Arrow stream writer and reader implementation questions

I was using the python arrow function record_batch and find that this does not exist in julia, but there is an arrow table function. However they don’t behave the same way.

I have a customer requirement (that I can neither explain nor defend) of using arrow to
make a stream writer and reader of io bytes rather than just use julia bytes
This works using python arrows and batch_record
batch = pa.record_batch([data_in],names=[‘f0’])

My question is, is the problem syntax (typing it wrong) or semantics (just not possible to do because functionality was not ported)?

my julia code is as follows:
blksz = (32 * 1024)

A = rand(ComplexF16, blksz)

B = reinterpret(Float16, A)

C = reinterpret(UInt8, B)

tbl = Arrow.Table(C)

→ MethodError: no method matching open(::Base.ReinterpretArray{UInt8, 1, ComplexF16, Vector{ComplexF16}, false}, ::String

Start by comparing Julia's Arrow.Table entry point with the Python record_batch usage shown in the issue, then reproduce the MethodError with the provided reinterpretation example. Determine whether stream readers and writers for IO bytes are supported; done should clarify the supported API or define the missing functionality needed.

Written by the indexing model from the issue text.