Update vendored DuckDB sources to 3a3967aa81

Author: duckdblabs-bot

src/duckdb/src/catalog/catalog_entry/view_catalog_entry.cpp

@@ -17,14 +17,19 @@ namespace duckdb {
 void ViewCatalogEntry::Initialize(CreateViewInfo &info) {
 	query = std::move(info.query);
 	this->aliases = info.aliases;
-	if (!info.types.empty() && !info.names.empty()) {
+	if (!info.types.empty()) {
 		bind_state = ViewBindState::BOUND;
 		view_columns = make_shared_ptr<ViewColumnInfo>();
 		view_columns->types = info.types;
 		view_columns->names = info.names;
-		if (info.types.size() != info.names.size()) {
-			throw InternalException("Error creating view %s - view types / names size mismatch (%d types, %d names)",
-			                        name, info.types.size(), info.names.size());
+		if (view_columns->names.empty()) {
+			// DuckDB v0.9.2 and below store their names in the "aliases" field
+			view_columns->names = info.aliases;
+		}
+		if (view_columns->types.size() != view_columns->names.size()) {
+			throw InvalidInputException(
+			    "Error creating view %s - view types / names size mismatch (%d types, %d names)", name,
+			    view_columns->types.size(), view_columns->names.size());
 		}
 	}
 	this->temporary = info.temporary;

src/duckdb/src/common/enum_util.cpp

@@ -1169,25 +1169,6 @@ CompressedMaterializationDirection EnumUtil::FromString<CompressedMaterializatio
 	return static_cast<CompressedMaterializationDirection>(StringUtil::StringToEnum(GetCompressedMaterializationDirectionValues(), 3, "CompressedMaterializationDirection", value));
 }
 
-const StringUtil::EnumStringLiteral *GetCompressionFunctionSetLoadResultValues() {
-	static constexpr StringUtil::EnumStringLiteral values[] {
-		{ static_cast<uint32_t>(CompressionFunctionSetLoadResult::ALREADY_LOADED_BEFORE_LOCK), "ALREADY_LOADED_BEFORE_LOCK" },
-		{ static_cast<uint32_t>(CompressionFunctionSetLoadResult::ALREADY_LOADED_AFTER_LOCK), "ALREADY_LOADED_AFTER_LOCK" },
-		{ static_cast<uint32_t>(CompressionFunctionSetLoadResult::LAZILY_LOADED), "LAZILY_LOADED" }
-	};
-	return values;
-}
-
-template<>
-const char* EnumUtil::ToChars<CompressionFunctionSetLoadResult>(CompressionFunctionSetLoadResult value) {
-	return StringUtil::EnumToString(GetCompressionFunctionSetLoadResultValues(), 3, "CompressionFunctionSetLoadResult", static_cast<uint32_t>(value));
-}
-
-template<>
-CompressionFunctionSetLoadResult EnumUtil::FromString<CompressionFunctionSetLoadResult>(const char *value) {
-	return static_cast<CompressionFunctionSetLoadResult>(StringUtil::StringToEnum(GetCompressionFunctionSetLoadResultValues(), 3, "CompressionFunctionSetLoadResult", value));
-}
-
 const StringUtil::EnumStringLiteral *GetCompressionTypeValues() {
 	static constexpr StringUtil::EnumStringLiteral values[] {
 		{ static_cast<uint32_t>(CompressionType::COMPRESSION_AUTO), "AUTO" },

src/duckdb/src/common/multi_file/multi_file_column_mapper.cpp

@@ -1035,8 +1035,20 @@ static unique_ptr<TableFilter> TryCastTableFilter(const TableFilter &global_filt
 }
 
 void SetIndexToZero(unique_ptr<Expression> &root_expr) {
+#ifdef DEBUG
+	optional_idx index;
+	ExpressionIterator::VisitExpressionMutable<BoundReferenceExpression>(root_expr, [&](BoundReferenceExpression &ref,
+	                                                                                    unique_ptr<Expression> &expr) {
+		if (index.IsValid() && index.GetIndex() != ref.index) {
+			throw InternalException("Expected an expression that only references a single column, but found multiple!");
+		}
+		index = ref.index;
+		ref.index = 0;
+	});
+#else
 	ExpressionIterator::VisitExpressionMutable<BoundReferenceExpression>(
 	    root_expr, [&](BoundReferenceExpression &ref, unique_ptr<Expression> &expr) { ref.index = 0; });
+#endif
 }
 
 bool CanPropagateCast(const MultiFileIndexMapping &mapping, const LogicalType &local_type,

src/duckdb/src/execution/operator/join/physical_iejoin.cpp

@@ -733,6 +733,9 @@ idx_t IEJoinUnion::JoinComplexBlocks(unsafe_vector<idx_t> &lsel, unsafe_vector<i
 	// 8. initialize join result as an empty list for tuple pairs
 	idx_t result_count = 0;
 
+	lsel.resize(STANDARD_VECTOR_SIZE);
+	rsel.resize(STANDARD_VECTOR_SIZE);
+
 	// 11. for(i←1 to n) do
 	while (i < n) {
 		// 13. for (j ← pos+eqOff to n) do
@@ -776,6 +779,9 @@ idx_t IEJoinUnion::JoinComplexBlocks(unsafe_vector<idx_t> &lsel, unsafe_vector<i
 		}
 	}
 
+	lsel.resize(result_count);
+	rsel.resize(result_count);
+
 	return result_count;
 }
 

src/duckdb/src/execution/physical_plan/plan_get.cpp

@@ -102,10 +102,21 @@ PhysicalOperator &PhysicalPlanGenerator::CreatePlan(LogicalGet &op) {
 		vector<unique_ptr<Expression>> select_list;
 		unique_ptr<Expression> unsupported_filter;
 		unordered_set<idx_t> to_remove;
+
+		virtual_column_map_t virtual_columns;
+		if (op.function.get_virtual_columns) {
+			virtual_columns = op.function.get_virtual_columns(context, op.bind_data.get());
+		}
 		for (auto &entry : table_filters->filters) {
 			auto column_id = column_ids[entry.first].GetPrimaryIndex();
-			auto &type = op.returned_types[column_id];
 			if (!op.function.supports_pushdown_type(*op.bind_data, column_id)) {
+				LogicalType column_type;
+				if (IsVirtualColumn(column_id)) {
+					auto &column = virtual_columns.at(column_id);
+					column_type = column.type;
+				} else {
+					column_type = op.returned_types[column_id];
+				}
 				idx_t column_id_filter = entry.first;
 				bool found_projection = false;
 				for (idx_t i = 0; i < projection_ids.size(); i++) {
@@ -119,7 +130,7 @@ PhysicalOperator &PhysicalPlanGenerator::CreatePlan(LogicalGet &op) {
 					projection_ids.push_back(entry.first);
 					column_id_filter = projection_ids.size() - 1;
 				}
-				auto column = make_uniq<BoundReferenceExpression>(type, column_id_filter);
+				auto column = make_uniq<BoundReferenceExpression>(column_type, column_id_filter);
 				select_list.push_back(entry.second->ToExpression(*column));
 				to_remove.insert(entry.first);
 			}
@@ -132,7 +143,12 @@ PhysicalOperator &PhysicalPlanGenerator::CreatePlan(LogicalGet &op) {
 			vector<LogicalType> filter_types;
 			for (auto &c : projection_ids) {
 				auto column_id = column_ids[c].GetPrimaryIndex();
-				filter_types.push_back(op.returned_types[column_id]);
+				if (IsVirtualColumn(column_id)) {
+					auto &column = virtual_columns.at(column_id);
+					filter_types.push_back(column.type);
+				} else {
+					filter_types.push_back(op.returned_types[column_id]);
+				}
 			}
 			filter = Make<PhysicalFilter>(filter_types, std::move(select_list), op.estimated_cardinality);
 		}

src/duckdb/src/function/compression_config.cpp

@@ -84,11 +84,31 @@ idx_t CompressionFunctionSet::GetCompressionIndex(CompressionType type) {
 }
 
 CompressionFunctionSet::CompressionFunctionSet() {
-	for (idx_t i = 0; i < PHYSICAL_TYPE_COUNT; i++) {
-		is_loaded[i] = false;
-	}
 	ResetDisabledMethods();
 	functions.resize(PHYSICAL_TYPE_COUNT);
+
+	static PhysicalType physical_types[PHYSICAL_TYPE_COUNT] = {
+	    PhysicalType::BOOL,    PhysicalType::UINT8,    PhysicalType::INT8,   PhysicalType::UINT16, PhysicalType::INT16,
+	    PhysicalType::UINT32,  PhysicalType::INT32,    PhysicalType::UINT64, PhysicalType::INT64,  PhysicalType::FLOAT,
+	    PhysicalType::DOUBLE,  PhysicalType::INTERVAL, PhysicalType::LIST,   PhysicalType::STRUCT, PhysicalType::ARRAY,
+	    PhysicalType::VARCHAR, PhysicalType::UINT128,  PhysicalType::INT128, PhysicalType::BIT,
+	};
+	for (idx_t type_index = 0; type_index < PHYSICAL_TYPE_COUNT; type_index++) {
+		const auto physical_type = physical_types[type_index];
+		const auto index = GetCompressionIndex(physical_type);
+		auto &function_list = functions[index];
+
+		for (idx_t method_index = 0; internal_compression_methods[method_index].get_function; method_index++) {
+			auto &method = internal_compression_methods[method_index];
+			if (!method.supports_type(physical_type)) {
+				// Not supported for this type.
+				continue;
+			}
+			// The type is supported.
+			// We create the function and insert it into the set of available functions.
+			function_list.push_back(method.get_function(physical_type));
+		}
+	}
 }
 
 bool EmitCompressionFunction(CompressionType type) {
@@ -113,14 +133,12 @@ bool EmitCompressionFunction(CompressionType type) {
 
 vector<reference<const CompressionFunction>>
 CompressionFunctionSet::GetCompressionFunctions(PhysicalType physical_type) {
-	LoadCompressionFunctions(physical_type);
-	auto index = GetCompressionIndex(physical_type);
+	const auto index = GetCompressionIndex(physical_type);
 	auto &function_list = functions[index];
 	vector<reference<const CompressionFunction>> result;
 	for (auto &entry : function_list) {
-		auto compression_index = GetCompressionIndex(entry.type);
+		const auto compression_index = GetCompressionIndex(entry.type);
 		if (is_disabled[compression_index]) {
-			// explicitly disabled
 			continue;
 		}
 		if (!EmitCompressionFunction(entry.type)) {
@@ -131,50 +149,22 @@ CompressionFunctionSet::GetCompressionFunctions(PhysicalType physical_type) {
 	return result;
 }
 
-CompressionFunctionSetLoadResult CompressionFunctionSet::LoadCompressionFunctions(PhysicalType physical_type) {
-	auto index = GetCompressionIndex(physical_type);
-	if (is_loaded[index]) {
-		return CompressionFunctionSetLoadResult::ALREADY_LOADED_BEFORE_LOCK;
-	}
-	// not loaded - try to load it
-	lock_guard<mutex> guard(lock);
-	// verify nobody loaded it in the mean-time
-	if (is_loaded[index]) {
-		return CompressionFunctionSetLoadResult::ALREADY_LOADED_AFTER_LOCK;
-	}
-	// actually perform the load
-	auto &function_list = functions[index];
-	for (idx_t i = 0; internal_compression_methods[i].get_function; i++) {
-		auto &method = internal_compression_methods[i];
-		if (!method.supports_type(physical_type)) {
-			// not supported for this type
-			continue;
-		}
-		// The type is supported. We create the function and insert it into the set of available functions.
-		function_list.push_back(method.get_function(physical_type));
-	}
-	is_loaded[index] = true;
-	return CompressionFunctionSetLoadResult::LAZILY_LOADED;
-}
-
-pair<CompressionFunctionSetLoadResult, optional_ptr<const CompressionFunction>>
+optional_ptr<const CompressionFunction>
 CompressionFunctionSet::GetCompressionFunction(CompressionType type, const PhysicalType physical_type) {
-	const auto load_result = LoadCompressionFunctions(physical_type);
-
 	const auto index = GetCompressionIndex(physical_type);
 	const auto &function_list = functions[index];
 	for (auto &function : function_list) {
 		if (function.type == type) {
-			return make_pair(load_result, &function);
+			return &function;
 		}
 	}
-	return make_pair(load_result, nullptr);
+	return nullptr;
 }
 
 void CompressionFunctionSet::SetDisabledCompressionMethods(const vector<CompressionType> &methods) {
 	ResetDisabledMethods();
 	for (auto &method : methods) {
-		auto idx = GetCompressionIndex(method);
+		const auto idx = GetCompressionIndex(method);
 		is_disabled[idx] = true;
 	}
 }
@@ -193,55 +183,6 @@ vector<CompressionType> CompressionFunctionSet::GetDisabledCompressionMethods()
 	return result;
 }
 
-string CompressionFunctionSet::GetDebugInfo() const {
-	static PhysicalType physical_types[PHYSICAL_TYPE_COUNT] = {
-	    PhysicalType::BOOL,    PhysicalType::UINT8,    PhysicalType::INT8,   PhysicalType::UINT16, PhysicalType::INT16,
-	    PhysicalType::UINT32,  PhysicalType::INT32,    PhysicalType::UINT64, PhysicalType::INT64,  PhysicalType::FLOAT,
-	    PhysicalType::DOUBLE,  PhysicalType::INTERVAL, PhysicalType::LIST,   PhysicalType::STRUCT, PhysicalType::ARRAY,
-	    PhysicalType::VARCHAR, PhysicalType::UINT128,  PhysicalType::INT128, PhysicalType::BIT,
-	};
-
-	vector<string> compression_type_debug_infos;
-	for (idx_t i = 0; i < COMPRESSION_TYPE_COUNT; i++) {
-		compression_type_debug_infos.push_back(
-		    StringUtil::Format("%llu: {compression type: %s, is disabled: %llu}", i,
-		                       EnumUtil::ToString(internal_compression_methods[i].type), is_disabled[i].load()));
-	}
-
-	lock_guard<mutex> guard(lock);
-	vector<string> physical_type_debug_infos;
-	for (idx_t index = 0; index < PHYSICAL_TYPE_COUNT; index++) {
-		const auto &physical_type = physical_types[index];
-		D_ASSERT(GetCompressionIndex(physical_type) == index);
-
-		idx_t out_of = 0;
-		for (idx_t i = 0; internal_compression_methods[i].get_function; i++) {
-			auto &method = internal_compression_methods[i];
-			if (method.supports_type(physical_type)) {
-				out_of++;
-			}
-		}
-
-		const auto &function_list = functions[index];
-		vector<string> function_list_debug_infos;
-		for (idx_t function_index = 0; function_index < function_list.size(); function_index++) {
-			auto &function = function_list[function_index];
-			function_list_debug_infos.push_back(StringUtil::Format("%llu: {compression type: %s, physical type: %s}",
-			                                                       function_index, EnumUtil::ToString(function.type),
-			                                                       EnumUtil::ToString(function.data_type)));
-		}
-
-		physical_type_debug_infos.push_back(StringUtil::Format(
-		    "%llu: {physical type: %s, loaded: %llu, loaded functions: %llu (out of: %llu)}\t\t%s", index,
-		    EnumUtil::ToString(physical_type), is_loaded[index].load(), function_list.size(), out_of,
-		    function_list_debug_infos.empty() ? "" : "\n\t\t" + StringUtil::Join(function_list_debug_infos, "\n\t\t")));
-	}
-
-	return StringUtil::Format("DEBUG INFO:\n - Compression types:\n\t%s\n\n - Physical types:\n\t%s",
-	                          StringUtil::Join(compression_type_debug_infos, "\n\t"),
-	                          StringUtil::Join(physical_type_debug_infos, "\n\t"));
-}
-
 vector<CompressionType> DBConfig::GetDisabledCompressionMethods() const {
 	return compression_functions->GetDisabledCompressionMethods();
 }
@@ -258,18 +199,17 @@ vector<reference<const CompressionFunction>> DBConfig::GetCompressionFunctions(c
 
 optional_ptr<const CompressionFunction> DBConfig::TryGetCompressionFunction(CompressionType type,
                                                                             const PhysicalType physical_type) const {
-	return compression_functions->GetCompressionFunction(type, physical_type).second;
+	return compression_functions->GetCompressionFunction(type, physical_type);
 }
 
 reference<const CompressionFunction> DBConfig::GetCompressionFunction(CompressionType type,
                                                                       const PhysicalType physical_type) const {
 	auto result = compression_functions->GetCompressionFunction(type, physical_type);
-	if (!result.second) {
+	if (!result) {
 		throw InternalException(
 		    "Could not find compression function \"%s\" for physical type \"%s\". Load result: %s. %s",
-		    EnumUtil::ToString(type), EnumUtil::ToString(physical_type), EnumUtil::ToString(result.first),
-		    compression_functions->GetDebugInfo());
+		    EnumUtil::ToString(type), EnumUtil::ToString(physical_type));
 	}
-	return *result.second;
+	return *result;
 }
 } // namespace duckdb

src/duckdb/src/function/table/system/duckdb_views.cpp

@@ -56,6 +56,9 @@ static unique_ptr<FunctionData> DuckDBViewsBind(ClientContext &context, TableFun
 	names.emplace_back("sql");
 	return_types.emplace_back(LogicalType::VARCHAR);
 
+	names.emplace_back("is_bound");
+	return_types.emplace_back(LogicalType::BOOLEAN);
+
 	return nullptr;
 }
 
@@ -135,15 +138,24 @@ void DuckDBViewsFunction(ClientContext &context, TableFunctionInput &data_p, Dat
 			case 10: {
 				// column_count, LogicalType::BIGINT
 				// make sure the view is bound so we know the columns it emits
-				view.BindView(context);
 				auto columns = view.GetColumnInfo();
-				output.SetValue(c, count, Value::BIGINT(NumericCast<int64_t>(columns->types.size())));
+				Value column_count;
+				if (columns) {
+					column_count = Value::BIGINT(NumericCast<int64_t>(columns->types.size()));
+				}
+				output.SetValue(c, count, column_count);
 				break;
 			}
 			case 11:
 				// sql, LogicalType::VARCHAR
 				output.SetValue(c, count, Value(view.ToSQL()));
 				break;
+			case 12: {
+				// is_bound, LogicalType::BOOLEAN
+				auto columns = view.GetColumnInfo();
+				output.SetValue(c, count, Value::BOOLEAN(columns.get()));
+				break;
+			}
 			default:
 				throw InternalException("Unsupported column index for duckdb_views");
 			}

src/duckdb/src/function/table/table_scan.cpp

@@ -622,9 +622,15 @@ bool TryScanIndex(ART &art, IndexEntry &entry, const ColumnList &column_list, Ta
 	vector<reference<ART>> arts_to_scan;
 	arts_to_scan.push_back(art);
 	if (entry.deleted_rows_in_use) {
+		if (entry.deleted_rows_in_use->GetIndexType() != ART::TYPE_NAME) {
+			throw InternalException("Concurrent changes made to a non-ART index");
+		}
 		arts_to_scan.push_back(entry.deleted_rows_in_use->Cast<ART>());
 	}
 	if (entry.added_data_during_checkpoint) {
+		if (entry.added_data_during_checkpoint->GetIndexType() != ART::TYPE_NAME) {
+			throw InternalException("Concurrent changes made to a non-ART index");
+		}
 		arts_to_scan.push_back(entry.added_data_during_checkpoint->Cast<ART>());
 	}
 

src/duckdb/src/function/table/version/pragma_version.cpp

@@ -1,5 +1,5 @@
 #ifndef DUCKDB_PATCH_VERSION
-#define DUCKDB_PATCH_VERSION "0-dev7526"
+#define DUCKDB_PATCH_VERSION "0-dev7605"
 #endif
 #ifndef DUCKDB_MINOR_VERSION
 #define DUCKDB_MINOR_VERSION 5
@@ -8,10 +8,10 @@
 #define DUCKDB_MAJOR_VERSION 1
 #endif
 #ifndef DUCKDB_VERSION
-#define DUCKDB_VERSION "v1.5.0-dev7526"
+#define DUCKDB_VERSION "v1.5.0-dev7605"
 #endif
 #ifndef DUCKDB_SOURCE_ID
-#define DUCKDB_SOURCE_ID "490e7a8f47"
+#define DUCKDB_SOURCE_ID "3a3967aa81"
 #endif
 #include "duckdb/function/table/system_functions.hpp"
 #include "duckdb/main/database.hpp"