`PyArray` ignores `suboffsets`

[nhz2]

Affects: PythonCall

Describe the bug
When constructing a PyArray the suboffsets are ignored. In almost all real world use cases I could find suboffsets is NULL and should be ignored, but in theory it should be checked and an error should be thrown saying it isn't supported. suboffsets allow having an array stored in PIL-style, but as the example below shows, if treated like a regular strided array we get garbage.

MWE generated by Claude Code:

using PythonCall

tb = pyimport("_testbuffer")
nd = tb.ndarray(
    pylist([10, 11, 12, 20, 21, 22]),
    shape=pylist([2, 3]),
    format="i",
    flags=tb.ND_PIL | tb.ND_WRITABLE,
)

mv = pybuiltins.memoryview(nd)
println("Python sees:")
println("  shape      = ", mv.shape)
println("  strides    = ", mv.strides)
println("  suboffsets = ", mv.suboffsets)
println("  tolist     = ", mv.tolist())

println()
println("PyArray sees:")
a = PyArray(nd; array=false, buffer=true)
@show typeof(a)
@show size(a)
@show a.strides
@show collect(a)

Python sees:
  shape      = (2, 3)
  strides    = (8, 4)
  suboffsets = (0, -1)
  tolist     = [[10, 11, 12], [20, 21, 22]]

PyArray sees:
typeof(a) = PyArray{Int32, 2, true, false, Int32}
size(a) = (2, 3)
a.strides = (8, 4)
collect(a) = Int32[1081565312 31157 1081565324; 1081565324 31157 10]

Your system
Julia Version 1.12.6
Commit 15346901f00 (2026-04-09 19:20 UTC)
Build Info:
Official https://julialang.org release
Platform Info:
OS: Linux (x86_64-linux-gnu)
CPU: 16 × Intel(R) Core(TM) i7-10700KF CPU @ 3.80GHz
WORD_SIZE: 64
LLVM: libLLVM-18.1.7 (ORCJIT, skylake)
GC: Built with stock GC
Threads: 1 default, 1 interactive, 1 GC (on 16 virtual cores)

Python Version 3.14.4
PythonCall Version: current main