On Sat, Mar 17, 2012 at 7:03 PM, Douglas Gregor <[email protected]> wrote: > > On Mar 17, 2012, at 5:00 PM, Aaron Ballman wrote: > >> On Thu, Mar 15, 2012 at 7:29 AM, Aaron Ballman <[email protected]> >> wrote: >>> On Wed, Mar 14, 2012 at 11:25 PM, J B <[email protected]> wrote: >>>> The reason I recommend an error is because the code will error out later >>>> without the .tlh (where it is requested via attributes or absence of). The >>>> errors later will be a direct result of the failure of #import to function >>>> properly. >>> >>> That's a valid reason to error in my book. I'll go down that route. >>> Thanks for the insights! >> >> This is the revised patch based on feedback from everyone. It causes >> a (shorter) error to happen when using #import in MS mode, and does >> not perform the actual include. It also eats the optional attributes >> so that parsing can continue. >> >> Again, this will fix PR 10727 > > +def err_pp_import_directive_ms : Error< > + "#import does not convert a type library to C++ classes in Microsoft > Mode">, > + InGroup<Microsoft>; > > This should really just say that this Microsoft feature is not implemented. > > - if (!LangOpts.ObjC1) // #import is standard for ObjC. > + if (LangOpts.MicrosoftMode) > + return HandleMicrosoftImportDirective(ImportTok); > + else if (!LangOpts.ObjC1) // #import is standard for ObjC. > Diag(ImportTok, diag::ext_pp_import_directive); > > I think we should do this the other way: if LangOpts.ObjC1, we treat it as an > Objective-C #import. Otherwise, we either error (if in Microsoft mode) or > give an extension warning (not in Microsoft mode).
Good points -- rectified in this patch. ~Aaron
MicrosoftImport.patch
Description: Binary data
_______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
