Remove build_struct|list_column

python/cudf/cudf/core/column/__init__.py

@@ -11,8 +11,6 @@
     as_column,
     build_categorical_column,
     build_column,
-    build_list_column,
-    build_struct_column,
     column_empty,
     column_empty_like,
     column_empty_like_same_mask,

python/cudf/cudf/core/column/column.py

@@ -1690,92 +1690,6 @@ def build_categorical_column(
     return cast("cudf.core.column.CategoricalColumn", result)
 
 
-def build_list_column(
-    indices: ColumnBase,
-    elements: ColumnBase,
-    mask: Optional[Buffer] = None,
-    size: Optional[int] = None,
-    offset: int = 0,
-    null_count: Optional[int] = None,
-) -> "cudf.core.column.ListColumn":
-    """
-    Build a ListColumn
-
-    Parameters
-    ----------
-    indices : ColumnBase
-        Column of list indices
-    elements : ColumnBase
-        Column of list elements
-    mask: Buffer
-        Null mask
-    size: int, optional
-    offset: int, optional
-    """
-    dtype = ListDtype(element_type=elements.dtype)
-    if size is None:
-        if indices.size == 0:
-            size = 0
-        else:
-            # one less because the last element of offsets is the number of
-            # bytes in the data buffer
-            size = indices.size - 1
-        size = size - offset
-
-    result = build_column(
-        data=None,
-        dtype=dtype,
-        mask=mask,
-        size=size,
-        offset=offset,
-        null_count=null_count,
-        children=(indices, elements),
-    )
-
-    return cast("cudf.core.column.ListColumn", result)
-
-
-def build_struct_column(
-    names: Sequence[str],
-    children: Tuple[ColumnBase, ...],
-    dtype: Optional[Dtype] = None,
-    mask: Optional[Buffer] = None,
-    size: Optional[int] = None,
-    offset: int = 0,
-    null_count: Optional[int] = None,
-) -> "cudf.core.column.StructColumn":
-    """
-    Build a StructColumn
-
-    Parameters
-    ----------
-    names : sequence of strings
-        Field names to map to children dtypes, must be strings.
-    children : tuple
-
-    mask: Buffer
-        Null mask
-    size: int, optional
-    offset: int, optional
-    """
-    if dtype is None:
-        dtype = StructDtype(
-            fields={name: col.dtype for name, col in zip(names, children)}
-        )
-
-    result = build_column(
-        data=None,
-        dtype=dtype,
-        mask=mask,
-        size=size,
-        offset=offset,
-        null_count=null_count,
-        children=children,
-    )
-
-    return cast("cudf.core.column.StructColumn", result)
-
-
 def _make_copy_replacing_NaT_with_null(column):
     """Return a copy with NaT values replaced with nulls."""
     if np.issubdtype(column.dtype, np.timedelta64):

python/cudf/cudf/core/column/lists.py

@@ -184,15 +184,16 @@ def _with_type_metadata(
         self: "cudf.core.column.ListColumn", dtype: Dtype
     ) -> "cudf.core.column.ListColumn":
         if isinstance(dtype, ListDtype):
-            return column.build_list_column(
-                indices=self.base_children[0],
-                elements=self.base_children[1]._with_type_metadata(
-                    dtype.element_type
-                ),
+            elements = self.base_children[1]._with_type_metadata(
+                dtype.element_type
+            )
+            return ListColumn(
+                dtype=dtype,
                 mask=self.base_mask,
                 size=self.size,
                 offset=self.offset,
                 null_count=self.null_count,
+                children=(self.base_children[0], elements),
             )
 
         return self

python/cudf/cudf/core/column/struct.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2020-2023, NVIDIA CORPORATION.
+# Copyright (c) 2020-2024, NVIDIA CORPORATION.
 from __future__ import annotations
 
 from functools import cached_property
@@ -9,7 +9,7 @@
 
 import cudf
 from cudf._typing import Dtype
-from cudf.core.column import ColumnBase, build_struct_column
+from cudf.core.column import ColumnBase
 from cudf.core.column.methods import ColumnMethods
 from cudf.core.dtypes import StructDtype
 from cudf.core.missing import NA
@@ -134,8 +134,9 @@ def _with_type_metadata(self: StructColumn, dtype: Dtype) -> StructColumn:
         if isinstance(dtype, IntervalDtype):
             return IntervalColumn.from_struct_column(self, closed=dtype.closed)
         elif isinstance(dtype, StructDtype):
-            return build_struct_column(
-                names=dtype.fields.keys(),
+            return StructColumn(
+                data=None,
+                dtype=dtype,
                 children=tuple(
                     self.base_children[i]._with_type_metadata(dtype.fields[f])
                     for i, f in enumerate(dtype.fields.keys())

python/cudf/cudf/core/dataframe.py

@@ -61,6 +61,7 @@
 from cudf.core.column import (
     CategoricalColumn,
     ColumnBase,
+    StructColumn,
     as_column,
     build_categorical_column,
     build_column,
@@ -7127,12 +7128,13 @@ def to_struct(self, name=None):
                 "requires field name to be string. Non-string column names "
                 "will be casted to string as the field name."
             )
-        field_names = [str(name) for name in self._data.names]
-
-        col = cudf.core.column.build_struct_column(
-            names=field_names,
+        fields = {str(name): col.dtype for name, col in self._data.items()}
+        col = StructColumn(
+            data=None,
+            dtype=cudf.StructDtype(fields=fields),
             children=tuple(col.copy(deep=True) for col in self._data.columns),
             size=len(self),
+            offset=0,
         )
         return cudf.Series._from_data(
             cudf.core.column_accessor.ColumnAccessor({name: col}),

python/cudf/cudf/core/groupby/groupby.py

@@ -25,7 +25,7 @@
 from cudf.api.types import is_bool_dtype, is_float_dtype, is_list_like
 from cudf.core._compat import PANDAS_LT_300
 from cudf.core.abc import Serializable
-from cudf.core.column.column import ColumnBase, as_column
+from cudf.core.column.column import ColumnBase, StructDtype, as_column
 from cudf.core.column_accessor import ColumnAccessor
 from cudf.core.join._join_helpers import _match_join_keys
 from cudf.core.mixins import Reducible, Scannable
@@ -2036,10 +2036,14 @@ def _cov_or_corr(self, func, method_name):
                 )
                 x, y = str(x), str(y)
 
-            column_pair_structs[(x, y)] = cudf.core.column.build_struct_column(
-                names=(x, y),
+            column_pair_structs[(x, y)] = cudf.core.column.StructColumn(
+                data=None,
+                dtype=StructDtype(
+                    fields={x: self.obj._data[x].dtype, y: self.obj._data[y]}
+                ),
                 children=(self.obj._data[x], self.obj._data[y]),
                 size=len(self.obj),
+                offset=0,
             )
 
         column_pair_groupby = cudf.DataFrame._from_data(