Lisandro Dalcin wrote: > On Tue, Oct 6, 2009 at 11:47 AM, Neal Becker > <ndbeck...@gmail.com> wrote: >>> >>> Yes, but C99 complexes could be used in C++ (for example, IBM >>> compilers seems to let you use C99 complexes in C++, look at the very >>> end of: >>> >> http://publib.boulder.ibm.com/infocenter/comphelp/v7v91/index.jsp?topic=/com.ibm.vacpp7a.doc/language/ref/clrc03complex_types.htm) >>> >> I'm pretty sure gcc won't allow this. >> > > OK. Now I've actually read the C++0x draft, and it seems that the C99 > complexes will not be supported (am I right here)? In such case, now > it is pretty clear what to do: by default, Cython uses the > struct-based complex, and with an appropriate directive (we should > rename c99_complex to something more language-agnostic, right?) Cython > switches to use C99 complexes in case of C compilation, and C++ > complexes in case of C++ compilation. >
Keep in mind that c++ complex != c99 complex. Specifically, you cannot just get/put values using: double y = z.imag <NO z.imag = 2 <NO The portable way is: double y = z.imag() or imag (z) <get the value z.imag(2) <set the value _______________________________________________ Cython-dev mailing list Cython-dev@codespeak.net http://codespeak.net/mailman/listinfo/cython-dev
