Remove cudf._lib.avro in favor of inlining pylicudf (#17319)

Contributes to #17317

Authors:
  - Matthew Roeschke (https://github.com/mroeschke)

Approvers:
  - Vyas Ramasubramani (https://github.com/vyasr)

URL: #17319

python/cudf/cudf/_lib/CMakeLists.txt

@@ -14,7 +14,6 @@
 
 set(cython_sources
     aggregation.pyx
-    avro.pyx
     binaryop.pyx
     column.pyx
     concat.pyx

python/cudf/cudf/_lib/__init__.py

@@ -2,7 +2,6 @@
 import numpy as np
 
 from . import (
-    avro,
     binaryop,
     concat,
     copying,

python/cudf/cudf/_lib/utils.pxd

@@ -11,7 +11,7 @@ from pylibcudf.libcudf.table.table cimport table, table_view
 cdef data_from_unique_ptr(
     unique_ptr[table] c_tbl, column_names, index_names=*)
 cdef data_from_pylibcudf_table(tbl, column_names, index_names=*)
-cdef data_from_pylibcudf_io(tbl_with_meta, column_names = *, index_names = *)
+cpdef data_from_pylibcudf_io(tbl_with_meta, column_names = *, index_names = *)
 cdef data_from_table_view(
     table_view tv, object owner, object column_names, object index_names=*)
 cdef table_view table_view_from_columns(columns) except *

python/cudf/cudf/_lib/utils.pyx

@@ -316,7 +316,7 @@ cdef data_from_pylibcudf_table(tbl, column_names, index_names=None):
         index_names
     )
 
-cdef data_from_pylibcudf_io(tbl_with_meta, column_names=None, index_names=None):
+cpdef data_from_pylibcudf_io(tbl_with_meta, column_names=None, index_names=None):
     """
     Unpacks the TableWithMetadata from libcudf I/O
     into a dict of columns and an Index (cuDF format)

python/cudf/cudf/io/avro.py

@@ -1,7 +1,9 @@
 # Copyright (c) 2019-2024, NVIDIA CORPORATION.
 
+import pylibcudf as plc
+
 import cudf
-from cudf import _lib as libcudf
+from cudf._lib.utils import data_from_pylibcudf_io
 from cudf.utils import ioutils
 
 
@@ -23,8 +25,19 @@ def read_avro(
         filepath_or_buffer, "read_avro"
     )
 
-    return cudf.DataFrame._from_data(
-        *libcudf.avro.read_avro(
-            filepath_or_buffer, columns, skiprows, num_rows
-        )
+    num_rows = -1 if num_rows is None else num_rows
+    skip_rows = 0 if skiprows is None else skiprows
+
+    if not isinstance(num_rows, int) or num_rows < -1:
+        raise TypeError("num_rows must be an int >= -1")
+    if not isinstance(skip_rows, int) or skip_rows < 0:
+        raise TypeError("skip_rows must be an int >= 0")
+
+    plc_result = plc.io.avro.read_avro(
+        plc.io.types.SourceInfo([filepath_or_buffer]),
+        columns,
+        skip_rows,
+        num_rows,
     )
+
+    return cudf.DataFrame._from_data(*data_from_pylibcudf_io(plc_result))