>> I will try this way..Thanks very much...
Erm, that won't do what you want it to do. That only stores 2001 values, ranging from -1000 to +1000.
>> it will automatically change to type in these ranges...
You can use the modulus operator:
std::cout<< 1253215%1001;
Make a class that stores two values, a real and an imaginary one. use the modulus operator to make sure their ranges are within the right values and you have it really.
This is obviously comming from a Pascal or Delphi background. In those languages you could do exactly what the OP describes.
This functionality is not built-in C or C++. The closest thing you can do is specify how many bits of a bitfiend a number takes up. But that gives a range that is always a power of two.
To do this you need to write a class, as others have began to show. I'll see what I can come up with shortly...
beene: That answer truly belongs on the WorseThanFailure (previously theDailyWTF) website. Go hide your shameful face.
>> beene: That answer truly belongs on the WorseThanFailure (previously theDailyWTF) website. Go hide your shameful face.
To be honest while it is not good code it's an interpretation of the OP's problem which was rather illucid to begin with, in my opinion, so he wrote the code to fit the problem as he saw it. You ever see that strip with the programmer, customer, and the product and the 6 different interpretations of it? It was very amusing, but I'm damned if I can find it now. I'll edit if I do.
Here you go...
Something like this should do what you want:Declare and use the variable like this:Code:#include <climits> template <int MINRANGE = INT_MIN, int MAXRANGE = INT_MAX > class rangedInteger { int value; public: operator int() const { return value; } rangedInteger() : value(MINRANGE) { } rangedInteger(int val) : value(val) { if (value < MINRANGE || value > MAXRANGE) throw std::exception("Out of range"); } rangedInteger(const rangedInteger &r) : value(r.value) { if (value < MINRANGE || value > MAXRANGE) throw std::exception("Out of range"); } const rangedInteger& operator = (const rangedInteger &r) { value = r.value; if (value < MINRANGE || value > MAXRANGE) throw std::exception("Out of range"); return *this; } };Edit: Fixed defaultsCode:rangedInteger<-1000, 1000> ri; ri = -397; ri = ri + 2000; // Will throw an exception
Last edited by iMalc; 03-23-2007 at 03:38 PM.
Unless you make the int unsigned, -1 is probably a bad choice for the default MAXRANGE. All positive numbers would fail. std::numeric_limits<int>::max() might be better?
Edit: Apparently std::numeric_limits<int>::max() is not a compile time constant. I guess INT_MAX would have to do it.
Last edited by Daved; 03-23-2007 at 03:28 PM.
I know I was harsh, I'm sure he knows I'm not that serious.Originally Posted by twomers
Do you mean the one about the tree with the swing? - seen it.
To be honest though, if you've ever used Pascal style languages, it is absolutely crystal clear what he is asking for from his initial description, and I believe the solution I just posted is exactly what he wants.
not quite
>> it will automatically change to type in these ranges...
so use a % in the constructor I'd say.
>> if you've ever used Pascal style languages, it is absolutely crystal clear what he is asking for from his initial description
I'm just uneducated thenI've never used Pascal. Well, I did. But It was a long time ago when I hated programming so my memories of it are "as whispers upon yesterday's breeze", as someone once put it.
Indeed. I'd make the minimum allowed value default to std::numeric_limits<int>::min() as well.
I just tried it with VC++ 2003 and it wouldn't allow it. Apparently numeric_limits<int>::max() is not a compile time constant. As I mentioned in an edit, INT_MAX and INT_MIN from <climits> would work.
I actually made the default max less than the default min in the thinking that it forces you to not leave it unspecified. But duh it's a template and you have to specify it to make it useful, so yeah INT_MAX and INT_MIN would be good, or you could just forget the defaults.
Actually, if one wants to use operators like +=, -=, *=, /=, >>=, <<= then you'd have to add those to the class too.
I'll leave the rest as an exersize for the reader.Code:const rangedInteger& operator += (const rangedInteger &r) { value += r.value; if (value < MINRANGE || value > MAXRANGE) throw std::exception("Out of range"); return *this; }
Last edited by iMalc; 03-23-2007 at 03:36 PM.
