void FunctionCoalesce::executeImpl(Block & block, const ColumnNumbers & arguments, size_t result)
{
/// coalesce(arg0, arg1, ..., argN) is essentially
/// multiIf(isNotNull(arg0), arg0, isNotNull(arg1), arg1, ..., isNotNull(argN), argN, NULL)
FunctionIsNotNull is_not_null;
ColumnNumbers multi_if_args;
Block temp_block = block;
for (size_t i = 0; i < arguments.size(); ++i)
{
size_t res_pos = temp_block.columns();
temp_block.insert({nullptr, std::make_shared<DataTypeUInt8>(), ""});
is_not_null.executeImpl(temp_block, {arguments[i]}, res_pos);
multi_if_args.push_back(res_pos);
multi_if_args.push_back(arguments[i]);
}
/// Argument corresponding to the fallback NULL value.
multi_if_args.push_back(temp_block.columns());
/// Append a fallback NULL column.
ColumnWithTypeAndName elem;
elem.column = std::make_shared<ColumnNull>(temp_block.rows(), Null());
elem.type = std::make_shared<DataTypeNull>();
elem.name = "NULL";
temp_block.insert(elem);
FunctionMultiIf{}.executeImpl(temp_block, multi_if_args, result);
block.safeGetByPosition(result).column = std::move(temp_block.safeGetByPosition(result).column);
}
void executeImpl(Block & block, const ColumnNumbers & args, size_t result, size_t input_rows_count) override
{
/** We will gather values from columns in branches to result column,
* depending on values of conditions.
*/
struct Instruction
{
const IColumn * condition = nullptr;
const IColumn * source = nullptr;
bool condition_always_true = false;
bool condition_is_nullable = false;
bool source_is_constant = false;
};
std::vector<Instruction> instructions;
instructions.reserve(args.size() / 2 + 1);
Columns converted_columns_holder;
converted_columns_holder.reserve(instructions.size());
const DataTypePtr & return_type = block.getByPosition(result).type;
for (size_t i = 0; i < args.size(); i += 2)
{
Instruction instruction;
size_t source_idx = i + 1;
if (source_idx == args.size())
{
/// The last, "else" branch can be treated as a branch with always true condition "else if (true)".
--source_idx;
instruction.condition_always_true = true;
}
else
{
const ColumnWithTypeAndName & cond_col = block.getByPosition(args[i]);
/// We skip branches that are always false.
/// If we encounter a branch that is always true, we can finish.
if (cond_col.column->onlyNull())
continue;
if (cond_col.column->isColumnConst())
{
Field value = typeid_cast<const ColumnConst &>(*cond_col.column).getField();
if (value.isNull())
continue;
if (value.get<UInt64>() == 0)
continue;
instruction.condition_always_true = true;
}
else
{
if (cond_col.column->isColumnNullable())
instruction.condition_is_nullable = true;
instruction.condition = cond_col.column.get();
}
}
const ColumnWithTypeAndName & source_col = block.getByPosition(args[source_idx]);
if (source_col.type->equals(*return_type))
{
instruction.source = source_col.column.get();
}
else
{
/// Cast all columns to result type.
converted_columns_holder.emplace_back(castColumn(source_col, return_type, context));
instruction.source = converted_columns_holder.back().get();
}
if (instruction.source && instruction.source->isColumnConst())
instruction.source_is_constant = true;
instructions.emplace_back(std::move(instruction));
if (instructions.back().condition_always_true)
break;
}
size_t rows = input_rows_count;
MutableColumnPtr res = return_type->createColumn();
for (size_t i = 0; i < rows; ++i)
{
for (const auto & instruction : instructions)
{
bool insert = false;
if (instruction.condition_always_true)
insert = true;
else if (!instruction.condition_is_nullable)
insert = static_cast<const ColumnUInt8 &>(*instruction.condition).getData()[i];
else
{
const ColumnNullable & condition_nullable = static_cast<const ColumnNullable &>(*instruction.condition);
const ColumnUInt8 & condition_nested = static_cast<const ColumnUInt8 &>(condition_nullable.getNestedColumn());
const NullMap & condition_null_map = condition_nullable.getNullMapData();
insert = !condition_null_map[i] && condition_nested.getData()[i];
}
if (insert)
{
if (!instruction.source_is_constant)
res->insertFrom(*instruction.source, i);
else
res->insertFrom(static_cast<const ColumnConst &>(*instruction.source).getDataColumn(), 0);
break;
}
}
}
block.getByPosition(result).column = std::move(res);
}
void FunctionArrayIntersect::executeImpl(Block & block, const ColumnNumbers & arguments, size_t result, size_t input_rows_count)
{
const auto & return_type = block.getByPosition(result).type;
auto return_type_array = checkAndGetDataType<DataTypeArray>(return_type.get());
if (!return_type)
throw Exception{"Return type for function " + getName() + " must be array.", ErrorCodes::LOGICAL_ERROR};
const auto & nested_return_type = return_type_array->getNestedType();
if (typeid_cast<const DataTypeNothing *>(nested_return_type.get()))
{
block.getByPosition(result).column = return_type->createColumnConstWithDefaultValue(input_rows_count);
return;
}
auto num_args = arguments.size();
DataTypes data_types;
data_types.reserve(num_args);
for (size_t i = 0; i < num_args; ++i)
data_types.push_back(block.getByPosition(arguments[i]).type);
auto return_type_with_nulls = getMostSubtype(data_types, true, true);
Columns columns = castColumns(block, arguments, return_type, return_type_with_nulls);
UnpackedArrays arrays = prepareArrays(columns);
ColumnPtr result_column;
auto not_nullable_nested_return_type = removeNullable(nested_return_type);
TypeListNumbers::forEach(NumberExecutor(arrays, not_nullable_nested_return_type, result_column));
using DateMap = ClearableHashMap<DataTypeDate::FieldType, size_t, DefaultHash<DataTypeDate::FieldType>,
HashTableGrower<INITIAL_SIZE_DEGREE>,
HashTableAllocatorWithStackMemory<(1ULL << INITIAL_SIZE_DEGREE) * sizeof(DataTypeDate::FieldType)>>;
using DateTimeMap = ClearableHashMap<DataTypeDateTime::FieldType, size_t, DefaultHash<DataTypeDateTime::FieldType>,
HashTableGrower<INITIAL_SIZE_DEGREE>,
HashTableAllocatorWithStackMemory<(1ULL << INITIAL_SIZE_DEGREE) * sizeof(DataTypeDateTime::FieldType)>>;
using StringMap = ClearableHashMap<StringRef, size_t, StringRefHash, HashTableGrower<INITIAL_SIZE_DEGREE>,
HashTableAllocatorWithStackMemory<(1ULL << INITIAL_SIZE_DEGREE) * sizeof(StringRef)>>;
if (!result_column)
{
auto column = not_nullable_nested_return_type->createColumn();
WhichDataType which(not_nullable_nested_return_type);
if (which.isDate())
result_column = execute<DateMap, ColumnVector<DataTypeDate::FieldType>, true>(arrays, std::move(column));
else if (which.isDateTime())
result_column = execute<DateTimeMap, ColumnVector<DataTypeDateTime::FieldType>, true>(arrays, std::move(column));
else if (which.isString())
result_column = execute<StringMap, ColumnString, false>(arrays, std::move(column));
else if (which.isFixedString())
result_column = execute<StringMap, ColumnFixedString, false>(arrays, std::move(column));
else
{
column = static_cast<const DataTypeArray &>(*return_type_with_nulls).getNestedType()->createColumn();
result_column = castRemoveNullable(execute<StringMap, IColumn, false>(arrays, std::move(column)), return_type);
}
}
block.getByPosition(result).column = std::move(result_column);
}
Columns FunctionArrayIntersect::castColumns(
Block & block, const ColumnNumbers & arguments, const DataTypePtr & return_type,
const DataTypePtr & return_type_with_nulls) const
{
size_t num_args = arguments.size();
Columns columns(num_args);
auto type_array = checkAndGetDataType<DataTypeArray>(return_type.get());
auto & type_nested = type_array->getNestedType();
auto type_not_nullable_nested = removeNullable(type_nested);
const bool is_numeric_or_string = isNumber(type_not_nullable_nested)
|| isDateOrDateTime(type_not_nullable_nested)
|| isStringOrFixedString(type_not_nullable_nested);
DataTypePtr nullable_return_type;
if (is_numeric_or_string)
{
auto type_nullable_nested = makeNullable(type_nested);
nullable_return_type = std::make_shared<DataTypeArray>(type_nullable_nested);
}
const bool nested_is_nullable = type_nested->isNullable();
for (size_t i = 0; i < num_args; ++i)
{
const ColumnWithTypeAndName & arg = block.getByPosition(arguments[i]);
auto & column = columns[i];
if (is_numeric_or_string)
{
/// Cast to Array(T) or Array(Nullable(T)).
if (nested_is_nullable)
{
if (arg.type->equals(*return_type))
column = arg.column;
else
column = castColumn(arg, return_type, context);
}
else
{
/// If result has array type Array(T) still cast Array(Nullable(U)) to Array(Nullable(T))
/// because cannot cast Nullable(T) to T.
if (arg.type->equals(*return_type) || arg.type->equals(*nullable_return_type))
column = arg.column;
else if (static_cast<const DataTypeArray &>(*arg.type).getNestedType()->isNullable())
column = castColumn(arg, nullable_return_type, context);
else
column = castColumn(arg, return_type, context);
}
}
else
{
/// return_type_with_nulls is the most common subtype with possible nullable parts.
if (arg.type->equals(*return_type_with_nulls))
column = arg.column;
else
column = castColumn(arg, return_type_with_nulls, context);
}
}
return columns;
}
void FunctionCoalesce::executeImpl(Block & block, const ColumnNumbers & arguments, size_t result)
{
/// coalesce(arg0, arg1, ..., argN) is essentially
/// multiIf(isNotNull(arg0), assumeNotNull(arg0), isNotNull(arg1), assumeNotNull(arg1), ..., argN)
/// with constant NULL arguments removed.
ColumnNumbers filtered_args;
filtered_args.reserve(arguments.size());
for (const auto & arg : arguments)
{
const auto & type = block.getByPosition(arg).type;
if (type->onlyNull())
continue;
filtered_args.push_back(arg);
if (!type->isNullable())
break;
}
FunctionIsNotNull is_not_null;
FunctionAssumeNotNull assume_not_null;
ColumnNumbers multi_if_args;
Block temp_block = block;
for (size_t i = 0; i < filtered_args.size(); ++i)
{
size_t res_pos = temp_block.columns();
bool is_last = i + 1 == filtered_args.size();
if (is_last)
{
multi_if_args.push_back(filtered_args[i]);
}
else
{
temp_block.insert({nullptr, std::make_shared<DataTypeUInt8>(), ""});
is_not_null.execute(temp_block, {filtered_args[i]}, res_pos);
temp_block.insert({nullptr, removeNullable(block.getByPosition(filtered_args[i]).type), ""});
assume_not_null.execute(temp_block, {filtered_args[i]}, res_pos + 1);
multi_if_args.push_back(res_pos);
multi_if_args.push_back(res_pos + 1);
}
}
/// If all arguments appeared to be NULL.
if (multi_if_args.empty())
{
block.getByPosition(result).column = block.getByPosition(result).type->createColumnConstWithDefaultValue(block.rows());
return;
}
if (multi_if_args.size() == 1)
{
block.getByPosition(result).column = block.getByPosition(multi_if_args.front()).column;
return;
}
FunctionMultiIf{context}.execute(temp_block, multi_if_args, result);
ColumnPtr res = std::move(temp_block.getByPosition(result).column);
/// if last argument is not nullable, result should be also not nullable
if (!block.getByPosition(multi_if_args.back()).column->isColumnNullable() && res->isColumnNullable())
res = static_cast<const ColumnNullable &>(*res).getNestedColumnPtr();
block.getByPosition(result).column = std::move(res);
}
void executeImpl(Block & block, const ColumnNumbers & arguments, size_t result, size_t input_rows_count) override
{
/// coalesce(arg0, arg1, ..., argN) is essentially
/// multiIf(isNotNull(arg0), assumeNotNull(arg0), isNotNull(arg1), assumeNotNull(arg1), ..., argN)
/// with constant NULL arguments removed.
ColumnNumbers filtered_args;
filtered_args.reserve(arguments.size());
for (const auto & arg : arguments)
{
const auto & type = block.getByPosition(arg).type;
if (type->onlyNull())
continue;
filtered_args.push_back(arg);
if (!type->isNullable())
break;
}
auto is_not_null = FunctionFactory::instance().get("isNotNull", context);
auto assume_not_null = FunctionFactory::instance().get("assumeNotNull", context);
auto multi_if = FunctionFactory::instance().get("multiIf", context);
ColumnNumbers multi_if_args;
Block temp_block = block;
for (size_t i = 0; i < filtered_args.size(); ++i)
{
size_t res_pos = temp_block.columns();
bool is_last = i + 1 == filtered_args.size();
if (is_last)
{
multi_if_args.push_back(filtered_args[i]);
}
else
{
temp_block.insert({nullptr, std::make_shared<DataTypeUInt8>(), ""});
is_not_null->build({temp_block.getByPosition(filtered_args[i])})->execute(temp_block, {filtered_args[i]}, res_pos, input_rows_count);
temp_block.insert({nullptr, removeNullable(block.getByPosition(filtered_args[i]).type), ""});
assume_not_null->build({temp_block.getByPosition(filtered_args[i])})->execute(temp_block, {filtered_args[i]}, res_pos + 1, input_rows_count);
multi_if_args.push_back(res_pos);
multi_if_args.push_back(res_pos + 1);
}
}
/// If all arguments appeared to be NULL.
if (multi_if_args.empty())
{
block.getByPosition(result).column = block.getByPosition(result).type->createColumnConstWithDefaultValue(input_rows_count);
return;
}
if (multi_if_args.size() == 1)
{
block.getByPosition(result).column = block.getByPosition(multi_if_args.front()).column;
return;
}
ColumnsWithTypeAndName multi_if_args_elems;
multi_if_args_elems.reserve(multi_if_args.size());
for (auto column_num : multi_if_args)
multi_if_args_elems.emplace_back(temp_block.getByPosition(column_num));
multi_if->build(multi_if_args_elems)->execute(temp_block, multi_if_args, result, input_rows_count);
ColumnPtr res = std::move(temp_block.getByPosition(result).column);
/// if last argument is not nullable, result should be also not nullable
if (!block.getByPosition(multi_if_args.back()).column->isColumnNullable() && res->isColumnNullable())
res = static_cast<const ColumnNullable &>(*res).getNestedColumnPtr();
block.getByPosition(result).column = std::move(res);
}
void executeImpl(Block & block, const ColumnNumbers & arguments, size_t result, size_t input_rows_count) override
{
size_t num_elements = arguments.size();
if (num_elements == 0)
{
/// We should return constant empty array.
block.getByPosition(result).column = block.getByPosition(result).type->createColumnConstWithDefaultValue(input_rows_count);
return;
}
const DataTypePtr & return_type = block.getByPosition(result).type;
const DataTypePtr & elem_type = static_cast<const DataTypeArray &>(*return_type).getNestedType();
size_t block_size = input_rows_count;
/** If part of columns have not same type as common type of all elements of array,
* then convert them to common type.
* If part of columns are constants,
* then convert them to full columns.
*/
Columns columns_holder(num_elements);
ColumnRawPtrs columns(num_elements);
for (size_t i = 0; i < num_elements; ++i)
{
const auto & arg = block.getByPosition(arguments[i]);
ColumnPtr preprocessed_column = arg.column;
if (!arg.type->equals(*elem_type))
preprocessed_column = castColumn(arg, elem_type, context);
preprocessed_column = preprocessed_column->convertToFullColumnIfConst();
columns_holder[i] = std::move(preprocessed_column);
columns[i] = columns_holder[i].get();
}
/// Create and fill the result array.
auto out = ColumnArray::create(elem_type->createColumn());
IColumn & out_data = out->getData();
IColumn::Offsets & out_offsets = out->getOffsets();
out_data.reserve(block_size * num_elements);
out_offsets.resize(block_size);
IColumn::Offset current_offset = 0;
for (size_t i = 0; i < block_size; ++i)
{
for (size_t j = 0; j < num_elements; ++j)
out_data.insertFrom(*columns[j], i);
current_offset += num_elements;
out_offsets[i] = current_offset;
}
block.getByPosition(result).column = std::move(out);
}
void FunctionArrayReduce::executeImpl(Block & block, const ColumnNumbers & arguments, size_t result, size_t input_rows_count)
{
IAggregateFunction & agg_func = *aggregate_function.get();
AlignedBuffer place_holder(agg_func.sizeOfData(), agg_func.alignOfData());
AggregateDataPtr place = place_holder.data();
std::unique_ptr<Arena> arena = agg_func.allocatesMemoryInArena() ? std::make_unique<Arena>() : nullptr;
size_t rows = input_rows_count;
/// Aggregate functions do not support constant columns. Therefore, we materialize them.
std::vector<ColumnPtr> materialized_columns;
const size_t num_arguments_columns = arguments.size() - 1;
std::vector<const IColumn *> aggregate_arguments_vec(num_arguments_columns);
const ColumnArray::Offsets * offsets = nullptr;
for (size_t i = 0; i < num_arguments_columns; ++i)
{
const IColumn * col = block.getByPosition(arguments[i + 1]).column.get();
const ColumnArray::Offsets * offsets_i = nullptr;
if (const ColumnArray * arr = checkAndGetColumn<ColumnArray>(col))
{
aggregate_arguments_vec[i] = &arr->getData();
offsets_i = &arr->getOffsets();
}
else if (const ColumnConst * const_arr = checkAndGetColumnConst<ColumnArray>(col))
{
materialized_columns.emplace_back(const_arr->convertToFullColumn());
const auto & arr = typeid_cast<const ColumnArray &>(*materialized_columns.back().get());
aggregate_arguments_vec[i] = &arr.getData();
offsets_i = &arr.getOffsets();
}
else
throw Exception("Illegal column " + col->getName() + " as argument of function " + getName(), ErrorCodes::ILLEGAL_COLUMN);
if (i == 0)
offsets = offsets_i;
else if (*offsets_i != *offsets)
throw Exception("Lengths of all arrays passed to " + getName() + " must be equal.",
ErrorCodes::SIZES_OF_ARRAYS_DOESNT_MATCH);
}
const IColumn ** aggregate_arguments = aggregate_arguments_vec.data();
MutableColumnPtr result_holder = block.getByPosition(result).type->createColumn();
IColumn & res_col = *result_holder;
/// AggregateFunction's states should be inserted into column using specific way
auto res_col_aggregate_function = typeid_cast<ColumnAggregateFunction *>(&res_col);
if (!res_col_aggregate_function && agg_func.isState())
throw Exception("State function " + agg_func.getName() + " inserts results into non-state column "
+ block.getByPosition(result).type->getName(), ErrorCodes::ILLEGAL_COLUMN);
ColumnArray::Offset current_offset = 0;
for (size_t i = 0; i < rows; ++i)
{
agg_func.create(place);
ColumnArray::Offset next_offset = (*offsets)[i];
try
{
for (size_t j = current_offset; j < next_offset; ++j)
agg_func.add(place, aggregate_arguments, j, arena.get());
if (!res_col_aggregate_function)
agg_func.insertResultInto(place, res_col);
else
res_col_aggregate_function->insertFrom(place);
}
catch (...)
{
agg_func.destroy(place);
throw;
}
agg_func.destroy(place);
current_offset = next_offset;
}
block.getByPosition(result).column = std::move(result_holder);
}