apacheGH-44006: [GLib][Parquet] Add `gparquet_arrow_file_writer_new_r…

…ow_group()` (apache#44039) ### Rationale for this change This is a low-level API to control how to write data. This is for advanced users. ### What changes are included in this PR? `gparquet_arrow_file_writer_write_chunked_array()` is also added to write a test for `gparquet_arrow_file_writer_new_row_group()`. ### Are these changes tested? Yes. ### Are there any user-facing changes? Yes. * GitHub Issue: apache#44006 Authored-by: Sutou Kouhei <[email protected]> Signed-off-by: Sutou Kouhei <[email protected]>
vibhatha · Sep 11, 2024 · 8556001 · 8556001
1 parent 2a793d6
commit 8556001
Show file tree

Hide file tree

Showing 3 changed files with 90 additions and 4 deletions.
diff --git a/c_glib/parquet-glib/arrow-file-writer.cpp b/c_glib/parquet-glib/arrow-file-writer.cpp
@@ -548,13 +548,57 @@ gparquet_arrow_file_writer_write_record_batch(GParquetArrowFileWriter *writer,
 gboolean
 gparquet_arrow_file_writer_write_table(GParquetArrowFileWriter *writer,
                                        GArrowTable *table,
-                                       guint64 chunk_size,
+                                       gsize chunk_size,
                                        GError **error)
 {
   auto parquet_arrow_file_writer = gparquet_arrow_file_writer_get_raw(writer);
   auto arrow_table = garrow_table_get_raw(table).get();
-  auto status = parquet_arrow_file_writer->WriteTable(*arrow_table, chunk_size);
-  return garrow_error_check(error, status, "[parquet][arrow][file-writer][write-table]");
+  return garrow::check(error,
+                       parquet_arrow_file_writer->WriteTable(*arrow_table, chunk_size),
+                       "[parquet][arrow][file-writer][write-table]");
+}
+
+/**
+ * gparquet_arrow_file_writer_new_row_group:
+ * @writer: A #GParquetArrowFileWriter.
+ * @chunk_size: The max number of rows in a row group.
+ * @error: (nullable): Return location for a #GError or %NULL.
+ *
+ * Returns: %TRUE on success, %FALSE if there was an error.
+ *
+ * Since: 18.0.0
+ */
+gboolean
+gparquet_arrow_file_writer_new_row_group(GParquetArrowFileWriter *writer,
+                                         gsize chunk_size,
+                                         GError **error)
+{
+  auto parquet_arrow_file_writer = gparquet_arrow_file_writer_get_raw(writer);
+  return garrow::check(error,
+                       parquet_arrow_file_writer->NewRowGroup(chunk_size),
+                       "[parquet][arrow][file-writer][new-row-group]");
+}
+
+/**
+ * gparquet_arrow_file_writer_write_chunked_array:
+ * @writer: A #GParquetArrowFileWriter.
+ * @chunked_array: A #GArrowChunkedArray to be written.
+ * @error: (nullable): Return location for a #GError or %NULL.
+ *
+ * Returns: %TRUE on success, %FALSE if there was an error.
+ *
+ * Since: 18.0.0
+ */
+gboolean
+gparquet_arrow_file_writer_write_chunked_array(GParquetArrowFileWriter *writer,
+                                               GArrowChunkedArray *chunked_array,
+                                               GError **error)
+{
+  auto parquet_arrow_file_writer = gparquet_arrow_file_writer_get_raw(writer);
+  auto arrow_chunked_array = garrow_chunked_array_get_raw(chunked_array);
+  return garrow::check(error,
+                       parquet_arrow_file_writer->WriteColumnChunk(arrow_chunked_array),
+                       "[parquet][arrow][file-writer][write-chunked-array]");
 }
 
 /**

diff --git a/c_glib/parquet-glib/arrow-file-writer.h b/c_glib/parquet-glib/arrow-file-writer.h
@@ -130,9 +130,21 @@ GPARQUET_AVAILABLE_IN_0_11
 gboolean
 gparquet_arrow_file_writer_write_table(GParquetArrowFileWriter *writer,
                                        GArrowTable *table,
-                                       guint64 chunk_size,
+                                       gsize chunk_size,
                                        GError **error);
 
+GPARQUET_AVAILABLE_IN_18_0
+gboolean
+gparquet_arrow_file_writer_new_row_group(GParquetArrowFileWriter *writer,
+                                         gsize chunk_size,
+                                         GError **error);
+
+GPARQUET_AVAILABLE_IN_18_0
+gboolean
+gparquet_arrow_file_writer_write_chunked_array(GParquetArrowFileWriter *writer,
+                                               GArrowChunkedArray *chunked_array,
+                                               GError **error);
+
 GPARQUET_AVAILABLE_IN_0_11
 gboolean
 gparquet_arrow_file_writer_close(GParquetArrowFileWriter *writer, GError **error);

diff --git a/c_glib/test/parquet/test-arrow-file-writer.rb b/c_glib/test/parquet/test-arrow-file-writer.rb
@@ -82,4 +82,34 @@ def test_write_table
       reader.unref
     end
   end
+
+  def test_write_chunked_array
+    schema = build_schema("enabled" => :boolean)
+    writer = Parquet::ArrowFileWriter.new(schema, @file.path)
+    writer.new_row_group(2)
+    chunked_array = Arrow::ChunkedArray.new([build_boolean_array([true, nil])])
+    writer.write_chunked_array(chunked_array)
+    writer.new_row_group(1)
+    chunked_array = Arrow::ChunkedArray.new([build_boolean_array([false])])
+    writer.write_chunked_array(chunked_array)
+    writer.close
+
+    reader = Parquet::ArrowFileReader.new(@file.path)
+    begin
+      reader.use_threads = true
+      assert_equal([
+                     2,
+                     build_table("enabled" => [
+                                   build_boolean_array([true, nil]),
+                                   build_boolean_array([false]),
+                                 ]),
+                   ],
+                   [
+                     reader.n_row_groups,
+                     reader.read_table,
+                   ])
+    ensure
+      reader.unref
+    end
+  end
 end