From 93313711fa8dc1eee613fabc5f10d74fbd6ee807 Mon Sep 17 00:00:00 2001
From: Vyas Ramasubramani <vyasr@nvidia.com>
Date: Thu, 17 Aug 2023 09:57:12 -0700
Subject: [PATCH 1/6] Add new functions for dtype extraction that operate
 directly on pylibcudf types

---
 python/cudf/cudf/_lib/column.pyx           |  4 +-
 python/cudf/cudf/_lib/pylibcudf/column.pxd |  5 ++
 python/cudf/cudf/_lib/pylibcudf/column.pyx | 24 +++++++++
 python/cudf/cudf/_lib/types.pyx            | 57 ++++++++++++++++++++++
 4 files changed, 88 insertions(+), 2 deletions(-)

diff --git a/python/cudf/cudf/_lib/column.pyx b/python/cudf/cudf/_lib/column.pyx
index 2b1fc14f398..50b56f1d484 100644
--- a/python/cudf/cudf/_lib/column.pyx
+++ b/python/cudf/cudf/_lib/column.pyx
@@ -40,6 +40,7 @@ from cudf._lib.types cimport (
 )
 
 from cudf._lib.null_mask import bitmask_allocation_size_bytes
+from cudf._lib.types import dtype_from_pylibcudf_column
 
 cimport cudf._lib.cpp.types as libcudf_types
 cimport cudf._lib.cpp.unary as libcudf_unary
@@ -607,8 +608,7 @@ cdef class Column:
         pylibcudf.Column
             A new pylibcudf.Column referencing the same data.
         """
-        # TODO: Rewrite utility for dtype conversion to not need a column view.
-        dtype = dtype_from_column_view(col.view())
+        dtype = dtype_from_pylibcudf_column(col)
 
         return cudf.core.column.build_column(
             data=as_buffer(col.data.obj) if col.data is not None else None,
diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pxd b/python/cudf/cudf/_lib/pylibcudf/column.pxd
index 2b08e6863a1..498714dc0c3 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pxd
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pxd
@@ -23,8 +23,13 @@ cdef class Column:
         size_type offset
         # children: List[Column]
         list children
+        size_type _num_children
 
     cdef column_view view(self) nogil
 
     @staticmethod
     cdef Column from_libcudf(unique_ptr[column] libcudf_col)
+
+    cpdef DataType type(self) noexcept
+    cpdef Column child(self, size_type index) noexcept
+    cpdef size_type num_children(self) noexcept
diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pyx b/python/cudf/cudf/_lib/pylibcudf/column.pyx
index be4eff4c49d..bfb010992c0 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pyx
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pyx
@@ -52,6 +52,7 @@ cdef class Column:
         self.null_count = null_count
         self.offset = offset
         self.children = children
+        self._num_children = len(children)
 
     cdef column_view view(self) nogil:
         """Generate a libcudf column_view to pass to libcudf algorithms.
@@ -133,3 +134,26 @@ cdef class Column:
             0,
             children,
         )
+
+    cpdef DataType type(self):
+        """The type of data in the column."""
+        return self.data_type
+
+    cpdef Column child(self, size_type index) noexcept:
+        """Get a child column of this column.
+
+        Parameters
+        ----------
+        index : size_type
+            The index of the child column to get.
+
+        Returns
+        -------
+        Column
+            The child column.
+        """
+        return self.children[index]
+
+    cpdef size_type num_children(self) noexcept:
+        """The number of children of this column."""
+        return self._num_children
diff --git a/python/cudf/cudf/_lib/types.pyx b/python/cudf/cudf/_lib/types.pyx
index c3eca8090a3..283ba8a8e72 100644
--- a/python/cudf/cudf/_lib/types.pyx
+++ b/python/cudf/cudf/_lib/types.pyx
@@ -129,6 +129,11 @@ LIBCUDF_TO_SUPPORTED_NUMPY_TYPES = {
     TypeId.STRUCT: np.dtype("object"),
 }
 
+PYLIBCUDF_TO_SUPPORTED_NUMPY_TYPES = {
+    pylibcudf.TypeId(k).value: v
+    for k, v in LIBCUDF_TO_SUPPORTED_NUMPY_TYPES.items()
+}
+
 duration_unit_map = {
     TypeId.DURATION_SECONDS: "s",
     TypeId.DURATION_MILLISECONDS: "ms",
@@ -275,3 +280,55 @@ cdef bool is_decimal_type_id(libcudf_types.type_id tid) except *:
         libcudf_types.type_id.DECIMAL64,
         libcudf_types.type_id.DECIMAL32,
     )
+
+
+def dtype_from_pylibcudf_lists_column(pylibcudf.Column col):
+    # TODO: Currently hardcoding the child column index for lists, should come
+    # up with a cleaner solution here.
+    child = col.child(1)
+
+    if child.type().id() == pylibcudf.TypeId.LIST:
+        return cudf.ListDtype(dtype_from_pylibcudf_lists_column(child))
+    elif child.type().id() == pylibcudf.TypeId.EMPTY:
+        return cudf.ListDtype("int8")
+    else:
+        return cudf.ListDtype(
+            dtype_from_pylibcudf_column(child)
+        )
+
+
+def dtype_from_pylibcudf_structs_column(pylibcudf.Column col):
+    fields = {
+        str(i): dtype_from_pylibcudf_column(col.child(i))
+        for i in range(col.num_children())
+    }
+    return cudf.StructDtype(fields)
+
+
+def dtype_from_pylibcudf_column(pylibcudf.Column col):
+    type_ = col.type()
+    tid = type_.id()
+
+    if tid == pylibcudf.TypeId.LIST:
+        return dtype_from_pylibcudf_lists_column(col)
+    elif tid == pylibcudf.TypeId.STRUCT:
+        return dtype_from_pylibcudf_structs_column(col)
+    elif tid == pylibcudf.TypeId.DECIMAL64:
+        return cudf.Decimal64Dtype(
+            precision=cudf.Decimal64Dtype.MAX_PRECISION,
+            scale=-type_.scale()
+        )
+    elif tid == pylibcudf.TypeId.DECIMAL32:
+        return cudf.Decimal32Dtype(
+            precision=cudf.Decimal32Dtype.MAX_PRECISION,
+            scale=-type_.scale()
+        )
+    elif tid == pylibcudf.TypeId.DECIMAL128:
+        return cudf.Decimal128Dtype(
+            precision=cudf.Decimal128Dtype.MAX_PRECISION,
+            scale=-type_.scale()
+        )
+    else:
+        return PYLIBCUDF_TO_SUPPORTED_NUMPY_TYPES[
+            <underlying_type_t_type_id>(tid)
+        ]

From 38149f82d6bc7871360c296c7775b615e037e159 Mon Sep 17 00:00:00 2001
From: Vyas Ramasubramani <vyasr@nvidia.com>
Date: Thu, 17 Aug 2023 10:45:13 -0700
Subject: [PATCH 2/6] Add accessor for list properties

---
 python/cudf/cudf/_lib/pylibcudf/column.pxd |  9 +++++++++
 python/cudf/cudf/_lib/pylibcudf/column.pyx | 19 ++++++++++++++++++-
 python/cudf/cudf/_lib/types.pyx            |  9 ++++-----
 3 files changed, 31 insertions(+), 6 deletions(-)

diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pxd b/python/cudf/cudf/_lib/pylibcudf/column.pxd
index 498714dc0c3..20d11350904 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pxd
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pxd
@@ -33,3 +33,12 @@ cdef class Column:
     cpdef DataType type(self) noexcept
     cpdef Column child(self, size_type index) noexcept
     cpdef size_type num_children(self) noexcept
+
+    cpdef list_view(self)
+
+
+cdef class ListColumnView:
+    """Accessor for methods of a Column that are specific to lists."""
+    cdef Column _column
+    cpdef child(self)
+    cpdef offsets(self)
diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pyx b/python/cudf/cudf/_lib/pylibcudf/column.pyx
index bfb010992c0..bb958bead9d 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pyx
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pyx
@@ -9,7 +9,7 @@ from cudf._lib.cpp.column.column cimport column, column_contents
 from cudf._lib.cpp.types cimport size_type
 
 from .gpumemoryview cimport gpumemoryview
-from .types cimport DataType
+from .types cimport DataType, TypeId
 from .utils cimport int_to_bitmask_ptr, int_to_void_ptr
 
 
@@ -157,3 +157,20 @@ cdef class Column:
     cpdef size_type num_children(self) noexcept:
         """The number of children of this column."""
         return self._num_children
+
+    cpdef list_view(self):
+        return ListColumnView(self)
+
+
+cdef class ListColumnView:
+    """Accessor for methods of a Column that are specific to lists."""
+    def __init__(self, Column col):
+        if col.type().id() != TypeId.LIST:
+            raise TypeError("Column is not a list type")
+        self._column = col
+
+    cpdef child(self):
+        return self._column.child(1)
+
+    cpdef offsets(self):
+        return self._column.child(1)
diff --git a/python/cudf/cudf/_lib/types.pyx b/python/cudf/cudf/_lib/types.pyx
index 283ba8a8e72..d6396157af4 100644
--- a/python/cudf/cudf/_lib/types.pyx
+++ b/python/cudf/cudf/_lib/types.pyx
@@ -283,13 +283,12 @@ cdef bool is_decimal_type_id(libcudf_types.type_id tid) except *:
 
 
 def dtype_from_pylibcudf_lists_column(pylibcudf.Column col):
-    # TODO: Currently hardcoding the child column index for lists, should come
-    # up with a cleaner solution here.
-    child = col.child(1)
+    child = col.list_view().child()
+    tid = child.type().id()
 
-    if child.type().id() == pylibcudf.TypeId.LIST:
+    if tid == pylibcudf.TypeId.LIST:
         return cudf.ListDtype(dtype_from_pylibcudf_lists_column(child))
-    elif child.type().id() == pylibcudf.TypeId.EMPTY:
+    elif tid == pylibcudf.TypeId.EMPTY:
         return cudf.ListDtype("int8")
     else:
         return cudf.ListDtype(

From 9f7bea6e2c9f10b371b8cffd836b054663e98ed3 Mon Sep 17 00:00:00 2001
From: Vyas Ramasubramani <vyasr@nvidia.com>
Date: Thu, 17 Aug 2023 11:43:46 -0700
Subject: [PATCH 3/6] Add method to access columns of a table

---
 python/cudf/cudf/_lib/pylibcudf/table.pxd | 4 +++-
 python/cudf/cudf/_lib/pylibcudf/table.pyx | 9 ++++++---
 python/cudf/cudf/_lib/utils.pyx           | 2 +-
 3 files changed, 10 insertions(+), 5 deletions(-)

diff --git a/python/cudf/cudf/_lib/pylibcudf/table.pxd b/python/cudf/cudf/_lib/pylibcudf/table.pxd
index 4f189f2c398..95f197b13eb 100644
--- a/python/cudf/cudf/_lib/pylibcudf/table.pxd
+++ b/python/cudf/cudf/_lib/pylibcudf/table.pxd
@@ -8,9 +8,11 @@ from cudf._lib.cpp.table.table_view cimport table_view
 
 cdef class Table:
     # List[pylibcudf.Column]
-    cdef object columns
+    cdef list _columns
 
     cdef table_view view(self) nogil
 
     @staticmethod
     cdef Table from_libcudf(unique_ptr[table] libcudf_tbl)
+
+    cpdef list columns(self)
diff --git a/python/cudf/cudf/_lib/pylibcudf/table.pyx b/python/cudf/cudf/_lib/pylibcudf/table.pyx
index db422dd420b..720f9815bd6 100644
--- a/python/cudf/cudf/_lib/pylibcudf/table.pyx
+++ b/python/cudf/cudf/_lib/pylibcudf/table.pyx
@@ -21,7 +21,7 @@ cdef class Table:
         The columns in this table.
     """
     def __init__(self, list columns):
-        self.columns = columns
+        self._columns = columns
 
     cdef table_view view(self) nogil:
         """Generate a libcudf table_view to pass to libcudf algorithms.
@@ -31,11 +31,11 @@ cdef class Table:
         (even direct pylibcudf Cython users).
         """
         # TODO: Make c_columns a class attribute that is updated along with
-        # self.columns whenever new columns are added or columns are removed.
+        # self._columns whenever new columns are added or columns are removed.
         cdef vector[column_view] c_columns
 
         with gil:
-            for col in self.columns:
+            for col in self._columns:
                 c_columns.push_back((<Column> col).view())
 
         return table_view(c_columns)
@@ -57,3 +57,6 @@ cdef class Table:
             Column.from_libcudf(move(c_columns[i]))
             for i in range(c_columns.size())
         ])
+
+    cpdef list columns(self):
+        return self._columns
diff --git a/python/cudf/cudf/_lib/utils.pyx b/python/cudf/cudf/_lib/utils.pyx
index 8907143c289..4815c705958 100644
--- a/python/cudf/cudf/_lib/utils.pyx
+++ b/python/cudf/cudf/_lib/utils.pyx
@@ -261,7 +261,7 @@ cdef columns_from_pylibcudf_table(pylibcudf.Table tbl):
         A list of columns.
     """
     cdef pylibcudf.Column plc
-    return [Column.from_pylibcudf(plc) for plc in tbl.columns]
+    return [Column.from_pylibcudf(plc) for plc in tbl.columns()]
 
 
 cdef data_from_unique_ptr(

From 984a67a6c512b1dbecc61e196b4acd86b4b79822 Mon Sep 17 00:00:00 2001
From: Vyas Ramasubramani <vyasr@nvidia.com>
Date: Thu, 17 Aug 2023 11:56:45 -0700
Subject: [PATCH 4/6] Make pylibcudf columns more opaque

---
 python/cudf/cudf/_lib/column.pyx           |  6 ++--
 python/cudf/cudf/_lib/pylibcudf/column.pxd | 16 +++++----
 python/cudf/cudf/_lib/pylibcudf/column.pyx | 38 +++++++++++++---------
 3 files changed, 35 insertions(+), 25 deletions(-)

diff --git a/python/cudf/cudf/_lib/column.pyx b/python/cudf/cudf/_lib/column.pyx
index 50b56f1d484..36a5ccd9140 100644
--- a/python/cudf/cudf/_lib/column.pyx
+++ b/python/cudf/cudf/_lib/column.pyx
@@ -611,10 +611,12 @@ cdef class Column:
         dtype = dtype_from_pylibcudf_column(col)
 
         return cudf.core.column.build_column(
-            data=as_buffer(col.data.obj) if col.data is not None else None,
+            data=as_buffer(col.data().obj) if col.data() is not None else None,
             dtype=dtype,
             size=col.size,
-            mask=as_buffer(col.mask.obj) if col.mask is not None else None,
+            mask=as_buffer(
+                col.null_mask().obj
+            ) if col.null_mask() is not None else None,
             offset=col.offset,
             null_count=col.null_count,
             children=tuple([
diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pxd b/python/cudf/cudf/_lib/pylibcudf/column.pxd
index 20d11350904..6d7d9f4019d 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pxd
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pxd
@@ -15,14 +15,14 @@ cdef class Column:
     # TODO: Should we document these attributes? Should we mark them readonly?
     cdef:
         # Core data
-        DataType data_type
-        size_type size
-        gpumemoryview data
-        gpumemoryview mask
-        size_type null_count
-        size_type offset
+        DataType _data_type
+        size_type _size
+        gpumemoryview _data
+        gpumemoryview _mask
+        size_type _null_count
+        size_type _offset
         # children: List[Column]
-        list children
+        list _children
         size_type _num_children
 
     cdef column_view view(self) nogil
@@ -33,6 +33,8 @@ cdef class Column:
     cpdef DataType type(self) noexcept
     cpdef Column child(self, size_type index) noexcept
     cpdef size_type num_children(self) noexcept
+    cpdef gpumemoryview data(self)
+    cpdef gpumemoryview null_mask(self)
 
     cpdef list_view(self)
 
diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pyx b/python/cudf/cudf/_lib/pylibcudf/column.pyx
index bb958bead9d..07c8d162afc 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pyx
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pyx
@@ -45,13 +45,13 @@ cdef class Column:
         gpumemoryview mask, size_type null_count, size_type offset,
         list children
     ):
-        self.data_type = data_type
-        self.size = size
-        self.data = data
-        self.mask = mask
-        self.null_count = null_count
-        self.offset = offset
-        self.children = children
+        self._data_type = data_type
+        self._size = size
+        self._data = data
+        self._mask = mask
+        self._null_count = null_count
+        self._offset = offset
+        self._children = children
         self._num_children = len(children)
 
     cdef column_view view(self) nogil:
@@ -64,16 +64,16 @@ cdef class Column:
         cdef const void * data = NULL
         cdef const bitmask_type * null_mask = NULL
 
-        if self.data is not None:
-            data = int_to_void_ptr(self.data.ptr)
-        if self.mask is not None:
-            null_mask = int_to_bitmask_ptr(self.mask.ptr)
+        if self._data is not None:
+            data = int_to_void_ptr(self._data.ptr)
+        if self._mask is not None:
+            null_mask = int_to_bitmask_ptr(self._mask.ptr)
 
         # TODO: Check if children can ever change. If not, this could be
         # computed once in the constructor and always be reused.
         cdef vector[column_view] c_children
         with gil:
-            if self.children is not None:
+            if self._children is not None:
                 for child in self.children:
                     # Need to cast to Column here so that Cython knows that
                     # `view` returns a typed object, not a Python object. We
@@ -87,8 +87,8 @@ cdef class Column:
                     c_children.push_back((<Column> child).view())
 
         return column_view(
-            self.data_type.c_obj, self.size, data, null_mask,
-            self.null_count, self.offset, c_children
+            self._data_type.c_obj, self._size, data, null_mask,
+            self._null_count, self._offset, c_children
         )
 
     @staticmethod
@@ -137,7 +137,7 @@ cdef class Column:
 
     cpdef DataType type(self):
         """The type of data in the column."""
-        return self.data_type
+        return self._data_type
 
     cpdef Column child(self, size_type index) noexcept:
         """Get a child column of this column.
@@ -152,7 +152,7 @@ cdef class Column:
         Column
             The child column.
         """
-        return self.children[index]
+        return self._children[index]
 
     cpdef size_type num_children(self) noexcept:
         """The number of children of this column."""
@@ -161,6 +161,12 @@ cdef class Column:
     cpdef list_view(self):
         return ListColumnView(self)
 
+    cpdef gpumemoryview data(self):
+        return self._data
+
+    cpdef gpumemoryview null_mask(self):
+        return self._mask
+
 
 cdef class ListColumnView:
     """Accessor for methods of a Column that are specific to lists."""

From c0816e9c8c7e84a861eacfa53ceb8442fd87995a Mon Sep 17 00:00:00 2001
From: Vyas Ramasubramani <vyasr@nvidia.com>
Date: Thu, 17 Aug 2023 11:58:03 -0700
Subject: [PATCH 5/6] Use pylibcudf only in pure Python mode

---
 python/cudf/cudf/_lib/column.pxd  |  3 +--
 python/cudf/cudf/_lib/column.pyx  | 13 +++++--------
 python/cudf/cudf/_lib/copying.pyx |  4 ++--
 python/cudf/cudf/_lib/types.pxd   |  3 +--
 python/cudf/cudf/_lib/types.pyx   |  9 ++++-----
 python/cudf/cudf/_lib/utils.pxd   |  3 +--
 python/cudf/cudf/_lib/utils.pyx   |  4 +---
 7 files changed, 15 insertions(+), 24 deletions(-)

diff --git a/python/cudf/cudf/_lib/column.pxd b/python/cudf/cudf/_lib/column.pxd
index fbdf6288538..7ffb55a6cc6 100644
--- a/python/cudf/cudf/_lib/column.pxd
+++ b/python/cudf/cudf/_lib/column.pxd
@@ -7,7 +7,6 @@ from libcpp.memory cimport unique_ptr
 
 from rmm._lib.device_buffer cimport device_buffer
 
-from cudf._lib cimport pylibcudf
 from cudf._lib.cpp.column.column cimport column
 from cudf._lib.cpp.column.column_view cimport column_view, mutable_column_view
 from cudf._lib.cpp.types cimport size_type
@@ -30,7 +29,7 @@ cdef class Column:
     cdef column_view _view(self, size_type null_count) except *
     cdef column_view view(self) except *
     cdef mutable_column_view mutable_view(self) except *
-    cpdef pylibcudf.Column to_pylibcudf(self, mode: Literal["read", "write"])
+    cpdef to_pylibcudf(self, mode: Literal["read", "write"])
 
     @staticmethod
     cdef Column from_unique_ptr(
diff --git a/python/cudf/cudf/_lib/column.pyx b/python/cudf/cudf/_lib/column.pyx
index 36a5ccd9140..a320f1c3425 100644
--- a/python/cudf/cudf/_lib/column.pyx
+++ b/python/cudf/cudf/_lib/column.pyx
@@ -11,9 +11,6 @@ import rmm
 import cudf
 import cudf._lib as libcudf
 from cudf._lib import pylibcudf
-
-from cudf._lib cimport pylibcudf
-
 from cudf.api.types import is_categorical_dtype, is_datetime64tz_dtype
 from cudf.core.buffer import (
     Buffer,
@@ -447,7 +444,7 @@ cdef class Column:
     # underlying buffers as exposed before this function can itself be exposed
     # publicly.  User requests to convert to pylibcudf must assume that the
     # data may be modified afterwards.
-    cpdef pylibcudf.Column to_pylibcudf(self, mode: Literal["read", "write"]):
+    cpdef to_pylibcudf(self, mode: Literal["read", "write"]):
         """Convert this Column to a pylibcudf.Column.
 
         This function will generate a pylibcudf Column pointing to the same
@@ -477,9 +474,9 @@ cdef class Column:
         else:
             col = self
 
-        cdef pylibcudf.DataType dtype = dtype_to_pylibcudf_type(col.dtype)
+        dtype = dtype_to_pylibcudf_type(col.dtype)
 
-        cdef pylibcudf.gpumemoryview data = None
+        data = None
         if col.base_data is not None:
             cai = cuda_array_interface_wrapper(
                 ptr=col.base_data.get_ptr(mode=mode),
@@ -488,7 +485,7 @@ cdef class Column:
             )
             data = pylibcudf.gpumemoryview(cai)
 
-        cdef pylibcudf.gpumemoryview mask = None
+        mask = None
         if self.nullable:
             # TODO: Are we intentionally use self's mask instead of col's?
             # Where is the mask stored for categoricals?
@@ -587,7 +584,7 @@ cdef class Column:
     #  TODO: Actually support exposed data pointers.
     @staticmethod
     def from_pylibcudf(
-        pylibcudf.Column col, bint data_ptr_exposed=False
+        col, bint data_ptr_exposed=False
     ):
         """Create a Column from a pylibcudf.Column.
 
diff --git a/python/cudf/cudf/_lib/copying.pyx b/python/cudf/cudf/_lib/copying.pyx
index 944a80158df..f57bc15ed57 100644
--- a/python/cudf/cudf/_lib/copying.pyx
+++ b/python/cudf/cudf/_lib/copying.pyx
@@ -11,9 +11,9 @@ from libcpp.vector cimport vector
 from rmm._lib.device_buffer cimport DeviceBuffer
 
 import cudf
+from cudf._lib import pylibcudf
 from cudf.core.buffer import Buffer, acquire_spill_lock, as_buffer
 
-from cudf._lib cimport pylibcudf
 from cudf._lib.column cimport Column
 
 from cudf._lib.scalar import as_device_scalar
@@ -174,7 +174,7 @@ def gather(
     Column gather_map,
     bool nullify=False
 ):
-    cdef pylibcudf.Table tbl = pylibcudf.copying.gather(
+    tbl = pylibcudf.copying.gather(
         pylibcudf.Table([col.to_pylibcudf(mode="read") for col in columns]),
         gather_map.to_pylibcudf(mode="read"),
         pylibcudf.copying.OutOfBoundsPolicy.NULLIFY if nullify
diff --git a/python/cudf/cudf/_lib/types.pxd b/python/cudf/cudf/_lib/types.pxd
index 1eeaa23c260..a95db84ceff 100644
--- a/python/cudf/cudf/_lib/types.pxd
+++ b/python/cudf/cudf/_lib/types.pxd
@@ -4,7 +4,6 @@ from libc.stdint cimport int32_t
 from libcpp cimport bool
 
 cimport cudf._lib.cpp.types as libcudf_types
-from cudf._lib cimport pylibcudf
 from cudf._lib.cpp.column.column_view cimport column_view
 from cudf._lib.cpp.lists.lists_column_view cimport lists_column_view
 
@@ -18,5 +17,5 @@ ctypedef bool underlying_type_t_null_policy
 cdef dtype_from_column_view(column_view cv)
 
 cdef libcudf_types.data_type dtype_to_data_type(dtype) except *
-cpdef pylibcudf.DataType dtype_to_pylibcudf_type(dtype)
+cpdef dtype_to_pylibcudf_type(dtype)
 cdef bool is_decimal_type_id(libcudf_types.type_id tid) except *
diff --git a/python/cudf/cudf/_lib/types.pyx b/python/cudf/cudf/_lib/types.pyx
index d6396157af4..8594e37ac4a 100644
--- a/python/cudf/cudf/_lib/types.pyx
+++ b/python/cudf/cudf/_lib/types.pyx
@@ -17,7 +17,6 @@ from cudf._lib.types cimport (
 
 import cudf
 from cudf._lib import pylibcudf
-from cudf._lib cimport pylibcudf
 
 size_type_dtype = np.dtype("int32")
 
@@ -257,7 +256,7 @@ cdef libcudf_types.data_type dtype_to_data_type(dtype) except *:
     else:
         return libcudf_types.data_type(tid)
 
-cpdef pylibcudf.DataType dtype_to_pylibcudf_type(dtype):
+cpdef dtype_to_pylibcudf_type(dtype):
     if cudf.api.types.is_list_dtype(dtype):
         return pylibcudf.DataType(pylibcudf.TypeId.LIST)
     elif cudf.api.types.is_struct_dtype(dtype):
@@ -282,7 +281,7 @@ cdef bool is_decimal_type_id(libcudf_types.type_id tid) except *:
     )
 
 
-def dtype_from_pylibcudf_lists_column(pylibcudf.Column col):
+def dtype_from_pylibcudf_lists_column(col):
     child = col.list_view().child()
     tid = child.type().id()
 
@@ -296,7 +295,7 @@ def dtype_from_pylibcudf_lists_column(pylibcudf.Column col):
         )
 
 
-def dtype_from_pylibcudf_structs_column(pylibcudf.Column col):
+def dtype_from_pylibcudf_structs_column(col):
     fields = {
         str(i): dtype_from_pylibcudf_column(col.child(i))
         for i in range(col.num_children())
@@ -304,7 +303,7 @@ def dtype_from_pylibcudf_structs_column(pylibcudf.Column col):
     return cudf.StructDtype(fields)
 
 
-def dtype_from_pylibcudf_column(pylibcudf.Column col):
+def dtype_from_pylibcudf_column(col):
     type_ = col.type()
     tid = type_.id()
 
diff --git a/python/cudf/cudf/_lib/utils.pxd b/python/cudf/cudf/_lib/utils.pxd
index f2cdc110b64..653fa8f2b8b 100644
--- a/python/cudf/cudf/_lib/utils.pxd
+++ b/python/cudf/cudf/_lib/utils.pxd
@@ -4,7 +4,6 @@ from libcpp.memory cimport unique_ptr
 from libcpp.string cimport string
 from libcpp.vector cimport vector
 
-from cudf._lib cimport pylibcudf
 from cudf._lib.cpp.column.column cimport column_view
 from cudf._lib.cpp.table.table cimport table, table_view
 
@@ -19,4 +18,4 @@ cdef table_view table_view_from_columns(columns) except *
 cdef table_view table_view_from_table(tbl, ignore_index=*) except*
 cdef columns_from_unique_ptr(unique_ptr[table] c_tbl)
 cdef columns_from_table_view(table_view tv, object owners)
-cdef columns_from_pylibcudf_table(pylibcudf.Table table)
+cdef columns_from_pylibcudf_table(tbl)
diff --git a/python/cudf/cudf/_lib/utils.pyx b/python/cudf/cudf/_lib/utils.pyx
index 4815c705958..03982a58517 100644
--- a/python/cudf/cudf/_lib/utils.pyx
+++ b/python/cudf/cudf/_lib/utils.pyx
@@ -11,7 +11,6 @@ from libcpp.string cimport string
 from libcpp.utility cimport move
 from libcpp.vector cimport vector
 
-from cudf._lib cimport pylibcudf
 from cudf._lib.column cimport Column
 from cudf._lib.cpp.column.column cimport column, column_view
 from cudf._lib.cpp.table.table cimport table
@@ -247,7 +246,7 @@ cdef columns_from_unique_ptr(
     return columns
 
 
-cdef columns_from_pylibcudf_table(pylibcudf.Table tbl):
+cdef columns_from_pylibcudf_table(tbl):
     """Convert a pylibcudf table into list of columns.
 
     Parameters
@@ -260,7 +259,6 @@ cdef columns_from_pylibcudf_table(pylibcudf.Table tbl):
     list[Column]
         A list of columns.
     """
-    cdef pylibcudf.Column plc
     return [Column.from_pylibcudf(plc) for plc in tbl.columns()]
 
 

From c18ced3f52884b0dfe4388d5279a896d84275c81 Mon Sep 17 00:00:00 2001
From: Vyas Ramasubramani <vyasr@nvidia.com>
Date: Thu, 17 Aug 2023 16:12:56 -0700
Subject: [PATCH 6/6] Fix remaining issues

---
 python/cudf/cudf/_lib/column.pyx                |  8 ++++----
 python/cudf/cudf/_lib/pylibcudf/column.pxd      | 10 +++++++---
 python/cudf/cudf/_lib/pylibcudf/column.pyx      | 17 ++++++++++++++---
 .../cudf/cudf/_lib/pylibcudf/gpumemoryview.pxd  |  7 +++++--
 4 files changed, 30 insertions(+), 12 deletions(-)

diff --git a/python/cudf/cudf/_lib/column.pyx b/python/cudf/cudf/_lib/column.pyx
index a320f1c3425..4db3761b1b8 100644
--- a/python/cudf/cudf/_lib/column.pyx
+++ b/python/cudf/cudf/_lib/column.pyx
@@ -610,15 +610,15 @@ cdef class Column:
         return cudf.core.column.build_column(
             data=as_buffer(col.data().obj) if col.data() is not None else None,
             dtype=dtype,
-            size=col.size,
+            size=col.size(),
             mask=as_buffer(
                 col.null_mask().obj
             ) if col.null_mask() is not None else None,
-            offset=col.offset,
-            null_count=col.null_count,
+            offset=col.offset(),
+            null_count=col.null_count(),
             children=tuple([
                 Column.from_pylibcudf(child)
-                for child in col.children
+                for child in col.children()
             ])
         )
 
diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pxd b/python/cudf/cudf/_lib/pylibcudf/column.pxd
index 6d7d9f4019d..2af87db5b03 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pxd
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pxd
@@ -30,11 +30,15 @@ cdef class Column:
     @staticmethod
     cdef Column from_libcudf(unique_ptr[column] libcudf_col)
 
-    cpdef DataType type(self) noexcept
-    cpdef Column child(self, size_type index) noexcept
-    cpdef size_type num_children(self) noexcept
+    cpdef DataType type(self)
+    cpdef Column child(self, size_type index)
+    cpdef size_type num_children(self)
+    cpdef size_type size(self)
+    cpdef size_type null_count(self)
+    cpdef size_type offset(self)
     cpdef gpumemoryview data(self)
     cpdef gpumemoryview null_mask(self)
+    cpdef list children(self)
 
     cpdef list_view(self)
 
diff --git a/python/cudf/cudf/_lib/pylibcudf/column.pyx b/python/cudf/cudf/_lib/pylibcudf/column.pyx
index 07c8d162afc..d9b2ca98ead 100644
--- a/python/cudf/cudf/_lib/pylibcudf/column.pyx
+++ b/python/cudf/cudf/_lib/pylibcudf/column.pyx
@@ -74,7 +74,7 @@ cdef class Column:
         cdef vector[column_view] c_children
         with gil:
             if self._children is not None:
-                for child in self.children:
+                for child in self._children:
                     # Need to cast to Column here so that Cython knows that
                     # `view` returns a typed object, not a Python object. We
                     # cannot use a typed variable for `child` because cdef
@@ -139,7 +139,7 @@ cdef class Column:
         """The type of data in the column."""
         return self._data_type
 
-    cpdef Column child(self, size_type index) noexcept:
+    cpdef Column child(self, size_type index):
         """Get a child column of this column.
 
         Parameters
@@ -154,7 +154,7 @@ cdef class Column:
         """
         return self._children[index]
 
-    cpdef size_type num_children(self) noexcept:
+    cpdef size_type num_children(self):
         """The number of children of this column."""
         return self._num_children
 
@@ -167,6 +167,17 @@ cdef class Column:
     cpdef gpumemoryview null_mask(self):
         return self._mask
 
+    cpdef size_type size(self):
+        return self._size
+
+    cpdef size_type offset(self):
+        return self._offset
+
+    cpdef size_type null_count(self):
+        return self._null_count
+
+    cpdef list children(self):
+        return self._children
 
 cdef class ListColumnView:
     """Accessor for methods of a Column that are specific to lists."""
diff --git a/python/cudf/cudf/_lib/pylibcudf/gpumemoryview.pxd b/python/cudf/cudf/_lib/pylibcudf/gpumemoryview.pxd
index 5e656744a8c..713697bd139 100644
--- a/python/cudf/cudf/_lib/pylibcudf/gpumemoryview.pxd
+++ b/python/cudf/cudf/_lib/pylibcudf/gpumemoryview.pxd
@@ -2,5 +2,8 @@
 
 
 cdef class gpumemoryview:
-    cdef Py_ssize_t ptr
-    cdef object obj
+    # TODO: Eventually probably want to make this opaque, but for now it's fine
+    # to treat this object as something like a POD struct
+    cdef readonly:
+        Py_ssize_t ptr
+        object obj