Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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
26 changes: 21 additions & 5 deletions src/iceberg/json_internal.cc
Original file line number Diff line number Diff line change
Expand Up @@ -206,6 +206,19 @@ Result<std::unique_ptr<SortField>> SortFieldFromJson(const nlohmann::json& json)
null_order);
}

Result<std::unique_ptr<SortOrder>> SortOrderFromJson(
const nlohmann::json& json, const std::shared_ptr<Schema>& current_schema) {
ICEBERG_ASSIGN_OR_RAISE(auto order_id, GetJsonValue<int32_t>(json, kOrderId));
ICEBERG_ASSIGN_OR_RAISE(auto fields, GetJsonValue<nlohmann::json>(json, kFields));

std::vector<SortField> sort_fields;
for (const auto& field_json : fields) {
ICEBERG_ASSIGN_OR_RAISE(auto sort_field, SortFieldFromJson(field_json));
sort_fields.push_back(std::move(*sort_field));
}
return SortOrder::Make(*current_schema, order_id, std::move(sort_fields));
}

Result<std::unique_ptr<SortOrder>> SortOrderFromJson(const nlohmann::json& json) {
ICEBERG_ASSIGN_OR_RAISE(auto order_id, GetJsonValue<int32_t>(json, kOrderId));
ICEBERG_ASSIGN_OR_RAISE(auto fields, GetJsonValue<nlohmann::json>(json, kFields));
Expand All @@ -215,7 +228,7 @@ Result<std::unique_ptr<SortOrder>> SortOrderFromJson(const nlohmann::json& json)
ICEBERG_ASSIGN_OR_RAISE(auto sort_field, SortFieldFromJson(field_json));
sort_fields.push_back(std::move(*sort_field));
}
return std::make_unique<SortOrder>(order_id, std::move(sort_fields));
return SortOrder::Make(order_id, std::move(sort_fields));
}

nlohmann::json ToJson(const SchemaField& field) {
Expand Down Expand Up @@ -919,9 +932,11 @@ Status ParsePartitionSpecs(const nlohmann::json& json, int8_t format_version,
///
/// \param[in] json The JSON object to parse.
/// \param[in] format_version The format version of the table.
/// \param[in] current_schema The current schema.
/// \param[out] default_sort_order_id The default sort order ID.
/// \param[out] sort_orders The list of sort orders.
Status ParseSortOrders(const nlohmann::json& json, int8_t format_version,
const std::shared_ptr<Schema>& current_schema,
int32_t& default_sort_order_id,
std::vector<std::shared_ptr<SortOrder>>& sort_orders) {
if (json.contains(kSortOrders)) {
Expand All @@ -930,7 +945,8 @@ Status ParseSortOrders(const nlohmann::json& json, int8_t format_version,
ICEBERG_ASSIGN_OR_RAISE(auto sort_order_array,
GetJsonValue<nlohmann::json>(json, kSortOrders));
for (const auto& sort_order_json : sort_order_array) {
ICEBERG_ASSIGN_OR_RAISE(auto sort_order, SortOrderFromJson(sort_order_json));
ICEBERG_ASSIGN_OR_RAISE(auto sort_order,
SortOrderFromJson(sort_order_json, current_schema));
sort_orders.push_back(std::move(sort_order));
}
} else {
Expand Down Expand Up @@ -1005,9 +1021,9 @@ Result<std::unique_ptr<TableMetadata>> TableMetadataFromJson(const nlohmann::jso
}
}

ICEBERG_RETURN_UNEXPECTED(ParseSortOrders(json, table_metadata->format_version,
table_metadata->default_sort_order_id,
table_metadata->sort_orders));
ICEBERG_RETURN_UNEXPECTED(ParseSortOrders(
json, table_metadata->format_version, current_schema,
table_metadata->default_sort_order_id, table_metadata->sort_orders));

if (json.contains(kProperties)) {
ICEBERG_ASSIGN_OR_RAISE(table_metadata->properties, FromJsonMap(json, kProperties));
Expand Down
3 changes: 2 additions & 1 deletion src/iceberg/json_internal.h
Original file line number Diff line number Diff line change
Expand Up @@ -69,10 +69,11 @@ ICEBERG_EXPORT nlohmann::json ToJson(const SortOrder& sort_order);
/// Each `SortField` will be parsed using the `SortFieldFromJson` function.
///
/// \param json The JSON object representing a `SortOrder`.
/// \param current_schema The current schema associated with the sort order.
/// \return An `expected` value containing either a `SortOrder` object or an error. If the
/// JSON is malformed or missing expected fields, an error will be returned.
ICEBERG_EXPORT Result<std::unique_ptr<SortOrder>> SortOrderFromJson(
const nlohmann::json& json);
const nlohmann::json& json, const std::shared_ptr<Schema>& current_schema);

/// \brief Convert an Iceberg Schema to JSON.
///
Expand Down
9 changes: 5 additions & 4 deletions src/iceberg/sort_order.cc
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@
#include "iceberg/sort_order.h"

#include <format>
#include <memory>
#include <optional>
#include <ranges>

Expand All @@ -36,8 +37,8 @@ SortOrder::SortOrder(int32_t order_id, std::vector<SortField> fields)
: order_id_(order_id), fields_(std::move(fields)) {}

const std::shared_ptr<SortOrder>& SortOrder::Unsorted() {
static const std::shared_ptr<SortOrder> unsorted =
std::make_shared<SortOrder>(kUnsortedOrderId, std::vector<SortField>{});
static const std::shared_ptr<SortOrder> unsorted = std::shared_ptr<SortOrder>(
new SortOrder(kUnsortedOrderId, std::vector<SortField>{}));
return unsorted;
}

Expand Down Expand Up @@ -113,7 +114,7 @@ Result<std::unique_ptr<SortOrder>> SortOrder::Make(const Schema& schema, int32_t
return InvalidArgument("Sort order must have at least one sort field");
}

auto sort_order = std::make_unique<SortOrder>(sort_id, std::move(fields));
auto sort_order = std::unique_ptr<SortOrder>(new SortOrder(sort_id, std::move(fields)));
ICEBERG_RETURN_UNEXPECTED(sort_order->Validate(schema));
return sort_order;
}
Expand All @@ -128,7 +129,7 @@ Result<std::unique_ptr<SortOrder>> SortOrder::Make(int32_t sort_id,
return InvalidArgument("Sort order must have at least one sort field");
}

return std::make_unique<SortOrder>(sort_id, std::move(fields));
return std::unique_ptr<SortOrder>(new SortOrder(sort_id, std::move(fields)));
}

} // namespace iceberg
8 changes: 6 additions & 2 deletions src/iceberg/sort_order.h
Original file line number Diff line number Diff line change
Expand Up @@ -41,8 +41,6 @@ class ICEBERG_EXPORT SortOrder : public util::Formattable {
static constexpr int32_t kUnsortedOrderId = 0;
static constexpr int32_t kInitialSortOrderId = 1;

SortOrder(int32_t order_id, std::vector<SortField> fields);

/// \brief Get an unsorted sort order singleton.
static const std::shared_ptr<SortOrder>& Unsorted();

Expand Down Expand Up @@ -95,6 +93,12 @@ class ICEBERG_EXPORT SortOrder : public util::Formattable {
std::vector<SortField> fields);

private:
/// \brief Constructs a SortOrder instance.
/// \param order_id The sort order id.
/// \param fields The sort fields.
/// \note Use the static Make methods to create SortOrder instances.
SortOrder(int32_t order_id, std::vector<SortField> fields);

/// \brief Compare two sort orders for equality.
bool Equals(const SortOrder& other) const;

Expand Down
20 changes: 12 additions & 8 deletions src/iceberg/test/json_internal_test.cc
Original file line number Diff line number Diff line change
Expand Up @@ -49,11 +49,6 @@ Result<std::unique_ptr<SortField>> FromJsonHelper(const nlohmann::json& json) {
return SortFieldFromJson(json);
}

template <>
Result<std::unique_ptr<SortOrder>> FromJsonHelper(const nlohmann::json& json) {
return SortOrderFromJson(json);
}

template <>
Result<std::unique_ptr<PartitionField>> FromJsonHelper(const nlohmann::json& json) {
return PartitionFieldFromJson(json);
Expand Down Expand Up @@ -107,17 +102,26 @@ TEST(JsonInternalTest, SortField) {
}

TEST(JsonInternalTest, SortOrder) {
auto schema = std::make_shared<Schema>(
std::vector<SchemaField>{SchemaField(5, "region", iceberg::string(), false),
SchemaField(7, "ts", iceberg::int64(), false)},
/*schema_id=*/100);
auto identity_transform = Transform::Identity();
SortField st_ts(5, identity_transform, SortDirection::kAscending, NullOrder::kFirst);
SortField st_bar(7, identity_transform, SortDirection::kDescending, NullOrder::kLast);
SortOrder sort_order(100, {st_ts, st_bar});

ICEBERG_UNWRAP_OR_FAIL(auto sort_order, SortOrder::Make(100, {st_ts, st_bar}));
EXPECT_TRUE(sort_order->Validate(*schema));
nlohmann::json expected_sort_order =
R"({"order-id":100,"fields":[
{"transform":"identity","source-id":5,"direction":"asc","null-order":"nulls-first"},
{"transform":"identity","source-id":7,"direction":"desc","null-order":"nulls-last"}]})"_json;

TestJsonConversion(sort_order, expected_sort_order);
auto json = ToJson(*sort_order);
EXPECT_EQ(expected_sort_order, json) << "JSON conversion mismatch.";

// Specialize FromJson based on type (T)
ICEBERG_UNWRAP_OR_FAIL(auto obj_ex, SortOrderFromJson(expected_sort_order, schema));
EXPECT_EQ(*sort_order, *obj_ex) << "Deserialized object mismatch.";
}

TEST(JsonInternalTest, PartitionField) {
Expand Down
28 changes: 19 additions & 9 deletions src/iceberg/test/metadata_serde_test.cc
Original file line number Diff line number Diff line change
Expand Up @@ -148,12 +148,17 @@ TEST(MetadataSerdeTest, DeserializeV2Valid) {
std::vector<PartitionField>{PartitionField(/*source_id=*/1, /*field_id=*/1000, "x",
Transform::Identity())});

auto expected_sort_order = std::make_shared<SortOrder>(
/*order_id=*/3,
std::vector<SortField>{SortField(/*source_id=*/2, Transform::Identity(),
SortDirection::kAscending, NullOrder::kFirst),
SortField(/*source_id=*/3, Transform::Bucket(4),
SortDirection::kDescending, NullOrder::kLast)});
ICEBERG_UNWRAP_OR_FAIL(
auto sort_order,
SortOrder::Make(*expected_schema_2,
/*order_id=*/3,
std::vector<SortField>{
SortField(/*source_id=*/2, Transform::Identity(),
SortDirection::kAscending, NullOrder::kFirst),
SortField(/*source_id=*/3, Transform::Bucket(4),
SortDirection::kDescending, NullOrder::kLast)}));

auto expected_sort_order = std::shared_ptr<SortOrder>(std::move(sort_order));

auto expected_snapshot_1 = std::make_shared<Snapshot>(Snapshot{
.snapshot_id = 3051729675574597004,
Expand Down Expand Up @@ -228,13 +233,18 @@ TEST(MetadataSerdeTest, DeserializeV2ValidMinimal) {
std::vector<PartitionField>{PartitionField(/*source_id=*/1, /*field_id=*/1000, "x",
Transform::Identity())});

auto expected_sort_order = std::make_shared<SortOrder>(
/*order_id=*/3, std::vector<SortField>{
ICEBERG_UNWRAP_OR_FAIL(
auto sort_order,
SortOrder::Make(*expected_schema,
/*order_id=*/3,
std::vector<SortField>{
SortField(/*source_id=*/2, Transform::Identity(),
SortDirection::kAscending, NullOrder::kFirst),
SortField(/*source_id=*/3, Transform::Bucket(4),
SortDirection::kDescending, NullOrder::kLast),
});
}));

auto expected_sort_order = std::shared_ptr<SortOrder>(std::move(sort_order));

TableMetadata expected{
.format_version = 2,
Expand Down
Loading
Loading