[C++] Improve FlattenRecursively by making it materialize fewer intermediate array values

[ZhangHuiGui]

Describe the enhancement requested

Improve FlattenRecursively by making it materialize fewer intermediate array values

This function is used to flatten a multi-level nested logical list Arrays.

There are multiple ways to batch the copying of data of deeply-nested lists without allocating a new Array per level.

For instance, if you are flattening a fixed_size_list<fixed_size_list<int32>, 2>, 3> without bitmaps on the internal arrays of values, you can gather blocks containing 2 * 3 32-bit integers from the innermost values array to a pre-allocated destination ArrayData.

Comes from : #41092 (comment)

Component(s)

C++

[ZhangHuiGui]

In this task, we can also add a micro-benchmark to ensure that our subsequent optimization is effective.

[github-actions]

This issue has been marked as stale because it has had no activity in the past 365 days. Please remove the stale label or comment below, or this issue will be closed in 14 days. If this improvement is still desired but has no current owner, please add the 'Status: needs champion' label.