On Tue, Feb 24, 2009 at 5:00 AM, James Reeves <weavejes...@googlemail.com> wrote: > > On Feb 24, 3:11 am, "Michel S." <michel.syl...@gmail.com> wrote: >> I started Quiche after taking a look at Fact, actually; the difference >> between what I'm proposing and Fact is that the latter is a standalone >> test framework, whereas the random-testing part of Quiche (property) >> and Fact (fact) could, IMHO, be built on top of test-is. > > I originally planned to build Fact on top of test-is, but there were > other parts of test-is I wanted to change, such as how test failures > were reported, so I created a separate library. > > I don't think there's much overlap between Fact and test-is, insofar > as a function to test a predicate against a sequence of values is > relatively simple to make. For test-is, I'd be tempted to create > something like this: > > (deftest foo > (for-all [x random-int, y random-int] > (> (+ x y) x)) >
> IMO, for-all is more descriptive than 'property', and fits better with > 'is' as a name. However, because this uses asserts, it doesn't fit in > with Fact's predicate system. for-all does look clearer than 'property', especially since the generating functions are made explicit (unlike in, say, Haskell's QC). > The only complicated part of Fact is its ability to generate random > test data from regular expressions, but I've already split that out > into a library called re-rand. This part could, perhaps, be merged into clojure-contrib? Thanks, -- miʃel salim • http://hircus.jaiku.com/ IUCS • msa...@cs.indiana.edu Fedora • sali...@fedoraproject.org MacPorts • hir...@macports.org --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -~----------~----~----~----~------~----~------~--~---
