c++ - 在 C++ 中重载运算符时如何防止整数歧义

Question

给定以下代码:

#include <string>
#include <sstream>

class SomeClass {
public:
    SomeClass& operator<<(long value) { return *this; }

    SomeClass& operator<<(unsigned long value) { return *this; }

    SomeClass& operator<<(float value) { return *this; }

    SomeClass& operator<<(double value) { return *this; }

    SomeClass& operator<<(long double value) { return *this; }

    SomeClass& operator<<(bool value);
    { return *this; }

    SomeClass& operator<<(const void* value);
    { return *this; }

    SomeClass& operator<<(::std::string aString) { return *this; }
};

int main() {
    SomeClass aClass;
    std::ostringstream aStream;

    aClass << 2;
    aClass << "Hello World";
    aClass << '\0';
    aClass << -2;

    aStream << 2;
    aStream << "Hello World";
    aStream << '\0';
    aStream << -2;

    return 0;
}

为什么是basic_ostream能够重载<<这么多整数类型没问题？然而，编译器为 SomeClass 的 << 运算符提供了不明确的重载错误？

例如:

../OverloadTest.cpp:65: error: ambiguous overload for ‘operator<<’ in ‘aClass << 2’
../OverloadTest.cpp:14: note: candidates are: SomeClass& SomeClass::operator<<(long int)   
../OverloadTest.cpp:19: note:                 SomeClass& SomeClass::operator<<(long unsigned int)
../OverloadTest.cpp:24: note:                 SomeClass& SomeClass::operator<<(float)    
../OverloadTest.cpp:29: note:                 SomeClass& SomeClass::operator<<(double)    
../OverloadTest.cpp:35: note:                 SomeClass& SomeClass::operator<<(long double)    
../OverloadTest.cpp:40: note:                 SomeClass& SomeClass::operator<<(bool)    
../OverloadTest.cpp:46: note:                 SomeClass& SomeClass::operator<<(const void*)     
../OverloadTest.cpp:51: note:                 SomeClass& SomeClass::operator<<(std::string) 

Answer 1

您至少需要为 int 添加重载(以及完成 unsigned int 的代码)。

SomeClass& operator<<(int value) { return *this; }

SomeClass& operator<<(unsigned int value) { return *this; }

当您调用时:aClass << 2;

文字 2类型为 int , 并没有重载完全吻合，编译器需要对数字进行转换，在转换中就是出现歧义的地方。 2用什么比较好, long或 unsigned long或 float或 double或 long double甚至 bool ，编译器不知道。