On Mon Aug 18 2014 at 11:20:14 PM Sean Silva <[email protected]> wrote:
> On Sun, Aug 17, 2014 at 3:06 PM, Manuel Klimek <[email protected]> wrote: > >> >> >> On Sun Aug 17 2014 at 11:43:39 PM Rafael Espíndola < >> [email protected]> wrote: >> >>> > Note that we do want to migrate (most of) clang-format's tests to lit >>> tests >>> > (because most of them are integration tests), but that's a different >>> topic. >>> >>> Well, that was the topic of my reply :-) >>> >>> When gtest is appropriate, do use it. My request was just to not build >>> another large test set that we want to migrate in the future. >>> >> >> Then we agree! :) Please give specific feedback if you see a gtest test >> that you think should be a lit test! >> >> > > In this case, it seems like you should write a unit test for the routine > that uses the offsets to actually do the renaming. Then you can write this > test as a lit test that checks that the relevant things get renamed. > Yes, the routine that finds what is under the cursor is unit tested, the rename is lit tested. I'm not sure how that differs from what you're describing. > Also, why are we committing code that is "grossly under-tested" in the > first place? (sorry if this was discussed in another thread) > a) it was a little tongue-in-cheek from my side (mainly meaning that I think we should focus the next steps on the testing first, before adding more features) b) I'm generally fine with incremental development in-tree (especially for leaf projects) > > -- Sean Silva > > >> >>> Cheers, >>> Rafael >>> >> >> _______________________________________________ >> 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
