fixup

Author: MarcoGorelli

pandas/core/interchange/buffer.py

@@ -79,13 +79,17 @@ def __repr__(self) -> str:
         )
 
 
-class PandasPyarrowBackedBuffer(Buffer):
+class PandasBufferPyarrow(Buffer):
     """
     Data in the buffer is guaranteed to be contiguous in memory.
     """
 
     def __init__(
-        self, chunked_array: pa.ChunkedArray, *, allow_copy: bool = True
+        self,
+        chunked_array: pa.ChunkedArray,
+        *,
+        is_validity: bool,
+        allow_copy: bool = True,
     ) -> None:
         """
         Handle pyarrow chunked arrays.
@@ -98,7 +102,10 @@ def __init__(
                     "Found multi-chunk pyarrow array, but `allow_copy` is False"
                 )
             arr = chunked_array.combine_chunks()
-        self._buffer = arr.buffers()[1]
+        if is_validity:
+            self._buffer = arr.buffers()[0]
+        else:
+            self._buffer = arr.buffers()[1]
         self._length = len(arr)
         self._dlpack = arr.__dlpack__
 
@@ -130,7 +137,7 @@ def __dlpack_device__(self) -> tuple[DlpackDeviceType, int | None]:
 
     def __repr__(self) -> str:
         return (
-            "PandasBuffer("
+            "PandasBuffer[pyarrow]("
             + str(
                 {
                     "bufsize": self.bufsize,

pandas/core/interchange/column.py

@@ -19,7 +19,7 @@
 from pandas.api.types import is_string_dtype
 from pandas.core.interchange.buffer import (
     PandasBuffer,
-    PandasPyarrowBackedBuffer,
+    PandasBufferPyarrow,
 )
 from pandas.core.interchange.dataframe_protocol import (
     Column,
@@ -311,7 +311,9 @@ def _get_data_buffer(
             arr = self._col.array
             if isinstance(self._col.dtype, ArrowDtype):
                 arr = self._col.array
-                buffer = PandasPyarrowBackedBuffer(arr._pa_array)
+                buffer = PandasBufferPyarrow(
+                    arr._pa_array, is_validity=False, allow_copy=self._allow_copy
+                )
                 return buffer, dtype
             if isinstance(self._col.dtype, BaseMaskedDtype):
                 np_arr = arr._data  # type: ignore[attr-defined]
@@ -364,7 +366,9 @@ def _get_validity_buffer(self) -> tuple[PandasBuffer, Any]:
             dtype = (DtypeKind.BOOL, 1, ArrowCTypes.BOOL, Endianness.NATIVE)
             if buf is None:
                 return buf, dtype
-            buffer = PandasPyarrowBackedBuffer(arr._pa_array)
+            buffer = PandasBufferPyarrow(
+                arr._pa_array, is_validity=True, allow_copy=self._allow_copy
+            )
             return buffer, dtype
 
         if isinstance(self._col.dtype, BaseMaskedDtype):