Add remaining string.char_types APIs to pylibcudf

cpp/include/cudf/strings/char_types/char_types.hpp

@@ -30,7 +30,7 @@ namespace strings {
  */
 
 /**
- * @brief Returns a boolean column identifying strings entries in which all
+ * @brief Returns a boolean column identifying string entries where all
  * characters are of the type specified.
  *
  * The output row entry will be set to false if the corresponding string element
@@ -105,7 +105,8 @@ std::unique_ptr<column> all_characters_of_type(
  *        `types_to_remove` will be filtered.
  * @param mr Device memory resource used to allocate the returned column's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches
- * @return New column of boolean results for each string
+ * @return New strings column with the specified characters filtered out and replaced with specified
+ * replacement string.
  */
 std::unique_ptr<column> filter_characters_of_type(
   strings_column_view const& input,

python/cudf/cudf/_lib/strings/char_types.pyx

@@ -1,50 +1,28 @@
 # Copyright (c) 2021-2024, NVIDIA CORPORATION.
 
-
 from libcpp cimport bool
-from libcpp.memory cimport unique_ptr
-from libcpp.utility cimport move
 
 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.char_types cimport (
-    all_characters_of_type as cpp_all_characters_of_type,
-    filter_characters_of_type as cpp_filter_characters_of_type,
-    string_character_types,
-)
-
 from cudf._lib.column cimport Column
-from cudf._lib.scalar cimport DeviceScalar
+
+from pylibcudf.strings import char_types
 
 
 @acquire_spill_lock()
 def filter_alphanum(Column source_strings, object py_repl, bool keep=True):
     """
     Returns a Column of strings keeping only alphanumeric character types.
     """
-
-    cdef DeviceScalar repl = py_repl.device_value
-
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-    cdef const string_scalar* scalar_repl = <const string_scalar*>(
-        repl.get_raw_ptr()
+    plc_column = char_types.filter_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.ALL_TYPES if keep
+        else char_types.StringCharacterTypes.ALPHANUM,
+        py_repl.device_value.c_value,
+        char_types.StringCharacterTypes.ALPHANUM if keep
+        else char_types.StringCharacterTypes.ALL_TYPES
     )
-
-    with nogil:
-        c_result = move(cpp_filter_characters_of_type(
-            source_view,
-            string_character_types.ALL_TYPES if keep
-            else string_character_types.ALPHANUM,
-            scalar_repl[0],
-            string_character_types.ALPHANUM if keep
-            else string_character_types.ALL_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -54,17 +32,12 @@ def is_decimal(Column source_strings):
     that contain only decimal characters -- those that can be used
     to extract base10 numbers.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.DECIMAL,
-            string_character_types.ALL_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.DECIMAL,
+        char_types.StringCharacterTypes.ALL_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -75,17 +48,12 @@ def is_alnum(Column source_strings):
 
     Equivalent to: is_alpha() or is_digit() or is_numeric() or is_decimal()
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.ALPHANUM,
-            string_character_types.ALL_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.ALPHANUM,
+        char_types.StringCharacterTypes.ALL_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -94,17 +62,12 @@ def is_alpha(Column source_strings):
     Returns a Column of boolean values with True for `source_strings`
     that contain only alphabetic characters.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.ALPHA,
-            string_character_types.ALL_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.ALPHA,
+        char_types.StringCharacterTypes.ALL_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -113,17 +76,12 @@ def is_digit(Column source_strings):
     Returns a Column of boolean values with True for `source_strings`
     that contain only decimal and digit characters.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.DIGIT,
-            string_character_types.ALL_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.DIGIT,
+        char_types.StringCharacterTypes.ALL_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -133,17 +91,12 @@ def is_numeric(Column source_strings):
     that contain only numeric characters. These include digit and
     numeric characters.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.NUMERIC,
-            string_character_types.ALL_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.NUMERIC,
+        char_types.StringCharacterTypes.ALL_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -152,17 +105,12 @@ def is_upper(Column source_strings):
     Returns a Column of boolean values with True for `source_strings`
     that contain only upper-case characters.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.UPPER,
-            string_character_types.CASE_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.UPPER,
+        char_types.StringCharacterTypes.CASE_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -171,17 +119,12 @@ def is_lower(Column source_strings):
     Returns a Column of boolean values with True for `source_strings`
     that contain only lower-case characters.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.LOWER,
-            string_character_types.CASE_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.LOWER,
+        char_types.StringCharacterTypes.CASE_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)
 
 
 @acquire_spill_lock()
@@ -190,14 +133,9 @@ def is_space(Column source_strings):
     Returns a Column of boolean values with True for `source_strings`
     that contains all characters which are spaces only.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-
-    with nogil:
-        c_result = move(cpp_all_characters_of_type(
-            source_view,
-            string_character_types.SPACE,
-            string_character_types.ALL_TYPES
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = char_types.all_characters_of_type(
+        source_strings.to_pylibcudf(mode="read"),
+        char_types.StringCharacterTypes.SPACE,
+        char_types.StringCharacterTypes.ALL_TYPES
+    )
+    return Column.from_pylibcudf(plc_column)

python/pylibcudf/pylibcudf/libcudf/strings/char_types.pxd

@@ -22,9 +22,6 @@ cdef extern from "cudf/strings/char_types/char_types.hpp" \
         CASE_TYPES
         ALL_TYPES
 
-cdef extern from "cudf/strings/char_types/char_types.hpp" \
-        namespace "cudf::strings" nogil:
-
     cdef unique_ptr[column] all_characters_of_type(
         column_view source_strings,
         string_character_types types,

python/pylibcudf/pylibcudf/strings/char_types.pxd

@@ -1,3 +1,19 @@
 # Copyright (c) 2024, NVIDIA CORPORATION.
 
+from pylibcudf.column cimport Column
 from pylibcudf.libcudf.strings.char_types cimport string_character_types
+from pylibcudf.scalar cimport Scalar
+
+
+cpdef Column all_characters_of_type(
+    Column source_strings,
+    string_character_types types,
+    string_character_types verify_types
+)
+
+cpdef Column filter_characters_of_type(
+    Column source_strings,
+    string_character_types types_to_remove,
+    Scalar replacement,
+    string_character_types types_to_keep
+)

python/pylibcudf/pylibcudf/strings/char_types.pyx

@@ -1,4 +1,93 @@
 # Copyright (c) 2024, NVIDIA CORPORATION.
 
+from libcpp.memory cimport unique_ptr
+from libcpp.utility cimport move
+from pylibcudf.column cimport Column
+from pylibcudf.libcudf.column.column cimport column
+from pylibcudf.libcudf.scalar.scalar cimport string_scalar
+from pylibcudf.libcudf.strings cimport char_types as cpp_char_types
+from pylibcudf.scalar cimport Scalar
+
+from cython.operator import dereference
 from pylibcudf.libcudf.strings.char_types import \
     string_character_types as StringCharacterTypes  # no-cython-lint
+
+
+cpdef Column all_characters_of_type(
+    Column source_strings,
+    string_character_types types,
+    string_character_types verify_types
+):
+    """
+    Identifies strings where all characters match the specified type.
+
+    Parameters
+    ----------
+    source_strings : Column
+        Strings instance for this operation
+    types : StringCharacterTypes
+        The character types to check in each string
+    verify_types : StringCharacterTypes
+        Only verify against these character types.
+
+    Returns
+    -------
+    Column
+        New column of boolean results for each string
+    """
+    cdef unique_ptr[column] c_result
+
+    with nogil:
+        c_result = move(
+            cpp_char_types.all_characters_of_type(
+                source_strings.view(),
+                types,
+                verify_types,
+            )
+        )
+
+    return Column.from_libcudf(move(c_result))
+
+cpdef Column filter_characters_of_type(
+    Column source_strings,
+    string_character_types types_to_remove,
+    Scalar replacement,
+    string_character_types types_to_keep
+):
+    """
+    Filter specific character types from a column of strings.
+
+    Parameters
+    ----------
+    source_strings : Column
+        Strings instance for this operation
+    types_to_remove : StringCharacterTypes
+        The character types to check in each string.
+    replacement : Scalar
+        The replacement character to use when removing characters
+    types_to_keep : StringCharacterTypes
+        Default `ALL_TYPES` means all characters of `types_to_remove`
+        will be filtered.
+
+    Returns
+    -------
+    Column
+        New column with the specified characters filtered out and
+        replaced with the specified replacement string.
+    """
+    cdef const string_scalar* c_replacement = <const string_scalar*>(
+        replacement.c_obj.get()
+    )
+    cdef unique_ptr[column] c_result
+
+    with nogil:
+        c_result = move(
+            cpp_char_types.filter_characters_of_type(
+                source_strings.view(),
+                types_to_remove,
+                dereference(c_replacement),
+                types_to_keep,
+            )
+        )
+
+    return Column.from_libcudf(move(c_result))