[arrow-nanoarrow] branch main updated: [C] Make sure the data buffer is non-null (#45)

[paleolimbot]

This is an automated email from the ASF dual-hosted git repository.

paleolimbot pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/arrow-nanoarrow.git


The following commit(s) were added to refs/heads/main by this push:
     new f218a6c  [C] Make sure the data buffer is non-null (#45)
f218a6c is described below

commit f218a6c2b0fdb92fc5ba26d811ce9f6b3d8d9e3e
Author: Dewey Dunnington <de...@fishandwhistle.net>
AuthorDate: Tue Sep 13 12:01:35 2022 -0300

    [C] Make sure the data buffer is non-null (#45)
    
    * make sure data buffer is non-null
    
    * free array in test
---
 src/nanoarrow/array.c       | 29 +++++++++++++++++++++++++++++
 src/nanoarrow/array_test.cc | 10 ++++++++++
 2 files changed, 39 insertions(+)

diff --git a/src/nanoarrow/array.c b/src/nanoarrow/array.c
index a84d6cc..05b50ed 100644
--- a/src/nanoarrow/array.c
+++ b/src/nanoarrow/array.c
@@ -347,6 +347,32 @@ ArrowErrorCode ArrowArrayReserve(struct ArrowArray* array,
   return NANOARROW_OK;
 }
 
+static ArrowErrorCode ArrowArrayFinalizeBuffers(struct ArrowArray* array) {
+  struct ArrowArrayPrivateData* private_data =
+      (struct ArrowArrayPrivateData*)array->private_data;
+
+  // The only buffer finalizing this currently does is make sure the data
+  // buffer for (Large)String|Binary is never NULL
+  switch (private_data->storage_type) {
+    case NANOARROW_TYPE_BINARY:
+    case NANOARROW_TYPE_STRING:
+    case NANOARROW_TYPE_LARGE_BINARY:
+    case NANOARROW_TYPE_LARGE_STRING:
+      if (ArrowArrayBuffer(array, 2)->data == NULL) {
+        ArrowBufferAppendUInt8(ArrowArrayBuffer(array, 2), 0);
+      }
+      break;
+    default:
+      break;
+  }
+
+  for (int64_t i = 0; i < array->n_children; i++) {
+    NANOARROW_RETURN_NOT_OK(ArrowArrayFinalizeBuffers(array->children[i]));
+  }
+
+  return NANOARROW_OK;
+}
+
 static void ArrowArrayFlushInternalPointers(struct ArrowArray* array) {
   struct ArrowArrayPrivateData* private_data =
       (struct ArrowArrayPrivateData*)array->private_data;
@@ -395,6 +421,9 @@ static ArrowErrorCode ArrowArrayCheckInternalBufferSizes(
 
 ArrowErrorCode ArrowArrayFinishBuilding(struct ArrowArray* array,
                                         struct ArrowError* error) {
+  // Even if the data buffer is size zero, the value needs to be non-null
+  NANOARROW_RETURN_NOT_OK(ArrowArrayFinalizeBuffers(array));
+
   // Make sure the value we get with array->buffers[i] is set to the actual
   // pointer (which may have changed from the original due to reallocation)
   ArrowArrayFlushInternalPointers(array);
diff --git a/src/nanoarrow/array_test.cc b/src/nanoarrow/array_test.cc
index e66fb31..1383401 100644
--- a/src/nanoarrow/array_test.cc
+++ b/src/nanoarrow/array_test.cc
@@ -363,6 +363,16 @@ TEST(ArrayTest, ArrayTestAppendToStringArray) {
       ArrayFromJSON(utf8(), "[\"1234\", null, null, \"56789\"]")));
 }
 
+TEST(ArrayTest, ArrayTestAppendEmptyToString) {
+  struct ArrowArray array;
+  ASSERT_EQ(ArrowArrayInit(&array, NANOARROW_TYPE_STRING), NANOARROW_OK);
+  ASSERT_EQ(ArrowArrayStartAppending(&array), NANOARROW_OK);
+  ASSERT_EQ(ArrowArrayAppendString(&array, ArrowCharView("")), NANOARROW_OK);
+  EXPECT_EQ(ArrowArrayFinishBuilding(&array, nullptr), NANOARROW_OK);
+  EXPECT_NE(array.buffers[2], nullptr);
+  array.release(&array);
+}
+
 TEST(ArrayTest, ArrayTestAppendToUInt64Array) {
   struct ArrowArray array;