GH-50249: [C++] Add ArrayBuilder::IsValid and ArrayBuilder::IsNull

[andishgar]

Rationale for this change

ArrayBuilder does not provide a public API to check whether an element is null.

What changes are included in this PR?

Add ArrayBuilder::IsNull and ArrayBuilder::IsValid.

Are these changes tested?

Yes, I ran the relevant unit tests.

Are there any user-facing changes?

Yes. This PR adds the public APIs ArrayBuilder::IsNull and ArrayBuilder::IsValid.

• GitHub Issue: [C++] Change BaseBinaryBuilder::GetView to return a view with a null data pointer for entries added via AppendNull #50249

[github-actions]

⚠️ GitHub issue #50249 has been automatically assigned in GitHub to PR creator.

[pitrou]

Thanks @andishgar , just a suggestion to update the docstring

[pitrou]

It's not exactly that it does not work, it's that it doesn't account for logical nulls.

Suggested change

	/// Note that this method does not work for types that do not have a
	/// top-level validity bitmap (Union and Run-End Encoded (RLE) types).
	/// Unlike Array::IsNull, this method does not account for logical nulls
	/// for types that do not have a top-level validity bitmap
	/// (Union and Run-End Encoded types).

[pitrou]

These three should return true perhaps?

[pitrou]

cc @zanmato1984 for opinions

[zanmato1984]

I don't think IsValid has a well-defined logical semantic for Union/RLE builders at the base ArrayBuilder level. Returning false is misleading since it makes every slot look null, while returning true only means there is no top-level null bitmap to invalidate the slot.

So I think we either need to narrow the API semantics/name to physical/top-level validity, e.g. IsPhysicalValid or IsTopLevelValid, or make these builders unsupported for this API instead of pretending to answer logical validity.