Skip to content

fix(arrow): release schema after conversion failure #862

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
WZhuo wants to merge 1 commit into
base: main
Choose a base branch
from
+22 −3
Open

fix(arrow): release schema after conversion failure #862

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions src/iceberg/arrow_c_data_guard_internal.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,6 +46,12 @@ class ICEBERG_EXPORT ArrowSchemaGuard {
explicit ArrowSchemaGuard(ArrowSchema* schema) : schema_(schema) {}
~ArrowSchemaGuard();

/// \brief Release the guard without calling ArrowSchemaRelease.
///
/// Call this when ownership of the underlying ArrowSchema has been
/// transferred elsewhere and the guard should not release it.
void Release() { schema_ = nullptr; }

private:
ArrowSchema* schema_;
};
Expand Down
6 changes: 4 additions & 2 deletions src/iceberg/schema_internal.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,6 +25,7 @@
#include <optional>
#include <string>

#include "iceberg/arrow_c_data_guard_internal.h"
#include "iceberg/constants.h"
#include "iceberg/schema.h"
#include "iceberg/type.h"
Expand Down Expand Up @@ -71,6 +72,7 @@ ArrowErrorCode ToArrowSchema(const Type& type, bool optional, std::string_view n
std::optional<int32_t> field_id, ArrowSchema* schema) {
ArrowBuffer metadata_buffer;
NANOARROW_RETURN_NOT_OK(ArrowMetadataBuilderInit(&metadata_buffer, nullptr));
internal::ArrowArrayBufferGuard metadata_buffer_guard(&metadata_buffer);
if (field_id.has_value()) {
NANOARROW_RETURN_NOT_OK(ArrowMetadataBuilderAppend(
&metadata_buffer, ArrowCharView(std::string(kParquetFieldIdKey).c_str()),
Expand Down Expand Up @@ -183,7 +185,6 @@ ArrowErrorCode ToArrowSchema(const Type& type, bool optional, std::string_view n
case TypeId::kVariant:
case TypeId::kGeometry:
case TypeId::kGeography:
ArrowBufferReset(&metadata_buffer);
return EINVAL;
}

Expand All @@ -193,7 +194,6 @@ ArrowErrorCode ToArrowSchema(const Type& type, bool optional, std::string_view n

NANOARROW_RETURN_NOT_OK(ArrowSchemaSetMetadata(
schema, reinterpret_cast<const char*>(metadata_buffer.data)));
ArrowBufferReset(&metadata_buffer);

if (optional) {
schema->flags |= ARROW_FLAG_NULLABLE;
Expand All @@ -214,6 +214,7 @@ Status ToArrowSchema(const Schema& schema, ArrowSchema* out) {
ICEBERG_RETURN_UNEXPECTED(CheckArrowCompatible(schema));

ArrowSchemaInit(out);
internal::ArrowSchemaGuard schema_guard(out);

if (ArrowErrorCode errorCode = ToArrowSchema(schema, /*optional=*/false, /*name=*/"",
/*field_id=*/std::nullopt, out);
Expand All @@ -222,6 +223,7 @@ Status ToArrowSchema(const Schema& schema, ArrowSchema* out) {
"Failed to convert Iceberg schema to Arrow schema, error code: {}", errorCode);
}

schema_guard.Release();
return {};
}

Expand Down
13 changes: 12 additions & 1 deletion src/iceberg/test/arrow_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -130,12 +130,23 @@ TEST(ToArrowSchemaTest, UnsupportedV3Types) {
Schema schema(
{SchemaField::MakeOptional(/*field_id=*/1, "unsupported", unsupported_type)},
/*schema_id=*/0);
ArrowSchema arrow_schema;
ArrowSchema arrow_schema{};
ASSERT_THAT(ToArrowSchema(schema, &arrow_schema),
HasErrorMessage("is not supported by Arrow conversion"));
EXPECT_EQ(arrow_schema.release, nullptr);
}
}

TEST(ToArrowSchemaTest, ReleasesSchemaOnConversionFailure) {
// fixed(0) passes CheckArrowCompatible but nanoarrow rejects its non-positive width.
Schema schema({SchemaField::MakeOptional(/*field_id=*/1, "invalid_fixed", fixed(0))},
/*schema_id=*/0);
ArrowSchema arrow_schema{};

EXPECT_THAT(ToArrowSchema(schema, &arrow_schema), IsError(ErrorKind::kInvalidSchema));
EXPECT_EQ(arrow_schema.release, nullptr);
}

namespace {

void CheckArrowField(const ::arrow::Field& field, ::arrow::Type::type type_id,
Expand Down
Loading