================
Comment at: include/clang/ASTMatchers/Dynamic/VariantValue.h:140
@@ +139,3 @@
+ };
+ union {
+ SingleRep *AsSingle;
----------------
Manuel Klimek wrote:
> Any reason not to use an interface with two implementations? This smells a
> lot like building our own inheritance implementation ...
These two could hide behind an interface, but I don't know how the interface
would work for the variadic case coming next (from D1427).
The basic idea for that one is that on getTypedMatcher<T> it will convert all
the inner matchers to Matcher<T>, then create a
VariadicOperatorMatcherInterface<T> with them and return it.
If I try to use interfaces, I will need two: one for the payload of the
matcher, and one for the operations from <T>. It can be done, but it might be
messier.
I'll give it a try to see how it looks.
http://llvm-reviews.chandlerc.com/D1446
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
