thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-21 07:01:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Updated usage of different types during IN query

Browse Source 
1. Added accurateCast function.
2. Use accurateCast in Set during execute.
3. Added accurateCast tests.
4. Updated select_in_different_types tests.
This commit is contained in:
Maksim Kita 2020-12-05 16:18:56 +03:00
parent 6f4bc77dbc
commit 0464859cfe
15 changed files with 294 additions and 231 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
docs/en/sql-reference/functions/type-conversion-functions.md
View File

@ -430,7 +430,28 @@ SELECT toTypeName(CAST(x, 'Nullable(UInt16)')) FROM t_null
- [cast_keep_nullable](../../operations/settings/settings.md#cast_keep_nullable) setting
## accurateCastOrNull(x, T) {#type_conversion_function-accurate-cast}
## accurateCast(x, T) {#type_conversion_function-accurate-cast}
Converts ‘x’ to the ‘t’ data type. The differente from cast(x, T) is that accurateCast
does not allow overflow of numeric types during cast if type value x does not fit
bounds of type T.
Example
``` sql
SELECT cast(-1, 'UInt8') as uint8;
```
``` text
```
```sql
SELECT accurateCast(-1, 'UInt8') as uint8;
```
``` text
```
## accurateCastOrNull(x, T) {#type_conversion_function-accurate-cast_or_null}
Converts ‘x’ to the ‘t’ data type. Always returns nullable type and returns NULL
if the casted value is not representable in the target type.

2
src/AggregateFunctions/AggregateFunctionAvg.h
View File

@ -33,7 +33,7 @@ struct AvgFraction
/// Allow division by zero as sometimes we need to return NaN.
/// Invoked only is either Numerator or Denominator are Decimal.
Float64 NO_SANITIZE_UNDEFINED divideIfAnyDecimal(UInt32 num_scale, UInt32 denom_scale [[maybe_unused]]) const
Float64 NO_SANITIZE_UNDEFINED divideIfAnyDecimal(UInt32 num_scale, UInt32 denom_scale [[maybe_unused]]) const
{
if constexpr (IsDecimalNumber<Numerator> && IsDecimalNumber<Denominator>)
{

19
src/Core/AccurateComparison.h
View File

@ -515,23 +515,34 @@ inline bool NO_SANITIZE_UNDEFINED convertNumeric(From value, To & result)
return true;
}
if constexpr (std::is_floating_point_v<From> && std::is_floating_point_v<To>) {
if constexpr (std::is_floating_point_v<From> && std::is_floating_point_v<To>)
{
/// Note that NaNs doesn't compare equal to anything, but they are still in range of any Float type.
if (isNaN(value))
{
result = value;
return true;
}
if (value == std::numeric_limits<From>::infinity())
{
result = std::numeric_limits<To>::infinity();
return true;
}
if (value == -std::numeric_limits<From>::infinity())
{
result = -std::numeric_limits<To>::infinity();
return true;
}
}
if (accurate::greaterOp(value, std::numeric_limits<To>::max())
|| accurate::greaterOp(std::numeric_limits<To>::min(), value))
|| accurate::greaterOp(std::numeric_limits<To>::lowest(), value))
{
return false;
}
result = static_cast<To>(value);
return equalsOp(value, result);
}
}

17
src/Functions/FunctionsConversion.cpp
View File

@ -1,21 +1,10 @@
#include <Functions/FunctionFactory.h>
#include <Functions/FunctionsConversion.h>
#include <Interpreters/Context.h>
namespace DB
{
FunctionOverloadResolverImplPtr CastOverloadResolver::create(const Context & context)
{
return createImpl(context.getSettingsRef().cast_keep_nullable);
}
FunctionOverloadResolverImplPtr AccurateCastOverloadResolver::create(const Context &)
{
return std::make_unique<AccurateCastOverloadResolver>();
}
void registerFunctionFixedString(FunctionFactory & factory);
void registerFunctionsConversion(FunctionFactory & factory)
@ -49,8 +38,10 @@ void registerFunctionsConversion(FunctionFactory & factory)
registerFunctionFixedString(factory);
factory.registerFunction<FunctionToUnixTimestamp>();
factory.registerFunction<CastOverloadResolver>(FunctionFactory::CaseInsensitive);
factory.registerFunction<AccurateCastOverloadResolver>();
factory.registerFunction<CastOverloadResolver<CastType::nonAccurate>>(FunctionFactory::CaseInsensitive);
factory.registerFunction<CastOverloadResolver<CastType::accurate>>();
factory.registerFunction<CastOverloadResolver<CastType::accurateOrNull>>();
factory.registerFunction<FunctionToUInt8OrZero>();
factory.registerFunction<FunctionToUInt16OrZero>();

317
src/Functions/FunctionsConversion.h
View File

@ -41,9 +41,10 @@
#include <Functions/FunctionsMiscellaneous.h>
#include <Functions/FunctionHelpers.h>
#include <Functions/DateTimeTransforms.h>
#include <Functions/toFixedString.h>
#include <DataTypes/DataTypeLowCardinality.h>
#include <Columns/ColumnLowCardinality.h>
#include <Functions/toFixedString.h>
#include <Interpreters/Context.h>
namespace DB
@ -70,6 +71,7 @@ namespace ErrorCodes
extern const int ILLEGAL_TYPE_OF_ARGUMENT;
extern const int NOT_IMPLEMENTED;
extern const int CANNOT_INSERT_NULL_IN_ORDINARY_COLUMN;
extern const int DECIMAL_OVERFLOW;
}
@ -95,15 +97,14 @@ inline UInt32 extractToDecimalScale(const ColumnWithTypeAndName & named_column)
/// Function toUnixTimestamp has exactly the same implementation as toDateTime of String type.
struct NameToUnixTimestamp { static constexpr auto name = "toUnixTimestamp"; };
struct AccurateAdditions
struct AccurateConvertStrategyAdditions
{
UInt32 scale { 0 };
};
enum class ConvertStrategy
struct AccurateOrNullConvertStrategyAdditions
{
NonAccurate,
Accurate
UInt32 scale { 0 };
};
/** Conversion of number types to each other, enums to numbers, dates and datetimes to numbers and back: done by straight assignment.
@ -117,11 +118,9 @@ struct ConvertImpl
template <typename Additions = void *>
static ColumnPtr NO_SANITIZE_UNDEFINED execute(
const ColumnsWithTypeAndName & arguments, const DataTypePtr & /*result_type*/, size_t /*input_rows_count*/,
const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type [[maybe_unused]], size_t /*input_rows_count*/,
Additions additions [[maybe_unused]] = Additions())
{
static constexpr auto convert_strategy
= std::is_same_v<Additions, AccurateAdditions> ? ConvertStrategy::Accurate : ConvertStrategy::NonAccurate;
const ColumnWithTypeAndName & named_from = arguments[0];
using ColVecFrom = typename FromDataType::ColumnType;
@ -150,7 +149,8 @@ struct ConvertImpl
if constexpr (IsDataTypeDecimal<ToDataType>)
{
UInt32 scale;
if constexpr (convert_strategy == ConvertStrategy::Accurate)
if constexpr (std::is_same_v<Additions, AccurateConvertStrategyAdditions>
|| std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>)
{
scale = additions.scale;
}
@ -171,7 +171,7 @@ struct ConvertImpl
ColumnUInt8::MutablePtr col_null_map_to;
ColumnUInt8::Container * vec_null_map_to [[maybe_unused]] = nullptr;
if constexpr (convert_strategy == ConvertStrategy::Accurate)
if constexpr (std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>)
{
col_null_map_to = ColumnUInt8::create(size, false);
vec_null_map_to = &col_null_map_to->getData();
@ -179,95 +179,91 @@ struct ConvertImpl
for (size_t i = 0; i < size; ++i)
{
if constexpr (IsDataTypeDecimal<FromDataType> || IsDataTypeDecimal<ToDataType>)
if constexpr ((is_big_int_v<FromFieldType> || is_big_int_v<ToFieldType>)
&& (std::is_same_v<FromFieldType, UInt128> || std::is_same_v<ToFieldType, UInt128>))
{
try
if constexpr (std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>)
(*vec_null_map_to)[i] = true;
else
throw Exception("Unexpected UInt128 to big int conversion", ErrorCodes::NOT_IMPLEMENTED);
}
else {
if constexpr (IsDataTypeDecimal<FromDataType> || IsDataTypeDecimal<ToDataType>)
{
if constexpr (IsDataTypeDecimal<FromDataType> && IsDataTypeDecimal<ToDataType>)
vec_to[i] = convertDecimals<FromDataType, ToDataType>(vec_from[i], vec_from.getScale(), vec_to.getScale());
else if constexpr (IsDataTypeDecimal<FromDataType> && IsDataTypeNumber<ToDataType>)
vec_to[i] = convertFromDecimal<FromDataType, ToDataType>(vec_from[i], vec_from.getScale());
else if constexpr (IsDataTypeNumber<FromDataType> && IsDataTypeDecimal<ToDataType>)
vec_to[i] = convertToDecimal<FromDataType, ToDataType>(vec_from[i], vec_to.getScale());
else
try
{
throw Exception("Unsupported data type in conversion function", ErrorCodes::CANNOT_CONVERT_TYPE);
if constexpr (IsDataTypeDecimal<FromDataType> && IsDataTypeDecimal<ToDataType>)
vec_to[i] = convertDecimals<FromDataType, ToDataType>(vec_from[i], vec_from.getScale(), vec_to.getScale());
else if constexpr (IsDataTypeDecimal<FromDataType> && IsDataTypeNumber<ToDataType>)
vec_to[i] = convertFromDecimal<FromDataType, ToDataType>(vec_from[i], vec_from.getScale());
else if constexpr (IsDataTypeNumber<FromDataType> && IsDataTypeDecimal<ToDataType>)
vec_to[i] = convertToDecimal<FromDataType, ToDataType>(vec_from[i], vec_to.getScale());
else
{
throw Exception("Unsupported data type in conversion function", ErrorCodes::CANNOT_CONVERT_TYPE);
}
}
catch (const Exception & exception)
{
if constexpr (std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>)
{
if (exception.code() == ErrorCodes::CANNOT_CONVERT_TYPE || exception.code() == ErrorCodes::DECIMAL_OVERFLOW)
(*vec_null_map_to)[i] = true;
else
throw exception;
}
else
throw exception;
}
}
catch (...)
{
/// Handle decimal overflow that propagated as exception
if constexpr (convert_strategy == ConvertStrategy::Accurate)
(*vec_null_map_to)[i] = true;
else
throw;
}
}
else if constexpr (is_big_int_v<FromFieldType> || is_big_int_v<ToFieldType>)
{
if constexpr (std::is_same_v<FromFieldType, UInt128> || std::is_same_v<ToFieldType, UInt128>)
{
if constexpr (convert_strategy == ConvertStrategy::Accurate)
(*vec_null_map_to)[i] = true;
else
throw Exception("Unexpected UInt128 to big int conversion", ErrorCodes::NOT_IMPLEMENTED);
}
/// If From Data is Nan or Inf, throw exception
else if (!isFinite(vec_from[i]))
{
if constexpr (convert_strategy == ConvertStrategy::Accurate)
vec_null_map_to[i] = true;
else
throw Exception("Unexpected inf or nan to big int conversion", ErrorCodes::NOT_IMPLEMENTED);
}
else
{
if constexpr (convert_strategy == ConvertStrategy::Accurate)
/// If From Data is Nan or Inf, throw exception
/// TODO: Probably this can be applied to all integers not just big integers
/// https://stackoverflow.com/questions/38795544/is-casting-of-infinity-to-integer-undefined
if constexpr (is_big_int_v<ToFieldType>)
{
if (accurate::greaterOp(vec_from[i], std::numeric_limits<ToFieldType>::max())
|| accurate::greaterOp(std::numeric_limits<ToFieldType>::min(), vec_from[i]))
if (!isFinite(vec_from[i]))
{
(*vec_null_map_to)[i] = true;
continue;
if constexpr (std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>)
{
vec_null_map_to[i] = true;
continue;
}
else
throw Exception("Unexpected inf or nan to big int conversion", ErrorCodes::NOT_IMPLEMENTED);
}
}
ToFieldType from = bigint_cast<ToFieldType>(vec_from[i]);
vec_to[i] = static_cast<ToFieldType>(from);
}
}
else if constexpr (std::is_same_v<ToFieldType, UInt128> && sizeof(FromFieldType) <= sizeof(UInt64))
{
if constexpr (convert_strategy == ConvertStrategy::Accurate)
{
if (accurate::greaterOp(vec_from[i], std::numeric_limits<ToFieldType>::max())
|| accurate::greaterOp(std::numeric_limits<ToFieldType>::min(), vec_from[i]))
if constexpr (std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>
|| std::is_same_v<Additions, AccurateConvertStrategyAdditions>)
{
(*vec_null_map_to)[i] = true;
continue;
bool convert_result = accurate::convertNumeric(vec_from[i], vec_to[i]);
if (!convert_result)
{
if (std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>)
{
(*vec_null_map_to)[i] = true;
}
else
{
throw Exception(
"Value in column " + named_from.column->getName() + " cannot be safely converted into type "
+ result_type->getName(),
ErrorCodes::CANNOT_CONVERT_TYPE);
}
}
}
else
{
vec_to[i] = static_cast<ToFieldType>(vec_from[i]);
}
}
UInt64 from = static_cast<UInt64>(vec_from[i]);
vec_to[i] = static_cast<ToFieldType>(from);
}
else
{
if constexpr (convert_strategy == ConvertStrategy::Accurate)
{
if (accurate::greaterOp(vec_from[i], std::numeric_limits<ToFieldType>::max())
|| accurate::greaterOp(std::numeric_limits<ToFieldType>::min(), vec_from[i]))
{
(*vec_null_map_to)[i] = true;
continue;
}
}
vec_to[i] = static_cast<ToFieldType>(vec_from[i]);
}
}
if constexpr (convert_strategy == ConvertStrategy::Accurate)
if constexpr (std::is_same_v<Additions, AccurateOrNullConvertStrategyAdditions>)
return ColumnNullable::create(std::move(col_to), std::move(col_null_map_to));
else
return col_to;
@ -2028,6 +2024,12 @@ private:
struct NameCast { static constexpr auto name = "CAST"; };
enum class CastType {
nonAccurate,
accurate,
accurateOrNull
};
class FunctionCast final : public IFunctionBaseImpl
{
public:
@ -2037,10 +2039,10 @@ public:
FunctionCast(const char * name_, MonotonicityForRange && monotonicity_for_range_
, const DataTypes & argument_types_, const DataTypePtr & return_type_
, std::optional<Diagnostic> diagnostic_, bool is_accurate_cast_or_null_)
, std::optional<Diagnostic> diagnostic_, CastType cast_type_)
: name(name_), monotonicity_for_range(std::move(monotonicity_for_range_))
, argument_types(argument_types_), return_type(return_type_), diagnostic(std::move(diagnostic_))
, is_accurate_cast_or_null(is_accurate_cast_or_null_)
, cast_type(cast_type_)
{
}
@ -2087,7 +2089,7 @@ private:
DataTypePtr return_type;
std::optional<Diagnostic> diagnostic;
bool is_accurate_cast_or_null;
CastType cast_type;
WrapperType createFunctionAdaptor(FunctionPtr function, const DataTypePtr & from_type) const
{
@ -2101,7 +2103,7 @@ private:
};
}
WrapperType createToNullableColumnWrapper() const
static WrapperType createToNullableColumnWrapper()
{
return [] (ColumnsWithTypeAndName &, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count)
{
@ -2116,7 +2118,8 @@ private:
{
TypeIndex from_type_index = from_type->getTypeId();
WhichDataType which(from_type_index);
bool can_apply_accurate_cast = is_accurate_cast_or_null && (which.isInt() || which.isUInt() || which.isFloat());
bool can_apply_accurate_cast = (cast_type == CastType::accurate || cast_type == CastType::accurateOrNull)
&& (which.isInt() || which.isUInt() || which.isFloat());
if (requested_result_is_nullable && checkAndGetDataType<DataTypeString>(from_type.get()))
{
@ -2131,10 +2134,9 @@ private:
return createFunctionAdaptor(function, from_type);
}
auto nullable_column_wrapper = createToNullableColumnWrapper();
bool is_accurate_cast = is_accurate_cast_or_null;
auto wrapper_cast_type = cast_type;
return [is_accurate_cast, nullable_column_wrapper, from_type_index, to_type]
return [wrapper_cast_type, from_type_index, to_type]
(ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *column_nullable, size_t input_rows_count)
{
ColumnPtr result_column;
@ -2145,8 +2147,17 @@ private:
if constexpr (IsDataTypeNumber<LeftDataType> && IsDataTypeNumber<RightDataType>)
{
result_column = ConvertImpl<LeftDataType, RightDataType, NameCast>::execute(
arguments, result_type, input_rows_count, AccurateAdditions());
if (wrapper_cast_type == CastType::accurate)
{
result_column = ConvertImpl<LeftDataType, RightDataType, NameCast>::execute(
arguments, result_type, input_rows_count, AccurateConvertStrategyAdditions());
}
else
{
result_column = ConvertImpl<LeftDataType, RightDataType, NameCast>::execute(
arguments, result_type, input_rows_count, AccurateOrNullConvertStrategyAdditions());
}
return true;
}
@ -2156,8 +2167,11 @@ private:
/// Additionally check if callOnIndexAndDataType wasn't called at all.
if (!res)
{
if (is_accurate_cast)
if (wrapper_cast_type == CastType::accurateOrNull)
{
auto nullable_column_wrapper = FunctionCast::createToNullableColumnWrapper();
return nullable_column_wrapper(arguments, result_type, column_nullable, input_rows_count);
}
else
{
throw Exception{"Conversion from " + std::string(getTypeName(from_type_index)) + " to " + to_type->getName() + " is not supported",
@ -2180,7 +2194,7 @@ private:
if (!isStringOrFixedString(from_type))
throw Exception{"CAST AS FixedString is only implemented for types String and FixedString", ErrorCodes::NOT_IMPLEMENTED};
bool exception_mode_null = is_accurate_cast_or_null;
bool exception_mode_null = cast_type == CastType::accurateOrNull;
return [exception_mode_null, N] (ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable *, size_t /*input_rows_count*/)
{
if (exception_mode_null)
@ -2202,17 +2216,16 @@ private:
|| which.isStringOrFixedString();
if (!ok)
{
if (is_accurate_cast_or_null)
if (cast_type == CastType::accurateOrNull)
return createToNullableColumnWrapper();
else
throw Exception{"Conversion from " + from_type->getName() + " to " + to_type->getName() + " is not supported",
ErrorCodes::CANNOT_CONVERT_TYPE};
}
auto nullable_column_wrapper = createToNullableColumnWrapper();
bool is_accurate_cast = is_accurate_cast_or_null;
auto wrapper_cast_type = cast_type;
return [is_accurate_cast, nullable_column_wrapper, type_index, scale, to_type]
return [wrapper_cast_type, type_index, scale, to_type]
(ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *column_nullable, size_t input_rows_count)
{
ColumnPtr result_column;
@ -2223,11 +2236,22 @@ private:
if constexpr (IsDataTypeDecimalOrNumber<LeftDataType> && IsDataTypeDecimalOrNumber<RightDataType>)
{
if (is_accurate_cast)
if (wrapper_cast_type == CastType::accurate)
{
AccurateAdditions additions;
AccurateConvertStrategyAdditions additions;
additions.scale = scale;
result_column = ConvertImpl<LeftDataType, RightDataType, NameCast>::execute(arguments, result_type, input_rows_count, additions);
result_column = ConvertImpl<LeftDataType, RightDataType, NameCast>::execute(
arguments, result_type, input_rows_count, additions);
return true;
}
else if (wrapper_cast_type == CastType::accurateOrNull)
{
AccurateOrNullConvertStrategyAdditions additions;
additions.scale = scale;
result_column = ConvertImpl<LeftDataType, RightDataType, NameCast>::execute(
arguments, result_type, input_rows_count, additions);
return true;
}
}
@ -2240,8 +2264,11 @@ private:
/// Additionally check if callOnIndexAndDataType wasn't called at all.
if (!res)
{
if (is_accurate_cast)
if (wrapper_cast_type == CastType::accurateOrNull)
{
auto nullable_column_wrapper = FunctionCast::createToNullableColumnWrapper();
return nullable_column_wrapper(arguments, result_type, column_nullable, input_rows_count);
}
else
throw Exception{"Conversion from " + std::string(getTypeName(type_index)) + " to " + to_type->getName() + " is not supported",
ErrorCodes::CANNOT_CONVERT_TYPE};
@ -2263,7 +2290,7 @@ private:
}
else
{
if (is_accurate_cast_or_null)
if (cast_type == CastType::accurateOrNull)
return createToNullableColumnWrapper();
else
throw Exception{"Conversion from " + from_type_untyped->getName() + " to " + to_type->getName() +
@ -2397,7 +2424,7 @@ private:
}
else
{
if (is_accurate_cast_or_null)
if (cast_type == CastType::accurateOrNull)
return createToNullableColumnWrapper();
else
throw Exception{"Conversion from " + from_type->getName() + " to " + to_type->getName() + " is not supported",
@ -2504,7 +2531,7 @@ private:
{
if (!to_nested->isNullable())
{
if (is_accurate_cast_or_null)
if (cast_type == CastType::accurateOrNull)
{
return createToNullableColumnWrapper();
}
@ -2750,7 +2777,7 @@ private:
break;
}
if (is_accurate_cast_or_null)
if (cast_type == CastType::accurateOrNull)
return createToNullableColumnWrapper();
else
throw Exception{"Conversion from " + from_type->getName() + " to " + to_type->getName() + " is not supported",
@ -2815,15 +2842,26 @@ template <typename DataType>
}
};
template<CastType cast_type>
class CastOverloadResolver : public IFunctionOverloadResolverImpl
{
public:
using MonotonicityForRange = FunctionCast::MonotonicityForRange;
using Diagnostic = FunctionCast::Diagnostic;
static constexpr auto name = "CAST";
static constexpr auto accurate_cast_name = "accurateCast";
static constexpr auto accurate_cast_or_null_name = "accurateCastOrNull";
static constexpr auto cast_name = "CAST";
static FunctionOverloadResolverImplPtr create(const Context & context);
static constexpr auto name =
cast_type == CastType::accurate ? accurate_cast_name :
(cast_type == CastType::accurateOrNull ? accurate_cast_or_null_name : cast_name);
static FunctionOverloadResolverImplPtr create(const Context & context)
{
return createImpl(context.getSettingsRef().cast_keep_nullable);
}
static FunctionOverloadResolverImplPtr createImpl(bool keep_nullable, std::optional<Diagnostic> diagnostic = {})
{
return std::make_unique<CastOverloadResolver>(keep_nullable, std::move(diagnostic));
@ -2840,8 +2878,6 @@ public:
ColumnNumbers getArgumentsThatAreAlwaysConstant() const override { return {1}; }
protected:
FunctionBaseImplPtr build(const ColumnsWithTypeAndName & arguments, const DataTypePtr & return_type) const override
{
DataTypes data_types(arguments.size());
@ -2850,7 +2886,7 @@ protected:
data_types[i] = arguments[i].type;
auto monotonicity = MonotonicityHelper::getMonotonicityInformation(arguments.front().type, return_type.get());
return std::make_unique<FunctionCast>(name, std::move(monotonicity), data_types, return_type, diagnostic, false);
return std::make_unique<FunctionCast>(name, std::move(monotonicity), data_types, return_type, diagnostic, cast_type);
}
DataTypePtr getReturnType(const ColumnsWithTypeAndName & arguments) const override
@ -2868,9 +2904,17 @@ protected:
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
DataTypePtr type = DataTypeFactory::instance().get(type_col->getValue<String>());
if (keep_nullable && arguments.front().type->isNullable())
if constexpr (cast_type == CastType::accurateOrNull)
{
return makeNullable(type);
return type;
}
else
{
if (keep_nullable && arguments.front().type->isNullable())
return makeNullable(type);
return type;
}
}
bool useDefaultImplementationForNulls() const override { return false; }
@ -2881,53 +2925,4 @@ private:
std::optional<Diagnostic> diagnostic;
};
class AccurateCastOverloadResolver : public IFunctionOverloadResolverImpl
{
public:
using MonotonicityForRange = FunctionCast::MonotonicityForRange;
using Diagnostic = FunctionCast::Diagnostic;
static constexpr auto name = "accurateCastOrNull";
static FunctionOverloadResolverImplPtr create(const Context & context);
String getName() const override { return name; }
size_t getNumberOfArguments() const override { return 2; }
ColumnNumbers getArgumentsThatAreAlwaysConstant() const override { return {1}; }
FunctionBaseImplPtr build(const ColumnsWithTypeAndName & arguments, const DataTypePtr & return_type) const override
{
DataTypes data_types(arguments.size());
for (size_t i = 0; i < arguments.size(); ++i)
data_types[i] = arguments[i].type;
auto monotonicity = MonotonicityHelper::getMonotonicityInformation(arguments.front().type, return_type.get());
return std::make_unique<FunctionCast>(name, std::move(monotonicity), data_types, return_type, std::optional<Diagnostic>(), true);
}
protected:
DataTypePtr getReturnType(const ColumnsWithTypeAndName & arguments) const override
{
const auto & column = arguments.back().column;
if (!column)
throw Exception("Second argument to " + getName() + " must be a constant string describing type."
" Instead there is non-constant column of type " + arguments.back().type->getName(),
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
const auto type_col = checkAndGetColumnConst<ColumnString>(column.get());
if (!type_col)
throw Exception("Second argument to " + getName() + " must be a constant string describing type."
" Instead there is a column with the following structure: " + column->dumpStructure(),
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
DataTypePtr type = DataTypeFactory::instance().get(type_col->getValue<String>());
return makeNullable(type);
}
bool useDefaultImplementationForNulls() const override { return false; }
bool useDefaultImplementationForLowCardinalityColumns() const override { return false; }
};
}

4
src/Interpreters/ActionsDAG.cpp
View File

@ -662,10 +662,10 @@ ActionsDAGPtr ActionsDAG::makeConvertingActions(
auto * right_arg = const_cast<Node *>(&actions_dag->addColumn(std::move(column), true));
auto * left_arg = src_node;
CastOverloadResolver::Diagnostic diagnostic = {src_node->result_name, res_elem.name};
FunctionCast::Diagnostic diagnostic = {src_node->result_name, res_elem.name};
FunctionOverloadResolverPtr func_builder_cast =
std::make_shared<FunctionOverloadResolverAdaptor>(
CastOverloadResolver::createImpl(false, std::move(diagnostic)));
CastOverloadResolver<CastType::nonAccurate>::createImpl(false, std::move(diagnostic)));
Inputs children = { left_arg, right_arg };
src_node = &actions_dag->addFunction(func_builder_cast, std::move(children), {}, true);

21
src/Interpreters/Set.cpp
View File

@ -25,6 +25,7 @@
#include <Interpreters/evaluateConstantExpression.h>
#include <Interpreters/NullableUtils.h>
#include <Interpreters/sortBlock.h>
#include <Interpreters/castColumn.h>
#include <Interpreters/Context.h>
#include <Storages/MergeTree/KeyCondition.h>
@ -32,8 +33,6 @@
#include <ext/range.h>
#include <DataTypes/DataTypeLowCardinality.h>
#include <iostream>
namespace DB
{
@ -257,15 +256,19 @@ ColumnPtr Set::execute(const Block & block, bool negative) const
for (size_t i = 0; i < num_key_columns; ++i)
{
/// TODO: Optimize making cast only if types different
/// TODO: This case SELECT '1' IN (SELECT 1); should not work but with AccurateCastOrNull it works
ColumnPtr result;
auto & column_before_cast = block.safeGetByPosition(i);
ColumnWithTypeAndName column
ColumnWithTypeAndName column_to_cast
= {column_before_cast.column->convertToFullColumnIfConst(), column_before_cast.type, column_before_cast.name};
auto accurate_cast = AccurateCastOverloadResolver().build({column}, data_types[i]);
auto accurate_cast_executable = accurate_cast->prepare({column});
auto casted_column = accurate_cast_executable->execute({column}, data_types[i], column.column->size());
materialized_columns.emplace_back() = casted_column;
if (!transform_null_in) {
result = castColumn<CastType::accurateOrNull>(column_to_cast, data_types[i]);
} else {
result = castColumn<CastType::accurate>(column_to_cast, data_types[i]);
}
materialized_columns.emplace_back() = result;
key_columns.emplace_back() = materialized_columns.back().get();
}

34
src/Interpreters/castColumn.cpp
View File

@ -1,34 +0,0 @@
#include <Core/Field.h>
#include <Interpreters/castColumn.h>
#include <Interpreters/ExpressionActions.h>
#include <DataTypes/DataTypeString.h>
#include <Functions/IFunctionAdaptors.h>
#include <Functions/FunctionsConversion.h>
namespace DB
{
ColumnPtr castColumn(const ColumnWithTypeAndName & arg, const DataTypePtr & type)
{
if (arg.type->equals(*type))
return arg.column;
ColumnsWithTypeAndName arguments
{
arg,
{
DataTypeString().createColumnConst(arg.column->size(), type->getName()),
std::make_shared<DataTypeString>(),
""
}
};
FunctionOverloadResolverPtr func_builder_cast =
std::make_shared<FunctionOverloadResolverAdaptor>(CastOverloadResolver::createImpl(false));
auto func_cast = func_builder_cast->build(arguments);
return func_cast->execute(arguments, type, arg.column->size());
}
}

35
src/Interpreters/castColumn.h
View File

@ -2,8 +2,41 @@
#include <Core/ColumnWithTypeAndName.h>
#include <Functions/FunctionsConversion.h>
namespace DB
{
ColumnPtr castColumn(const ColumnWithTypeAndName & arg, const DataTypePtr & type);
template<CastType cast_type = CastType::nonAccurate>
ColumnPtr castColumn(const ColumnWithTypeAndName & arg, const DataTypePtr & type)
{
if (arg.type->equals(*type))
return arg.column;
ColumnsWithTypeAndName arguments
{
arg,
{
DataTypeString().createColumnConst(arg.column->size(), type->getName()),
std::make_shared<DataTypeString>(),
""
}
};
FunctionOverloadResolverPtr func_builder_cast =
std::make_shared<FunctionOverloadResolverAdaptor>(CastOverloadResolver<cast_type>::createImpl(false));
auto func_cast = func_builder_cast->build(arguments);
if constexpr (cast_type == CastType::accurateOrNull)
{
return func_cast->execute(arguments, makeNullable(type), arg.column->size());
}
else
{
return func_cast->execute(arguments, type, arg.column->size());
}
}
}

1
src/Interpreters/ya.make
View File

@ -143,7 +143,6 @@ SRCS(
TreeRewriter.cpp
addMissingDefaults.cpp
addTypeConversionToAST.cpp
castColumn.cpp
convertFieldToType.cpp
createBlockSelector.cpp
evaluateConstantExpression.cpp

2
src/Storages/MergeTree/KeyCondition.cpp
View File

@ -1137,7 +1137,7 @@ bool KeyCondition::tryParseAtomFromAST(const ASTPtr & node, const Context & cont
ColumnsWithTypeAndName arguments{
{nullptr, key_expr_type, ""}, {DataTypeString().createColumnConst(1, common_type->getName()), common_type, ""}};
FunctionOverloadResolverPtr func_builder_cast
= std::make_shared<FunctionOverloadResolverAdaptor>(CastOverloadResolver::createImpl(false));
= std::make_shared<FunctionOverloadResolverAdaptor>(CastOverloadResolver<CastType::nonAccurate>::createImpl(false));
auto func_cast = func_builder_cast->build(arguments);
/// If we know the given range only contains one value, then we treat all functions as positive monotonic.

5
tests/queries/0_stateless/01600_select_in_different_types.reference
View File

@ -1,6 +1,11 @@
1
0
1
1
2
2
1
1
1
0
0

13
tests/queries/0_stateless/01600_select_in_different_types.sql
View File

@ -1,5 +1,7 @@
-- SELECT 1 IN (SELECT -1)
-- SELECT -1 IN (SELECT 1)
SELECT 1 IN (SELECT 1);
SELECT -1 IN (SELECT 1);
DROP TABLE IF EXISTS select_in_test;
CREATE TABLE select_in_test(value UInt8) ENGINE=TinyLog;
INSERT INTO select_in_test VALUES (1), (2), (3);
@ -24,5 +26,10 @@ SELECT value FROM select_in_test WHERE value IN (SELECT 2);
DROP TABLE select_in_test;
SELECT 1 IN (1);
-- Right now this working because of accurate cast. Need to discuss.
SELECT '1' IN (SELECT 1);
SELECT 1 IN (SELECT 1) SETTINGS transform_null_in = 1;
SELECT 1 IN (SELECT 'a') SETTINGS transform_null_in = 1;
SELECT 'a' IN (SELECT 1) SETTINGS transform_null_in = 1; -- { serverError 6 }
SELECT 1 IN (SELECT -1) SETTINGS transform_null_in = 1;
SELECT -1 IN (SELECT 1) SETTINGS transform_null_in = 1; -- { serverError 70 }

8
tests/queries/0_stateless/01601_accurate_cast.reference Normal file
View File

@ -0,0 +1,8 @@
5
5
5
5
5
5
1.000000000
12

24
tests/queries/0_stateless/01601_accurate_cast.sql Normal file
View File

@ -0,0 +1,24 @@
SELECT accurateCast(-1, 'UInt8'); -- { serverError 70 }
SELECT accurateCast(5, 'UInt8');
SELECT accurateCast(257, 'UInt8'); -- { serverError 70 }
SELECT accurateCast(-1, 'UInt16'); -- { serverError 70 }
SELECT accurateCast(5, 'UInt16');
SELECT accurateCast(65536, 'UInt16'); -- { serverError 70 }
SELECT accurateCast(-1, 'UInt32'); -- { serverError 70 }
SELECT accurateCast(5, 'UInt32');
SELECT accurateCast(4294967296, 'UInt32'); -- { serverError 70 }
SELECT accurateCast(-1, 'UInt64'); -- { serverError 70 }
SELECT accurateCast(5, 'UInt64');
SELECT accurateCast(-1, 'UInt256'); -- { serverError 70 }
SELECT accurateCast(5, 'UInt256');
SELECT accurateCast(-129, 'Int8'); -- { serverError 70 }
SELECT accurateCast(5, 'Int8');
SELECT accurateCast(128, 'Int8'); -- { serverError 70 }
SELECT accurateCast(10, 'Decimal32(9)'); -- { serverError 407 }
SELECT accurateCast(1, 'Decimal32(9)');
SELECT accurateCast(-10, 'Decimal32(9)'); -- { serverError 407 }
SELECT accurateCast('123', 'FixedString(2)'); -- { serverError 131 }
SELECT accurateCast('12', 'FixedString(2)');