CGAL users discussion list

[Andreas Fabri <>]

No way to help you for something else than the latest release of CGAL.
Unless you pay well ===>:-

andreas

On 20/05/2010 13:21, JohnHans wrote:
> Hi, Sebastien Loriot, Andreas Fabri,
>
> Thank you very much for your soon reply.
>
> Actually, I would move an application routine from Linux to Windows, which
> had already been built successfully on CGAL 3.3.1 by senior fellow
> apprentice on Linux. When I built it based on Qt4, CGAL-3.3.1 within VS2005
> on Windows, there are many C3861 errors (identifier not found) for lroundf,
> isnan and so on, which usually are defined in Linux. So I try to create a
> MathFix.h as following:
>
> MathFix.h:
> #include<math.h>
> #include<stdlib.h>
> #include<limits>
> #include<float.h>
>
> #ifndef M_PI
> const double piDouble = 3.14159265358979323846;
> const float piFloat = 3.14159265358979323846f;
> #else
> const double piDouble = M_PI;
> const float piFloat = static_cast<float>(M_PI);
> #endif
>
> #ifndef M_PI_4
> const double piOverFourDouble = 0.785398163397448309616;
> const float piOverFourFloat = 0.785398163397448309616f;
> #else
> const double piOverFourDouble = M_PI_4;
> const float piOverFourFloat = static_cast<float>(M_PI_4);
> #endif
>
> inline long lround(double num) { return static_cast<long>(num>  0 ? num +
> 0.5 : ceil(num - 0.5)); }
> inline long lroundf(float num) { return static_cast<long>(num>  0 ? num +
> 0.5f : ceilf(num - 0.5f)); }
> inline double round(double num) { return num>  0 ? floor(num + 0.5) :
> ceil(num - 0.5); }
> inline float roundf(float num) { return num>  0 ? floorf(num + 0.5f) :
> ceilf(num - 0.5f); }
> inline double trunc(double num) { return num>  0 ? floor(num) : ceil(num); }
>
> inline bool isinf(double num) { return !_finite(num)&&  !_isnan(num); }
> inline bool isnan(double num) { return !!_isnan(num); }
> inline bool signbit(double num) { return _copysign(1.0, num)<  0; }
>
> inline double nextafter(double x, double y) { return _nextafter(x, y); }
> inline float nextafterf(float x, float y) { return x>  y ? x - FLT_EPSILON :
> x + FLT_EPSILON; }
>
> inline double copysign(double x, double y) { return _copysign(x, y); }
> inline int isfinite(double x) { return _finite(x); }
>
> inline double wtf_atan2(double x, double y)
> {
>      double posInf = std::numeric_limits<double>::infinity();
>      double negInf = -std::numeric_limits<double>::infinity();
>      double nan = std::numeric_limits<double>::quiet_NaN();
>
>      double result = nan;
>
>      if (x == posInf&&  y == posInf)
>          result = piOverFourDouble;
>      else if (x == posInf&&  y == negInf)
>          result = 3 * piOverFourDouble;
>      else if (x == negInf&&  y == posInf)
>          result = -piOverFourDouble;
>      else if (x == negInf&&  y == negInf)
>          result = -3 * piOverFourDouble;
>      else
>          result = ::atan2(x, y);
>
>      return result;
> }
>
> inline double wtf_fmod(double x, double y) { return (!isinf(x)&&  isinf(y))
> ? x : fmod(x, y); }
>
> inline double wtf_pow(double x, double y) { return y == 0 ? 1 : pow(x, y); }
>
> #define atan2(x, y) wtf_atan2(x, y)
> #define fmod(x, y) wtf_fmod(x, y)
> #define pow(x, y) wtf_pow(x, y)
>
> After MathFix.h was included in that application routine, I tried to build
> it based on Qt4, CGAL3.3.1 (and later CGAL 3.5.1) within VS2005 on Windows
> and got the conflicts between CGAL with VS2005 as I described in the first
> message.
>
> BTW, I just installed CGAL3.3.1 and later CGAL 3.5.1 on Windows and have not
> built them yet.
>
> What’s wrong with me? What’s wrong with MathFix.h? Is there another way to
> solve the problem of isnan, lroundf and so on? Or the conflicts?
>
> I have spent more than 2 weeks on these problems, could you, someone please
> instruct me to solve it or give me some advice?
>
> Thank you very much in advance.
>
> Best wishes,
>
> John Hans