• "KarL" <none nowhere.org> Jun 24 2002
• Heinz Saathoff <hsaat bre.ipnet.de> Jun 24 2002
• "Walter" <walter digitalmars.com> Jul 01 2002

Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

The following code snippet illustrates what I intended to do:
bool bPatterns[sizeof(short)*8];
void getMaskpattern(unsigned short mask){  unsigned short i;  for =
(i=3D0; i<sizeof(short) * 8; i++)    bPatterns[i] =3D mask & (1 << i);}
Took me a while to realise that the bool behavious is as if
#define bool char
And therefore the higher bits of mask is not "boolean copied" to
the array. The compiler did not warn me about it for "type
conversions".

The fix is of course changing the assignment to
    bPatterns[i] =3D (mask & (1 << i)) !=3D 0;
Question:-
What should be the correct behaviour for bool in the example above?

DMC++ version 8.28n

KarL schrieb...
 
 The fix is of course changing the assignment to
     bPatterns[i] = (mask & (1 << i)) != 0;
 Question:-
 What should be the correct behaviour for bool in the example above?
 
 DMC++ version 8.28n


The standard says that there should be a boolean conversion. Therefore 
the explicit != 0 should not be neccessary. Seems like a bug in DMC.

Regards,
	Heinz

Thanks. I'll check it out. -Walter

"Heinz Saathoff" <hsaat bre.ipnet.de> wrote in message
news:MPG.1780f0b4d4554bcc9896a8 news.digitalmars.com...
 KarL schrieb...
 The fix is of course changing the assignment to
     bPatterns[i] = (mask & (1 << i)) != 0;
 Question:-
 What should be the correct behaviour for bool in the example above?

 DMC++ version 8.28n


 The standard says that there should be a boolean conversion. Therefore
 the explicit != 0 should not be neccessary. Seems like a bug in DMC.

 Regards,
 Heinz