Merge remote-tracking branch 'upstream/branch-24.12' into pylibcudf/w…

…rappers/decimal

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_booleans.rst

@@ -0,0 +1,6 @@
+================
+convert_booleans
+================
+
+.. automodule:: pylibcudf.strings.convert.convert_booleans
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_datetime.rst

@@ -0,0 +1,6 @@
+================
+convert_datetime
+================
+
+.. automodule:: pylibcudf.strings.convert.convert_datetime
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_durations.rst

@@ -0,0 +1,6 @@
+=================
+convert_durations
+=================
+
+.. automodule:: pylibcudf.strings.convert.convert_durations
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_fixed_point.rst

@@ -0,0 +1,6 @@
+===================
+convert_fixed_point
+===================
+
+.. automodule:: pylibcudf.strings.convert.convert_fixed_point
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_floats.rst

@@ -0,0 +1,6 @@
+==============
+convert_floats
+==============
+
+.. automodule:: pylibcudf.strings.convert.convert_floats
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_ipv4.rst

@@ -0,0 +1,6 @@
+============
+convert_ipv4
+============
+
+.. automodule:: pylibcudf.strings.convert.convert_ipv4
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_lists.rst

@@ -0,0 +1,6 @@
+=============
+convert_lists
+=============
+
+.. automodule:: pylibcudf.strings.convert.convert_lists
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/convert_urls.rst

@@ -0,0 +1,6 @@
+============
+convert_urls
+============
+
+.. automodule:: pylibcudf.strings.convert.convert_urls
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/convert/index.rst

@@ -0,0 +1,14 @@
+convert
+=======
+
+.. toctree::
+    :maxdepth: 1
+
+    convert_booleans
+    convert_datetime
+    convert_durations
+    convert_fixed_point
+    convert_floats
+    convert_ipv4
+    convert_lists
+    convert_urls

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/index.rst

@@ -21,3 +21,9 @@ strings
     split
     strip
     wrap
+
+.. toctree::
+    :maxdepth: 2
+    :caption: Subpackages
+
+    convert/index.rst

python/cudf/cudf/_lib/strings/convert/convert_lists.pyx

@@ -1,23 +1,13 @@
 # Copyright (c) 2021-2024, NVIDIA CORPORATION.
 
-from libcpp.memory cimport unique_ptr
-from libcpp.utility cimport move
+import pylibcudf as plc
 
 from cudf.core.buffer import acquire_spill_lock
 
-from pylibcudf.libcudf.column.column cimport column
-from pylibcudf.libcudf.column.column_view cimport column_view
-from pylibcudf.libcudf.scalar.scalar cimport string_scalar
-from pylibcudf.libcudf.strings.convert.convert_lists cimport (
-    format_list_column as cpp_format_list_column,
-)
-
 from cudf._lib.column cimport Column
 
 from cudf._lib.scalar import as_device_scalar
 
-from cudf._lib.scalar cimport DeviceScalar
-
 
 @acquire_spill_lock()
 def format_list_column(Column source_list, Column separators):
@@ -34,19 +24,9 @@ def format_list_column(Column source_list, Column separators):
     -------
     Formatted strings column
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_list.view()
-    cdef column_view separators_view = separators.view()
-    # Use 'None' as null-replacement string
-    cdef DeviceScalar str_na_rep = as_device_scalar("None")
-    cdef const string_scalar* string_scalar_na_rep = <const string_scalar*>(
-        str_na_rep.get_raw_ptr())
-
-    with nogil:
-        c_result = move(cpp_format_list_column(
-            source_view, string_scalar_na_rep[0], separators_view
-        ))
-
-    return Column.from_unique_ptr(
-        move(c_result)
+    plc_column = plc.strings.convert.convert_lists.format_list_column(
+        source_list.to_pylibcudf(mode="read"),
+        as_device_scalar("None").c_value,
+        separators.to_pylibcudf(mode="read"),
     )
+    return Column.from_pylibcudf(plc_column)

python/pylibcudf/pylibcudf/libcudf/strings/convert/convert_lists.pxd

@@ -9,6 +9,6 @@ cdef extern from "cudf/strings/convert/convert_lists.hpp" namespace \
         "cudf::strings" nogil:
 
     cdef unique_ptr[column] format_list_column(
-        column_view input_col,
+        column_view input,
         string_scalar na_rep,
         column_view separators) except +

python/pylibcudf/pylibcudf/strings/convert/CMakeLists.txt

@@ -12,8 +12,9 @@
 # the License.
 # =============================================================================
 
-set(cython_sources convert_booleans.pyx convert_datetime.pyx convert_durations.pyx
-                   convert_fixed_point.pyx convert_floats.pyx convert_ipv4.pyx convert_urls.pyx
+set(cython_sources
+    convert_booleans.pyx convert_datetime.pyx convert_durations.pyx convert_fixed_point.pyx
+    convert_floats.pyx convert_ipv4.pyx convert_lists.pyx convert_urls.pyx
 )
 
 set(linked_libraries cudf::cudf)

python/pylibcudf/pylibcudf/strings/convert/__init__.pxd

@@ -6,5 +6,6 @@ from . cimport (
     convert_fixed_point,
     convert_floats,
     convert_ipv4,
+    convert_lists,
     convert_urls,
 )

python/pylibcudf/pylibcudf/strings/convert/__init__.py

@@ -6,5 +6,6 @@
     convert_fixed_point,
     convert_floats,
     convert_ipv4,
+    convert_lists,
     convert_urls,
 )

python/pylibcudf/pylibcudf/strings/convert/convert_fixed_point.pyx

@@ -15,7 +15,7 @@ cpdef Column to_fixed_point(Column input, DataType output_type):
     Returns a new fixed-point column parsing decimal values from the
     provided strings column.
 
-    For details, see :cpp:details:`cudf::strings::to_fixed_point`
+    For details, see :cpp:func:`cudf::strings::to_fixed_point`
 
     Parameters
     ----------
@@ -47,7 +47,7 @@ cpdef Column from_fixed_point(Column input):
     Returns a new strings column converting the fixed-point values
     into a strings column.
 
-    For details, see :cpp:details:`cudf::strings::from_fixed_point`
+    For details, see :cpp:func:`cudf::strings::from_fixed_point`
 
     Parameters
     ----------
@@ -75,7 +75,7 @@ cpdef Column is_fixed_point(Column input, DataType decimal_type=None):
     Returns a boolean column identifying strings in which all
     characters are valid for conversion to fixed-point.
 
-    For details, see :cpp:details:`cudf::strings::is_fixed_point`
+    For details, see :cpp:func:`cudf::strings::is_fixed_point`
 
     Parameters
     ----------

python/pylibcudf/pylibcudf/strings/convert/convert_lists.pxd

@@ -0,0 +1,11 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+from pylibcudf.column cimport Column
+from pylibcudf.scalar cimport Scalar
+
+
+cpdef Column format_list_column(
+    Column input,
+    Scalar na_rep=*,
+    Column separators=*
+)

python/pylibcudf/pylibcudf/strings/convert/convert_lists.pyx

@@ -0,0 +1,72 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+from libcpp.memory cimport unique_ptr
+from libcpp.utility cimport move
+from pylibcudf.column cimport Column
+from pylibcudf.column_factories cimport make_empty_column
+from pylibcudf.libcudf.column.column cimport column
+from pylibcudf.libcudf.scalar.scalar cimport string_scalar
+from pylibcudf.libcudf.scalar.scalar_factories cimport (
+    make_string_scalar as cpp_make_string_scalar,
+)
+from pylibcudf.libcudf.strings.convert cimport (
+    convert_lists as cpp_convert_lists,
+)
+from pylibcudf.scalar cimport Scalar
+from pylibcudf.types cimport type_id
+
+from cython.operator import dereference
+
+
+cpdef Column format_list_column(
+    Column input,
+    Scalar na_rep=None,
+    Column separators=None
+):
+    """
+    Convert a list column of strings into a formatted strings column.
+
+    For details, see :cpp:func`cudf::strings::format_list_column`
+
+    Parameters
+    ----------
+    input : Column
+        Lists column to format
+
+    na_rep : Scalar
+        Replacement string for null elements.
+        Default, empty string
+
+    separators : Column
+        Strings to use for enclosing list components and separating elements.
+        Default, ``,``, ``[``, ``]``
+
+    Returns
+    -------
+    Column
+        New strings column
+    """
+    cdef unique_ptr[column] c_result
+
+    if na_rep is None:
+        na_rep = Scalar.from_libcudf(
+            cpp_make_string_scalar("".encode())
+        )
+
+    cdef const string_scalar* c_na_rep = <const string_scalar*>(
+        na_rep.c_obj.get()
+    )
+
+    if separators is None:
+        separators = make_empty_column(type_id.STRING)
+
+    with nogil:
+        c_result = move(
+            cpp_convert_lists.format_list_column(
+                input.view(),
+                dereference(c_na_rep),
+                separators.view()
+            )
+        )
+
+    return Column.from_libcudf(move(c_result))

python/pylibcudf/pylibcudf/tests/test_string_convert_lists.py

@@ -0,0 +1,21 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+import pyarrow as pa
+import pylibcudf as plc
+import pytest
+from utils import assert_column_eq
+
+
+@pytest.mark.parametrize("na_rep", [None, pa.scalar("")])
+@pytest.mark.parametrize("separators", [None, pa.array([",", "[", "]"])])
+def test_format_list_column(na_rep, separators):
+    arr = pa.array([["1", "A"], None])
+    result = plc.strings.convert.convert_lists.format_list_column(
+        plc.interop.from_arrow(arr),
+        na_rep if na_rep is None else plc.interop.from_arrow(na_rep),
+        separators
+        if separators is None
+        else plc.interop.from_arrow(separators),
+    )
+    expected = pa.array(["[1,A]", ""])
+    assert_column_eq(result, expected)