Patch dbus-binding-tool

Matt Hoosier mwhoosier at gmail.com
Mon Aug 28 13:03:16 PDT 2006


That would be preferable, but aren't the generated functions always
marked 'static'? It would be nice if the client glue prototypes could
be generated in a different file from the function implementations.

[Havoc: sorry for the duplicate; I accidentally replied only to you originally.]

On 8/26/06, Havoc Pennington <hp at redhat.com> wrote:
>
>
> Matt Hoosier wrote:
> > What about the following version that would adjust to the language
> > under which it's compiled?
> >
> > WRITE_OR_LOSE ("#ifdef __cplusplus\n");
> > WRITE_OR_LOSE ("  DBusGASyncData *data = static_cast<DBusGAsyncData*>
> > (user_data);\n");
> > WRITE_OR_LOSE ("#else\n");
> > WRITE_OR_LOSE ("  DBusGAsyncData *data = (DBusGASyncData *) user_data;\n");
> > WRITE_OR_LOSE ("#endif\n");
> > WRITE_OR_LOSE ("  GError *error = NULL;\n");
> >
>
> If there are simple fixes to allow C++ compilation then fine, but I
> don't think it's worth making a mess about... it's a pretty odd thing to
> do, every platform I know of you could use a C compiler and then link
> the C files into a C++ program.
>
> Havoc
>
>


More information about the dbus mailing list