On Mar 18, 2014, at 10:19 AM, Dmitri Gribenko <[email protected]> wrote:
> On Tue, Mar 18, 2014 at 4:52 PM, Ben Langmuir <[email protected]> wrote: >> Thanks Argyrios. Richard & Dmitri: just to confirm - are you okay with me >> committing this now, or did either of you intend to review the patch first? > > + // For frameworks, the preferred spelling is Modules/module.modulemap. > + // Otherwise it is just module.modulemap. > > ... Otherwise it is just "module.map" at the framework root. WIll do. > > Apart from that, two comments: > > 1. There is also the "module_private.map" thing. Do we want to rename > that file as well? That is done in this patch as well - see loadModuleMapFileImpl. If we find ‘module.modulemap’ we will look for ‘module.private.modulemap’. If we find ‘module.map’, we will look for ‘module_private.map’. > > 2. I don't immediately see how DirectoryHasModuleMap interacts with > frameworks now. Does it contain only root framework directories or > 'Module' subdirectories as well? Could you add a sub framework test > to test this logic in hasModuleMap? Hmm, I didn’t think about this. Right now only the ‘Modules’ subdirectory will be in DirectoryHasModuleMap unless we are going through hasModuleMap, in which case I believe they will both be there. This is probably safe, as we will not load the module map more than once, but it would be better to also add the framework directory so that we detect this earlier - I’ll make that happen and add the test case you suggest. Ben > > Dmitri > > -- > main(i,j){for(i=2;;i++){for(j=2;j<i;j++){if(!(i%j)){j=0;break;}}if > (j){printf("%d\n",i);}}} /*Dmitri Gribenko <[email protected]>*/ _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
