AddingDefaultsBlockInputStream fixed types optimisation

dbms/src/DataStreams/AddingDefaultsBlockInputStream.cpp

@@ -1,19 +1,33 @@
-#include <Columns/ColumnsNumber.h>
-#include <Columns/ColumnsCommon.h>
-#include <Columns/ColumnConst.h>
-#include <Columns/FilterDescription.h>
-#include <Interpreters/ExpressionActions.h>
 #include <Common/typeid_cast.h>
+#include <Functions/FunctionHelpers.h>
+#include <Interpreters/ExpressionActions.h>
 #include <Interpreters/evaluateMissingDefaults.h>
 #include <DataStreams/AddingDefaultsBlockInputStream.h>
 
+#include <Columns/ColumnsNumber.h>
+#include <Columns/ColumnsCommon.h>
+#include <Columns/ColumnDecimal.h>
+#include <Columns/ColumnConst.h>
+#include <Columns/FilterDescription.h>
+
+#include <DataTypes/DataTypesNumber.h>
+#include <DataTypes/DataTypesDecimal.h>
+#include <DataTypes/DataTypeDate.h>
+#include <DataTypes/DataTypeDateTime.h>
+#include <DataTypes/DataTypeEnum.h>
+#include <DataTypes/DataTypeUUID.h>
+#include <DataTypes/DataTypeString.h>
+#include <DataTypes/DataTypeFixedString.h>
+
 
 namespace DB
 {
 
 namespace ErrorCodes
 {
+    extern const int LOGICAL_ERROR;
     extern const int SIZES_OF_COLUMNS_DOESNT_MATCH;
+    extern const int TYPE_MISMATCH;
 }
 
 
@@ -61,30 +75,25 @@ Block AddingDefaultsBlockInputStream::readImpl()
             continue;
 
         size_t block_column_position = res.getPositionByName(column_name);
-        const ColumnWithTypeAndName & column_read = res.getByPosition(block_column_position);
+        ColumnWithTypeAndName & column_read = res.getByPosition(block_column_position);
-
-        if (column_read.column->size() != column_def.column->size())
-            throw Exception("Mismach column sizes while adding defaults", ErrorCodes::SIZES_OF_COLUMNS_DOESNT_MATCH);
-
         const auto & defaults_mask = delayed_defaults.getDefaultsBitmask(block_column_position);
+
+        checkCalculated(column_read, column_def, defaults_mask.size());
+
         if (!defaults_mask.empty())
         {
-            MutableColumnPtr column_mixed = column_read.column->cloneEmpty();
+            /// TODO: FixedString
-
+            if (isColumnedAsNumber(column_read.type) || isDecimal(column_read.type))
-            for (size_t row_idx = 0; row_idx < column_read.column->size(); ++row_idx)
             {
-                if (row_idx < defaults_mask.size() && defaults_mask[row_idx])
+                MutableColumnPtr column_mixed = (*std::move(column_read.column)).mutate();
-                {
+                mixNumberColumns(column_read.type->getTypeId(), column_mixed, column_def.column, defaults_mask);
-                    if (column_def.column->isColumnConst())
+                column_read.column = std::move(column_mixed);
-                        column_mixed->insert((*column_def.column)[row_idx]);
+            }
-                    else
+            else
-                        column_mixed->insertFrom(*column_def.column, row_idx);
+            {
-                }
+                MutableColumnPtr column_mixed = mixColumns(column_read, column_def, defaults_mask);
-                else
+                mixed_columns.emplace(block_column_position, std::move(column_mixed));
-                    column_mixed->insertFrom(*column_read.column, row_idx);
             }
-
-            mixed_columns.emplace(std::make_pair(block_column_position, std::move(column_mixed)));
         }
     }
 
@@ -104,4 +113,95 @@ Block AddingDefaultsBlockInputStream::readImpl()
     return res;
 }
 
+void AddingDefaultsBlockInputStream::checkCalculated(const ColumnWithTypeAndName & col_read,
+                                                     const ColumnWithTypeAndName & col_defaults,
+                                                     size_t defaults_needed) const
+{
+    size_t column_size = col_read.column->size();
+
+    if (column_size != col_defaults.column->size())
+        throw Exception("Mismach column sizes while adding defaults", ErrorCodes::SIZES_OF_COLUMNS_DOESNT_MATCH);
+
+    if (column_size < defaults_needed)
+        throw Exception("Unexpected defaults count", ErrorCodes::SIZES_OF_COLUMNS_DOESNT_MATCH);
+
+    if (col_read.type->getTypeId() != col_defaults.type->getTypeId())
+        throw Exception("Mismach column types while adding defaults", ErrorCodes::TYPE_MISMATCH);
+}
+
+void AddingDefaultsBlockInputStream::mixNumberColumns(TypeIndex type_idx, MutableColumnPtr & column_mixed, const ColumnPtr & column_defs,
+                                                      const BlockMissingValues::RowsBitMask & defaults_mask) const
+{
+    auto call = [&](const auto & types) -> bool
+    {
+        using Types = std::decay_t<decltype(types)>;
+        using DataType = typename Types::LeftType;
+
+        if constexpr (!std::is_same_v<DataType, DataTypeString> && !std::is_same_v<DataType, DataTypeFixedString>)
+        {
+            using FieldType = typename DataType::FieldType;
+            using ColVecType = std::conditional_t<IsDecimalNumber<FieldType>, ColumnDecimal<FieldType>, ColumnVector<FieldType>>;
+
+            auto col_read = typeid_cast<ColVecType *>(column_mixed.get());
+            if (!col_read)
+                return false;
+
+            typename ColVecType::Container & dst = col_read->getData();
+
+            if (auto const_col_defs = checkAndGetColumnConst<ColVecType>(column_defs.get()))
+            {
+                FieldType value = checkAndGetColumn<ColVecType>(const_col_defs->getDataColumnPtr().get())->getData()[0];
+
+                for (size_t i = 0; i < defaults_mask.size(); ++i)
+                    if (defaults_mask[i])
+                        dst[i] = value;
+
+                return true;
+            }
+            else if (auto col_defs = checkAndGetColumn<ColVecType>(column_defs.get()))
+            {
+                auto & src = col_defs->getData();
+                for (size_t i = 0; i < defaults_mask.size(); ++i)
+                    if (defaults_mask[i])
+                        dst[i] = src[i];
+
+                return true;
+            }
+        }
+
+        return false;
+    };
+
+    if (!callOnIndexAndDataType<void>(type_idx, call))
+        throw Exception("Unexpected type on mixNumberColumns", ErrorCodes::LOGICAL_ERROR);
+}
+
+MutableColumnPtr AddingDefaultsBlockInputStream::mixColumns(const ColumnWithTypeAndName & col_read,
+                                                            const ColumnWithTypeAndName & col_defaults,
+                                                            const BlockMissingValues::RowsBitMask & defaults_mask) const
+{
+    size_t column_size = col_read.column->size();
+    size_t defaults_needed = defaults_mask.size();
+
+    MutableColumnPtr column_mixed = col_read.column->cloneEmpty();
+
+    for (size_t i = 0; i < defaults_needed; ++i)
+    {
+        if (defaults_mask[i])
+        {
+            if (col_defaults.column->isColumnConst())
+                column_mixed->insert((*col_defaults.column)[i]);
+            else
+                column_mixed->insertFrom(*col_defaults.column, i);
+        }
+        else
+            column_mixed->insertFrom(*col_read.column, i);
+    }
+
+    for (size_t i = defaults_needed; i < column_size; ++i)
+        column_mixed->insertFrom(*col_read.column, i);
+
+    return column_mixed;
+}
+
 }

dbms/src/DataStreams/AddingDefaultsBlockInputStream.h

@@ -27,6 +27,12 @@ private:
     Block header;
     const ColumnDefaults column_defaults;
     const Context & context;
+
+    void checkCalculated(const ColumnWithTypeAndName & col_read, const ColumnWithTypeAndName & col_defaults, size_t needed) const;
+    MutableColumnPtr mixColumns(const ColumnWithTypeAndName & col_read, const ColumnWithTypeAndName & col_defaults,
+                                const BlockMissingValues::RowsBitMask & defaults_mask) const;
+    void mixNumberColumns(TypeIndex type_idx, MutableColumnPtr & col_mixed, const ColumnPtr & col_defaults,
+                          const BlockMissingValues::RowsBitMask & defaults_mask) const;
 };
 
 }

dbms/src/DataTypes/IDataType.h

@@ -512,6 +512,13 @@ inline bool isNumber(const T & data_type)
     return which.isInt() || which.isUInt() || which.isFloat();
 }
 
+template <typename T>
+inline bool isColumnedAsNumber(const T & data_type)
+{
+    WhichDataType which(data_type);
+    return which.isInt() || which.isUInt() || which.isFloat() || which.isDateOrDateTime() || which.isUUID();
+}
+
 template <typename T>
 inline bool isString(const T & data_type)
 {