wgtmac commented on code in PR #801:
URL: https://github.com/apache/iceberg-cpp/pull/801#discussion_r3789239382
##########
src/iceberg/inspect/metadata_table.h:
##########
@@ -20,46 +20,108 @@
#pragma once
/// \file iceberg/inspect/metadata_table.h
-/// \brief Define base APIs for metadata tables.
+/// \brief Base APIs for inspecting Iceberg metadata tables.
+#include <concepts>
#include <memory>
+#include <string>
+#include <utility>
+#include <variant>
+#include "iceberg/arrow_c_data.h"
#include "iceberg/iceberg_export.h"
#include "iceberg/result.h"
-#include "iceberg/table_identifier.h"
#include "iceberg/type_fwd.h"
+#include "iceberg/util/timepoint.h"
namespace iceberg {
-/// \brief Base class for Iceberg metadata tables.
+/// \brief Base interface for an Iceberg metadata table.
class ICEBERG_EXPORT MetadataTable {
public:
+ /// \brief Supported metadata table kinds.
enum class Kind {
kSnapshots,
kHistory,
};
- static Result<std::unique_ptr<MetadataTable>> Make(std::shared_ptr<Table>
table,
- Kind kind);
+ /// \brief Maximum number of rows emitted in each Arrow batch.
+ static constexpr int64_t kBatchSize = 1024;
+
+ /// \brief Create a metadata table of the requested concrete type.
+ ///
+ /// \tparam MetadataTableType Concrete class derived from MetadataTable.
+ /// \param table Source table whose metadata will be exposed.
+ /// \return The constructed metadata table, or an error.
+ template <typename MetadataTableType>
+ requires std::derived_from<MetadataTableType, MetadataTable>
+ static Result<std::unique_ptr<MetadataTableType>>
Make(std::shared_ptr<Table> table) {
+ return MetadataTableType::Make(std::move(table));
+ }
virtual ~MetadataTable();
+ /// \brief Return this metadata table's kind.
virtual Kind kind() const noexcept = 0;
- const TableIdentifier& name() const { return identifier_; }
+ /// \brief Return the schema of rows emitted by scans.
+ virtual const std::shared_ptr<Schema>& schema() const = 0;
+
+ /// \brief Return the source table whose metadata is exposed.
+ const std::shared_ptr<Table>& source_table() const;
- const std::shared_ptr<Schema>& schema() const { return schema_; }
+ /// \brief Return whether this metadata table supports time travel.
+ virtual bool supports_time_travel() const noexcept;
- const std::shared_ptr<Table>& source_table() const { return source_table_; }
+ /// \brief Scan the metadata table without time travel.
+ ///
+ /// The caller owns the returned stream and must release it with
+ /// ArrowArrayStreamRelease.
+ virtual Result<ArrowArrayStream> Scan() = 0;
protected:
- explicit MetadataTable(std::shared_ptr<Table> source_table, TableIdentifier
identifier,
- std::shared_ptr<Schema> schema);
+ explicit MetadataTable(std::shared_ptr<Table> source_table);
private:
- TableIdentifier identifier_;
- std::shared_ptr<Schema> schema_;
std::shared_ptr<Table> source_table_;
};
+/// \brief Snapshot selection parameters for a time-travel scan.
+struct SnapshotSelection {
+ /// \brief Select the current snapshot, a snapshot ID, or an as-of timestamp.
+ ///
+ /// std::monostate selects the current snapshot.
+ std::variant<std::monostate, int64_t, TimePointMs> snapshot;
+
+ /// \brief Resolve the snapshot relative to this branch or tag.
+ ///
+ /// An empty string uses the main branch.
+ std::string ref_name;
+};
+
+/// \brief Base interface for metadata tables that support time travel.
+class ICEBERG_EXPORT TimeTravelMetadataTable : public MetadataTable {
+ public:
+ ~TimeTravelMetadataTable() override;
+
+ /// \brief Return true because this interface supports time travel.
+ bool supports_time_travel() const noexcept final;
+
+ /// \brief Scan using the current snapshot on the main branch.
+ Result<ArrowArrayStream> Scan() final;
+
+ /// \brief Scan using the requested snapshot selection.
+ ///
+ /// \param snapshot_selection Snapshot ID, timestamp, and optional ref
selection.
+ /// \return An Arrow stream containing the metadata table rows, or an error.
+ Result<ArrowArrayStream> Scan(const SnapshotSelection& snapshot_selection);
Review Comment:
Is it better to remove this extra layer of `TimeTravelMetadataTable`? We can
check `supports_time_travel()` and by default `Scan(const SnapshotSelection&
snapshot_selection)` can just return `NotSupported`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
