.net运算符重载

Question

Why is it required on some operators Eg the comparison operators.

<   <=   >   >=   =  <>

You have to also implement their complement also?

<  >=

What is restrict your object implementing just

<

if that has specific meaning to it, and

>= 

doesn't

Note: The workaround I've is to stick the Obsolete attribute on the unwanted operator.

Answer 1

Operator overloading in .Net (at least for C# and VB) is intentionally less flexible than operator overloading in C++.

One reason that Java went so far as to remove overloading entirely is that too many developers abused overloading to do clever-looking things that confused other developers and broke core assumptions of generic algorithms that expect to be able to, say, reasonably compare two instances of a user type.

While .Net brought back some flexibility, the designers were careful to only allow overloading of certain operators with the intent that they only be used for their customary meanings.

The requirement to implement comparison operators symmetrically is a nod to that intention.

If you want to do more aggressive operator overloading, I'd encourage you to consider using F#, where you have access to a much broader palette of symbols with which to draw clever operator symbols.

Answer 2

Presumably because if you overload one side of an operator, you are implementing some custom rule, and the default operator would not match your rule, and thus produce odd behaviours.

For example, if your definition of 'int >' was 'return (x > 5);' (for some reason...) and you didn't overlaod the inverse rule for the 'int <' operator, then;

4 > 1 would be false (by your rule), but 4 < 1 would also be false (by the default rule).

Which is illogical, Jim.