Update vendored DuckDB sources to 9d02a50

Author: duckdblabs-bot

src/duckdb/extension/parquet/parquet_reader.cpp

@@ -778,12 +778,13 @@ void ParquetReader::PrepareRowGroupBuffer(ParquetReaderScanState &state, idx_t c
 			FilterPropagateResult prune_result;
 			// TODO we might not have stats but STILL a bloom filter so move this up
 			// check the bloom filter if present
-			if (!column_reader.Type().IsNested() &&
+			bool is_generated_column = column_reader.FileIdx() >= group.columns.size();
+			if (!column_reader.Type().IsNested() && !is_generated_column &&
 			    ParquetStatisticsUtils::BloomFilterSupported(column_reader.Type().id()) &&
 			    ParquetStatisticsUtils::BloomFilterExcludes(filter, group.columns[column_reader.FileIdx()].meta_data,
 			                                                *state.thrift_file_proto, allocator)) {
 				prune_result = FilterPropagateResult::FILTER_ALWAYS_FALSE;
-			} else if (column_reader.Type().id() == LogicalTypeId::VARCHAR &&
+			} else if (column_reader.Type().id() == LogicalTypeId::VARCHAR && !is_generated_column &&
 			           group.columns[column_reader.FileIdx()].meta_data.statistics.__isset.min_value &&
 			           group.columns[column_reader.FileIdx()].meta_data.statistics.__isset.max_value) {
 

src/duckdb/src/common/arrow/arrow_converter.cpp

@@ -73,6 +73,7 @@ void SetArrowMapFormat(DuckDBArrowSchemaHolder &root_holder, ArrowSchema &child,
 	InitializeChild(root_holder.nested_children.back()[0], root_holder);
 	child.children = &root_holder.nested_children_ptr.back()[0];
 	child.children[0]->name = "entries";
+	child.children[0]->flags = 0; // Set the 'entries' field to non-nullable
 	SetArrowFormat(root_holder, **child.children, ListType::GetChildType(type), options, context);
 }
 

src/duckdb/src/common/types/column/column_data_allocator.cpp

@@ -54,12 +54,7 @@ ColumnDataAllocator::~ColumnDataAllocator() {
 	for (auto &block : blocks) {
 		block.handle->SetDestroyBufferUpon(DestroyBufferUpon::UNPIN);
 	}
-	const auto data_size = SizeInBytes();
 	blocks.clear();
-	if (Allocator::SupportsFlush() &&
-	    data_size > alloc.buffer_manager->GetBufferPool().GetAllocatorBulkDeallocationFlushThreshold()) {
-		Allocator::FlushAll();
-	}
 }
 
 BufferHandle ColumnDataAllocator::Pin(uint32_t block_id) {

src/duckdb/src/common/types/row/tuple_data_segment.cpp

@@ -119,10 +119,6 @@ TupleDataSegment::~TupleDataSegment() {
 	}
 	pinned_row_handles.clear();
 	pinned_heap_handles.clear();
-	if (Allocator::SupportsFlush() && allocator &&
-	    data_size > allocator->GetBufferManager().GetBufferPool().GetAllocatorBulkDeallocationFlushThreshold()) {
-		Allocator::FlushAll();
-	}
 	allocator.reset();
 }
 

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

@@ -638,7 +638,6 @@ void JoinFilterPushdownInfo::PushInFilter(const JoinFilterPushdownFilter &info,
 
 	// generate the OR filter
 	auto in_filter = make_uniq<InFilter>(std::move(in_list));
-	in_filter->origin_is_hash_join = true;
 
 	// we push the OR filter as an OptionalFilter so that we can use it for zonemap pruning only
 	// the IN-list is expensive to execute otherwise

src/duckdb/src/execution/radix_partitioned_hashtable.cpp

@@ -281,12 +281,12 @@ idx_t RadixHTConfig::GetRadixBits() const {
 }
 
 void RadixHTConfig::SetRadixBitsInternal(const idx_t radix_bits_p, bool external) {
-	if (sink_radix_bits >= radix_bits_p || sink.any_combined) {
+	if (sink_radix_bits > radix_bits_p || sink.any_combined) {
 		return;
 	}
 
 	auto guard = sink.Lock();
-	if (sink_radix_bits >= radix_bits_p || sink.any_combined) {
+	if (sink_radix_bits > radix_bits_p || sink.any_combined) {
 		return;
 	}
 

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

@@ -24,6 +24,7 @@
 #include "duckdb/planner/expression/bound_constant_expression.hpp"
 #include "duckdb/planner/expression/bound_comparison_expression.hpp"
 #include "duckdb/planner/filter/conjunction_filter.hpp"
+#include "duckdb/common/types/value_map.hpp"
 
 namespace duckdb {
 
@@ -361,56 +362,137 @@ unique_ptr<GlobalTableFunctionState> DuckIndexScanInitGlobal(ClientContext &cont
 	return std::move(g_state);
 }
 
-void ExtractInFilter(unique_ptr<TableFilter> &filter, BoundColumnRefExpression &bound_ref,
-                     unique_ptr<vector<unique_ptr<Expression>>> &filter_expressions) {
-	// Special-handling of IN filters.
-	// They are part of a CONJUNCTION_AND.
-	if (filter->filter_type != TableFilterType::CONJUNCTION_AND) {
-		return;
+void ExtractExpressionsFromValues(value_set_t &unique_values, BoundColumnRefExpression &bound_ref,
+                                  vector<unique_ptr<Expression>> &expressions) {
+	for (const auto &value : unique_values) {
+		auto bound_constant = make_uniq<BoundConstantExpression>(value);
+		auto filter_expr = make_uniq<BoundComparisonExpression>(ExpressionType::COMPARE_EQUAL, bound_ref.Copy(),
+		                                                        std::move(bound_constant));
+		expressions.push_back(std::move(filter_expr));
 	}
+}
 
-	auto &and_filter = filter->Cast<ConjunctionAndFilter>();
-	auto &children = and_filter.child_filters;
-	if (children.empty()) {
-		return;
+void ExtractIn(InFilter &filter, BoundColumnRefExpression &bound_ref, vector<unique_ptr<Expression>> &expressions) {
+	// Eliminate any duplicates.
+	value_set_t unique_values;
+	for (const auto &value : filter.values) {
+		if (unique_values.find(value) == unique_values.end()) {
+			unique_values.insert(value);
+		}
 	}
-	if (children[0]->filter_type != TableFilterType::OPTIONAL_FILTER) {
+	ExtractExpressionsFromValues(unique_values, bound_ref, expressions);
+}
+
+void ExtractConjunctionAnd(ConjunctionAndFilter &filter, BoundColumnRefExpression &bound_ref,
+                           vector<unique_ptr<Expression>> &expressions) {
+	if (filter.child_filters.empty()) {
 		return;
 	}
 
-	auto &optional_filter = children[0]->Cast<OptionalFilter>();
-	auto &child = optional_filter.child_filter;
-	if (child->filter_type != TableFilterType::IN_FILTER) {
+	// Extract the CONSTANT_COMPARISON and IN_FILTER children.
+	vector<reference<ConstantFilter>> comparisons;
+	vector<reference<InFilter>> in_filters;
+
+	for (idx_t i = 0; i < filter.child_filters.size(); i++) {
+		if (filter.child_filters[i]->filter_type == TableFilterType::CONSTANT_COMPARISON) {
+			auto &comparison = filter.child_filters[i]->Cast<ConstantFilter>();
+			comparisons.push_back(comparison);
+			continue;
+		}
+
+		if (filter.child_filters[i]->filter_type == TableFilterType::OPTIONAL_FILTER) {
+			auto &optional_filter = filter.child_filters[i]->Cast<OptionalFilter>();
+			if (!optional_filter.child_filter) {
+				return;
+			}
+			if (optional_filter.child_filter->filter_type != TableFilterType::IN_FILTER) {
+				// No support for other optional filter types yet.
+				return;
+			}
+			auto &in_filter = optional_filter.child_filter->Cast<InFilter>();
+			in_filters.push_back(in_filter);
+			continue;
+		}
+
+		// No support for other filter types than CONSTANT_COMPARISON and IN_FILTER in CONJUNCTION_AND yet.
 		return;
 	}
 
-	auto &in_filter = child->Cast<InFilter>();
-	if (!in_filter.origin_is_hash_join) {
+	// No support for other CONJUNCTION_AND cases yet.
+	if (in_filters.empty()) {
 		return;
 	}
 
-	// They are all on the same column, so we can split them.
-	for (const auto &value : in_filter.values) {
-		auto bound_constant = make_uniq<BoundConstantExpression>(value);
-		auto filter_expr = make_uniq<BoundComparisonExpression>(ExpressionType::COMPARE_EQUAL, bound_ref.Copy(),
-		                                                        std::move(bound_constant));
-		filter_expressions->push_back(std::move(filter_expr));
+	// Get the combined unique values of the IN filters.
+	value_set_t unique_values;
+	for (idx_t filter_idx = 0; filter_idx < in_filters.size(); filter_idx++) {
+		auto &in_filter = in_filters[filter_idx].get();
+		for (idx_t value_idx = 0; value_idx < in_filter.values.size(); value_idx++) {
+			auto &value = in_filter.values[value_idx];
+			if (unique_values.find(value) != unique_values.end()) {
+				continue;
+			}
+			unique_values.insert(value);
+		}
+	}
+
+	// Extract all qualifying values.
+	for (auto value_it = unique_values.begin(); value_it != unique_values.end();) {
+		bool qualifies = true;
+		for (idx_t comp_idx = 0; comp_idx < comparisons.size(); comp_idx++) {
+			if (!comparisons[comp_idx].get().Compare(*value_it)) {
+				qualifies = false;
+				value_it = unique_values.erase(value_it);
+				break;
+			}
+		}
+		if (qualifies) {
+			value_it++;
+		}
+	}
+
+	ExtractExpressionsFromValues(unique_values, bound_ref, expressions);
+}
+
+void ExtractFilter(TableFilter &filter, BoundColumnRefExpression &bound_ref,
+                   vector<unique_ptr<Expression>> &expressions) {
+	switch (filter.filter_type) {
+	case TableFilterType::OPTIONAL_FILTER: {
+		auto &optional_filter = filter.Cast<OptionalFilter>();
+		if (!optional_filter.child_filter) {
+			return;
+		}
+		return ExtractFilter(*optional_filter.child_filter, bound_ref, expressions);
+	}
+	case TableFilterType::IN_FILTER: {
+		auto &in_filter = filter.Cast<InFilter>();
+		ExtractIn(in_filter, bound_ref, expressions);
+		return;
+	}
+	case TableFilterType::CONJUNCTION_AND: {
+		auto &conjunction_and = filter.Cast<ConjunctionAndFilter>();
+		ExtractConjunctionAnd(conjunction_and, bound_ref, expressions);
+		return;
+	}
+	default:
+		return;
 	}
 }
 
-unique_ptr<vector<unique_ptr<Expression>>> ExtractFilters(const ColumnDefinition &col, unique_ptr<TableFilter> &filter,
-                                                          idx_t storage_idx) {
+vector<unique_ptr<Expression>> ExtractFilterExpressions(const ColumnDefinition &col, unique_ptr<TableFilter> &filter,
+                                                        idx_t storage_idx) {
 	ColumnBinding binding(0, storage_idx);
 	auto bound_ref = make_uniq<BoundColumnRefExpression>(col.Name(), col.Type(), binding);
 
-	auto filter_expressions = make_uniq<vector<unique_ptr<Expression>>>();
-	ExtractInFilter(filter, *bound_ref, filter_expressions);
+	vector<unique_ptr<Expression>> expressions;
+	ExtractFilter(*filter, *bound_ref, expressions);
 
-	if (filter_expressions->empty()) {
+	// Attempt matching the top-level filter to the index expression.
+	if (expressions.empty()) {
 		auto filter_expr = filter->ToExpression(*bound_ref);
-		filter_expressions->push_back(std::move(filter_expr));
+		expressions.push_back(std::move(filter_expr));
 	}
-	return filter_expressions;
+	return expressions;
 }
 
 bool TryScanIndex(ART &art, const ColumnList &column_list, TableFunctionInitInput &input, TableFilterSet &filter_set,
@@ -453,8 +535,8 @@ bool TryScanIndex(ART &art, const ColumnList &column_list, TableFunctionInitInpu
 		return false;
 	}
 
-	auto filter_expressions = ExtractFilters(col, filter->second, storage_index.GetIndex());
-	for (const auto &filter_expr : *filter_expressions) {
+	auto expressions = ExtractFilterExpressions(col, filter->second, storage_index.GetIndex());
+	for (const auto &filter_expr : expressions) {
 		auto scan_state = art.TryInitializeScan(*index_expr, *filter_expr);
 		if (!scan_state) {
 			return false;

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

@@ -1,5 +1,5 @@
 #ifndef DUCKDB_PATCH_VERSION
-#define DUCKDB_PATCH_VERSION "4-dev4648"
+#define DUCKDB_PATCH_VERSION "4-dev4679"
 #endif
 #ifndef DUCKDB_MINOR_VERSION
 #define DUCKDB_MINOR_VERSION 1
@@ -8,10 +8,10 @@
 #define DUCKDB_MAJOR_VERSION 1
 #endif
 #ifndef DUCKDB_VERSION
-#define DUCKDB_VERSION "v1.1.4-dev4648"
+#define DUCKDB_VERSION "v1.1.4-dev4679"
 #endif
 #ifndef DUCKDB_SOURCE_ID
-#define DUCKDB_SOURCE_ID "09cef57892"
+#define DUCKDB_SOURCE_ID "0024e5d4be"
 #endif
 #include "duckdb/function/table/system_functions.hpp"
 #include "duckdb/main/database.hpp"

src/duckdb/src/include/duckdb/planner/filter/constant_filter.hpp

@@ -27,6 +27,7 @@ class ConstantFilter : public TableFilter {
 	Value constant;
 
 public:
+	bool Compare(const Value &value) const;
 	FilterPropagateResult CheckStatistics(BaseStatistics &stats) override;
 	string ToString(const string &column_name) override;
 	bool Equals(const TableFilter &other) const override;

src/duckdb/src/include/duckdb/planner/filter/in_filter.hpp

@@ -19,10 +19,8 @@ class InFilter : public TableFilter {
 
 public:
 	explicit InFilter(vector<Value> values);
-	InFilter(vector<Value> values, bool origin_is_hash_join);
 
 	vector<Value> values;
-	bool origin_is_hash_join;
 
 public:
 	FilterPropagateResult CheckStatistics(BaseStatistics &stats) override;