PR10124

[Ted Kremenek]

On Apr 24, 2013, at 12:26 AM, Nikola Smiljanic <popizdeh at gmail.com> wrote:

> But won't you have to change this code if you change the docs to say "returns zero if the set contains the specified cursor"
> 
>  CXCursorSet_Impl *setImpl = unpackCXCursorSet(set);
>   if (!setImpl)
>     return 0;
> 
> ???

True, but the behavior is weird.  If we have an invalid CXCursorSet, it means that it contains any cursor.  That just seems wrong.  I'd rather fix the API than perpetuate the breakage.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20130424/26cde25a/attachment.html>