Add pandas-2.3.0 support

conda/environments/all_cuda-128_arch-aarch64.yaml

@@ -64,7 +64,7 @@ dependencies:
 - openpyxl
 - packaging
 - pandas
-- pandas>=2.0,<2.2.4dev0
+- pandas>=2.0,<2.4.0dev0
 - pandoc
 - polars>=1.25,<1.31
 - pre-commit

conda/environments/all_cuda-128_arch-x86_64.yaml

@@ -65,7 +65,7 @@ dependencies:
 - openpyxl
 - packaging
 - pandas
-- pandas>=2.0,<2.2.4dev0
+- pandas>=2.0,<2.4.0dev0
 - pandoc
 - polars>=1.25,<1.31
 - pre-commit

conda/recipes/cudf/recipe.yaml

@@ -67,7 +67,7 @@ requirements:
   run:
     - python
     - typing_extensions >=4.0.0
-    - pandas >=2.0,<2.2.4dev0
+    - pandas >=2.0,<2.4.0dev0
     - cupy >=12.0.0
     - numba-cuda >=0.14.0,<0.15.0a0
     - numba >=0.59.1,<0.62.0a0

conda/recipes/pylibcudf/recipe.yaml

@@ -66,7 +66,7 @@ requirements:
   run:
     - python
     - typing_extensions >=4.0.0
-    - pandas >=2.0,<2.2.4dev0
+    - pandas >=2.0,<2.4.0dev0
     - numpy >=1.23,<3.0a0
     - pyarrow>=14.0.0,<20.0.0a0
     - libcudf =${{ version }}

dependencies.yaml

@@ -638,7 +638,7 @@ dependencies:
         packages:
           - fsspec>=0.6.0
           - &numpy numpy>=1.23,<3.0a0
-          - pandas>=2.0,<2.2.4dev0
+          - pandas>=2.0,<2.4.0dev0
   run_pylibcudf:
     common:
       - output_types: [conda, requirements, pyproject]
@@ -776,7 +776,7 @@ dependencies:
             packages:
               - *numba-cuda-dep
               - *numba-dep
-              - pandas==2.2.3
+              - pandas==2.3.0
           - matrix:
             packages:
       - output_types: conda

python/cudf/cudf/core/_compat.py

@@ -1,13 +1,14 @@
-# Copyright (c) 2020-2024, NVIDIA CORPORATION.
+# Copyright (c) 2020-2025, NVIDIA CORPORATION.
 
 import pandas as pd
 from packaging import version
 
-PANDAS_CURRENT_SUPPORTED_VERSION = version.parse("2.2.3")
+PANDAS_CURRENT_SUPPORTED_VERSION = version.parse("2.3.0")
 PANDAS_VERSION = version.parse(pd.__version__)
 
 
 PANDAS_GE_210 = PANDAS_VERSION >= version.parse("2.1.0")
 PANDAS_GT_214 = PANDAS_VERSION > version.parse("2.1.4")
 PANDAS_GE_220 = PANDAS_VERSION >= version.parse("2.2.0")
+PANDAS_GE_230 = PANDAS_VERSION >= version.parse("2.3.0")
 PANDAS_LT_300 = PANDAS_VERSION < version.parse("3.0.0")

python/cudf/cudf/core/accessors/string.py

@@ -13,6 +13,7 @@
 import pylibcudf as plc
 
 import cudf
+from cudf.api.extensions import no_default
 from cudf.api.types import is_integer, is_scalar
 from cudf.core.accessors.base_accessor import BaseAccessor
 from cudf.core.column.column import ColumnBase, as_column
@@ -622,7 +623,7 @@ def contains(
         pat: str | Sequence,
         case: bool = True,
         flags: int = 0,
-        na=np.nan,
+        na=no_default,
         regex: bool = True,
     ) -> Series | Index:
         r"""
@@ -734,7 +735,18 @@ def contains(
             The `flags` parameter currently only supports re.DOTALL and
             re.MULTILINE.
         """
-        if na is not np.nan:
+        if (
+            na is not no_default
+            and not pd.isna(na)
+            and not isinstance(na, bool)
+        ):
+            # GH#59561
+            warnings.warn(
+                "Allowing a non-bool 'na' in obj.str.contains is deprecated "
+                "and will raise in a future version.",
+                FutureWarning,
+            )
+        if na not in {no_default, np.nan}:
             raise NotImplementedError("`na` parameter is not yet supported")
         if regex and isinstance(pat, re.Pattern):
             flags = pat.flags & ~re.U

python/cudf/cudf/pandas/_wrappers/pandas.py

@@ -659,14 +659,40 @@ def Index__setattr__(self, name, value):
         pd.core.arrays.string_arrow.ArrowStringArrayNumpySemantics,
         fast_to_slow=_Unusable(),
         slow_to_fast=_Unusable(),
+        additional_attributes={
+            "_pa_array": _FastSlowAttribute("_pa_array", private=True),
+        },
     )
 
+if cudf.core._compat.PANDAS_GE_230:
+    StringArrayNumpySemantics = make_final_proxy_type(
+        "StringArrayNumpySemantics",
+        _Unusable,
+        pd.core.arrays.string_.StringArrayNumpySemantics,
+        bases=(StringArray,),
+        fast_to_slow=_Unusable(),
+        slow_to_fast=_Unusable(),
+    )
+
+
 ArrowStringArray = make_final_proxy_type(
     "ArrowStringArray",
     _Unusable,
     pd.core.arrays.string_arrow.ArrowStringArray,
     fast_to_slow=_Unusable(),
     slow_to_fast=_Unusable(),
+    additional_attributes={
+        "_pa_array": _FastSlowAttribute("_pa_array", private=True),
+    },
+)
+
+
+StorageExtensionDtype = make_final_proxy_type(
+    "StorageExtensionDtype",
+    _Unusable,
+    pd.core.dtypes.base.StorageExtensionDtype,
+    fast_to_slow=_Unusable(),
+    slow_to_fast=_Unusable(),
 )
 
 StringDtype = make_final_proxy_type(
@@ -1851,6 +1877,9 @@ def holiday_calendar_factory_wrapper(*args, **kwargs):
     pd.arrays.ArrowExtensionArray,
     fast_to_slow=_Unusable(),
     slow_to_fast=_Unusable(),
+    additional_attributes={
+        "_pa_array": _FastSlowAttribute("_pa_array", private=True),
+    },
 )