Optimize array_sort with nested-list-aware sort kernel #21041
Description
Is your feature request related to a problem or challenge?
array_sort currently invokes an Arrow kernel for every row and assembles the results. While #21006 makes this a bit faster, it is still relatively slow.
If we instead wrote a custom sort kernel that operates on the entire ListArray and sorted it directly into a preallocated output buffer, that would likely be much faster.
Idea per @Dandandan in #21006
Describe the solution you'd like
No response
Describe alternatives you've considered
No response
Additional context
No response