#pragma once #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include namespace DB { namespace ErrorCodes { extern const int ATTEMPT_TO_READ_AFTER_EOF; extern const int CANNOT_PARSE_NUMBER; extern const int CANNOT_READ_ARRAY_FROM_TEXT; extern const int CANNOT_PARSE_INPUT_ASSERTION_FAILED; extern const int CANNOT_PARSE_QUOTED_STRING; extern const int CANNOT_PARSE_ESCAPE_SEQUENCE; extern const int CANNOT_PARSE_DATE; extern const int CANNOT_PARSE_DATETIME; extern const int CANNOT_PARSE_TEXT; extern const int CANNOT_PARSE_UUID; extern const int TOO_FEW_ARGUMENTS_FOR_FUNCTION; extern const int LOGICAL_ERROR; extern const int TYPE_MISMATCH; extern const int CANNOT_CONVERT_TYPE; extern const int ILLEGAL_COLUMN; extern const int NUMBER_OF_ARGUMENTS_DOESNT_MATCH; extern const int ILLEGAL_TYPE_OF_ARGUMENT; extern const int NOT_IMPLEMENTED; extern const int CANNOT_INSERT_NULL_IN_ORDINARY_COLUMN; } /** Type conversion functions. * toType - conversion in "natural way"; */ inline UInt32 extractToDecimalScale(const ColumnWithTypeAndName & named_column) { const auto * arg_type = named_column.type.get(); bool ok = checkAndGetDataType(arg_type) || checkAndGetDataType(arg_type) || checkAndGetDataType(arg_type) || checkAndGetDataType(arg_type); if (!ok) throw Exception("Illegal type of toDecimal() scale " + named_column.type->getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); Field field; named_column.column->get(0, field); return field.get(); } /// Function toUnixTimestamp has exactly the same implementation as toDateTime of String type. struct NameToUnixTimestamp { static constexpr auto name = "toUnixTimestamp"; }; struct AccurateConvertStrategyAdditions { UInt32 scale { 0 }; }; struct AccurateOrNullConvertStrategyAdditions { UInt32 scale { 0 }; }; struct ConvertDefaultBehaviorTag {}; struct ConvertReturnNullOnErrorTag {}; /** Conversion of number types to each other, enums to numbers, dates and datetimes to numbers and back: done by straight assignment. * (Date is represented internally as number of days from some day; DateTime - as unix timestamp) */ template struct ConvertImpl { using FromFieldType = typename FromDataType::FieldType; using ToFieldType = typename ToDataType::FieldType; template static ColumnPtr NO_SANITIZE_UNDEFINED execute( const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type [[maybe_unused]], size_t input_rows_count, Additions additions [[maybe_unused]] = Additions()) { const ColumnWithTypeAndName & named_from = arguments[0]; using ColVecFrom = typename FromDataType::ColumnType; using ColVecTo = typename ToDataType::ColumnType; if (std::is_same_v) { if (isDateOrDate32(named_from.type)) throw Exception("Illegal type " + named_from.type->getName() + " of first argument of function " + Name::name, ErrorCodes::ILLEGAL_COLUMN); } if constexpr ((IsDataTypeDecimal || IsDataTypeDecimal) && !(std::is_same_v || std::is_same_v)) { if constexpr (!IsDataTypeDecimalOrNumber || !IsDataTypeDecimalOrNumber) { throw Exception("Illegal column " + named_from.column->getName() + " of first argument of function " + Name::name, ErrorCodes::ILLEGAL_COLUMN); } } if (const ColVecFrom * col_from = checkAndGetColumn(named_from.column.get())) { typename ColVecTo::MutablePtr col_to = nullptr; if constexpr (IsDataTypeDecimal) { UInt32 scale; if constexpr (std::is_same_v || std::is_same_v) { scale = additions.scale; } else { scale = additions; } col_to = ColVecTo::create(0, scale); } else col_to = ColVecTo::create(); const auto & vec_from = col_from->getData(); auto & vec_to = col_to->getData(); vec_to.resize(input_rows_count); ColumnUInt8::MutablePtr col_null_map_to; ColumnUInt8::Container * vec_null_map_to [[maybe_unused]] = nullptr; if constexpr (std::is_same_v) { col_null_map_to = ColumnUInt8::create(input_rows_count, false); vec_null_map_to = &col_null_map_to->getData(); } bool result_is_bool = isBool(result_type); for (size_t i = 0; i < input_rows_count; ++i) { if constexpr (std::is_same_v) { if (result_is_bool) { vec_to[i] = vec_from[i] != FromFieldType(0); continue; } } if constexpr (std::is_same_v != std::is_same_v) { throw Exception("Conversion between numeric types and UUID is not supported", ErrorCodes::NOT_IMPLEMENTED); } else { if constexpr (IsDataTypeDecimal || IsDataTypeDecimal) { if constexpr (std::is_same_v) { ToFieldType result; bool convert_result = false; if constexpr (IsDataTypeDecimal && IsDataTypeDecimal) convert_result = tryConvertDecimals(vec_from[i], col_from->getScale(), col_to->getScale(), result); else if constexpr (IsDataTypeDecimal && IsDataTypeNumber) convert_result = tryConvertFromDecimal(vec_from[i], col_from->getScale(), result); else if constexpr (IsDataTypeNumber && IsDataTypeDecimal) convert_result = tryConvertToDecimal(vec_from[i], col_to->getScale(), result); if (convert_result) vec_to[i] = result; else { vec_to[i] = static_cast(0); (*vec_null_map_to)[i] = true; } } else { if constexpr (IsDataTypeDecimal && IsDataTypeDecimal) vec_to[i] = convertDecimals(vec_from[i], col_from->getScale(), col_to->getScale()); else if constexpr (IsDataTypeDecimal && IsDataTypeNumber) vec_to[i] = convertFromDecimal(vec_from[i], col_from->getScale()); else if constexpr (IsDataTypeNumber && IsDataTypeDecimal) vec_to[i] = convertToDecimal(vec_from[i], col_to->getScale()); else throw Exception("Unsupported data type in conversion function", ErrorCodes::CANNOT_CONVERT_TYPE); } } else { /// If From Data is Nan or Inf and we convert to integer type, throw exception if constexpr (std::is_floating_point_v && !std::is_floating_point_v) { if (!isFinite(vec_from[i])) { if constexpr (std::is_same_v) { vec_to[i] = 0; (*vec_null_map_to)[i] = true; continue; } else throw Exception("Unexpected inf or nan to integer conversion", ErrorCodes::CANNOT_CONVERT_TYPE); } } if constexpr (std::is_same_v || std::is_same_v) { bool convert_result = accurate::convertNumeric(vec_from[i], vec_to[i]); if (!convert_result) { if (std::is_same_v) { vec_to[i] = 0; (*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(vec_from[i]); } } } } if constexpr (std::is_same_v) return ColumnNullable::create(std::move(col_to), std::move(col_null_map_to)); else return col_to; } else throw Exception("Illegal column " + named_from.column->getName() + " of first argument of function " + Name::name, ErrorCodes::ILLEGAL_COLUMN); } }; /** Conversion of DateTime to Date: throw off time component. */ template struct ConvertImpl : DateTimeTransformImpl {}; /** Conversion of DateTime to Date32: throw off time component. */ template struct ConvertImpl : DateTimeTransformImpl {}; /** Conversion of Date to DateTime: adding 00:00:00 time component. */ struct ToDateTimeImpl { static constexpr auto name = "toDateTime"; static inline UInt32 execute(UInt16 d, const DateLUTImpl & time_zone) { return time_zone.fromDayNum(DayNum(d)); } static inline Int64 execute(Int32 d, const DateLUTImpl & time_zone) { return time_zone.fromDayNum(ExtendedDayNum(d)); } static inline UInt32 execute(UInt32 dt, const DateLUTImpl & /*time_zone*/) { return dt; } // TODO: return UInt32 ??? static inline Int64 execute(Int64 dt64, const DateLUTImpl & /*time_zone*/) { return dt64; } }; template struct ConvertImpl : DateTimeTransformImpl {}; template struct ConvertImpl : DateTimeTransformImpl {}; /// Implementation of toDate function. template struct ToDateTransform32Or64 { static constexpr auto name = "toDate"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl & time_zone) { // since converting to Date, no need in values outside of default LUT range. return (from < DATE_LUT_MAX_DAY_NUM) ? from : time_zone.toDayNum(std::min(time_t(from), time_t(0xFFFFFFFF))); } }; template struct ToDateTransform32Or64Signed { static constexpr auto name = "toDate"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl & time_zone) { // TODO: decide narrow or extended range based on FromType /// The function should be monotonic (better for query optimizations), so we saturate instead of overflow. if (from < 0) return 0; return (from < DATE_LUT_MAX_DAY_NUM) ? from : time_zone.toDayNum(std::min(time_t(from), time_t(0xFFFFFFFF))); } }; template struct ToDateTransform8Or16Signed { static constexpr auto name = "toDate"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl &) { if (from < 0) return 0; return from; } }; template struct ConvertImpl : DateTimeTransformImpl> {}; /// Implementation of toDate32 function. template struct ToDate32Transform32Or64 { static constexpr auto name = "toDate32"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl & time_zone) { return (from < DATE_LUT_MAX_EXTEND_DAY_NUM) ? from : time_zone.toDayNum(std::min(time_t(from), time_t(0xFFFFFFFF))); } }; template struct ToDate32Transform32Or64Signed { static constexpr auto name = "toDate32"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl & time_zone) { static const Int32 daynum_min_offset = -static_cast(DateLUT::instance().getDayNumOffsetEpoch()); if (from < daynum_min_offset) return daynum_min_offset; return (from < DATE_LUT_MAX_EXTEND_DAY_NUM) ? from : time_zone.toDayNum(std::min(time_t(from), time_t(0xFFFFFFFF))); } }; template struct ToDate32Transform8Or16Signed { static constexpr auto name = "toDate32"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl &) { return from; } }; /** Special case of converting Int8, Int16, (U)Int32 or (U)Int64 (and also, for convenience, * Float32, Float64) to Date. If the number is negative, saturate it to unix epoch time. If the * number is less than 65536, then it is treated as DayNum, and if it's greater or equals to 65536, * then treated as unix timestamp. If the number exceeds UInt32, saturate to MAX_UINT32 then as DayNum. * It's a bit illogical, as we actually have two functions in one. * But allows to support frequent case, * when user write toDate(UInt32), expecting conversion of unix timestamp to Date. * (otherwise such usage would be frequent mistake). */ template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ToDateTimeTransform64 { static constexpr auto name = "toDateTime"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl &) { return std::min(time_t(from), time_t(0xFFFFFFFF)); } }; template struct ToDateTimeTransformSigned { static constexpr auto name = "toDateTime"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl &) { if (from < 0) return 0; return from; } }; template struct ToDateTimeTransform64Signed { static constexpr auto name = "toDateTime"; static inline NO_SANITIZE_UNDEFINED ToType execute(const FromType & from, const DateLUTImpl &) { if (from < 0) return 0; return std::min(time_t(from), time_t(0xFFFFFFFF)); } }; /** Special case of converting Int8, Int16, Int32 or (U)Int64 (and also, for convenience, Float32, * Float64) to DateTime. If the number is negative, saturate it to unix epoch time. If the number * exceeds UInt32, saturate to MAX_UINT32. */ template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; const time_t LUT_MIN_TIME = -1420070400l; // 1925-01-01 UTC const time_t LUT_MAX_TIME = 9877248000l; // 2282-12-31 UTC /** Conversion of numeric to DateTime64 */ template struct ToDateTime64TransformUnsigned { static constexpr auto name = "toDateTime64"; const DateTime64::NativeType scale_multiplier = 1; ToDateTime64TransformUnsigned(UInt32 scale = 0) : scale_multiplier(DecimalUtils::scaleMultiplier(scale)) {} inline NO_SANITIZE_UNDEFINED DateTime64::NativeType execute(FromType from, const DateLUTImpl &) const { from = std::min(from, LUT_MAX_TIME); return DecimalUtils::decimalFromComponentsWithMultiplier(from, 0, scale_multiplier); } }; template struct ToDateTime64TransformSigned { static constexpr auto name = "toDateTime64"; const DateTime64::NativeType scale_multiplier = 1; ToDateTime64TransformSigned(UInt32 scale = 0) : scale_multiplier(DecimalUtils::scaleMultiplier(scale)) {} inline NO_SANITIZE_UNDEFINED DateTime64::NativeType execute(FromType from, const DateLUTImpl &) const { from = std::max(from, LUT_MIN_TIME); from = std::min(from, LUT_MAX_TIME); return DecimalUtils::decimalFromComponentsWithMultiplier(from, 0, scale_multiplier); } }; template struct ToDateTime64TransformFloat { static constexpr auto name = "toDateTime64"; const UInt32 scale = 1; ToDateTime64TransformFloat(UInt32 scale_ = 0) : scale(scale_) {} inline NO_SANITIZE_UNDEFINED DateTime64::NativeType execute(FromType from, const DateLUTImpl &) const { if (from < 0) return 0; from = std::min(from, FromType(0xFFFFFFFF)); return convertToDecimal(from, scale); } }; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; /** Conversion of DateTime64 to Date or DateTime: discards fractional part. */ template struct FromDateTime64Transform { static constexpr auto name = Transform::name; const DateTime64::NativeType scale_multiplier = 1; FromDateTime64Transform(UInt32 scale) : scale_multiplier(DecimalUtils::scaleMultiplier(scale)) {} inline auto execute(DateTime64::NativeType dt, const DateLUTImpl & time_zone) const { const auto c = DecimalUtils::splitWithScaleMultiplier(DateTime64(dt), scale_multiplier); return Transform::execute(static_cast(c.whole), time_zone); } }; /** Conversion of DateTime64 to Date or DateTime: discards fractional part. */ template struct ConvertImpl : DateTimeTransformImpl> {}; template struct ConvertImpl : DateTimeTransformImpl> {}; struct ToDateTime64Transform { static constexpr auto name = "toDateTime64"; const DateTime64::NativeType scale_multiplier = 1; ToDateTime64Transform(UInt32 scale = 0) : scale_multiplier(DecimalUtils::scaleMultiplier(scale)) {} inline DateTime64::NativeType execute(UInt16 d, const DateLUTImpl & time_zone) const { const auto dt = ToDateTimeImpl::execute(d, time_zone); return execute(dt, time_zone); } inline DateTime64::NativeType execute(Int32 d, const DateLUTImpl & time_zone) const { const auto dt = ToDateTimeImpl::execute(d, time_zone); return DecimalUtils::decimalFromComponentsWithMultiplier(dt, 0, scale_multiplier); } inline DateTime64::NativeType execute(UInt32 dt, const DateLUTImpl & /*time_zone*/) const { return DecimalUtils::decimalFromComponentsWithMultiplier(dt, 0, scale_multiplier); } }; /** Conversion of Date or DateTime to DateTime64: add zero sub-second part. */ template struct ConvertImpl : DateTimeTransformImpl {}; template struct ConvertImpl : DateTimeTransformImpl {}; template struct ConvertImpl : DateTimeTransformImpl {}; /** Transformation of numbers, dates, datetimes to strings: through formatting. */ template struct FormatImpl { template static ReturnType execute(const typename DataType::FieldType x, WriteBuffer & wb, const DataType *, const DateLUTImpl *) { writeText(x, wb); return ReturnType(true); } }; template <> struct FormatImpl { template static ReturnType execute(const DataTypeDate::FieldType x, WriteBuffer & wb, const DataTypeDate *, const DateLUTImpl *) { writeDateText(DayNum(x), wb); return ReturnType(true); } }; template <> struct FormatImpl { template static ReturnType execute(const DataTypeDate::FieldType x, WriteBuffer & wb, const DataTypeDate32 *, const DateLUTImpl *) { writeDateText(ExtendedDayNum(x), wb); return ReturnType(true); } }; template <> struct FormatImpl { template static ReturnType execute(const DataTypeDateTime::FieldType x, WriteBuffer & wb, const DataTypeDateTime *, const DateLUTImpl * time_zone) { writeDateTimeText(x, wb, *time_zone); return ReturnType(true); } }; template <> struct FormatImpl { template static ReturnType execute(const DataTypeDateTime64::FieldType x, WriteBuffer & wb, const DataTypeDateTime64 * type, const DateLUTImpl * time_zone) { writeDateTimeText(DateTime64(x), type->getScale(), wb, *time_zone); return ReturnType(true); } }; template struct FormatImpl> { template static ReturnType execute(const FieldType x, WriteBuffer & wb, const DataTypeEnum * type, const DateLUTImpl *) { static constexpr bool throw_exception = std::is_same_v; if constexpr (throw_exception) { writeString(type->getNameForValue(x), wb); } else { StringRef res; bool is_ok = type->getNameForValue(x, res); if (is_ok) writeString(res, wb); return ReturnType(is_ok); } } }; template struct FormatImpl> { template static ReturnType execute(const FieldType x, WriteBuffer & wb, const DataTypeDecimal * type, const DateLUTImpl *) { writeText(x, type->getScale(), wb, false); return ReturnType(true); } }; /// DataTypeEnum to DataType free conversion template struct ConvertImpl, DataTypeNumber, Name, ConvertDefaultBehaviorTag> { static ColumnPtr execute(const ColumnsWithTypeAndName & arguments, const DataTypePtr &, size_t /*input_rows_count*/) { return arguments[0].column; } }; static ColumnUInt8::MutablePtr copyNullMap(ColumnPtr col) { ColumnUInt8::MutablePtr null_map = nullptr; if (const auto * col_null = checkAndGetColumn(col.get())) { null_map = ColumnUInt8::create(); null_map->insertRangeFrom(col_null->getNullMapColumn(), 0, col_null->size()); } return null_map; } template struct ConvertImpl, DataTypeString>, Name, ConvertDefaultBehaviorTag> { using FromFieldType = typename FromDataType::FieldType; using ColVecType = ColumnVectorOrDecimal; static ColumnPtr execute(const ColumnsWithTypeAndName & arguments, const DataTypePtr &, size_t /*input_rows_count*/) { ColumnUInt8::MutablePtr null_map = copyNullMap(arguments[0].column); const auto & col_with_type_and_name = columnGetNested(arguments[0]); const auto & type = static_cast(*col_with_type_and_name.type); const DateLUTImpl * time_zone = nullptr; /// For argument of DateTime type, second argument with time zone could be specified. if constexpr (std::is_same_v || std::is_same_v) { auto non_null_args = createBlockWithNestedColumns(arguments); time_zone = &extractTimeZoneFromFunctionArguments(non_null_args, 1, 0); } if (const auto col_from = checkAndGetColumn(col_with_type_and_name.column.get())) { auto col_to = ColumnString::create(); const typename ColVecType::Container & vec_from = col_from->getData(); ColumnString::Chars & data_to = col_to->getChars(); ColumnString::Offsets & offsets_to = col_to->getOffsets(); size_t size = vec_from.size(); if constexpr (std::is_same_v) data_to.resize(size * (strlen("YYYY-MM-DD") + 1)); else if constexpr (std::is_same_v) data_to.resize(size * (strlen("YYYY-MM-DD") + 1)); else if constexpr (std::is_same_v) data_to.resize(size * (strlen("YYYY-MM-DD hh:mm:ss") + 1)); else if constexpr (std::is_same_v) data_to.resize(size * (strlen("YYYY-MM-DD hh:mm:ss.") + col_from->getScale() + 1)); else data_to.resize(size * 3); /// Arbitrary offsets_to.resize(size); WriteBufferFromVector write_buffer(data_to); if (null_map) { for (size_t i = 0; i < size; ++i) { bool is_ok = FormatImpl::template execute(vec_from[i], write_buffer, &type, time_zone); null_map->getData()[i] |= !is_ok; writeChar(0, write_buffer); offsets_to[i] = write_buffer.count(); } } else { for (size_t i = 0; i < size; ++i) { FormatImpl::template execute(vec_from[i], write_buffer, &type, time_zone); writeChar(0, write_buffer); offsets_to[i] = write_buffer.count(); } } write_buffer.finalize(); if (null_map) return ColumnNullable::create(std::move(col_to), std::move(null_map)); return col_to; } else throw Exception("Illegal column " + arguments[0].column->getName() + " of first argument of function " + Name::name, ErrorCodes::ILLEGAL_COLUMN); } }; /// Generic conversion of any type to String or FixedString via serialization to text. template struct ConvertImplGenericToString { static ColumnPtr execute(const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, size_t /*input_rows_count*/) { static_assert(std::is_same_v || std::is_same_v, "Can be used only to serialize to ColumnString or ColumnFixedString"); ColumnUInt8::MutablePtr null_map = copyNullMap(arguments[0].column); const auto & col_with_type_and_name = columnGetNested(arguments[0]); const IDataType & type = *col_with_type_and_name.type; const IColumn & col_from = *col_with_type_and_name.column; size_t size = col_from.size(); auto col_to = result_type->createColumn(); { ColumnStringHelpers::WriteHelper write_helper( assert_cast(*col_to), size); auto & write_buffer = write_helper.getWriteBuffer(); FormatSettings format_settings; auto serialization = type.getDefaultSerialization(); for (size_t i = 0; i < size; ++i) { serialization->serializeText(col_from, i, write_buffer, format_settings); write_helper.rowWritten(); } write_helper.finalize(); } if (result_type->isNullable() && null_map) return ColumnNullable::create(std::move(col_to), std::move(null_map)); return col_to; } }; /** Conversion of time_t to UInt16, Int32, UInt32 */ template void convertFromTime(typename DataType::FieldType & x, time_t & time) { x = time; } template <> inline void convertFromTime(DataTypeDate::FieldType & x, time_t & time) { if (unlikely(time < 0)) x = 0; else if (unlikely(time > 0xFFFF)) x = 0xFFFF; else x = time; } template <> inline void convertFromTime(DataTypeDate32::FieldType & x, time_t & time) { x = time; } template <> inline void convertFromTime(DataTypeDateTime::FieldType & x, time_t & time) { if (unlikely(time < 0)) x = 0; else if (unlikely(time > 0xFFFFFFFF)) x = 0xFFFFFFFF; else x = time; } /** Conversion of strings to numbers, dates, datetimes: through parsing. */ template void parseImpl(typename DataType::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { readText(x, rb); } template <> inline void parseImpl(DataTypeDate::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { DayNum tmp(0); readDateText(tmp, rb); x = tmp; } template <> inline void parseImpl(DataTypeDate32::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { ExtendedDayNum tmp(0); readDateText(tmp, rb); x = tmp; } // NOTE: no need of extra overload of DateTime64, since readDateTimeText64 has different signature and that case is explicitly handled in the calling code. template <> inline void parseImpl(DataTypeDateTime::FieldType & x, ReadBuffer & rb, const DateLUTImpl * time_zone) { time_t time = 0; readDateTimeText(time, rb, *time_zone); convertFromTime(x, time); } template <> inline void parseImpl(DataTypeUUID::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { UUID tmp; readUUIDText(tmp, rb); x = tmp.toUnderType(); } template bool tryParseImpl(typename DataType::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { if constexpr (std::is_floating_point_v) return tryReadFloatText(x, rb); else /*if constexpr (is_integer_v)*/ return tryReadIntText(x, rb); } template <> inline bool tryParseImpl(DataTypeDate::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { DayNum tmp(0); if (!tryReadDateText(tmp, rb)) return false; x = tmp; return true; } template <> inline bool tryParseImpl(DataTypeDate32::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { ExtendedDayNum tmp(0); if (!tryReadDateText(tmp, rb)) { return false; } x = tmp; return true; } template <> inline bool tryParseImpl(DataTypeDateTime::FieldType & x, ReadBuffer & rb, const DateLUTImpl * time_zone) { time_t tmp = 0; if (!tryReadDateTimeText(tmp, rb, *time_zone)) return false; x = tmp; return true; } template <> inline bool tryParseImpl(DataTypeUUID::FieldType & x, ReadBuffer & rb, const DateLUTImpl *) { UUID tmp; if (!tryReadUUIDText(tmp, rb)) return false; x = tmp.toUnderType(); return true; } /** Throw exception with verbose message when string value is not parsed completely. */ [[noreturn]] inline void throwExceptionForIncompletelyParsedValue(ReadBuffer & read_buffer, const DataTypePtr result_type) { const IDataType & to_type = *result_type; WriteBufferFromOwnString message_buf; message_buf << "Cannot parse string " << quote << String(read_buffer.buffer().begin(), read_buffer.buffer().size()) << " as " << to_type.getName() << ": syntax error"; if (read_buffer.offset()) message_buf << " at position " << read_buffer.offset() << " (parsed just " << quote << String(read_buffer.buffer().begin(), read_buffer.offset()) << ")"; else message_buf << " at begin of string"; // Currently there are no functions toIPv{4,6}Or{Null,Zero} if (isNativeNumber(to_type) && !(to_type.getName() == "IPv4" || to_type.getName() == "IPv6")) message_buf << ". Note: there are to" << to_type.getName() << "OrZero and to" << to_type.getName() << "OrNull functions, which returns zero/NULL instead of throwing exception."; throw Exception(message_buf.str(), ErrorCodes::CANNOT_PARSE_TEXT); } enum class ConvertFromStringExceptionMode { Throw, /// Throw exception if value cannot be parsed. Zero, /// Fill with zero or default if value cannot be parsed. Null /// Return ColumnNullable with NULLs when value cannot be parsed. }; enum class ConvertFromStringParsingMode { Normal, BestEffort, /// Only applicable for DateTime. Will use sophisticated method, that is slower. BestEffortUS }; template struct ConvertThroughParsing { static_assert(std::is_same_v || std::is_same_v, "ConvertThroughParsing is only applicable for String or FixedString data types"); static constexpr bool to_datetime64 = std::is_same_v; // using ToFieldType = typename ToDataType::FieldType; static bool isAllRead(ReadBuffer & in) { /// In case of FixedString, skip zero bytes at end. if constexpr (std::is_same_v) while (!in.eof() && *in.position() == 0) ++in.position(); if (in.eof()) return true; /// Special case, that allows to parse string with DateTime as Date. if (std::is_same_v && (in.buffer().size()) == strlen("YYYY-MM-DD hh:mm:ss")) return true; return false; } template static ColumnPtr execute(const ColumnsWithTypeAndName & arguments, const DataTypePtr & res_type, size_t input_rows_count, Additions additions [[maybe_unused]] = Additions()) { using ColVecTo = typename ToDataType::ColumnType; const DateLUTImpl * local_time_zone [[maybe_unused]] = nullptr; const DateLUTImpl * utc_time_zone [[maybe_unused]] = nullptr; /// For conversion to DateTime type, second argument with time zone could be specified. if constexpr (std::is_same_v || to_datetime64) { const auto result_type = removeNullable(res_type); // Time zone is already figured out during result type resolution, no need to do it here. if (const auto dt_col = checkAndGetDataType(result_type.get())) local_time_zone = &dt_col->getTimeZone(); else { local_time_zone = &extractTimeZoneFromFunctionArguments(arguments, 1, 0); } if constexpr (parsing_mode == ConvertFromStringParsingMode::BestEffort || parsing_mode == ConvertFromStringParsingMode::BestEffortUS) utc_time_zone = &DateLUT::instance("UTC"); } const IColumn * col_from = arguments[0].column.get(); const ColumnString * col_from_string = checkAndGetColumn(col_from); const ColumnFixedString * col_from_fixed_string = checkAndGetColumn(col_from); if (std::is_same_v && !col_from_string) throw Exception("Illegal column " + col_from->getName() + " of first argument of function " + Name::name, ErrorCodes::ILLEGAL_COLUMN); if (std::is_same_v && !col_from_fixed_string) throw Exception("Illegal column " + col_from->getName() + " of first argument of function " + Name::name, ErrorCodes::ILLEGAL_COLUMN); size_t size = input_rows_count; typename ColVecTo::MutablePtr col_to = nullptr; if constexpr (IsDataTypeDecimal) { UInt32 scale = additions; if constexpr (to_datetime64) { ToDataType check_bounds_in_ctor(scale, local_time_zone ? local_time_zone->getTimeZone() : String{}); } else { ToDataType check_bounds_in_ctor(ToDataType::maxPrecision(), scale); } col_to = ColVecTo::create(size, scale); } else col_to = ColVecTo::create(size); typename ColVecTo::Container & vec_to = col_to->getData(); ColumnUInt8::MutablePtr col_null_map_to; ColumnUInt8::Container * vec_null_map_to [[maybe_unused]] = nullptr; if constexpr (exception_mode == ConvertFromStringExceptionMode::Null) { col_null_map_to = ColumnUInt8::create(size); vec_null_map_to = &col_null_map_to->getData(); } const ColumnString::Chars * chars = nullptr; const IColumn::Offsets * offsets = nullptr; size_t fixed_string_size = 0; if constexpr (std::is_same_v) { chars = &col_from_string->getChars(); offsets = &col_from_string->getOffsets(); } else { chars = &col_from_fixed_string->getChars(); fixed_string_size = col_from_fixed_string->getN(); } size_t current_offset = 0; for (size_t i = 0; i < size; ++i) { size_t next_offset = std::is_same_v ? (*offsets)[i] : (current_offset + fixed_string_size); size_t string_size = std::is_same_v ? next_offset - current_offset - 1 : fixed_string_size; ReadBufferFromMemory read_buffer(&(*chars)[current_offset], string_size); if constexpr (exception_mode == ConvertFromStringExceptionMode::Throw) { if constexpr (parsing_mode == ConvertFromStringParsingMode::BestEffort) { if constexpr (to_datetime64) { DateTime64 res = 0; parseDateTime64BestEffort(res, col_to->getScale(), read_buffer, *local_time_zone, *utc_time_zone); vec_to[i] = res; } else { time_t res; parseDateTimeBestEffort(res, read_buffer, *local_time_zone, *utc_time_zone); convertFromTime(vec_to[i], res); } } else if constexpr (parsing_mode == ConvertFromStringParsingMode::BestEffortUS) { if constexpr (to_datetime64) { DateTime64 res = 0; parseDateTime64BestEffortUS(res, col_to->getScale(), read_buffer, *local_time_zone, *utc_time_zone); vec_to[i] = res; } else { time_t res; parseDateTimeBestEffortUS(res, read_buffer, *local_time_zone, *utc_time_zone); convertFromTime(vec_to[i], res); } } else { if constexpr (to_datetime64) { DateTime64 value = 0; readDateTime64Text(value, col_to->getScale(), read_buffer, *local_time_zone); vec_to[i] = value; } else if constexpr (IsDataTypeDecimal) SerializationDecimal::readText( vec_to[i], read_buffer, ToDataType::maxPrecision(), col_to->getScale()); else { parseImpl(vec_to[i], read_buffer, local_time_zone); } } if (!isAllRead(read_buffer)) throwExceptionForIncompletelyParsedValue(read_buffer, res_type); } else { bool parsed; if constexpr (parsing_mode == ConvertFromStringParsingMode::BestEffort) { if constexpr (to_datetime64) { DateTime64 res = 0; parsed = tryParseDateTime64BestEffort(res, col_to->getScale(), read_buffer, *local_time_zone, *utc_time_zone); vec_to[i] = res; } else { time_t res; parsed = tryParseDateTimeBestEffort(res, read_buffer, *local_time_zone, *utc_time_zone); convertFromTime(vec_to[i],res); } } else if constexpr (parsing_mode == ConvertFromStringParsingMode::BestEffortUS) { time_t res; parsed = tryParseDateTimeBestEffortUS(res, read_buffer, *local_time_zone, *utc_time_zone); convertFromTime(vec_to[i],res); } else { if constexpr (to_datetime64) { DateTime64 value = 0; parsed = tryReadDateTime64Text(value, col_to->getScale(), read_buffer, *local_time_zone); vec_to[i] = value; } else if constexpr (IsDataTypeDecimal) parsed = SerializationDecimal::tryReadText( vec_to[i], read_buffer, ToDataType::maxPrecision(), col_to->getScale()); else parsed = tryParseImpl(vec_to[i], read_buffer, local_time_zone); } if (!isAllRead(read_buffer)) parsed = false; if (!parsed) { if constexpr (std::is_same_v) { vec_to[i] = -static_cast(DateLUT::instance().getDayNumOffsetEpoch()); } else { vec_to[i] = static_cast(0); } } if constexpr (exception_mode == ConvertFromStringExceptionMode::Null) (*vec_null_map_to)[i] = !parsed; } current_offset = next_offset; } if constexpr (exception_mode == ConvertFromStringExceptionMode::Null) return ColumnNullable::create(std::move(col_to), std::move(col_null_map_to)); else return col_to; } }; template struct ConvertImpl, DataTypeString>, ToDataType, Name, ConvertDefaultBehaviorTag> : ConvertThroughParsing {}; template struct ConvertImpl, DataTypeFixedString>, ToDataType, Name, ConvertDefaultBehaviorTag> : ConvertThroughParsing {}; template struct ConvertImpl, DataTypeString>, ToDataType, Name, ConvertReturnNullOnErrorTag> : ConvertThroughParsing {}; template struct ConvertImpl, DataTypeFixedString>, ToDataType, Name, ConvertReturnNullOnErrorTag> : ConvertThroughParsing {}; /// Generic conversion of any type from String. Used for complex types: Array and Tuple or types with custom serialization. template struct ConvertImplGenericFromString { static ColumnPtr execute(ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count) { static_assert(std::is_same_v || std::is_same_v, "Can be used only to parse from ColumnString or ColumnFixedString"); const IColumn & col_from = *arguments[0].column; const IDataType & data_type_to = *result_type; if (const StringColumnType * col_from_string = checkAndGetColumn(&col_from)) { auto res = data_type_to.createColumn(); IColumn & column_to = *res; column_to.reserve(input_rows_count); FormatSettings format_settings; auto serialization = data_type_to.getDefaultSerialization(); for (size_t i = 0; i < input_rows_count; ++i) { const auto & val = col_from_string->getDataAt(i); ReadBufferFromMemory read_buffer(val.data, val.size); serialization->deserializeWholeText(column_to, read_buffer, format_settings); if (!read_buffer.eof()) throwExceptionForIncompletelyParsedValue(read_buffer, result_type); } return res; } else throw Exception("Illegal column " + arguments[0].column->getName() + " of first argument of conversion function from string", ErrorCodes::ILLEGAL_COLUMN); } }; template <> struct ConvertImpl : ConvertImpl {}; template <> struct ConvertImpl : ConvertImpl {}; /** If types are identical, just take reference to column. */ template struct ConvertImpl, T, Name, ConvertDefaultBehaviorTag> { template static ColumnPtr execute(const ColumnsWithTypeAndName & arguments, const DataTypePtr &, size_t /*input_rows_count*/, Additions additions [[maybe_unused]] = Additions()) { return arguments[0].column; } }; template struct ConvertImpl { template static ColumnPtr execute(const ColumnsWithTypeAndName & arguments, const DataTypePtr &, size_t /*input_rows_count*/, Additions additions [[maybe_unused]] = Additions()) { return arguments[0].column; } }; /** Conversion from FixedString to String. * Cutting sequences of zero bytes from end of strings. */ template struct ConvertImpl { static ColumnPtr execute(const ColumnsWithTypeAndName & arguments, const DataTypePtr &, size_t /*input_rows_count*/) { if (const ColumnFixedString * col_from = checkAndGetColumn(arguments[0].column.get())) { auto col_to = ColumnString::create(); const ColumnFixedString::Chars & data_from = col_from->getChars(); ColumnString::Chars & data_to = col_to->getChars(); ColumnString::Offsets & offsets_to = col_to->getOffsets(); size_t size = col_from->size(); size_t n = col_from->getN(); data_to.resize(size * (n + 1)); /// + 1 - zero terminator offsets_to.resize(size); size_t offset_from = 0; size_t offset_to = 0; for (size_t i = 0; i < size; ++i) { size_t bytes_to_copy = n; while (bytes_to_copy > 0 && data_from[offset_from + bytes_to_copy - 1] == 0) --bytes_to_copy; memcpy(&data_to[offset_to], &data_from[offset_from], bytes_to_copy); offset_from += n; offset_to += bytes_to_copy; data_to[offset_to] = 0; ++offset_to; offsets_to[i] = offset_to; } data_to.resize(offset_to); return col_to; } else throw Exception("Illegal column " + arguments[0].column->getName() + " of first argument of function " + Name::name, ErrorCodes::ILLEGAL_COLUMN); } }; /// Declared early because used below. struct NameToDate { static constexpr auto name = "toDate"; }; struct NameToDate32 { static constexpr auto name = "toDate32"; }; struct NameToDateTime { static constexpr auto name = "toDateTime"; }; struct NameToDateTime32 { static constexpr auto name = "toDateTime32"; }; struct NameToDateTime64 { static constexpr auto name = "toDateTime64"; }; struct NameToString { static constexpr auto name = "toString"; }; struct NameToDecimal32 { static constexpr auto name = "toDecimal32"; }; struct NameToDecimal64 { static constexpr auto name = "toDecimal64"; }; struct NameToDecimal128 { static constexpr auto name = "toDecimal128"; }; struct NameToDecimal256 { static constexpr auto name = "toDecimal256"; }; #define DEFINE_NAME_TO_INTERVAL(INTERVAL_KIND) \ struct NameToInterval ## INTERVAL_KIND \ { \ static constexpr auto name = "toInterval" #INTERVAL_KIND; \ static constexpr auto kind = IntervalKind::INTERVAL_KIND; \ }; DEFINE_NAME_TO_INTERVAL(Second) DEFINE_NAME_TO_INTERVAL(Minute) DEFINE_NAME_TO_INTERVAL(Hour) DEFINE_NAME_TO_INTERVAL(Day) DEFINE_NAME_TO_INTERVAL(Week) DEFINE_NAME_TO_INTERVAL(Month) DEFINE_NAME_TO_INTERVAL(Quarter) DEFINE_NAME_TO_INTERVAL(Year) #undef DEFINE_NAME_TO_INTERVAL struct NameParseDateTimeBestEffort; struct NameParseDateTimeBestEffortOrZero; struct NameParseDateTimeBestEffortOrNull; template static inline bool isDateTime64(const ColumnsWithTypeAndName & arguments) { if constexpr (std::is_same_v) return true; else if constexpr (std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v) { return (arguments.size() == 2 && isUnsignedInteger(arguments[1].type)) || arguments.size() == 3; } return false; } template class FunctionConvert : public IFunction { public: using Monotonic = MonotonicityImpl; static constexpr auto name = Name::name; static constexpr bool to_decimal = std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v; static constexpr bool to_datetime64 = std::is_same_v; static constexpr bool to_string_or_fixed_string = std::is_same_v || std::is_same_v; static constexpr bool to_date_or_datetime = std::is_same_v || std::is_same_v || std::is_same_v; static FunctionPtr create(ContextPtr) { return std::make_shared(); } static FunctionPtr create() { return std::make_shared(); } String getName() const override { return name; } bool isVariadic() const override { return true; } size_t getNumberOfArguments() const override { return 0; } bool isInjective(const ColumnsWithTypeAndName &) const override { return std::is_same_v; } bool isSuitableForShortCircuitArgumentsExecution(const DataTypesWithConstInfo & arguments) const override { /// TODO: We can make more optimizations here. return !(to_date_or_datetime && isNumber(*arguments[0].type)); } using DefaultReturnTypeGetter = std::function; static DataTypePtr getReturnTypeDefaultImplementationForNulls(const ColumnsWithTypeAndName & arguments, const DefaultReturnTypeGetter & getter) { NullPresence null_presence = getNullPresense(arguments); if (null_presence.has_null_constant) { return makeNullable(std::make_shared()); } if (null_presence.has_nullable) { auto nested_columns = Block(createBlockWithNestedColumns(arguments)); auto return_type = getter(ColumnsWithTypeAndName(nested_columns.begin(), nested_columns.end())); return makeNullable(return_type); } return getter(arguments); } DataTypePtr getReturnTypeImpl(const ColumnsWithTypeAndName & arguments) const override { auto getter = [&] (const auto & args) { return getReturnTypeImplRemovedNullable(args); }; auto res = getReturnTypeDefaultImplementationForNulls(arguments, getter); to_nullable = res->isNullable(); checked_return_type = true; return res; } DataTypePtr getReturnTypeImplRemovedNullable(const ColumnsWithTypeAndName & arguments) const { FunctionArgumentDescriptors mandatory_args = {{"Value", nullptr, nullptr, nullptr}}; FunctionArgumentDescriptors optional_args; if constexpr (to_decimal) { mandatory_args.push_back({"scale", &isNativeInteger, &isColumnConst, "const Integer"}); } if (!to_decimal && isDateTime64(arguments)) { mandatory_args.push_back({"scale", &isNativeInteger, &isColumnConst, "const Integer"}); } // toString(DateTime or DateTime64, [timezone: String]) if ((std::is_same_v && !arguments.empty() && (isDateTime64(arguments[0].type) || isDateTime(arguments[0].type))) // toUnixTimestamp(value[, timezone : String]) || std::is_same_v // toDate(value[, timezone : String]) || std::is_same_v // TODO: shall we allow timestamp argument for toDate? DateTime knows nothing about timezones and this argument is ignored below. // toDate(value[, timezone : String]) || std::is_same_v // toDateTime(value[, timezone: String]) || std::is_same_v // toDateTime64(value, scale : Integer[, timezone: String]) || std::is_same_v) { optional_args.push_back({"timezone", &isString, &isColumnConst, "const String"}); } validateFunctionArgumentTypes(*this, arguments, mandatory_args, optional_args); if constexpr (std::is_same_v) { return std::make_shared(Name::kind); } else if constexpr (to_decimal) { UInt64 scale = extractToDecimalScale(arguments[1]); if constexpr (std::is_same_v) return createDecimalMaxPrecision(scale); else if constexpr (std::is_same_v) return createDecimalMaxPrecision(scale); else if constexpr (std::is_same_v) return createDecimalMaxPrecision(scale); else if constexpr (std::is_same_v) return createDecimalMaxPrecision(scale); throw Exception("Unexpected branch in code of conversion function: it is a bug.", ErrorCodes::LOGICAL_ERROR); } else { // Optional second argument with time zone for DateTime. UInt8 timezone_arg_position = 1; UInt32 scale [[maybe_unused]] = DataTypeDateTime64::default_scale; // DateTime64 requires more arguments: scale and timezone. Since timezone is optional, scale should be first. if (isDateTime64(arguments)) { timezone_arg_position += 1; scale = static_cast(arguments[1].column->get64(0)); if (to_datetime64 || scale != 0) /// toDateTime('xxxx-xx-xx xx:xx:xx', 0) return DateTime return std::make_shared(scale, extractTimeZoneNameFromFunctionArguments(arguments, timezone_arg_position, 0)); return std::make_shared(extractTimeZoneNameFromFunctionArguments(arguments, timezone_arg_position, 0)); } if constexpr (std::is_same_v) return std::make_shared(extractTimeZoneNameFromFunctionArguments(arguments, timezone_arg_position, 0)); else if constexpr (std::is_same_v) throw Exception("Unexpected branch in code of conversion function: it is a bug.", ErrorCodes::LOGICAL_ERROR); else return std::make_shared(); } } /// Function actually uses default implementation for nulls, /// but we need to know if return type is Nullable or not, /// so we use checked_return_type only to intercept the first call to getReturnTypeImpl(...). bool useDefaultImplementationForNulls() const override { bool to_nullable_string = to_nullable && std::is_same_v; return checked_return_type && !to_nullable_string; } bool useDefaultImplementationForConstants() const override { return true; } ColumnNumbers getArgumentsThatAreAlwaysConstant() const override { if constexpr (std::is_same_v) return {2}; return {1}; } bool canBeExecutedOnDefaultArguments() const override { return false; } ColumnPtr executeImpl(const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, size_t input_rows_count) const override { try { return executeInternal(arguments, result_type, input_rows_count); } catch (Exception & e) { /// More convenient error message. if (e.code() == ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF) { e.addMessage("Cannot parse " + result_type->getName() + " from " + arguments[0].type->getName() + ", because value is too short"); } else if (e.code() == ErrorCodes::CANNOT_PARSE_NUMBER || e.code() == ErrorCodes::CANNOT_READ_ARRAY_FROM_TEXT || e.code() == ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED || e.code() == ErrorCodes::CANNOT_PARSE_QUOTED_STRING || e.code() == ErrorCodes::CANNOT_PARSE_ESCAPE_SEQUENCE || e.code() == ErrorCodes::CANNOT_PARSE_DATE || e.code() == ErrorCodes::CANNOT_PARSE_DATETIME || e.code() == ErrorCodes::CANNOT_PARSE_UUID) { e.addMessage("Cannot parse " + result_type->getName() + " from " + arguments[0].type->getName()); } throw; } } bool hasInformationAboutMonotonicity() const override { return Monotonic::has(); } Monotonicity getMonotonicityForRange(const IDataType & type, const Field & left, const Field & right) const override { return Monotonic::get(type, left, right); } private: mutable bool checked_return_type = false; mutable bool to_nullable = false; ColumnPtr executeInternal(const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, size_t input_rows_count) const { if (arguments.empty()) throw Exception{"Function " + getName() + " expects at least 1 argument", ErrorCodes::TOO_FEW_ARGUMENTS_FOR_FUNCTION}; if (result_type->onlyNull()) return result_type->createColumnConstWithDefaultValue(input_rows_count); const DataTypePtr from_type = removeNullable(arguments[0].type); ColumnPtr result_column; auto call = [&](const auto & types, const auto & tag) -> bool { using Types = std::decay_t; using LeftDataType = typename Types::LeftType; using RightDataType = typename Types::RightType; using SpecialTag = std::decay_t; if constexpr (IsDataTypeDecimal) { if constexpr (std::is_same_v) { /// Account for optional timezone argument. if (arguments.size() != 2 && arguments.size() != 3) throw Exception{"Function " + getName() + " expects 2 or 3 arguments for DataTypeDateTime64.", ErrorCodes::TOO_FEW_ARGUMENTS_FOR_FUNCTION}; } else if (arguments.size() != 2) { throw Exception{"Function " + getName() + " expects 2 arguments for Decimal.", ErrorCodes::TOO_FEW_ARGUMENTS_FOR_FUNCTION}; } const ColumnWithTypeAndName & scale_column = arguments[1]; UInt32 scale = extractToDecimalScale(scale_column); result_column = ConvertImpl::execute(arguments, result_type, input_rows_count, scale); } else if constexpr (IsDataTypeDateOrDateTime && std::is_same_v) { const auto * dt64 = assert_cast(arguments[0].type.get()); result_column = ConvertImpl::execute(arguments, result_type, input_rows_count, dt64->getScale()); } else if constexpr (IsDataTypeDecimalOrNumber && IsDataTypeDecimalOrNumber) { using LeftT = typename LeftDataType::FieldType; using RightT = typename RightDataType::FieldType; static constexpr bool bad_left = is_decimal || std::is_floating_point_v || is_big_int_v || is_signed_v; static constexpr bool bad_right = is_decimal || std::is_floating_point_v || is_big_int_v || is_signed_v; /// Disallow int vs UUID conversion (but support int vs UInt128 conversion) if constexpr ((bad_left && std::is_same_v) || (bad_right && std::is_same_v)) { throw Exception("Wrong UUID conversion", ErrorCodes::CANNOT_CONVERT_TYPE); } else { result_column = ConvertImpl::execute(arguments, result_type, input_rows_count); } } else { result_column = ConvertImpl::execute(arguments, result_type, input_rows_count); } return true; }; if (isDateTime64(arguments)) { /// For toDateTime('xxxx-xx-xx xx:xx:xx.00', 2[, 'timezone']) we need to it convert to DateTime64 const ColumnWithTypeAndName & scale_column = arguments[1]; UInt32 scale = extractToDecimalScale(scale_column); if (to_datetime64 || scale != 0) /// When scale = 0, the data type is DateTime otherwise the data type is DateTime64 { if (!callOnIndexAndDataType(from_type->getTypeId(), call, ConvertDefaultBehaviorTag{})) throw Exception("Illegal type " + arguments[0].type->getName() + " of argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); return result_column; } } if constexpr (std::is_same_v) { if (from_type->getCustomSerialization()) return ConvertImplGenericToString::execute(arguments, result_type, input_rows_count); } bool done; if constexpr (to_string_or_fixed_string) { done = callOnIndexAndDataType(from_type->getTypeId(), call, ConvertDefaultBehaviorTag{}); } else { /// We should use ConvertFromStringExceptionMode::Null mode when converting from String (or FixedString) /// to Nullable type, to avoid 'value is too short' error on attempt to parse empty string from NULL values. if (to_nullable && WhichDataType(from_type).isStringOrFixedString()) done = callOnIndexAndDataType(from_type->getTypeId(), call, ConvertReturnNullOnErrorTag{}); else done = callOnIndexAndDataType(from_type->getTypeId(), call, ConvertDefaultBehaviorTag{}); } if (!done) { /// Generic conversion of any type to String. if (std::is_same_v) { return ConvertImplGenericToString::execute(arguments, result_type, input_rows_count); } else throw Exception("Illegal type " + arguments[0].type->getName() + " of argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); } return result_column; } }; /** Function toTOrZero (where T is number of date or datetime type): * try to convert from String to type T through parsing, * if cannot parse, return default value instead of throwing exception. * Function toTOrNull will return Nullable type with NULL when cannot parse. * NOTE Also need to implement tryToUnixTimestamp with timezone. */ template class FunctionConvertFromString : public IFunction { public: static constexpr auto name = Name::name; static constexpr bool to_decimal = std::is_same_v> || std::is_same_v> || std::is_same_v> || std::is_same_v>; static constexpr bool to_datetime64 = std::is_same_v; static FunctionPtr create(ContextPtr) { return std::make_shared(); } static FunctionPtr create() { return std::make_shared(); } String getName() const override { return name; } bool isVariadic() const override { return true; } bool isSuitableForShortCircuitArgumentsExecution(const DataTypesWithConstInfo & /*arguments*/) const override { return true; } size_t getNumberOfArguments() const override { return 0; } bool useDefaultImplementationForConstants() const override { return true; } bool canBeExecutedOnDefaultArguments() const override { return false; } ColumnNumbers getArgumentsThatAreAlwaysConstant() const override { return {1}; } DataTypePtr getReturnTypeImpl(const ColumnsWithTypeAndName & arguments) const override { DataTypePtr res; if (isDateTime64(arguments)) { validateFunctionArgumentTypes(*this, arguments, FunctionArgumentDescriptors{{"string", &isStringOrFixedString, nullptr, "String or FixedString"}}, // optional FunctionArgumentDescriptors{ {"precision", &isUInt8, isColumnConst, "const UInt8"}, {"timezone", &isStringOrFixedString, isColumnConst, "const String or FixedString"}, }); UInt64 scale = to_datetime64 ? DataTypeDateTime64::default_scale : 0; if (arguments.size() > 1) scale = extractToDecimalScale(arguments[1]); const auto timezone = extractTimeZoneNameFromFunctionArguments(arguments, 2, 0); res = scale == 0 ? res = std::make_shared(timezone) : std::make_shared(scale, timezone); } else { if ((arguments.size() != 1 && arguments.size() != 2) || (to_decimal && arguments.size() != 2)) throw Exception("Number of arguments for function " + getName() + " doesn't match: passed " + toString(arguments.size()) + ", should be 1 or 2. Second argument only make sense for DateTime (time zone, optional) and Decimal (scale).", ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH); if (!isStringOrFixedString(arguments[0].type)) { if (this->getName().find("OrZero") != std::string::npos || this->getName().find("OrNull") != std::string::npos) throw Exception("Illegal type " + arguments[0].type->getName() + " of first argument of function " + getName() + ". Conversion functions with postfix 'OrZero' or 'OrNull' should take String argument", ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); else throw Exception("Illegal type " + arguments[0].type->getName() + " of first argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); } if (arguments.size() == 2) { if constexpr (std::is_same_v) { if (!isString(arguments[1].type)) throw Exception("Illegal type " + arguments[1].type->getName() + " of 2nd argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); } else if constexpr (to_decimal) { if (!isInteger(arguments[1].type)) throw Exception("Illegal type " + arguments[1].type->getName() + " of 2nd argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); if (!arguments[1].column) throw Exception("Second argument for function " + getName() + " must be constant", ErrorCodes::ILLEGAL_COLUMN); } else { throw Exception("Number of arguments for function " + getName() + " doesn't match: passed " + toString(arguments.size()) + ", should be 1. Second argument makes sense only for DateTime and Decimal.", ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH); } } if constexpr (std::is_same_v) res = std::make_shared(extractTimeZoneNameFromFunctionArguments(arguments, 1, 0)); else if constexpr (std::is_same_v) throw Exception("LOGICAL ERROR: It is a bug.", ErrorCodes::LOGICAL_ERROR); else if constexpr (to_decimal) { UInt64 scale = extractToDecimalScale(arguments[1]); res = createDecimalMaxPrecision(scale); if (!res) throw Exception("Something wrong with toDecimalNNOrZero() or toDecimalNNOrNull()", ErrorCodes::LOGICAL_ERROR); } else res = std::make_shared(); } if constexpr (exception_mode == ConvertFromStringExceptionMode::Null) res = std::make_shared(res); return res; } template ColumnPtr executeInternal(const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, size_t input_rows_count, UInt32 scale = 0) const { const IDataType * from_type = arguments[0].type.get(); if (checkAndGetDataType(from_type)) { return ConvertThroughParsing::execute( arguments, result_type, input_rows_count, scale); } else if (checkAndGetDataType(from_type)) { return ConvertThroughParsing::execute( arguments, result_type, input_rows_count, scale); } return nullptr; } ColumnPtr executeImpl(const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, size_t input_rows_count) const override { ColumnPtr result_column; if constexpr (to_decimal) result_column = executeInternal(arguments, result_type, input_rows_count, assert_cast(*removeNullable(result_type)).getScale()); else { if (isDateTime64(arguments)) { UInt64 scale = to_datetime64 ? DataTypeDateTime64::default_scale : 0; if (arguments.size() > 1) scale = extractToDecimalScale(arguments[1]); if (scale == 0) result_column = executeInternal(arguments, result_type, input_rows_count); else { result_column = executeInternal(arguments, result_type, input_rows_count, static_cast(scale)); } } else { result_column = executeInternal(arguments, result_type, input_rows_count); } } if (!result_column) throw Exception("Illegal type " + arguments[0].type->getName() + " of argument of function " + getName() + ". Only String or FixedString argument is accepted for try-conversion function." + " For other arguments, use function without 'orZero' or 'orNull'.", ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); return result_column; } }; /// Monotonicity. struct PositiveMonotonicity { static bool has() { return true; } static IFunction::Monotonicity get(const IDataType &, const Field &, const Field &) { return { .is_monotonic = true }; } }; struct UnknownMonotonicity { static bool has() { return false; } static IFunction::Monotonicity get(const IDataType &, const Field &, const Field &) { return { }; } }; template struct ToNumberMonotonicity { static bool has() { return true; } static UInt64 divideByRangeOfType(UInt64 x) { if constexpr (sizeof(T) < sizeof(UInt64)) return x >> (sizeof(T) * 8); else return 0; } static IFunction::Monotonicity get(const IDataType & type, const Field & left, const Field & right) { if (!type.isValueRepresentedByNumber()) return {}; /// If type is same, the conversion is always monotonic. /// (Enum has separate case, because it is different data type) if (checkAndGetDataType>(&type) || checkAndGetDataType>(&type)) return { .is_monotonic = true, .is_always_monotonic = true }; /// Float cases. /// When converting to Float, the conversion is always monotonic. if constexpr (std::is_floating_point_v) return { .is_monotonic = true, .is_always_monotonic = true }; /// If converting from Float, for monotonicity, arguments must fit in range of result type. if (WhichDataType(type).isFloat()) { if (left.isNull() || right.isNull()) return {}; Float64 left_float = left.get(); Float64 right_float = right.get(); if (left_float >= static_cast(std::numeric_limits::min()) && left_float <= static_cast(std::numeric_limits::max()) && right_float >= static_cast(std::numeric_limits::min()) && right_float <= static_cast(std::numeric_limits::max())) return { .is_monotonic = true }; return {}; } /// Integer cases. const bool from_is_unsigned = type.isValueRepresentedByUnsignedInteger(); const bool to_is_unsigned = is_unsigned_v; const size_t size_of_from = type.getSizeOfValueInMemory(); const size_t size_of_to = sizeof(T); const bool left_in_first_half = left.isNull() ? from_is_unsigned : (left.get() >= 0); const bool right_in_first_half = right.isNull() ? !from_is_unsigned : (right.get() >= 0); /// Size of type is the same. if (size_of_from == size_of_to) { if (from_is_unsigned == to_is_unsigned) return { .is_monotonic = true, .is_always_monotonic = true }; if (left_in_first_half == right_in_first_half) return { .is_monotonic = true }; return {}; } /// Size of type is expanded. if (size_of_from < size_of_to) { if (from_is_unsigned == to_is_unsigned) return { .is_monotonic = true, .is_always_monotonic = true }; if (!to_is_unsigned) return { .is_monotonic = true, .is_always_monotonic = true }; /// signed -> unsigned. If arguments from the same half, then function is monotonic. if (left_in_first_half == right_in_first_half) return { .is_monotonic = true }; return {}; } /// Size of type is shrunk. if (size_of_from > size_of_to) { /// Function cannot be monotonic on unbounded ranges. if (left.isNull() || right.isNull()) return {}; /// Function cannot be monotonic when left and right are not on the same ranges. if (divideByRangeOfType(left.get()) != divideByRangeOfType(right.get())) return {}; if (to_is_unsigned) return { .is_monotonic = true }; else { // If To is signed, it's possible that the signedness is different after conversion. So we check it explicitly. const bool is_monotonic = (T(left.get()) >= 0) == (T(right.get()) >= 0); return { .is_monotonic = is_monotonic }; } } __builtin_unreachable(); } }; struct ToDateMonotonicity { static bool has() { return true; } static IFunction::Monotonicity get(const IDataType & type, const Field & left, const Field & right) { auto which = WhichDataType(type); if (which.isDateOrDate32() || which.isDateTime() || which.isDateTime64() || which.isInt8() || which.isInt16() || which.isUInt8() || which.isUInt16()) return { .is_monotonic = true, .is_always_monotonic = true }; else if ( (which.isUInt() && ((left.isNull() || left.get() < 0xFFFF) && (right.isNull() || right.get() >= 0xFFFF))) || (which.isInt() && ((left.isNull() || left.get() < 0xFFFF) && (right.isNull() || right.get() >= 0xFFFF))) || (which.isFloat() && ((left.isNull() || left.get() < 0xFFFF) && (right.isNull() || right.get() >= 0xFFFF))) || !type.isValueRepresentedByNumber()) return {}; else return { .is_monotonic = true, .is_always_monotonic = true }; } }; struct ToDateTimeMonotonicity { static bool has() { return true; } static IFunction::Monotonicity get(const IDataType & type, const Field &, const Field &) { if (type.isValueRepresentedByNumber()) return { .is_monotonic = true, .is_always_monotonic = true }; else return {}; } }; /** The monotonicity for the `toString` function is mainly determined for test purposes. * It is doubtful that anyone is looking to optimize queries with conditions `toString(CounterID) = 34`. */ struct ToStringMonotonicity { static bool has() { return true; } static IFunction::Monotonicity get(const IDataType & type, const Field & left, const Field & right) { IFunction::Monotonicity positive{ .is_monotonic = true }; IFunction::Monotonicity not_monotonic; const auto * type_ptr = &type; if (const auto * low_cardinality_type = checkAndGetDataType(type_ptr)) type_ptr = low_cardinality_type->getDictionaryType().get(); /// `toString` function is monotonous if the argument is Date or Date32 or DateTime or String, or non-negative numbers with the same number of symbols. if (checkDataTypes(type_ptr)) return positive; if (left.isNull() || right.isNull()) return {}; if (left.getType() == Field::Types::UInt64 && right.getType() == Field::Types::UInt64) { return (left.get() == 0 && right.get() == 0) || (floor(log10(left.get())) == floor(log10(right.get()))) ? positive : not_monotonic; } if (left.getType() == Field::Types::Int64 && right.getType() == Field::Types::Int64) { return (left.get() == 0 && right.get() == 0) || (left.get() > 0 && right.get() > 0 && floor(log10(left.get())) == floor(log10(right.get()))) ? positive : not_monotonic; } return not_monotonic; } }; struct NameToUInt8 { static constexpr auto name = "toUInt8"; }; struct NameToUInt16 { static constexpr auto name = "toUInt16"; }; struct NameToUInt32 { static constexpr auto name = "toUInt32"; }; struct NameToUInt64 { static constexpr auto name = "toUInt64"; }; struct NameToUInt128 { static constexpr auto name = "toUInt128"; }; struct NameToUInt256 { static constexpr auto name = "toUInt256"; }; struct NameToInt8 { static constexpr auto name = "toInt8"; }; struct NameToInt16 { static constexpr auto name = "toInt16"; }; struct NameToInt32 { static constexpr auto name = "toInt32"; }; struct NameToInt64 { static constexpr auto name = "toInt64"; }; struct NameToInt128 { static constexpr auto name = "toInt128"; }; struct NameToInt256 { static constexpr auto name = "toInt256"; }; struct NameToFloat32 { static constexpr auto name = "toFloat32"; }; struct NameToFloat64 { static constexpr auto name = "toFloat64"; }; struct NameToUUID { static constexpr auto name = "toUUID"; }; using FunctionToUInt8 = FunctionConvert>; using FunctionToUInt16 = FunctionConvert>; using FunctionToUInt32 = FunctionConvert>; using FunctionToUInt64 = FunctionConvert>; using FunctionToUInt128 = FunctionConvert>; using FunctionToUInt256 = FunctionConvert>; using FunctionToInt8 = FunctionConvert>; using FunctionToInt16 = FunctionConvert>; using FunctionToInt32 = FunctionConvert>; using FunctionToInt64 = FunctionConvert>; using FunctionToInt128 = FunctionConvert>; using FunctionToInt256 = FunctionConvert>; using FunctionToFloat32 = FunctionConvert>; using FunctionToFloat64 = FunctionConvert>; using FunctionToDate = FunctionConvert; using FunctionToDate32 = FunctionConvert; using FunctionToDateTime = FunctionConvert; using FunctionToDateTime32 = FunctionConvert; using FunctionToDateTime64 = FunctionConvert; using FunctionToUUID = FunctionConvert>; using FunctionToString = FunctionConvert; using FunctionToUnixTimestamp = FunctionConvert>; using FunctionToDecimal32 = FunctionConvert, NameToDecimal32, UnknownMonotonicity>; using FunctionToDecimal64 = FunctionConvert, NameToDecimal64, UnknownMonotonicity>; using FunctionToDecimal128 = FunctionConvert, NameToDecimal128, UnknownMonotonicity>; using FunctionToDecimal256 = FunctionConvert, NameToDecimal256, UnknownMonotonicity>; template struct FunctionTo; template <> struct FunctionTo { using Type = FunctionToUInt8; }; template <> struct FunctionTo { using Type = FunctionToUInt16; }; template <> struct FunctionTo { using Type = FunctionToUInt32; }; template <> struct FunctionTo { using Type = FunctionToUInt64; }; template <> struct FunctionTo { using Type = FunctionToUInt128; }; template <> struct FunctionTo { using Type = FunctionToUInt256; }; template <> struct FunctionTo { using Type = FunctionToInt8; }; template <> struct FunctionTo { using Type = FunctionToInt16; }; template <> struct FunctionTo { using Type = FunctionToInt32; }; template <> struct FunctionTo { using Type = FunctionToInt64; }; template <> struct FunctionTo { using Type = FunctionToInt128; }; template <> struct FunctionTo { using Type = FunctionToInt256; }; template <> struct FunctionTo { using Type = FunctionToFloat32; }; template <> struct FunctionTo { using Type = FunctionToFloat64; }; template <> struct FunctionTo { using Type = FunctionToDate; }; template <> struct FunctionTo { using Type = FunctionToDate32; }; template <> struct FunctionTo { using Type = FunctionToDateTime; }; template <> struct FunctionTo { using Type = FunctionToDateTime64; }; template <> struct FunctionTo { using Type = FunctionToUUID; }; template <> struct FunctionTo { using Type = FunctionToString; }; template <> struct FunctionTo { using Type = FunctionToFixedString; }; template <> struct FunctionTo> { using Type = FunctionToDecimal32; }; template <> struct FunctionTo> { using Type = FunctionToDecimal64; }; template <> struct FunctionTo> { using Type = FunctionToDecimal128; }; template <> struct FunctionTo> { using Type = FunctionToDecimal256; }; template struct FunctionTo> : FunctionTo> { }; struct NameToUInt8OrZero { static constexpr auto name = "toUInt8OrZero"; }; struct NameToUInt16OrZero { static constexpr auto name = "toUInt16OrZero"; }; struct NameToUInt32OrZero { static constexpr auto name = "toUInt32OrZero"; }; struct NameToUInt64OrZero { static constexpr auto name = "toUInt64OrZero"; }; struct NameToUInt128OrZero { static constexpr auto name = "toUInt128OrZero"; }; struct NameToUInt256OrZero { static constexpr auto name = "toUInt256OrZero"; }; struct NameToInt8OrZero { static constexpr auto name = "toInt8OrZero"; }; struct NameToInt16OrZero { static constexpr auto name = "toInt16OrZero"; }; struct NameToInt32OrZero { static constexpr auto name = "toInt32OrZero"; }; struct NameToInt64OrZero { static constexpr auto name = "toInt64OrZero"; }; struct NameToInt128OrZero { static constexpr auto name = "toInt128OrZero"; }; struct NameToInt256OrZero { static constexpr auto name = "toInt256OrZero"; }; struct NameToFloat32OrZero { static constexpr auto name = "toFloat32OrZero"; }; struct NameToFloat64OrZero { static constexpr auto name = "toFloat64OrZero"; }; struct NameToDateOrZero { static constexpr auto name = "toDateOrZero"; }; struct NameToDate32OrZero { static constexpr auto name = "toDate32OrZero"; }; struct NameToDateTimeOrZero { static constexpr auto name = "toDateTimeOrZero"; }; struct NameToDateTime64OrZero { static constexpr auto name = "toDateTime64OrZero"; }; struct NameToDecimal32OrZero { static constexpr auto name = "toDecimal32OrZero"; }; struct NameToDecimal64OrZero { static constexpr auto name = "toDecimal64OrZero"; }; struct NameToDecimal128OrZero { static constexpr auto name = "toDecimal128OrZero"; }; struct NameToDecimal256OrZero { static constexpr auto name = "toDecimal256OrZero"; }; struct NameToUUIDOrZero { static constexpr auto name = "toUUIDOrZero"; }; using FunctionToUInt8OrZero = FunctionConvertFromString; using FunctionToUInt16OrZero = FunctionConvertFromString; using FunctionToUInt32OrZero = FunctionConvertFromString; using FunctionToUInt64OrZero = FunctionConvertFromString; using FunctionToUInt128OrZero = FunctionConvertFromString; using FunctionToUInt256OrZero = FunctionConvertFromString; using FunctionToInt8OrZero = FunctionConvertFromString; using FunctionToInt16OrZero = FunctionConvertFromString; using FunctionToInt32OrZero = FunctionConvertFromString; using FunctionToInt64OrZero = FunctionConvertFromString; using FunctionToInt128OrZero = FunctionConvertFromString; using FunctionToInt256OrZero = FunctionConvertFromString; using FunctionToFloat32OrZero = FunctionConvertFromString; using FunctionToFloat64OrZero = FunctionConvertFromString; using FunctionToDateOrZero = FunctionConvertFromString; using FunctionToDate32OrZero = FunctionConvertFromString; using FunctionToDateTimeOrZero = FunctionConvertFromString; using FunctionToDateTime64OrZero = FunctionConvertFromString; using FunctionToDecimal32OrZero = FunctionConvertFromString, NameToDecimal32OrZero, ConvertFromStringExceptionMode::Zero>; using FunctionToDecimal64OrZero = FunctionConvertFromString, NameToDecimal64OrZero, ConvertFromStringExceptionMode::Zero>; using FunctionToDecimal128OrZero = FunctionConvertFromString, NameToDecimal128OrZero, ConvertFromStringExceptionMode::Zero>; using FunctionToDecimal256OrZero = FunctionConvertFromString, NameToDecimal256OrZero, ConvertFromStringExceptionMode::Zero>; using FunctionToUUIDOrZero = FunctionConvertFromString; struct NameToUInt8OrNull { static constexpr auto name = "toUInt8OrNull"; }; struct NameToUInt16OrNull { static constexpr auto name = "toUInt16OrNull"; }; struct NameToUInt32OrNull { static constexpr auto name = "toUInt32OrNull"; }; struct NameToUInt64OrNull { static constexpr auto name = "toUInt64OrNull"; }; struct NameToUInt128OrNull { static constexpr auto name = "toUInt128OrNull"; }; struct NameToUInt256OrNull { static constexpr auto name = "toUInt256OrNull"; }; struct NameToInt8OrNull { static constexpr auto name = "toInt8OrNull"; }; struct NameToInt16OrNull { static constexpr auto name = "toInt16OrNull"; }; struct NameToInt32OrNull { static constexpr auto name = "toInt32OrNull"; }; struct NameToInt64OrNull { static constexpr auto name = "toInt64OrNull"; }; struct NameToInt128OrNull { static constexpr auto name = "toInt128OrNull"; }; struct NameToInt256OrNull { static constexpr auto name = "toInt256OrNull"; }; struct NameToFloat32OrNull { static constexpr auto name = "toFloat32OrNull"; }; struct NameToFloat64OrNull { static constexpr auto name = "toFloat64OrNull"; }; struct NameToDateOrNull { static constexpr auto name = "toDateOrNull"; }; struct NameToDate32OrNull { static constexpr auto name = "toDate32OrNull"; }; struct NameToDateTimeOrNull { static constexpr auto name = "toDateTimeOrNull"; }; struct NameToDateTime64OrNull { static constexpr auto name = "toDateTime64OrNull"; }; struct NameToDecimal32OrNull { static constexpr auto name = "toDecimal32OrNull"; }; struct NameToDecimal64OrNull { static constexpr auto name = "toDecimal64OrNull"; }; struct NameToDecimal128OrNull { static constexpr auto name = "toDecimal128OrNull"; }; struct NameToDecimal256OrNull { static constexpr auto name = "toDecimal256OrNull"; }; struct NameToUUIDOrNull { static constexpr auto name = "toUUIDOrNull"; }; using FunctionToUInt8OrNull = FunctionConvertFromString; using FunctionToUInt16OrNull = FunctionConvertFromString; using FunctionToUInt32OrNull = FunctionConvertFromString; using FunctionToUInt64OrNull = FunctionConvertFromString; using FunctionToUInt128OrNull = FunctionConvertFromString; using FunctionToUInt256OrNull = FunctionConvertFromString; using FunctionToInt8OrNull = FunctionConvertFromString; using FunctionToInt16OrNull = FunctionConvertFromString; using FunctionToInt32OrNull = FunctionConvertFromString; using FunctionToInt64OrNull = FunctionConvertFromString; using FunctionToInt128OrNull = FunctionConvertFromString; using FunctionToInt256OrNull = FunctionConvertFromString; using FunctionToFloat32OrNull = FunctionConvertFromString; using FunctionToFloat64OrNull = FunctionConvertFromString; using FunctionToDateOrNull = FunctionConvertFromString; using FunctionToDate32OrNull = FunctionConvertFromString; using FunctionToDateTimeOrNull = FunctionConvertFromString; using FunctionToDateTime64OrNull = FunctionConvertFromString; using FunctionToDecimal32OrNull = FunctionConvertFromString, NameToDecimal32OrNull, ConvertFromStringExceptionMode::Null>; using FunctionToDecimal64OrNull = FunctionConvertFromString, NameToDecimal64OrNull, ConvertFromStringExceptionMode::Null>; using FunctionToDecimal128OrNull = FunctionConvertFromString, NameToDecimal128OrNull, ConvertFromStringExceptionMode::Null>; using FunctionToDecimal256OrNull = FunctionConvertFromString, NameToDecimal256OrNull, ConvertFromStringExceptionMode::Null>; using FunctionToUUIDOrNull = FunctionConvertFromString; struct NameParseDateTimeBestEffort { static constexpr auto name = "parseDateTimeBestEffort"; }; struct NameParseDateTimeBestEffortOrZero { static constexpr auto name = "parseDateTimeBestEffortOrZero"; }; struct NameParseDateTimeBestEffortOrNull { static constexpr auto name = "parseDateTimeBestEffortOrNull"; }; struct NameParseDateTimeBestEffortUS { static constexpr auto name = "parseDateTimeBestEffortUS"; }; struct NameParseDateTimeBestEffortUSOrZero { static constexpr auto name = "parseDateTimeBestEffortUSOrZero"; }; struct NameParseDateTimeBestEffortUSOrNull { static constexpr auto name = "parseDateTimeBestEffortUSOrNull"; }; struct NameParseDateTime32BestEffort { static constexpr auto name = "parseDateTime32BestEffort"; }; struct NameParseDateTime32BestEffortOrZero { static constexpr auto name = "parseDateTime32BestEffortOrZero"; }; struct NameParseDateTime32BestEffortOrNull { static constexpr auto name = "parseDateTime32BestEffortOrNull"; }; struct NameParseDateTime64BestEffort { static constexpr auto name = "parseDateTime64BestEffort"; }; struct NameParseDateTime64BestEffortOrZero { static constexpr auto name = "parseDateTime64BestEffortOrZero"; }; struct NameParseDateTime64BestEffortOrNull { static constexpr auto name = "parseDateTime64BestEffortOrNull"; }; using FunctionParseDateTimeBestEffort = FunctionConvertFromString< DataTypeDateTime, NameParseDateTimeBestEffort, ConvertFromStringExceptionMode::Throw, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTimeBestEffortOrZero = FunctionConvertFromString< DataTypeDateTime, NameParseDateTimeBestEffortOrZero, ConvertFromStringExceptionMode::Zero, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTimeBestEffortOrNull = FunctionConvertFromString< DataTypeDateTime, NameParseDateTimeBestEffortOrNull, ConvertFromStringExceptionMode::Null, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTimeBestEffortUS = FunctionConvertFromString< DataTypeDateTime, NameParseDateTimeBestEffortUS, ConvertFromStringExceptionMode::Throw, ConvertFromStringParsingMode::BestEffortUS>; using FunctionParseDateTimeBestEffortUSOrZero = FunctionConvertFromString< DataTypeDateTime, NameParseDateTimeBestEffortUSOrZero, ConvertFromStringExceptionMode::Zero, ConvertFromStringParsingMode::BestEffortUS>; using FunctionParseDateTimeBestEffortUSOrNull = FunctionConvertFromString< DataTypeDateTime, NameParseDateTimeBestEffortUSOrNull, ConvertFromStringExceptionMode::Null, ConvertFromStringParsingMode::BestEffortUS>; using FunctionParseDateTime32BestEffort = FunctionConvertFromString< DataTypeDateTime, NameParseDateTime32BestEffort, ConvertFromStringExceptionMode::Throw, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTime32BestEffortOrZero = FunctionConvertFromString< DataTypeDateTime, NameParseDateTime32BestEffortOrZero, ConvertFromStringExceptionMode::Zero, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTime32BestEffortOrNull = FunctionConvertFromString< DataTypeDateTime, NameParseDateTime32BestEffortOrNull, ConvertFromStringExceptionMode::Null, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTime64BestEffort = FunctionConvertFromString< DataTypeDateTime64, NameParseDateTime64BestEffort, ConvertFromStringExceptionMode::Throw, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTime64BestEffortOrZero = FunctionConvertFromString< DataTypeDateTime64, NameParseDateTime64BestEffortOrZero, ConvertFromStringExceptionMode::Zero, ConvertFromStringParsingMode::BestEffort>; using FunctionParseDateTime64BestEffortOrNull = FunctionConvertFromString< DataTypeDateTime64, NameParseDateTime64BestEffortOrNull, ConvertFromStringExceptionMode::Null, ConvertFromStringParsingMode::BestEffort>; class ExecutableFunctionCast : public IExecutableFunction { public: using WrapperType = std::function; struct Diagnostic { std::string column_from; std::string column_to; }; explicit ExecutableFunctionCast( WrapperType && wrapper_function_, const char * name_, std::optional diagnostic_) : wrapper_function(std::move(wrapper_function_)), name(name_), diagnostic(std::move(diagnostic_)) {} String getName() const override { return name; } protected: ColumnPtr executeImpl(const ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, size_t input_rows_count) const override { /// drop second argument, pass others ColumnsWithTypeAndName new_arguments{arguments.front()}; if (arguments.size() > 2) new_arguments.insert(std::end(new_arguments), std::next(std::begin(arguments), 2), std::end(arguments)); try { return wrapper_function(new_arguments, result_type, nullptr, input_rows_count); } catch (Exception & e) { if (diagnostic) e.addMessage("while converting source column " + backQuoteIfNeed(diagnostic->column_from) + " to destination column " + backQuoteIfNeed(diagnostic->column_to)); throw; } } bool useDefaultImplementationForNulls() const override { return false; } bool useDefaultImplementationForConstants() const override { return true; } bool useDefaultImplementationForLowCardinalityColumns() const override { return false; } ColumnNumbers getArgumentsThatAreAlwaysConstant() const override { return {1}; } private: WrapperType wrapper_function; const char * name; std::optional diagnostic; }; struct CastName { static constexpr auto name = "CAST"; }; struct CastInternalName { static constexpr auto name = "_CAST"; }; enum class CastType { nonAccurate, accurate, accurateOrNull }; class FunctionCastBase : public IFunctionBase { public: using MonotonicityForRange = std::function; using Diagnostic = ExecutableFunctionCast::Diagnostic; }; template class FunctionCast final : public FunctionCastBase { public: using WrapperType = std::function; FunctionCast(const char * cast_name_ , MonotonicityForRange && monotonicity_for_range_ , const DataTypes & argument_types_ , const DataTypePtr & return_type_ , std::optional diagnostic_ , CastType cast_type_ , bool cast_ipv4_ipv6_default_on_conversion_error_) : cast_name(cast_name_), monotonicity_for_range(std::move(monotonicity_for_range_)) , argument_types(argument_types_), return_type(return_type_), diagnostic(std::move(diagnostic_)) , cast_type(cast_type_) , cast_ipv4_ipv6_default_on_conversion_error(cast_ipv4_ipv6_default_on_conversion_error_) { } const DataTypes & getArgumentTypes() const override { return argument_types; } const DataTypePtr & getResultType() const override { return return_type; } ExecutableFunctionPtr prepare(const ColumnsWithTypeAndName & /*sample_columns*/) const override { try { return std::make_unique( prepareUnpackDictionaries(getArgumentTypes()[0], getResultType()), cast_name, diagnostic); } catch (Exception & e) { if (diagnostic) e.addMessage("while converting source column " + backQuoteIfNeed(diagnostic->column_from) + " to destination column " + backQuoteIfNeed(diagnostic->column_to)); throw; } } String getName() const override { return cast_name; } bool isDeterministic() const override { return true; } bool isDeterministicInScopeOfQuery() const override { return true; } bool isSuitableForShortCircuitArgumentsExecution(const DataTypesWithConstInfo & /*arguments*/) const override { return true; } bool hasInformationAboutMonotonicity() const override { return static_cast(monotonicity_for_range); } Monotonicity getMonotonicityForRange(const IDataType & type, const Field & left, const Field & right) const override { return monotonicity_for_range(type, left, right); } private: const char * cast_name; MonotonicityForRange monotonicity_for_range; DataTypes argument_types; DataTypePtr return_type; std::optional diagnostic; CastType cast_type; bool cast_ipv4_ipv6_default_on_conversion_error; static WrapperType createFunctionAdaptor(FunctionPtr function, const DataTypePtr & from_type) { auto function_adaptor = std::make_unique(function)->build({ColumnWithTypeAndName{nullptr, from_type, ""}}); return [function_adaptor] (ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count) { return function_adaptor->execute(arguments, result_type, input_rows_count); }; } static WrapperType createToNullableColumnWrapper() { return [] (ColumnsWithTypeAndName &, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count) { ColumnPtr res = result_type->createColumn(); ColumnUInt8::Ptr col_null_map_to = ColumnUInt8::create(input_rows_count, true); return ColumnNullable::create(res->cloneResized(input_rows_count), std::move(col_null_map_to)); }; } template WrapperType createWrapper(const DataTypePtr & from_type, const ToDataType * const to_type, bool requested_result_is_nullable) const { TypeIndex from_type_index = from_type->getTypeId(); WhichDataType which(from_type_index); 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(from_type.get())) { /// In case when converting to Nullable type, we apply different parsing rule, /// that will not throw an exception but return NULL in case of malformed input. FunctionPtr function = FunctionConvertFromString::create(); return createFunctionAdaptor(function, from_type); } else if (!can_apply_accurate_cast) { FunctionPtr function = FunctionTo::Type::create(); return createFunctionAdaptor(function, from_type); } auto wrapper_cast_type = cast_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; auto res = callOnIndexAndDataType(from_type_index, [&](const auto & types) -> bool { using Types = std::decay_t; using LeftDataType = typename Types::LeftType; using RightDataType = typename Types::RightType; if constexpr (IsDataTypeNumber && IsDataTypeNumber) { if (wrapper_cast_type == CastType::accurate) { result_column = ConvertImpl::execute( arguments, result_type, input_rows_count, AccurateConvertStrategyAdditions()); } else { result_column = ConvertImpl::execute( arguments, result_type, input_rows_count, AccurateOrNullConvertStrategyAdditions()); } return true; } return false; }); /// Additionally check if callOnIndexAndDataType wasn't called at all. if (!res) { 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(ErrorCodes::CANNOT_CONVERT_TYPE, "Conversion from {} to {} is not supported", from_type_index, to_type->getName()); } } return result_column; }; } template WrapperType createBoolWrapper(const DataTypePtr & from_type, const ToDataType * const to_type, bool requested_result_is_nullable) const { if (checkAndGetDataType(from_type.get())) { return &ConvertImplGenericFromString::execute; } return createWrapper(from_type, to_type, requested_result_is_nullable); } WrapperType createUInt8ToUInt8Wrapper(const DataTypePtr from_type, const DataTypePtr to_type) const { return [from_type, to_type] (ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable *, size_t /*input_rows_count*/) -> ColumnPtr { if (isBool(from_type) || !isBool(to_type)) return arguments.front().column; /// Special case when we convert UInt8 column to Bool column. /// both columns have type UInt8, but we shouldn't use identity wrapper, /// because Bool column can contain only 0 and 1. auto res_column = to_type->createColumn(); const auto & data_from = checkAndGetColumn(arguments[0].column.get())->getData(); auto & data_to = assert_cast(res_column.get())->getData(); data_to.resize(data_from.size()); for (size_t i = 0; i != data_from.size(); ++i) data_to[i] = static_cast(data_from[i]); return res_column; }; } static WrapperType createStringWrapper(const DataTypePtr & from_type) { FunctionPtr function = FunctionToString::create(); return createFunctionAdaptor(function, from_type); } WrapperType createFixedStringWrapper(const DataTypePtr & from_type, const size_t N) const { if (!isStringOrFixedString(from_type)) throw Exception{"CAST AS FixedString is only implemented for types String and FixedString", ErrorCodes::NOT_IMPLEMENTED}; 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) return FunctionToFixedString::executeForN(arguments, N); else return FunctionToFixedString::executeForN(arguments, N); }; } template std::enable_if_t, WrapperType> createDecimalWrapper(const DataTypePtr & from_type, const ToDataType * to_type, bool requested_result_is_nullable) const { TypeIndex type_index = from_type->getTypeId(); UInt32 scale = to_type->getScale(); WhichDataType which(type_index); bool ok = which.isNativeInt() || which.isNativeUInt() || which.isDecimal() || which.isFloat() || which.isDateOrDate32() || which.isDateTime() || which.isDateTime64() || which.isStringOrFixedString(); if (!ok) { 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 wrapper_cast_type = cast_type; return [wrapper_cast_type, type_index, scale, to_type, requested_result_is_nullable] (ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *column_nullable, size_t input_rows_count) { ColumnPtr result_column; auto res = callOnIndexAndDataType(type_index, [&](const auto & types) -> bool { using Types = std::decay_t; using LeftDataType = typename Types::LeftType; using RightDataType = typename Types::RightType; if constexpr (IsDataTypeDecimalOrNumber && IsDataTypeDecimalOrNumber && !std::is_same_v) { if (wrapper_cast_type == CastType::accurate) { AccurateConvertStrategyAdditions additions; additions.scale = scale; result_column = ConvertImpl::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::execute( arguments, result_type, input_rows_count, additions); return true; } } else if constexpr (std::is_same_v) { if (requested_result_is_nullable) { /// Consistent with CAST(Nullable(String) AS Nullable(Numbers)) /// In case when converting to Nullable type, we apply different parsing rule, /// that will not throw an exception but return NULL in case of malformed input. result_column = ConvertImpl::execute( arguments, result_type, input_rows_count, scale); return true; } } result_column = ConvertImpl::execute(arguments, result_type, input_rows_count, scale); return true; }); /// Additionally check if callOnIndexAndDataType wasn't called at all. if (!res) { 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(ErrorCodes::CANNOT_CONVERT_TYPE, "Conversion from {} to {} is not supported", type_index, to_type->getName()); } return result_column; }; } WrapperType createAggregateFunctionWrapper(const DataTypePtr & from_type_untyped, const DataTypeAggregateFunction * to_type) const { /// Conversion from String through parsing. if (checkAndGetDataType(from_type_untyped.get())) { return &ConvertImplGenericFromString::execute; } else { if (cast_type == CastType::accurateOrNull) return createToNullableColumnWrapper(); else throw Exception{"Conversion from " + from_type_untyped->getName() + " to " + to_type->getName() + " is not supported", ErrorCodes::CANNOT_CONVERT_TYPE}; } } WrapperType createArrayWrapper(const DataTypePtr & from_type_untyped, const DataTypeArray & to_type) const { /// Conversion from String through parsing. if (checkAndGetDataType(from_type_untyped.get())) { return &ConvertImplGenericFromString::execute; } const auto * from_type = checkAndGetDataType(from_type_untyped.get()); const auto * from_type_map = checkAndGetDataType(from_type_untyped.get()); /// Convert from Map if (from_type_map) from_type = checkAndGetDataType(from_type_map->getNestedType().get()); if (!from_type) { throw Exception(ErrorCodes::TYPE_MISMATCH, "CAST AS Array can only be performed between same-dimensional Array, Map or String types"); } DataTypePtr from_nested_type = from_type->getNestedType(); /// In query SELECT CAST([] AS Array(Array(String))) from type is Array(Nothing) bool from_empty_array = isNothing(from_nested_type); if (from_type->getNumberOfDimensions() != to_type.getNumberOfDimensions() && !from_empty_array) throw Exception(ErrorCodes::TYPE_MISMATCH, "CAST AS Array can only be performed between same-dimensional array types"); const DataTypePtr & to_nested_type = to_type.getNestedType(); /// Prepare nested type conversion const auto nested_function = prepareUnpackDictionaries(from_nested_type, to_nested_type); return [nested_function, from_nested_type, to_nested_type]( ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable * nullable_source, size_t /*input_rows_count*/) -> ColumnPtr { const auto & argument_column = arguments.front(); const ColumnArray * col_array = nullptr; if (const ColumnMap * col_map = checkAndGetColumn(argument_column.column.get())) col_array = &col_map->getNestedColumn(); else col_array = checkAndGetColumn(argument_column.column.get()); if (col_array) { /// create columns for converting nested column containing original and result columns ColumnsWithTypeAndName nested_columns{{ col_array->getDataPtr(), from_nested_type, "" }}; /// convert nested column auto result_column = nested_function(nested_columns, to_nested_type, nullable_source, nested_columns.front().column->size()); /// set converted nested column to result return ColumnArray::create(result_column, col_array->getOffsetsPtr()); } else { throw Exception(ErrorCodes::LOGICAL_ERROR, "Illegal column {} for function CAST AS Array", argument_column.column->getName()); } }; } using ElementWrappers = std::vector; ElementWrappers getElementWrappers(const DataTypes & from_element_types, const DataTypes & to_element_types) const { ElementWrappers element_wrappers; element_wrappers.reserve(from_element_types.size()); /// Create conversion wrapper for each element in tuple for (size_t i = 0; i < from_element_types.size(); ++i) { const DataTypePtr & from_element_type = from_element_types[i]; const DataTypePtr & to_element_type = to_element_types[i]; element_wrappers.push_back(prepareUnpackDictionaries(from_element_type, to_element_type)); } return element_wrappers; } WrapperType createTupleWrapper(const DataTypePtr & from_type_untyped, const DataTypeTuple * to_type) const { /// Conversion from String through parsing. if (checkAndGetDataType(from_type_untyped.get())) { return &ConvertImplGenericFromString::execute; } const auto * from_type = checkAndGetDataType(from_type_untyped.get()); if (!from_type) throw Exception{"CAST AS Tuple can only be performed between tuple types or from String.\nLeft type: " + from_type_untyped->getName() + ", right type: " + to_type->getName(), ErrorCodes::TYPE_MISMATCH}; if (from_type->getElements().size() != to_type->getElements().size()) throw Exception{"CAST AS Tuple can only be performed between tuple types with the same number of elements or from String.\n" "Left type: " + from_type->getName() + ", right type: " + to_type->getName(), ErrorCodes::TYPE_MISMATCH}; const auto & from_element_types = from_type->getElements(); const auto & to_element_types = to_type->getElements(); auto element_wrappers = getElementWrappers(from_element_types, to_element_types); return [element_wrappers, from_element_types, to_element_types] (ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable * nullable_source, size_t input_rows_count) -> ColumnPtr { const auto * col = arguments.front().column.get(); size_t tuple_size = from_element_types.size(); const ColumnTuple & column_tuple = typeid_cast(*col); Columns converted_columns(tuple_size); /// invoke conversion for each element for (size_t i = 0; i < tuple_size; ++i) { ColumnsWithTypeAndName element = {{column_tuple.getColumns()[i], from_element_types[i], "" }}; converted_columns[i] = element_wrappers[i](element, to_element_types[i], nullable_source, input_rows_count); } return ColumnTuple::create(converted_columns); }; } /// The case of: tuple([key1, key2, ..., key_n], [value1, value2, ..., value_n]) WrapperType createTupleToMapWrapper(const DataTypes & from_kv_types, const DataTypes & to_kv_types) const { return [element_wrappers = getElementWrappers(from_kv_types, to_kv_types), from_kv_types, to_kv_types] (ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable * nullable_source, size_t /*input_rows_count*/) -> ColumnPtr { const auto * col = arguments.front().column.get(); const auto & column_tuple = assert_cast(*col); Columns offsets(2); Columns converted_columns(2); for (size_t i = 0; i < 2; ++i) { const auto & column_array = assert_cast(column_tuple.getColumn(i)); ColumnsWithTypeAndName element = {{column_array.getDataPtr(), from_kv_types[i], ""}}; converted_columns[i] = element_wrappers[i](element, to_kv_types[i], nullable_source, (element[0].column)->size()); offsets[i] = column_array.getOffsetsPtr(); } const auto & keys_offsets = assert_cast(*offsets[0]).getData(); const auto & values_offsets = assert_cast(*offsets[1]).getData(); if (keys_offsets != values_offsets) throw Exception(ErrorCodes::TYPE_MISMATCH, "CAST AS Map can only be performed from tuple of arrays with equal sizes."); return ColumnMap::create(converted_columns[0], converted_columns[1], offsets[0]); }; } WrapperType createMapToMapWrrapper(const DataTypes & from_kv_types, const DataTypes & to_kv_types) const { return [element_wrappers = getElementWrappers(from_kv_types, to_kv_types), from_kv_types, to_kv_types] (ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable * nullable_source, size_t /*input_rows_count*/) -> ColumnPtr { const auto * col = arguments.front().column.get(); const auto & column_map = typeid_cast(*col); const auto & nested_data = column_map.getNestedData(); Columns converted_columns(2); for (size_t i = 0; i < 2; ++i) { ColumnsWithTypeAndName element = {{nested_data.getColumnPtr(i), from_kv_types[i], ""}}; converted_columns[i] = element_wrappers[i](element, to_kv_types[i], nullable_source, (element[0].column)->size()); } return ColumnMap::create(converted_columns[0], converted_columns[1], column_map.getNestedColumn().getOffsetsPtr()); }; } /// The case of: [(key1, value1), (key2, value2), ...] WrapperType createArrayToMapWrrapper(const DataTypes & from_kv_types, const DataTypes & to_kv_types) const { return [element_wrappers = getElementWrappers(from_kv_types, to_kv_types), from_kv_types, to_kv_types] (ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable * nullable_source, size_t /*input_rows_count*/) -> ColumnPtr { const auto * col = arguments.front().column.get(); const auto & column_array = typeid_cast(*col); const auto & nested_data = typeid_cast(column_array.getData()); Columns converted_columns(2); for (size_t i = 0; i < 2; ++i) { ColumnsWithTypeAndName element = {{nested_data.getColumnPtr(i), from_kv_types[i], ""}}; converted_columns[i] = element_wrappers[i](element, to_kv_types[i], nullable_source, (element[0].column)->size()); } return ColumnMap::create(converted_columns[0], converted_columns[1], column_array.getOffsetsPtr()); }; } WrapperType createMapWrapper(const DataTypePtr & from_type_untyped, const DataTypeMap * to_type) const { if (const auto * from_tuple = checkAndGetDataType(from_type_untyped.get())) { if (from_tuple->getElements().size() != 2) throw Exception{"CAST AS Map from tuple requeires 2 elements.\n" "Left type: " + from_tuple->getName() + ", right type: " + to_type->getName(), ErrorCodes::TYPE_MISMATCH}; DataTypes from_kv_types; const auto & to_kv_types = to_type->getKeyValueTypes(); for (const auto & elem : from_tuple->getElements()) { const auto * type_array = checkAndGetDataType(elem.get()); if (!type_array) throw Exception(ErrorCodes::TYPE_MISMATCH, "CAST AS Map can only be performed from tuples of array. Got: {}", from_tuple->getName()); from_kv_types.push_back(type_array->getNestedType()); } return createTupleToMapWrapper(from_kv_types, to_kv_types); } else if (const auto * from_array = typeid_cast(from_type_untyped.get())) { const auto * nested_tuple = typeid_cast(from_array->getNestedType().get()); if (!nested_tuple || nested_tuple->getElements().size() != 2) throw Exception{"CAST AS Map from array requeires nested tuple of 2 elements.\n" "Left type: " + from_array->getName() + ", right type: " + to_type->getName(), ErrorCodes::TYPE_MISMATCH}; return createArrayToMapWrrapper(nested_tuple->getElements(), to_type->getKeyValueTypes()); } else if (const auto * from_type = checkAndGetDataType(from_type_untyped.get())) { return createMapToMapWrrapper(from_type->getKeyValueTypes(), to_type->getKeyValueTypes()); } else { throw Exception{"Unsupported types to CAST AS Map\n" "Left type: " + from_type_untyped->getName() + ", right type: " + to_type->getName(), ErrorCodes::TYPE_MISMATCH}; } } template WrapperType createEnumWrapper(const DataTypePtr & from_type, const DataTypeEnum * to_type) const { using EnumType = DataTypeEnum; using Function = typename FunctionTo::Type; if (const auto * from_enum8 = checkAndGetDataType(from_type.get())) checkEnumToEnumConversion(from_enum8, to_type); else if (const auto * from_enum16 = checkAndGetDataType(from_type.get())) checkEnumToEnumConversion(from_enum16, to_type); if (checkAndGetDataType(from_type.get())) return createStringToEnumWrapper(); else if (checkAndGetDataType(from_type.get())) return createStringToEnumWrapper(); else if (isNativeNumber(from_type) || isEnum(from_type)) { auto function = Function::create(); return createFunctionAdaptor(function, from_type); } else { 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}; } } template void checkEnumToEnumConversion(const EnumTypeFrom * from_type, const EnumTypeTo * to_type) const { const auto & from_values = from_type->getValues(); const auto & to_values = to_type->getValues(); using ValueType = std::common_type_t; using NameValuePair = std::pair; using EnumValues = std::vector; EnumValues name_intersection; std::set_intersection(std::begin(from_values), std::end(from_values), std::begin(to_values), std::end(to_values), std::back_inserter(name_intersection), [] (auto && from, auto && to) { return from.first < to.first; }); for (const auto & name_value : name_intersection) { const auto & old_value = name_value.second; const auto & new_value = to_type->getValue(name_value.first); if (old_value != new_value) throw Exception{"Enum conversion changes value for element '" + name_value.first + "' from " + toString(old_value) + " to " + toString(new_value), ErrorCodes::CANNOT_CONVERT_TYPE}; } } template WrapperType createStringToEnumWrapper() const { const char * function_name = cast_name; return [function_name] ( ColumnsWithTypeAndName & arguments, const DataTypePtr & res_type, const ColumnNullable * nullable_col, size_t /*input_rows_count*/) { const auto & first_col = arguments.front().column.get(); const auto & result_type = typeid_cast(*res_type); const ColumnStringType * col = typeid_cast(first_col); if (col && nullable_col && nullable_col->size() != col->size()) throw Exception("ColumnNullable is not compatible with original", ErrorCodes::LOGICAL_ERROR); if (col) { const auto size = col->size(); auto res = result_type.createColumn(); auto & out_data = static_cast(*res).getData(); out_data.resize(size); auto default_enum_value = result_type.getValues().front().second; if (nullable_col) { for (size_t i = 0; i < size; ++i) { if (!nullable_col->isNullAt(i)) out_data[i] = result_type.getValue(col->getDataAt(i)); else out_data[i] = default_enum_value; } } else { for (size_t i = 0; i < size; ++i) out_data[i] = result_type.getValue(col->getDataAt(i)); } return res; } else throw Exception{"Unexpected column " + first_col->getName() + " as first argument of function " + function_name, ErrorCodes::LOGICAL_ERROR}; }; } static WrapperType createIdentityWrapper(const DataTypePtr &) { return [] (ColumnsWithTypeAndName & arguments, const DataTypePtr &, const ColumnNullable *, size_t /*input_rows_count*/) { return arguments.front().column; }; } static WrapperType createNothingWrapper(const IDataType * to_type) { ColumnPtr res = to_type->createColumnConstWithDefaultValue(1); return [res] (ColumnsWithTypeAndName &, const DataTypePtr &, const ColumnNullable *, size_t input_rows_count) { /// Column of Nothing type is trivially convertible to any other column return res->cloneResized(input_rows_count)->convertToFullColumnIfConst(); }; } WrapperType prepareUnpackDictionaries(const DataTypePtr & from_type, const DataTypePtr & to_type) const { const auto * from_low_cardinality = typeid_cast(from_type.get()); const auto * to_low_cardinality = typeid_cast(to_type.get()); const auto & from_nested = from_low_cardinality ? from_low_cardinality->getDictionaryType() : from_type; const auto & to_nested = to_low_cardinality ? to_low_cardinality->getDictionaryType() : to_type; if (from_type->onlyNull()) { if (!to_nested->isNullable()) { if (cast_type == CastType::accurateOrNull) { return createToNullableColumnWrapper(); } else { throw Exception{"Cannot convert NULL to a non-nullable type", ErrorCodes::CANNOT_CONVERT_TYPE}; } } return [](ColumnsWithTypeAndName &, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count) { return result_type->createColumnConstWithDefaultValue(input_rows_count)->convertToFullColumnIfConst(); }; } bool skip_not_null_check = false; if (from_low_cardinality && from_nested->isNullable() && !to_nested->isNullable()) /// Disable check for dictionary. Will check that column doesn't contain NULL in wrapper below. skip_not_null_check = true; auto wrapper = prepareRemoveNullable(from_nested, to_nested, skip_not_null_check); if (!from_low_cardinality && !to_low_cardinality) return wrapper; return [wrapper, from_low_cardinality, to_low_cardinality, skip_not_null_check] (ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable * nullable_source, size_t input_rows_count) -> ColumnPtr { ColumnsWithTypeAndName args = {arguments[0]}; auto & arg = args.front(); auto res_type = result_type; ColumnPtr converted_column; ColumnPtr res_indexes; /// For some types default can't be casted (for example, String to Int). In that case convert column to full. bool src_converted_to_full_column = false; { auto tmp_rows_count = input_rows_count; if (to_low_cardinality) res_type = to_low_cardinality->getDictionaryType(); if (from_low_cardinality) { const auto * col_low_cardinality = typeid_cast(arguments[0].column.get()); if (skip_not_null_check && col_low_cardinality->containsNull()) throw Exception{"Cannot convert NULL value to non-Nullable type", ErrorCodes::CANNOT_INSERT_NULL_IN_ORDINARY_COLUMN}; arg.column = col_low_cardinality->getDictionary().getNestedColumn(); arg.type = from_low_cardinality->getDictionaryType(); /// TODO: Make map with defaults conversion. src_converted_to_full_column = !removeNullable(arg.type)->equals(*removeNullable(res_type)); if (src_converted_to_full_column) arg.column = arg.column->index(col_low_cardinality->getIndexes(), 0); else res_indexes = col_low_cardinality->getIndexesPtr(); tmp_rows_count = arg.column->size(); } /// Perform the requested conversion. converted_column = wrapper(args, res_type, nullable_source, tmp_rows_count); } if (to_low_cardinality) { auto res_column = to_low_cardinality->createColumn(); auto * col_low_cardinality = typeid_cast(res_column.get()); if (from_low_cardinality && !src_converted_to_full_column) { col_low_cardinality->insertRangeFromDictionaryEncodedColumn(*converted_column, *res_indexes); } else col_low_cardinality->insertRangeFromFullColumn(*converted_column, 0, converted_column->size()); return res_column; } else if (!src_converted_to_full_column) return converted_column->index(*res_indexes, 0); else return converted_column; }; } WrapperType prepareRemoveNullable(const DataTypePtr & from_type, const DataTypePtr & to_type, bool skip_not_null_check) const { /// Determine whether pre-processing and/or post-processing must take place during conversion. bool source_is_nullable = from_type->isNullable(); bool result_is_nullable = to_type->isNullable(); auto wrapper = prepareImpl(removeNullable(from_type), removeNullable(to_type), result_is_nullable); if (result_is_nullable) { return [wrapper, source_is_nullable] (ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count) -> ColumnPtr { /// Create a temporary columns on which to perform the operation. const auto & nullable_type = static_cast(*result_type); const auto & nested_type = nullable_type.getNestedType(); ColumnsWithTypeAndName tmp_args; if (source_is_nullable) tmp_args = createBlockWithNestedColumns(arguments); else tmp_args = arguments; const ColumnNullable * nullable_source = nullptr; /// Add original ColumnNullable for createStringToEnumWrapper() if (source_is_nullable) { if (arguments.size() != 1) throw Exception("Invalid number of arguments", ErrorCodes::LOGICAL_ERROR); nullable_source = typeid_cast(arguments.front().column.get()); } /// Perform the requested conversion. auto tmp_res = wrapper(tmp_args, nested_type, nullable_source, input_rows_count); /// May happen in fuzzy tests. For debug purpose. if (!tmp_res) throw Exception("Couldn't convert " + arguments[0].type->getName() + " to " + nested_type->getName() + " in " + " prepareRemoveNullable wrapper.", ErrorCodes::LOGICAL_ERROR); return wrapInNullable(tmp_res, arguments, nested_type, input_rows_count); }; } else if (source_is_nullable) { /// Conversion from Nullable to non-Nullable. return [wrapper, skip_not_null_check] (ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count) -> ColumnPtr { auto tmp_args = createBlockWithNestedColumns(arguments); auto nested_type = removeNullable(result_type); /// Check that all values are not-NULL. /// Check can be skipped in case if LowCardinality dictionary is transformed. /// In that case, correctness will be checked beforehand. if (!skip_not_null_check) { const auto & col = arguments[0].column; const auto & nullable_col = assert_cast(*col); const auto & null_map = nullable_col.getNullMapData(); if (!memoryIsZero(null_map.data(), null_map.size())) throw Exception{"Cannot convert NULL value to non-Nullable type", ErrorCodes::CANNOT_INSERT_NULL_IN_ORDINARY_COLUMN}; } const ColumnNullable * nullable_source = typeid_cast(arguments.front().column.get()); return wrapper(tmp_args, nested_type, nullable_source, input_rows_count); }; } else return wrapper; } /// 'from_type' and 'to_type' are nested types in case of Nullable. /// 'requested_result_is_nullable' is true if CAST to Nullable type is requested. WrapperType prepareImpl(const DataTypePtr & from_type, const DataTypePtr & to_type, bool requested_result_is_nullable) const { if (from_type->equals(*to_type) && to_type->getCustomName() && to_type->getCustomName()->getName() != "IPv6") { if (isUInt8(from_type)) return createUInt8ToUInt8Wrapper(from_type, to_type); return createIdentityWrapper(from_type); } else if (WhichDataType(from_type).isNothing()) return createNothingWrapper(to_type.get()); WrapperType ret; auto make_default_wrapper = [&](const auto & types) -> bool { using Types = std::decay_t; using ToDataType = typename Types::LeftType; if constexpr ( std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v || std::is_same_v) { ret = createWrapper(from_type, checkAndGetDataType(to_type.get()), requested_result_is_nullable); return true; } if constexpr (std::is_same_v) { if (isBool(to_type)) ret = createBoolWrapper(from_type, checkAndGetDataType(to_type.get()), requested_result_is_nullable); else ret = createWrapper(from_type, checkAndGetDataType(to_type.get()), requested_result_is_nullable); return true; } if constexpr ( std::is_same_v || std::is_same_v) { ret = createEnumWrapper(from_type, checkAndGetDataType(to_type.get())); return true; } if constexpr ( std::is_same_v> || std::is_same_v> || std::is_same_v> || std::is_same_v> || std::is_same_v) { ret = createDecimalWrapper(from_type, checkAndGetDataType(to_type.get()), requested_result_is_nullable); return true; } return false; }; bool cast_ipv4_ipv6_default_on_conversion_error_value = cast_ipv4_ipv6_default_on_conversion_error; auto make_custom_serialization_wrapper = [&, cast_ipv4_ipv6_default_on_conversion_error_value](const auto & types) -> bool { using Types = std::decay_t; using ToDataType = typename Types::RightType; using FromDataType = typename Types::LeftType; if constexpr (WhichDataType(FromDataType::type_id).isStringOrFixedString()) { if (to_type->getCustomSerialization() && to_type->getCustomName()) { if (to_type->getCustomName()->getName() == "IPv4") { ret = [cast_ipv4_ipv6_default_on_conversion_error_value]( ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *, size_t) -> ColumnPtr { if (!WhichDataType(result_type).isUInt32()) throw Exception(ErrorCodes::TYPE_MISMATCH, "Wrong result type {}. Expected UInt32", result_type->getName()); if (cast_ipv4_ipv6_default_on_conversion_error_value) return convertToIPv4(arguments[0].column); else return convertToIPv4(arguments[0].column); }; return true; } if (to_type->getCustomName()->getName() == "IPv6") { ret = [cast_ipv4_ipv6_default_on_conversion_error_value]( ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *, size_t) -> ColumnPtr { if (!WhichDataType(result_type).isFixedString()) throw Exception( ErrorCodes::TYPE_MISMATCH, "Wrong result type {}. Expected FixedString", result_type->getName()); if (cast_ipv4_ipv6_default_on_conversion_error_value) return convertToIPv6(arguments[0].column); else return convertToIPv6(arguments[0].column); }; return true; } ret = &ConvertImplGenericFromString::execute; return true; } } if constexpr (WhichDataType(ToDataType::type_id).isStringOrFixedString()) { if (from_type->getCustomSerialization()) { ret = [](ColumnsWithTypeAndName & arguments, const DataTypePtr & result_type, const ColumnNullable *, size_t input_rows_count) -> ColumnPtr { return ConvertImplGenericToString::execute(arguments, result_type, input_rows_count); }; return true; } } return false; }; if (callOnTwoTypeIndexes(from_type->getTypeId(), to_type->getTypeId(), make_custom_serialization_wrapper)) return ret; if (callOnIndexAndDataType(to_type->getTypeId(), make_default_wrapper)) return ret; switch (to_type->getTypeId()) { case TypeIndex::String: return createStringWrapper(from_type); case TypeIndex::FixedString: return createFixedStringWrapper(from_type, checkAndGetDataType(to_type.get())->getN()); case TypeIndex::Array: return createArrayWrapper(from_type, static_cast(*to_type)); case TypeIndex::Tuple: return createTupleWrapper(from_type, checkAndGetDataType(to_type.get())); case TypeIndex::Map: return createMapWrapper(from_type, checkAndGetDataType(to_type.get())); case TypeIndex::AggregateFunction: return createAggregateFunctionWrapper(from_type, checkAndGetDataType(to_type.get())); default: break; } 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}; } }; class MonotonicityHelper { public: using MonotonicityForRange = FunctionCastBase::MonotonicityForRange; template static auto monotonicityForType(const DataType * const) { return FunctionTo::Type::Monotonic::get; } static MonotonicityForRange getMonotonicityInformation(const DataTypePtr & from_type, const IDataType * to_type) { if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (isEnum(from_type)) { if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); if (const auto * type = checkAndGetDataType(to_type)) return monotonicityForType(type); } /// other types like Null, FixedString, Array and Tuple have no monotonicity defined return {}; } }; }