feat: Refresh method support for table

Author: shuxu.li

src/iceberg/table.cc

@@ -21,16 +21,36 @@
 
 #include <algorithm>
 
+#include "iceberg/catalog.h"
 #include "iceberg/partition_spec.h"
 #include "iceberg/schema.h"
 #include "iceberg/sort_order.h"
 #include "iceberg/table_metadata.h"
 #include "iceberg/table_scan.h"
+#include "iceberg/util/macros.h"
 
 namespace iceberg {
 
 const std::string& Table::uuid() const { return metadata_->table_uuid; }
 
+Status Table::Refresh() {
+  if (!catalog_) {
+    return InvalidArgument("Cannot refresh table metadata without a catalog");
+  }
+
+  ICEBERG_ASSIGN_OR_RAISE(auto fresh, catalog_->LoadTable(identifier_));
+  if (metadata_location_ != fresh->metadata_location_) {
+    metadata_ = std::move(fresh->metadata_);
+    metadata_location_ = std::move(fresh->metadata_location_);
+    io_ = std::move(fresh->io_);
+
+    schemas_map_.reset();
+    partition_spec_map_.reset();
+    sort_orders_map_.reset();
+  }
+  return {};
+}
+
 Result<std::shared_ptr<Schema>> Table::schema() const { return metadata_->Schema(); }
 
 const std::shared_ptr<std::unordered_map<int32_t, std::shared_ptr<Schema>>>&

src/iceberg/table.h

@@ -57,6 +57,9 @@ class ICEBERG_EXPORT Table {
   /// \brief Returns the UUID of the table
   const std::string& uuid() const;
 
+  /// \brief Refresh the current table metadata
+  Status Refresh();
+
   /// \brief Return the schema for this table, return NotFoundError if not found
   Result<std::shared_ptr<Schema>> schema() const;
 
@@ -114,9 +117,11 @@ class ICEBERG_EXPORT Table {
   const std::shared_ptr<FileIO>& io() const;
 
  private:
+  friend class Table;
+
   const TableIdentifier identifier_;
   std::shared_ptr<TableMetadata> metadata_;
-  const std::string metadata_location_;
+  std::string metadata_location_;
   std::shared_ptr<FileIO> io_;
   std::shared_ptr<Catalog> catalog_;
 

test/in_memory_catalog_test.cc

@@ -115,6 +115,30 @@ TEST_F(InMemoryCatalogTest, RegisterTable) {
   ASSERT_EQ(table.value()->location(), "s3://bucket/test/location");
 }
 
+TEST_F(InMemoryCatalogTest, RefreshTable) {
+  TableIdentifier tableIdent{.ns = {}, .name = "t1"};
+
+  std::unique_ptr<TableMetadata> metadata;
+  ASSERT_NO_FATAL_FAILURE(ReadTableMetadata("TableMetadataV2Valid.json", &metadata));
+
+  auto table_location = GenerateTestTableLocation(tableIdent.name);
+  auto metadata_location = std::format("{}v0.metadata.json", table_location);
+  auto status = TableMetadataUtil::Write(*file_io_, metadata_location, *metadata);
+  EXPECT_THAT(status, IsOk());
+
+  auto table = catalog_->RegisterTable(tableIdent, metadata_location);
+  EXPECT_THAT(table, IsOk());
+  ASSERT_TRUE(table.value()->current_snapshot().has_value());
+  ASSERT_EQ(table.value()->current_snapshot().value()->snapshot_id, 3055729675574597004);
+
+  // Now we don't support commit method in catalog, so here only test refresh with the
+  // same version
+  status = table.value()->Refresh();
+  EXPECT_THAT(status, IsOk());
+  ASSERT_TRUE(table.value()->current_snapshot().has_value());
+  ASSERT_EQ(table.value()->current_snapshot().value()->snapshot_id, 3055729675574597004);
+}
+
 TEST_F(InMemoryCatalogTest, DropTable) {
   TableIdentifier tableIdent{.ns = {}, .name = "t1"};
   auto result = catalog_->DropTable(tableIdent, false);