ci: fix memory leak and enable sanitizer ci

Author: wgtmac

.github/workflows/sanitizer_test.yml

@@ -50,9 +50,9 @@ jobs:
       - name: Run Tests
         working-directory: build
         env:
-          ASAN_OPTIONS: log_path=out.log:detect_leaks=1:symbolize=1:strict_string_checks=1:halt_on_error=0:detect_container_overflow=0
+          ASAN_OPTIONS: log_path=out.log:detect_leaks=1:symbolize=1:strict_string_checks=1:halt_on_error=1:detect_container_overflow=0
           LSAN_OPTIONS: suppressions=${{ github.workspace }}/.github/lsan-suppressions.txt
-          UBSAN_OPTIONS: log_path=out.log:halt_on_error=0:print_stacktrace=1:suppressions=${{ github.workspace }}/.github/ubsan-suppressions.txt
+          UBSAN_OPTIONS: log_path=out.log:halt_on_error=1:print_stacktrace=1:suppressions=${{ github.workspace }}/.github/ubsan-suppressions.txt
         run: |
           ctest --output-on-failure
       - name: Save the test output

test/arrow_test.cc

@@ -35,6 +35,42 @@
 
 namespace iceberg {
 
+namespace {
+
+// A RAII guard for ArrowSchema.
+struct ArrowSchemaGuard {
+  ArrowSchema* schema;
+
+  ~ArrowSchemaGuard() {
+    if (schema != nullptr) {
+      ReleaseRecursively(schema);
+    }
+  }
+
+ private:
+  void ReleaseRecursively(ArrowSchema* schema_ptr) {
+    if (schema_ptr == nullptr) {
+      return;
+    }
+
+    // First, recursively release all children
+    if (schema_ptr->children != nullptr && schema_ptr->n_children > 0) {
+      for (int64_t i = 0; i < schema_ptr->n_children; ++i) {
+        if (schema_ptr->children[i] != nullptr) {
+          ReleaseRecursively(schema_ptr->children[i]);
+        }
+      }
+    }
+
+    // Then release this schema if it has a release callback
+    if (schema_ptr->release != nullptr) {
+      schema_ptr->release(schema_ptr);
+    }
+  }
+};
+
+}  // namespace
+
 TEST(ArrowCDataTest, CheckArrowSchemaAndArrayByNanoarrow) {
   auto [schema, array] = internal::CreateExampleArrowSchemaAndArrayByNanoarrow();
 
@@ -86,7 +122,8 @@ TEST_P(ToArrowSchemaTest, PrimitiveType) {
                       : SchemaField::MakeRequired(kFieldId, std::string(kFieldName),
                                                   param.iceberg_type)},
       /*schema_id=*/0);
-  ArrowSchema arrow_schema;
+  ArrowSchema arrow_schema{};
+  ArrowSchemaGuard guard{&arrow_schema};
   ASSERT_THAT(ToArrowSchema(schema, &arrow_schema), IsOk());
 
   auto imported_schema = ::arrow::ImportSchema(&arrow_schema).ValueOrDie();
@@ -170,7 +207,8 @@ TEST(ToArrowSchemaTest, StructType) {
                                            struct_type)},
                 /*schema_id=*/0);
 
-  ArrowSchema arrow_schema;
+  ArrowSchema arrow_schema{};
+  ArrowSchemaGuard guard{&arrow_schema};
   ASSERT_THAT(ToArrowSchema(schema, &arrow_schema), IsOk());
 
   auto imported_schema = ::arrow::ImportSchema(&arrow_schema).ValueOrDie();
@@ -202,7 +240,8 @@ TEST(ToArrowSchemaTest, ListType) {
       {SchemaField::MakeRequired(kListFieldId, std::string(kListFieldName), list_type)},
       /*schema_id=*/0);
 
-  ArrowSchema arrow_schema;
+  ArrowSchema arrow_schema{};
+  ArrowSchemaGuard guard{&arrow_schema};
   ASSERT_THAT(ToArrowSchema(schema, &arrow_schema), IsOk());
 
   auto imported_schema = ::arrow::ImportSchema(&arrow_schema).ValueOrDie();
@@ -237,7 +276,8 @@ TEST(ToArrowSchemaTest, MapType) {
       {SchemaField::MakeRequired(kFieldId, std::string(kMapFieldName), map_type)},
       /*schema_id=*/0);
 
-  ArrowSchema arrow_schema;
+  ArrowSchema arrow_schema{};
+  ArrowSchemaGuard guard{&arrow_schema};
   ASSERT_THAT(ToArrowSchema(schema, &arrow_schema), IsOk());
 
   auto imported_schema = ::arrow::ImportSchema(&arrow_schema).ValueOrDie();
@@ -278,7 +318,8 @@ TEST_P(FromArrowSchemaTest, PrimitiveType) {
           {std::string(kFieldIdKey), std::to_string(kFieldId)}});
   auto arrow_schema = ::arrow::schema({::arrow::field(
       std::string(kFieldName), param.arrow_type, param.optional, std::move(metadata))});
-  ArrowSchema exported_schema;
+  ArrowSchema exported_schema{};
+  ArrowSchemaGuard guard{&exported_schema};
   ASSERT_TRUE(::arrow::ExportSchema(*arrow_schema, &exported_schema).ok());
 
   auto type_result = FromArrowSchema(exported_schema, /*schema_id=*/1);
@@ -353,7 +394,8 @@ TEST(FromArrowSchemaTest, StructType) {
       ::arrow::key_value_metadata(std::unordered_map<std::string, std::string>{
           {std::string(kFieldIdKey), std::to_string(kStructFieldId)}}));
   auto arrow_schema = ::arrow::schema({struct_field});
-  ArrowSchema exported_schema;
+  ArrowSchema exported_schema{};
+  ArrowSchemaGuard guard{&exported_schema};
   ASSERT_TRUE(::arrow::ExportSchema(*arrow_schema, &exported_schema).ok());
 
   auto schema_result = FromArrowSchema(exported_schema, /*schema_id=*/0);
@@ -403,7 +445,8 @@ TEST(FromArrowSchemaTest, ListType) {
           {std::string(kFieldIdKey), std::to_string(kListFieldId)}}));
   auto arrow_schema = ::arrow::schema({list_field});
 
-  ArrowSchema exported_schema;
+  ArrowSchema exported_schema{};
+  ArrowSchemaGuard guard{&exported_schema};
   ASSERT_TRUE(::arrow::ExportSchema(*arrow_schema, &exported_schema).ok());
 
   auto schema_result = FromArrowSchema(exported_schema, /*schema_id=*/0);
@@ -453,7 +496,8 @@ TEST(FromArrowSchemaTest, MapType) {
           {std::string(kFieldIdKey), std::to_string(kFieldId)}}));
   auto arrow_schema = ::arrow::schema({map_field});
 
-  ArrowSchema exported_schema;
+  ArrowSchema exported_schema{};
+  ArrowSchemaGuard guard{&exported_schema};
   ASSERT_TRUE(::arrow::ExportSchema(*arrow_schema, &exported_schema).ok());
 
   auto schema_result = FromArrowSchema(exported_schema, /*schema_id=*/0);