Merge branch 'branch-25.02' into fea/plc/io/streams

python/cudf/cudf/io/orc.py

@@ -240,15 +240,27 @@ def read_orc(
         elif not isinstance(num_rows, int) or num_rows < -1:
             raise TypeError("num_rows must be an int >= -1")
 
-        tbl_w_meta = plc.io.orc.read_orc(
-            plc.io.SourceInfo(filepaths_or_buffers),
-            columns,
-            stripes,
-            skiprows,
-            num_rows,
-            use_index,
-            dtype_to_pylibcudf_type(cudf.dtype(timestamp_type)),
+        options = (
+            plc.io.orc.OrcReaderOptions.builder(
+                plc.io.types.SourceInfo(filepaths_or_buffers)
+            )
+            .use_index(use_index)
+            .build()
         )
+        if num_rows >= 0:
+            options.set_num_rows(num_rows)
+        if skiprows >= 0:
+            options.set_skip_rows(skiprows)
+        if stripes is not None and len(stripes) > 0:
+            options.set_stripes(stripes)
+        if timestamp_type is not None:
+            options.set_timestamp_type(
+                dtype_to_pylibcudf_type(cudf.dtype(timestamp_type))
+            )
+        if columns is not None and len(columns) > 0:
+            options.set_columns(columns)
+
+        tbl_w_meta = plc.io.orc.read_orc(options)
 
         if isinstance(columns, list) and len(columns) == 0:
             # When `columns=[]`, index needs to be

python/pylibcudf/pylibcudf/io/orc.pxd

@@ -1,5 +1,5 @@
 # Copyright (c) 2024, NVIDIA CORPORATION.
-from libc.stdint cimport uint64_t
+from libc.stdint cimport uint64_t, int64_t
 from libcpp cimport bool
 from libcpp.optional cimport optional
 from libcpp.string cimport string
@@ -19,6 +19,8 @@ from pylibcudf.libcudf.io.orc_metadata cimport (
 )
 from pylibcudf.libcudf.io.orc cimport (
     orc_chunked_writer,
+    orc_reader_options,
+    orc_reader_options_builder,
     orc_writer_options,
     orc_writer_options_builder,
     chunked_orc_writer_options,
@@ -34,18 +36,23 @@ from pylibcudf.libcudf.io.types cimport (
 from rmm._cuda.stream cimport Stream
 
 
-cpdef TableWithMetadata read_orc(
-    SourceInfo source_info,
-    list columns = *,
-    list stripes = *,
-    size_type skip_rows = *,
-    size_type nrows = *,
-    bool use_index = *,
-    bool use_np_dtypes = *,
-    DataType timestamp_type = *,
-    list decimal128_columns = *,
-    Stream stream = *,
-)
+cdef class OrcReaderOptions:
+    cdef orc_reader_options c_obj
+    cdef SourceInfo source
+    cpdef void set_num_rows(self, int64_t nrows)
+    cpdef void set_skip_rows(self, int64_t skip_rows)
+    cpdef void set_stripes(self, list stripes)
+    cpdef void set_decimal128_columns(self, list val)
+    cpdef void set_timestamp_type(self, DataType type_)
+    cpdef void set_columns(self, list col_names)
+
+cdef class OrcReaderOptionsBuilder:
+    cdef orc_reader_options_builder c_obj
+    cdef SourceInfo source
+    cpdef OrcReaderOptionsBuilder use_index(self, bool use)
+    cpdef OrcReaderOptions build(self)
+
+cpdef TableWithMetadata read_orc(OrcReaderOptions options)
 
 cdef class OrcColumnStatistics:
     cdef optional[uint64_t] number_of_values_c

python/pylibcudf/pylibcudf/io/orc.pyi

@@ -1,6 +1,8 @@
 # Copyright (c) 2024, NVIDIA CORPORATION.
 
-from typing import Any, Self
+from typing import Any
+
+from typing_extensions import Self
 
 from rmm._cuda.stream import Stream
 
@@ -13,20 +15,21 @@ from pylibcudf.io.types import (
     TableWithMetadata,
 )
 from pylibcudf.table import Table
-from pylibcudf.types import DataType
 
-def read_orc(
-    source_info: SourceInfo,
-    columns: list[str] | None = None,
-    stripes: list[list[int]] | None = None,
-    skip_rows: int = 0,
-    nrows: int = -1,
-    use_index: bool = True,
-    use_np_dtypes: bool = True,
-    timestamp_type: DataType | None = None,
-    decimal128_columns: list[str] | None = None,
-    stream: Stream = Stream(),
-) -> TableWithMetadata: ...
+class OrcReaderOptions:
+    def set_num_rows(self, nrows: int) -> None: ...
+    def set_skip_rows(self, skip_rows: int) -> None: ...
+    def set_stripes(self, stripes: list[list[int]]) -> None: ...
+    def set_decimal128_columns(self, val: list[str]) -> None: ...
+    def set_columns(self, col_names: list[str]) -> None: ...
+    @staticmethod
+    def builder(source: SourceInfo) -> OrcReaderOptionsBuilder: ...
+
+class OrcReaderOptionsBuilder:
+    def use_index(self, use: bool) -> Self: ...
+    def build(self) -> OrcReaderOptions: ...
+
+def read_orc(options: OrcReaderOptions) -> TableWithMetadata: ...
 
 class OrcColumnStatistics:
     def __init__(self): ...