I don't see the FunctorKey constructor you are talking about. I assume you mean something like Point2DKey(std::string const &)?
Oops, I actually meant the one that takes a SubSchema (which is just a combination of a Schema and string: it's what you get when you say schema[string]).
I still see a failure in testPsfFlux.py. Did you check out the
DM-433 ticket branch? I don't see it on the master branch, which probably means that I caused it, but the problem was very odd (having to do with the bounding boxes not matching). SInce this was your test, I thought you might be able to spot it more easily.
I didn't realize there was a branch for meas_base (this is one reason we typically paste the output of "git diff --stat" when requesting a review). Now that I'm trying that, I'm seeing all the tests fail. Are there any other branches I need to set up? In the meantime, I'll rebuild everything against the latest afw; maybe that's the problem.
The problem with sourceMatch only appears when the default version is 1. You have to change BaseTable.cc to see it.
Ah, ok. So it's not a problem right now, but it will be again in the future when we change the default version.
I'm not sure how to fix this problem with Swig. In C++, the parameter signatures disambiguate the defineSlot methods. For now, I am going to rename the methods which take FunctorKey parameters.
As I said, I don't think we need to fix it. I'd prefer to just remove those methods, and just leave the ones that take strings. Do you need them for anything?