Add a function to get memory size of array slice

[askoa]

Which issue does this PR close?

Closes #3407

Rationale for this change

See issue description.

What changes are included in this PR?

Function to calculate buffer size of array slice.

Are there any user-facing changes?

User will get a new API that'll calculate the buffer size of array slice.

[tustvold]

It is worth noting that this may overestimate for sliced ListArray and StructArray, but I'm not sure if we are being smart when writing those yet. Either way I think this is a good start

@viirya may know??

[viirya]

For StructArray, its layout is empty so this basically computes on all its children. Looks like when slicing StructArray, we already did correctly for its child data? Then seems okay.

I'm not sure if there is an issue on ListArray. But seems it is same as other types where have one offset buffer + child data?

[tustvold]

The way we propagate the offsets into children is not guaranteed, and is actually an incorrect special case I wish to remove, but yes ignoring this edge case, the logic here is fine.

As for lists, theoretically the IPC writer could find the min offset in the list slice, and then shift everything down - as is done for StringArray, etc... Not sure if we do this yet...

[ursabot]

Benchmark runs are scheduled for baseline = b4abb75 and contender = a8276c0. a8276c0 is a master commit associated with this PR. Results will be available as each benchmark for each run completes.
Conbench compare runs links:
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ec2-t3-xlarge-us-east-2] ec2-t3-xlarge-us-east-2
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on test-mac-arm] test-mac-arm
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-i9-9960x] ursa-i9-9960x
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-thinkcentre-m75q] ursa-thinkcentre-m75q
Buildkite builds:
Supported benchmarks:
ec2-t3-xlarge-us-east-2: Supported benchmark langs: Python, R. Runs only benchmarks with cloud = True
test-mac-arm: Supported benchmark langs: C++, Python, R
ursa-i9-9960x: Supported benchmark langs: Python, R, JavaScript
ursa-thinkcentre-m75q: Supported benchmark langs: C++, Java

[viirya]

As for lists, theoretically the IPC writer could find the min offset in the list slice, and then shift everything down - as is done for StringArray, etc... Not sure if we do this yet...

Oh, I see. I remember I did something on buffer slice in IPC writer. I may need to re-look it to see if we do it for list slice.

[alamb]

Wow -- thank you @askoa . This is great