On Thu, Aug 19, 2010 at 8:55 AM, Brad King <brad.k...@kitware.com> wrote:
> On 08/19/2010 07:23 AM, David Cole wrote: > > Here is a sampling of the output of processing vtkWin32OutputWindow.h > > with gccxml: > > > > <Method id="_17" name="IsTypeOf" returns="_64" static="1" context="_9" > > access="public" mangled="_ZN20vtkWin32OutputWindow8IsTypeOfEPKc" > > demangled="vtkWin32OutputWindow::IsTypeOf(char const*)" location="f1:35" > > file="f1" line="35" endline="35" inline="1"> > > <Argument name="type" type="_7" location="f1:35" file="f1" > line="35"/> > > </Method> > > > > There is clearly a static="1" attribute on each static method, just as I > > would expect. > > This is a class static method which is a different use of the "static" > keyword. It affects the interface of the function (whether there is > an implicit 'this' argument). > > > On Thu, Aug 19, 2010 at 12:17 AM, Óscar Fuentes <o...@wanadoo.es > > <mailto:o...@wanadoo.es>> wrote: > > > > Qt have something like this on their headers: > > > > inline static int foo(int blah) { .... } > > This use of "static" affects linkage but not the function interface. > Ah. I was assuming the context of this line was inside a class definition.... > > > `inline' is reported, but not `static'. > > > > Is this a limitation of gcc-xml or a bug? > > It was never intended to work. It may be possible to implement but > I do not know where GCC's parser keeps this information. Take a look > at GCC/gcc/tree.h, GCC/gcc/cp/cp-tree.h, and GCC/gcc/cp/xml.c if you > want to try it. > > -Brad >
_______________________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Follow this link to subscribe/unsubscribe: http://www.gccxml.org/mailman/listinfo/gccxml
