GH-48241: [Python] Scalar inferencing doesn't infer UUID

[tadeja]

Rationale for this change

This closes #48241, #44224 and #43855.
Currently uuid.UUID objects are not inferred/converted automatically in PyArrow, requiring users to explicitly specify the type.

What changes are included in this PR?

Adding support for Python's uuid.UUID objects in PyArrow's type inference and conversion.

Are these changes tested?

Yes, added test_uuid_scalar_from_python() and test_uuid_array_from_python() in test_extension.py.

Are there any user-facing changes?

Users can now pass Python uuid.UUID objects directly to PyArrow functions like pa.scalar() and pa.array() without specifying the type;

import uuid
import pyarrow as pa

pa.scalar(uuid.uuid4())

<pyarrow.UuidScalar: UUID('958174b9-3a5c-4cdd-8fc5-d51a2fc55784')>

pa.array([uuid.uuid4()])

<pyarrow.lib.UuidArray object at 0x1217725f0>
[
73611FD81F764A209C8B9CDBADDA1F53
]

• GitHub Issue: [Python] Scalar inferencing doesn't infer UUID #48241

[tadeja]

@AlenkaF Would you recommend any good place to document this UUID change? - I see @amoeba indicated the need for documentation in his draft pull request #44242.
ӇƛƤƤƳ_ƝЄƜ_ƳЄƛƦ:)

[AlenkaF]

Happy New Year! ❤️

I would suggest adding the documentation to the Extending PyArrow page under the Canonical extension types section as a separate subsection next to Fixed size tensor one.

[tadeja]

@AlenkaF, @rok do you have the chance to review this one - should enable multiple UUID use-cases.

[rok]

Looks good to me. Two minor nits.

[rok]

Looks good. Just minor suggestions for comments.

[rok]

@pitrou could you take a look at this PR? Especially cython change could use your expertise.

[tadeja]

@pitrou, any wise thoughts on changes here?

[pitrou]

We're duplicating code here between different module imports. It would be really nice to write something like this:

struct UuidModuleData {
  PyObject* UUID_class = nullptr;
};

UuidModuleData* InitUuidStaticData() {
  static ModuleOnceRunner runner("uuid");
  return runner.Run([&](OwnedRef module) -> UuidModuleData {
    UuidModuleData data;
    OwnedRef ref;
    if (ImportFromModule(module.obj(), "UUID", &ref).ok()) {
      data.UUID_class = ref.obj();
    }
    return data;
  });
}

struct ModuleOnceRunner {
  std::string module_name;
#ifdef Py_GIL_DISABLED
  std::once_flag initialized;
#else
  bool initialized = false;
#endif

  template <typename Func>
  auto Run(Func&& func) -> decltype(func(OwnedRef()) {
    using RetType = decltype(func(OwnedRef());
    RetType ret{};
    auto wrapper_func = [&]() {
      OwnerRef module;
      if (ImportModule("uuid", &module).ok()) {
        ret = func(std::move(module));
      }
    };
#ifdef Py_GIL_DISABLED
    std::call_once(initialized, wrapper_func);
#else
    if (!initialized) {
      initialized = true;
      wrapper_func();
    }
#endif
    return ret;
  };
};

I think @rok can help.

[pitrou]

Does MakeConverter support extension types here? I see that we only unwrap the extension type in the inference path above.