Could we change the Preprocessor to accept an optional pointer to a 
ModuleLoader ?

This will eliminate the whole "VoidModuleLoader" business and the hack of 
having the ASTUnit "implement" a ModuleLoader just so we have something to pass 
to the Preprocessor when loading an AST file.

-Argyrios

On Nov 5, 2012, at 11:22 AM, Kim Gräsman <[email protected]> wrote:

> Hi Chandler,
> 
> Sure, I can't add much more to that end, but I agree that this is probably 
> better off in a library shared between unit tests.
> 
> - Kim
> 
> On Mon, Nov 5, 2012 at 12:39 PM, Chandler Carruth <[email protected]> 
> wrote:
> I like the direction of this, but I think it should go into a
> unittest-specific library.
> 
> We don't currently have any such library infrastructure, but honestly,
> I think we need it and we should get it in place sooner rather than
> later.
> 
> Adding Manuel as he may have ideas here.
> 
> On Sun, Nov 4, 2012 at 11:27 AM, Kim Gräsman <[email protected]> wrote:
> > Hi committers,
> >
> > I contributed a small patch a few weeks ago, adding unittests for
> > PPCallbacks::InclusionDirective.
> >
> > I was bugged by the proliferation of stub ModuleLoader implementations but
> > didn't want to do anything about it in that context.
> >
> > Here's a patch to move VoidModulerLoader into ModuleLoader.h and remove all
> > redundant implementations.
> >
> > Could someone review and commit if this is acceptable?
> >
> > Thanks,
> > - Kim
> >
> > _______________________________________________
> > cfe-commits mailing list
> > [email protected]
> > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
> >
> 
> _______________________________________________
> cfe-commits mailing list
> [email protected]
> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to