-in function collect_list(collect_set) with reduction. Here is the example:

> SELECT collect_list(col) FROM VALUES (1), (2), (1) AS tab(col);
 [1,2,1]
> SELECT collect_set(col) FROM VALUES (1), (2), (1) AS tab(col);
 [1,2]

So, we need a method to produce scalars (in ListType) from corresponding input columns. It matches to the semantic of reduce. Or is there alternative approach to achieve this? Perhaps we can achieve the goal more directly with a method like make_list_scalar_from_column. And we also need to take care of null values according to null policy.

[github-actions]

This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.

[github-actions]

This issue has been labeled inactive-90d due to no recent activity in the past 90 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.