Finish convert_datetime

python/cudf/cudf/_lib/string_casting.pyx

@@ -3,9 +3,6 @@
 from cudf._lib.column cimport Column
 
 from cudf._lib.scalar import as_device_scalar
-
-from cudf._lib.scalar cimport DeviceScalar
-
 from cudf._lib.types import SUPPORTED_NUMPY_TO_LIBCUDF_TYPES
 
 from libcpp.memory cimport unique_ptr
@@ -14,10 +11,6 @@ from libcpp.utility cimport move
 
 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_datetime cimport (
-    is_timestamp as cpp_is_timestamp,
-)
 from pylibcudf.libcudf.strings.convert.convert_floats cimport (
     from_floats as cpp_from_floats,
     to_floats as cpp_to_floats,
@@ -460,11 +453,10 @@ def int2timestamp(
     A Column with date-time represented in string format
 
     """
-    cdef string c_timestamp_format = format.encode("UTF-8")
     return Column.from_pylibcudf(
         plc.strings.convert.convert_datetime.from_timestamps(
             input_col.to_pylibcudf(mode="read"),
-            c_timestamp_format,
+            format,
             names.to_pylibcudf(mode="read")
         )
     )
@@ -485,12 +477,11 @@ def timestamp2int(Column input_col, dtype, format):
 
     """
     dtype = dtype_to_pylibcudf_type(dtype)
-    cdef string c_timestamp_format = format.encode('UTF-8')
     return Column.from_pylibcudf(
         plc.strings.convert.convert_datetime.to_timestamps(
             input_col.to_pylibcudf(mode="read"),
             dtype,
-            c_timestamp_format
+            format
         )
     )
 
@@ -512,16 +503,11 @@ def istimestamp(Column input_col, str format):
     """
     if input_col.size == 0:
         return cudf.core.column.column_empty(0, dtype=cudf.dtype("bool"))
-    cdef column_view input_column_view = input_col.view()
-    cdef string c_timestamp_format = <string>str(format).encode('UTF-8')
-    cdef unique_ptr[column] c_result
-    with nogil:
-        c_result = move(
-            cpp_is_timestamp(
-                input_column_view,
-                c_timestamp_format))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = plc.strings.convert.convert_datetime.is_timestamp(
+        input_col.to_pylibcudf(mode="read"),
+        format
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 def timedelta2int(Column input_col, dtype, format):

python/pylibcudf/pylibcudf/strings/convert/convert_datetime.pxd

@@ -8,16 +8,16 @@ from pylibcudf.types cimport DataType
 cpdef Column to_timestamps(
     Column input,
     DataType timestamp_type,
-    const string& format
+    str format
 )
 
 cpdef Column from_timestamps(
-    Column input,
-    const string& format,
+    Column timestamps,
+    str format,
     Column input_strings_names
 )
 
 cpdef Column is_timestamp(
     Column input,
-    const string& format,
+    str format,
 )

python/pylibcudf/pylibcudf/strings/convert/convert_datetime.pyx

@@ -15,42 +15,108 @@ from pylibcudf.types import DataType
 cpdef Column to_timestamps(
     Column input,
     DataType timestamp_type,
-    const string& format
+    str format
 ):
+    """
+    Returns a new timestamp column converting a strings column into
+    timestamps using the provided format pattern.
+
+    For details, see cpp:`cudf::strings::to_timestamps`.
+
+    Parameters
+    ----------
+    input : Column
+        Strings instance for this operation.
+
+    timestamp_type : DataType
+        The timestamp type used for creating the output column.
+
+    format : str
+        String specifying the timestamp format in strings.
+
+    Returns
+    -------
+    Column
+        New datetime column
+    """
     cdef unique_ptr[column] c_result
+    cdef string c_format = format.encode()
     with nogil:
         c_result = cpp_convert_datetime.to_timestamps(
             input.view(),
             timestamp_type.c_obj,
-            format
+            c_format
         )
 
     return Column.from_libcudf(move(c_result))
 
 cpdef Column from_timestamps(
     Column timestamps,
-    const string& format,
+    str format,
     Column input_strings_names
 ):
+    """
+    Returns a new strings column converting a timestamp column into
+    strings using the provided format pattern.
+
+    For details, see cpp:`cudf::strings::from_timestamps`.
+
+    Parameters
+    ----------
+    timestamps : Column
+        Timestamp values to convert
+
+    format : str
+        The string specifying output format.
+
+    input_strings_names : Column
+        The string names to use for weekdays ("%a", "%A") and months ("%b", "%B").
+
+    Returns
+    -------
+    Column
+        New strings column with formatted timestamps.
+    """
     cdef unique_ptr[column] c_result
+    cdef string c_format = format.encode()
     with nogil:
         c_result = cpp_convert_datetime.from_timestamps(
-            input.view(),
-            format,
+            timestamps.view(),
+            c_format,
             input_strings_names.view()
         )
 
     return Column.from_libcudf(move(c_result))
 
 cpdef Column is_timestamp(
     Column input,
-    const string& format
+    str format
 ):
+    """
+    Verifies the given strings column can be parsed to timestamps
+    using the provided format pattern.
+
+    For details, see cpp:`cudf::strings::is_timestamp`.
+
+    Parameters
+    ----------
+    input : Column
+        Strings instance for this operation.
+
+    format : str
+        String specifying the timestamp format in strings.
+
+    Returns
+    -------
+    Column
+        New bool column.
+    """
     cdef unique_ptr[column] c_result
+    cdef string c_format = format.encode()
     with nogil:
         c_result = cpp_convert_datetime.is_timestamp(
             input.view(),
-            format
+            c_format
         )
 
     return Column.from_libcudf(move(c_result))

python/pylibcudf/pylibcudf/tests/test_string_convert_datetime.py

@@ -0,0 +1,46 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+import datetime
+
+import pyarrow as pa
+import pyarrow.compute as pc
+import pylibcudf as plc
+import pytest
+from utils import assert_column_eq
+
+
+@pytest.fixture
+def fmt():
+    return "%Y-%m-%dT%H:%M:%S"
+
+
+def test_to_timestamp(fmt):
+    arr = pa.array(["2020-01-01T01:01:01", None])
+    result = plc.strings.convert.convert_datetime.to_timestamps(
+        plc.interop.from_arrow(arr),
+        plc.DataType(plc.TypeId.TIMESTAMP_SECONDS),
+        fmt,
+    )
+    expected = pc.strptime(arr, fmt, "s")
+    assert_column_eq(result, expected)
+
+
+def test_from_timestamp(fmt):
+    arr = pa.array([datetime.datetime(2020, 1, 1, 1, 1, 1), None])
+    result = plc.strings.convert.convert_datetime.from_timestamps(
+        plc.interop.from_arrow(arr),
+        fmt,
+        plc.interop.from_arrow(pa.array([], type=pa.string())),
+    )
+    # pc.strftime will add the extra %f
+    expected = pa.array(["2020-01-01T01:01:01", None])
+    assert_column_eq(result, expected)
+
+
+def test_is_timestamp(fmt):
+    arr = pa.array(["2020-01-01T01:01:01", None, "2020-01-01"])
+    result = plc.strings.convert.convert_datetime.is_timestamp(
+        plc.interop.from_arrow(arr),
+        fmt,
+    )
+    expected = pa.array([True, None, False])
+    assert_column_eq(result, expected)