(arrow-nanoarrow) branch main updated: fix: Reject oversized binary view values (#947)

[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 ecba4756 fix: Reject oversized binary view values (#947)
ecba4756 is described below

commit ecba475603d82755856fd1b89bbaaa7fa43d8393
Author: Efe <gokdemirefe1...@gmail.com>
AuthorDate: Tue Sep 29 05:08:27 2026 +0300

    fix: Reject oversized binary view values (#947)
    
    ## Summary
    
    - Reject binary and string view values larger than `INT32_MAX` before
    narrowing the length into the Arrow binary-view representation.
    - Add a regression test covering the overflow path without allocating a
    multi-gigabyte buffer.
    
    Fixes #938
    
    ## Testing
    
    - `cmake -S . -B /tmp/arrow-nanoarrow-build-938
    -DNANOARROW_BUILD_TESTS=ON` — passed.
    - `cmake --build /tmp/arrow-nanoarrow-build-938 --parallel 2` — passed.
    - `ctest --test-dir /tmp/arrow-nanoarrow-build-938 --output-on-failure`
    — passed (287/287; 3 thread-safety tests skipped by project
    configuration).
    - `git diff --check` — passed.
    
    AI assistance was used during implementation; the submitted changes were
    reviewed and validated against the issue and the project test suite.
    
    Signed-off-by: Efe Gökdemir <e...@rexcode.co.uk>
---
 src/nanoarrow/common/array_test.cc  | 13 +++++++++++++
 src/nanoarrow/common/inline_array.h |  4 ++++
 2 files changed, 17 insertions(+)

diff --git a/src/nanoarrow/common/array_test.cc 
b/src/nanoarrow/common/array_test.cc
index 02550a7e..cd9ac88e 100644
--- a/src/nanoarrow/common/array_test.cc
+++ b/src/nanoarrow/common/array_test.cc
@@ -1267,6 +1267,19 @@ TEST(ArrayTest, ArrayTestAppendToBinaryArrayErrors) {
   ArrowArrayRelease(&array);
 }
 
+TEST(ArrayTest, ArrayTestAppendToBinaryViewArrayErrors) {
+  struct ArrowArray array;
+
+  ASSERT_EQ(ArrowArrayInitFromType(&array, NANOARROW_TYPE_BINARY_VIEW), 
NANOARROW_OK);
+  EXPECT_EQ(ArrowArrayStartAppending(&array), NANOARROW_OK);
+  struct ArrowBufferView item;
+  item.data.as_char = "";
+  item.size_bytes = static_cast<int64_t>(INT32_MAX) + 1;
+  EXPECT_EQ(ArrowArrayAppendBytes(&array, item), EOVERFLOW);
+
+  ArrowArrayRelease(&array);
+}
+
 TEST(ArrayTest, ArrayTestAppendToIntervalArrayYearMonth) {
   struct ArrowArray array;
 
diff --git a/src/nanoarrow/common/inline_array.h 
b/src/nanoarrow/common/inline_array.h
index 9bfc6dbd..eaec776e 100644
--- a/src/nanoarrow/common/inline_array.h
+++ b/src/nanoarrow/common/inline_array.h
@@ -584,6 +584,10 @@ static inline ArrowErrorCode ArrowArrayAppendBytes(struct 
ArrowArray* array,
 
   if (private_data->storage_type == NANOARROW_TYPE_STRING_VIEW ||
       private_data->storage_type == NANOARROW_TYPE_BINARY_VIEW) {
+    if (value.size_bytes > INT32_MAX) {
+      return EOVERFLOW;
+    }
+
     struct ArrowBuffer* data_buffer = ArrowArrayBuffer(array, 1);
     union ArrowBinaryView bvt;
     bvt.inlined.size = (int32_t)value.size_bytes;