Hi all,

I was playing with D operator overloading and I found the opCmp function a bit strange. Why the type returned is an int? It would not have to return a bit?

And how should I code an opCmp overload? I tried the some coding (atached)
and it worked well, but I am not sure if the opCmp overloading (at the end)
is Ok.

Thanks,

Bruno.

Bruno A. Costa wrote:

>Hi all,
>
>I was playing with D operator overloading and I found the opCmp function a
>bit strange. Why the type returned is an int? It would not have to return a
>bit?
>  
>

opCmp returns 3 states

0: equal
-1: less then
1: more then


>And how should I code an opCmp overload? I tried the some coding (atached)
>and it worked well, but I am not sure if the opCmp overloading (at the end)
>is Ok.
>

>	int opCmp (Byte bt) {
>		return (_value - bt.value);
>	}
>
looks fine to me.


-- 
-Anderson: http://badmama.com.au/~anderson/

In article <c75vi8$2q14$1@digitaldaemon.com>, Bruno A. Costa says...
>
>--nextPart1697271.eHQoPkZ5kg
>Content-Type: text/plain; charset=iso-8859-1
>Content-Transfer-Encoding: 8Bit
>
>Hi all,
>
>I was playing with D operator overloading and I found the opCmp function a bit strange. Why the type returned is an int? It would not have to return a bit?

0 means equal, <0 means one is lower, >0 means the other is lower.
(it's common practice)

>
>And how should I code an opCmp overload? I tried the some coding (atached)
>and it worked well, but I am not sure if the opCmp overloading (at the end)
>is Ok.

I fell on that trap once so I should help now.
you must override int opComp(Object) not create a new int opComp(Byte)

So, are you doing an all OO lib? or just what you need?

Ant

Ant wrote:

>In article <c75vi8$2q14$1@digitaldaemon.com>, Bruno A. Costa says...
>  
>
>>--nextPart1697271.eHQoPkZ5kg
>>Content-Type: text/plain; charset=iso-8859-1
>>Content-Transfer-Encoding: 8Bit
>>
>>Hi all,
>>
>>I was playing with D operator overloading and I found the opCmp function a
>>bit strange. Why the type returned is an int? It would not have to return a
>>bit?
>>    
>>
>
>0 means equal, <0 means one is lower, >0 means the other is lower.
>(it's common practice)
>
>  
>
>>And how should I code an opCmp overload? I tried the some coding (atached)
>>and it worked well, but I am not sure if the opCmp overloading (at the end)
>>is Ok.
>>    
>>
>
>I fell on that trap once so I should help now.
>you must override int opComp(Object) not create a new int opComp(Byte)
>  
>
class A
{
   byte val;

   int opCmp(A _a)
   {
     return val - _a.val;
   }   }


int main ( char [] [] args )
{
   A a = new A;
   A b = new A;
     a.val = 10;
   b.val = 11;
     printf("%d < %d = %d\n", a.val, b.val, b < a);
}

This works.  What are you talking about?  You do need to do such things if you want to use sort with arrays (which I think is a bug). Other then that it should work fine.

>So, are you doing an all OO lib? or just what you need?
>
>Ant
>
>
>
>  
>


-- 
-Anderson: http://badmama.com.au/~anderson/

In article <c766ek$32u$1@digitaldaemon.com>, J Anderson says...
>
>Ant wrote:
>
>>In article <c75vi8$2q14$1@digitaldaemon.com>, Bruno A. Costa says...
>> 
>>
>>>--nextPart1697271.eHQoPkZ5kg
>>>Content-Type: text/plain; charset=iso-8859-1
>>>Content-Transfer-Encoding: 8Bit
>>>
>>>Hi all,
>>>
>>>I was playing with D operator overloading and I found the opCmp function a bit strange. Why the type returned is an int? It would not have to return a bit?
>>> 
>>>
>>
>>0 means equal, <0 means one is lower, >0 means the other is lower.
>>(it's common practice)
>>
>> 
>>
>>>And how should I code an opCmp overload? I tried the some coding (atached)
>>>and it worked well, but I am not sure if the opCmp overloading (at the end)
>>>is Ok.
>>> 
>>>
>>
>>I fell on that trap once so I should help now.
>>you must override int opComp(Object) not create a new int opComp(Byte)
>> 
>>
>class A
>{
>    byte val;
>
>    int opCmp(A _a)
>    {
>      return val - _a.val;
>    }
>}
>
>
>int main ( char [] [] args )
>{
>    A a = new A;
>    A b = new A;
> 
>    a.val = 10;
>    b.val = 11;
> 
>    printf("%d < %d = %d\n", a.val, b.val, b < a);
>}
>
>This works.  What are you talking about?
Sorry, my mistake, I was talking about:

http://www.digitalmars.com/d/phobos.html#object

int cmp(Object obj)

Compare with another Object obj. Returns: <0 for (this < obj) =0 for (this == obj)
>0 for (this > obj)

Ant

Ant wrote:

> In article <c75vi8$2q14$1@digitaldaemon.com>, Bruno A. Costa says...
>>
>>--nextPart1697271.eHQoPkZ5kg
>>Content-Type: text/plain; charset=iso-8859-1
>>Content-Transfer-Encoding: 8Bit
>>
>>Hi all,
>>
>>I was playing with D operator overloading and I found the opCmp function a bit strange. Why the type returned is an int? It would not have to return a bit?
> 
> 0 means equal, <0 means one is lower, >0 means the other is lower.
> (it's common practice)
> 
>>
>>And how should I code an opCmp overload? I tried the some coding (atached)
>>and it worked well, but I am not sure if the opCmp overloading (at the
>>end) is Ok.
> 
> I fell on that trap once so I should help now.
> you must override int opComp(Object) not create a new int opComp(Byte)
> 
> So, are you doing an all OO lib? or just what you need?
> 
> Ant

I just wanted to learn D, so I started to write some easy wrappers to primitive types. But an OO lib would be a very nice thing to do.

Bruno.