On Nov 28, 2010, at 6:48 PM, Craig Silverstein wrote:
> } I'd be at least interested in seeing it.
> 
> We developed this functionality as a test for RecursiveASTVisitor.h.  As
> I understand it, the reason we haven't tried to submit it upstream yet
> is that it's still quite incomplete.  (It will probably need to be
> refactored into one test per kind of AST node, rather than a single
> monolithic test like we have now.)
> 
> This is all Zhanyong's work, so I'll let him speak to how practical it
> would be to get something upstream.  I also don't want so share the test
> code itself without his ok, but I've attached the actual test file
> below.  It's at least enough to show what the output looks like, and how
> one might do FileCheck tests with it.

Okay.  I'm not sure there's actually as much room for re-use here as
Chandler thinks;  I want to dump a *lot* more information than this,
and I'm not sure it makes sense to force your testing framework to deal
with that.  Thanks, though.

That said, having a RAV testing framework in-tree doesn't seem like a
bad thing.

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

Reply via email to