I'd go with any large file in the source distribution of Coq. In fact, a good
test might be to clone the Coq repo, reindent the whole standard library, check
these changes in, then reindent again with your changes and look at the git
diff. I expect the standard library + the Coq test suite to
I have a few tentative changes to coq-smie.el.
Does someone have some kind of test-suite somewhere against which I can
run my new code to try and avoid regressions?
Ideally, it should be fully automated, including fixing my bugs and
outputting a Coq proof that the result is correct, but I'll
I agree. One more argument: opam has problems and having Coq delivered only
by opam is risky these days. So having a reliable fallback is good. Even if
it is a bit outdated.
P
Le sam. 1 déc. 2018 à 03:53, Stefan Monnier a
écrit :
> > Now that you have mentioned it, we consider this pretty
