Apply suggestions from code review

Co-authored-by: Hugo van Kemenade <1324225+hugovk@users.noreply.github.com>

.github/workflows/test-windows.yml

@@ -35,17 +35,11 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: [ "3.10", "3.11", "3.12", "3.13" ]
+        python-version: ["pypy3.11", "pypy3.10", "3.10", "3.11", "3.12", "3.13", "3.14"]
         architecture: ["x64"]
         os: ["windows-latest"]
-        pyarrow: ["true"]
         include:
             # Test the oldest Python on 32-bit
-            - { python-version: "3.9", architecture: "x86", os: "windows-2019", pyarrow: "false" }
-            # test the non-pyarrow capable ones
-            - { python-version: "3.14", architecture: "x64", os: "windows-latest", pyarrow: "false" }
-            - { python-version: "pypy3.11", architecture: "x64", os: "windows-latest", pyarrow: "false" }
-            - { python-version: "pypy3.10", architecture: "x64", os: "windows-latest", pyarrow: "false" }
     timeout-minutes: 30
 
     name: Python ${{ matrix.python-version }} (${{ matrix.architecture }})
@@ -92,10 +86,9 @@ jobs:
       run: |
         python3 -m pip install PyQt6
 
-    - name: Install PyArrow Test Dependency
+    - name: Install PyArrow dependency
-      if: "matrix.pyarrow == 'true'"
       run: |
-        python3 -m pip install --only-binary=:all: pyarrow
+        python3 -m pip install --only-binary=:all: pyarrow || true
 
     - name: Install dependencies
       id: install

Tests/test_arrow.py

@@ -4,9 +4,7 @@ import pytest
 
 from PIL import Image
 
-from .helper import (
+from .helper import hopper
-    hopper,
-)
 
 
 @pytest.mark.parametrize(

docs/reference/arrow_support.rst

@@ -4,71 +4,72 @@
 Arrow Support
 =============
 
-Arrow is an in memory data exchange format that is the spritual
+`Arrow <https://arrow.apache.org/>`__
-successor to the numpy array interface. It provides for zero copy
+is an in-memory data exchange format that is the spiritual
-access to columnar data, which in our case is Image data.
+successor to the NumPy array interface. It provides for zero-copy
+access to columnar data, which in our case is ``Image`` data.
 
-The goal with Arrow is to provide native zero-copy interop with any
+The goal with Arrow is to provide native zero-copy interoperability
-arrow provider or consumer in the Python ecosystem.
+with any Arrow provider or consumer in the Python ecosystem.
 
-.. warning:: Zero-copy does not mean zero allocation -- The internal
+.. warning:: Zero-copy does not mean zero allocation -- the internal
   memory layout of Pillow images contains an allocation for row
   pointers, so there is a non-zero, but significantly smaller than a
-  full copy memory cost to reading an arrow image.
+  full-copy memory cost to reading an Arrow image.
 
 
 Data Formats
 ============
 
-Pillow currently supports exporting arrow images in all modes
+Pillow currently supports exporting Arrow images in all modes
 **except** for ``BGR;15``, ``BGR;16`` and ``BGR;24``. This is due to
-line length packing in these modes making for non-continuous memory.
+line-length packing in these modes making for non-continuous memory.
 
-For single band images, the exported array is width*height elements,
+For single-band images, the exported array is width*height elements,
-with each pixel corresponding to the appropriate arrow type.
+with each pixel corresponding to the appropriate Arrow type.
 
-For multiband images, the exported array is width*height fixed length
+For multiband images, the exported array is width*height fixed-length
-4 element arrays of uint8. This is memory compatible with the raw
+four-element arrays of uint8. This is memory compatible with the raw
-image storage of 4 bytes per pixel.
+image storage of four bytes per pixel.
 
-Mode ``1`` images are exported as 1 uint8 byte/pixel, as this is
+Mode ``1`` images are exported as one uint8 byte/pixel, as this is
 consistent with the internal storage.
 
 Pillow will accept, but not produce, one other format. For any
-multichannel image with 32 bit storage per pixel, Pillow will accept
+multichannel image with 32-bit storage per pixel, Pillow will accept
 an array of width*height int32 elements, which will then be
-interpreted using the mode specific interpretation of the bytes.
+interpreted using the mode-specific interpretation of the bytes.
 
-The image mode must match the arrow band format when reading single
+The image mode must match the Arrow band format when reading single
-channel images
+channel images.
 
 Memory Allocator
 ================
 
 Pillow's default memory allocator, the :ref:`block_allocator`,
-allocates up to a 16MB block for images by default. Larger images
+allocates up to a 16 MB block for images by default. Larger images
 overflow into additional blocks. Arrow requires a single continuous
 memory allocation, so images allocated in multiple blocks cannot be
-exported in the arrow format.
+exported in the Arrow format.
 
 To enable the single block allocator::
 
   from PIL import Image
   Image.core.set_use_block_allocator(1)
 
-Note that this is a global setting, not a per image setting.
+Note that this is a global setting, not a per-image setting.
 
 Unsupported Features
 ====================
 
-* Table/Dataframe protocol. We currently support a single array.
+* Table/dataframe protocol. We support a single array.
 * Null markers, producing or consuming. Null values are inferred from
   the mode. e.g. RGB images are stored in the first three bytes of
-  each 32 bit pixel, and the last byte is an implied null.
+  each 32-bit pixel, and the last byte is an implied null.
-* Schema Negotiation. There is an optional schema for the requested
+* Schema negotiation. There is an optional schema for the requested
-  datatype in the arrow source interface. We currently ignore that
+  datatype in the Arrow source interface. We ignore that
   parameter.
-* Array Metadata.
+* Array metadata.
 
 Internal Details
 ================
@@ -76,12 +77,12 @@ Internal Details
 Python Arrow C interface:
 https://arrow.apache.org/docs/format/CDataInterface/PyCapsuleInterface.html
 
-The memory that is exported from the arrow interface is shared -- not
+The memory that is exported from the Arrow interface is shared -- not
 copied, so the lifetime of the memory allocation is no longer strictly
-tied to the life of the python object.
+tied to the life of the Python object.
 
 The core imaging struct now has a refcount associated with it, and the
-lifetime of the core image struct is now divorced from the python
+lifetime of the core image struct is now divorced from the Python
 image object. Creating an arrow reference to the image increments the
 refcount, and the imaging struct is only released when the refcount
-reaches 0.
+reaches zero.

src/PIL/Image.py

@@ -3210,16 +3210,9 @@ class SupportsArrowArrayInterface(Protocol):
     data interface.
     """
 
-    # Sorry, no types for you until pre-commit.ci stops changing stringy
+    def __arrow_c_array__(
-    # PyCapsule type to an unimportable value type, which then fails lint.
+        self, requested_schema: "PyCapsule" = None  # type: ignore[name-defined]  # noqa: F821, UP037
-    # PyCapsules are not importable, and only available in the C-API layer.
+    ) -> tuple["PyCapsule", "PyCapsule"]:  # type: ignore[name-defined]  # noqa: F821, UP037
-    # def __arrow_c_array__(
-    #     self, requested_schema: 'PyCapsule' = None
-    # ) -> tuple['PyCapsule', 'PyCapsule']:
-    #     raise NotImplementedError()
-
-    # old not typed definition.
-    def __arrow_c_array__(self, requested_schema=None):
         raise NotImplementedError()
 
 
@@ -3312,7 +3305,7 @@ def fromarray(obj: SupportsArrayInterface, mode: str | None = None) -> Image:
 
 
 def fromarrow(obj: SupportsArrowArrayInterface, mode, size) -> Image:
-    """Creates an image with zero copy shared memory from an object exporting
+    """Creates an image with zero-copy shared memory from an object exporting
     the arrow_c_array interface protocol::
 
       from PIL import Image
@@ -3323,7 +3316,7 @@ def fromarrow(obj: SupportsArrowArrayInterface, mode, size) -> Image:
     If the data representation of the ``obj`` is not compatible with
     Pillow internal storage, a ValueError is raised.
 
-    Pillow images can also be converted to arrow objects::
+    Pillow images can also be converted to Arrow objects::
 
       from PIL import Image
       import pyarrow as pa
@@ -3339,13 +3332,13 @@ def fromarrow(obj: SupportsArrowArrayInterface, mode, size) -> Image:
     :param size: Image size. This must match the storage of the arrow object.
     :returns: An Image Object
 
-    Note that according to the arrow spec, both the producer and the
+    Note that according to the Arrow spec, both the producer and the
     consumer should consider the exported array to be immutable, as
     unsynchronized updates will potentially cause inconsistent data.
 
     See: :ref:`arrow-support` for more detailed information
 
-    .. versionadded:: 11.2
+    .. versionadded:: 11.2.0
 
     """
     if not hasattr(obj, "__arrow_c_array__"):

src/_imaging.c

@@ -240,14 +240,14 @@ ArrowError(int err) {
         return ImagingError_MemoryError();
     }
     if (err == IMAGING_ARROW_INCOMPATIBLE_MODE) {
-        return ImagingError_ValueError("Incompatible Pillow mode for Arrow Array");
+        return ImagingError_ValueError("Incompatible Pillow mode for Arrow array");
     }
     if (err == IMAGING_ARROW_MEMORY_LAYOUT) {
         return ImagingError_ValueError(
             "Image is in multiple array blocks, use imaging_new_block for zero copy"
         );
     }
-    return ImagingError_ValueError("Unknown Error");
+    return ImagingError_ValueError("Unknown error");
 }
 
 void
@@ -313,7 +313,7 @@ _new_arrow(PyObject *self, PyObject *args) {
         PyImagingNew(ImagingNewArrow(mode, xsize, ysize, schema_capsule, array_capsule)
         );
     if (!ret) {
-        return ImagingError_ValueError("Invalid arrow array mode or size mismatch");
+        return ImagingError_ValueError("Invalid Arrow array mode or size mismatch");
     }
     return ret;
 }