#pragma once #include #include #include #include #include #include namespace DB { /** Search and replace functions in strings: * * position(haystack, needle) - the normal search for a substring in a string, returns the position (in bytes) of the found substring starting with 1, or 0 if no substring is found. * positionUTF8(haystack, needle) - the same, but the position is calculated at code points, provided that the string is encoded in UTF-8. * positionCaseInsensitive(haystack, needle) * positionCaseInsensitiveUTF8(haystack, needle) * * like(haystack, pattern) - search by the regular expression LIKE; Returns 0 or 1. Case-insensitive, but only for Latin. * notLike(haystack, pattern) * * match(haystack, pattern) - search by regular expression re2; Returns 0 or 1. * * Applies regexp re2 and pulls: * - the first subpattern, if the regexp has a subpattern; * - the zero subpattern (the match part, otherwise); * - if not match - an empty string. * extract(haystack, pattern) * * replaceOne(haystack, pattern, replacement) - replacing the pattern with the specified rules, only the first occurrence. * replaceAll(haystack, pattern, replacement) - replacing the pattern with the specified rules, all occurrences. * * replaceRegexpOne(haystack, pattern, replacement) - replaces the pattern with the specified regexp, only the first occurrence. * replaceRegexpAll(haystack, pattern, replacement) - replaces the pattern with the specified type, all occurrences. * * Warning! At this point, the arguments needle, pattern, n, replacement must be constants. */ template class FunctionsStringSearch : public IFunction { public: static constexpr auto name = Name::name; static FunctionPtr create(const Context &) { return std::make_shared(); } String getName() const override { return name; } size_t getNumberOfArguments() const override { return 2; } DataTypePtr getReturnTypeImpl(const DataTypes & arguments) const override { if (!arguments[0]->isString()) throw Exception( "Illegal type " + arguments[0]->getName() + " of argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); if (!arguments[1]->isString()) throw Exception( "Illegal type " + arguments[1]->getName() + " of argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); return std::make_shared>(); } void executeImpl(Block & block, const ColumnNumbers & arguments, size_t result) override { using ResultType = typename Impl::ResultType; const ColumnPtr & column_haystack = block.getByPosition(arguments[0]).column; const ColumnPtr & column_needle = block.getByPosition(arguments[1]).column; const ColumnConst * col_haystack_const = typeid_cast(&*column_haystack); const ColumnConst * col_needle_const = typeid_cast(&*column_needle); if (col_haystack_const && col_needle_const) { ResultType res{}; Impl::constant_constant(col_haystack_const->getValue(), col_needle_const->getValue(), res); block.getByPosition(result).column = block.getByPosition(result).type->createColumnConst(col_haystack_const->size(), toField(res)); return; } auto col_res = ColumnVector::create(); block.getByPosition(result).column = col_res; typename ColumnVector::Container_t & vec_res = col_res->getData(); vec_res.resize(column_haystack->size()); const ColumnString * col_haystack_vector = checkAndGetColumn(&*column_haystack); const ColumnString * col_needle_vector = checkAndGetColumn(&*column_needle); if (col_haystack_vector && col_needle_vector) Impl::vector_vector(col_haystack_vector->getChars(), col_haystack_vector->getOffsets(), col_needle_vector->getChars(), col_needle_vector->getOffsets(), vec_res); else if (col_haystack_vector && col_needle_const) Impl::vector_constant(col_haystack_vector->getChars(), col_haystack_vector->getOffsets(), col_needle_const->getValue(), vec_res); else if (col_haystack_const && col_needle_vector) Impl::constant_vector(col_haystack_const->getValue(), col_needle_vector->getChars(), col_needle_vector->getOffsets(), vec_res); else throw Exception("Illegal columns " + block.getByPosition(arguments[0]).column->getName() + " and " + block.getByPosition(arguments[1]).column->getName() + " of arguments of function " + getName(), ErrorCodes::ILLEGAL_COLUMN); } }; template class FunctionsStringSearchToString : public IFunction { public: static constexpr auto name = Name::name; static FunctionPtr create(const Context &) { return std::make_shared(); } String getName() const override { return name; } size_t getNumberOfArguments() const override { return 2; } bool useDefaultImplementationForConstants() const override { return true; } ColumnNumbers getArgumentsThatAreAlwaysConstant() const override { return {1}; } DataTypePtr getReturnTypeImpl(const DataTypes & arguments) const override { if (!arguments[0]->isString()) throw Exception( "Illegal type " + arguments[0]->getName() + " of argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); if (!arguments[1]->isString()) throw Exception( "Illegal type " + arguments[1]->getName() + " of argument of function " + getName(), ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT); return std::make_shared(); } void executeImpl(Block & block, const ColumnNumbers & arguments, size_t result) override { const ColumnPtr column = block.getByPosition(arguments[0]).column; const ColumnPtr column_needle = block.getByPosition(arguments[1]).column; const ColumnConst * col_needle = typeid_cast(&*column_needle); if (!col_needle) throw Exception("Second argument of function " + getName() + " must be constant string.", ErrorCodes::ILLEGAL_COLUMN); if (const ColumnString * col = checkAndGetColumn(column.get())) { auto col_res = ColumnString::create(); ColumnString::Chars_t & vec_res = col_res->getChars(); ColumnString::Offsets_t & offsets_res = col_res->getOffsets(); Impl::vector(col->getChars(), col->getOffsets(), col_needle->getValue(), vec_res, offsets_res); block.getByPosition(result).column = std::move(col_res); } else throw Exception( "Illegal column " + block.getByPosition(arguments[0]).column->getName() + " of argument of function " + getName(), ErrorCodes::ILLEGAL_COLUMN); } }; }