vyasr commented on code in PR #406:
URL: https://github.com/apache/arrow-nanoarrow/pull/406#discussion_r1534426874


##########
examples/cmake-scenarios/src/library.cpp:
##########
@@ -0,0 +1,39 @@
+// Licensed to the Apache Software Foundation (ASF) under one
+// or more contributor license agreements.  See the NOTICE file
+// distributed with this work for additional information
+// regarding copyright ownership.  The ASF licenses this file
+// to you under the Apache License, Version 2.0 (the
+// "License"); you may not use this file except in compliance
+// with the License.  You may obtain a copy of the License at
+//
+//   http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing,
+// software distributed under the License is distributed on an
+// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, either express or implied.  See the License for the
+// specific language governing permissions and limitations
+// under the License.
+
+#include "nanoarrow/nanoarrow.h"
+#include "nanoarrow/nanoarrow.hpp"
+
+#include "library.hpp"
+
+std::optional<std::pair<std::unique_ptr<ArrowArray>, 
std::unique_ptr<ArrowSchema>>>
+make_simple_array() {
+  nanoarrow::UniqueArray tmp_array;
+  nanoarrow::UniqueSchema tmp_schema;
+  auto result = make_simple_array(tmp_array.get(), tmp_schema.get());
+  if (result != 0) {
+    if (tmp_array.get()->release) tmp_array.get()->release(tmp_array.get());
+    if (tmp_schema.get()->release) tmp_schema.get()->release(tmp_schema.get());
+    return std::nullopt;
+  }
+  auto ret_array = std::make_unique<ArrowArray>();

Review Comment:
   OK yeah so in that example the library interface is a C++ interface, so it's 
fine to return a `UniqueSchema`. That makes sense, I was just saying that I had 
written my example with the assumption that you'd typically want to code this 
kind of thing with nanoarrow assuming that the other library could be a C 
library consuming it.
   
   I guess the core issue here is that I wrote an example that's a very odd 
hybrid of C and C++. I made some assumptions that we'd end up interfacing with 
a C library, but at the moment the functions use C++ interfaces. That's also 
clear from my second question about lifetimes. I think what you're getting at 
is that because I'm using a `unique_ptr` to the array, you'd need to set a 
custom deleter on that object so that when the pointer goes out of scope the 
right deleter is called, which is completely true. I was just assuming that in 
practice this object would need to be released and the underlying raw pointer 
would need to be passed back, at which point the real consumer would be in 
charge of calling the release callback as per the C data interface contract, 
but you're totally right that as the code is written now there's a potential 
memory leak because I'm handing back a `unique_ptr` that could go out of scope 
and delete the array without actually freeing its underlying buffers.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to