Thanks for that Rob, the not-a-mock changes are now pushed. As for the issues you're facing with TS features - #1 works for me. The reason I added the fuzzier steps (instead of saying Then I should get 7 results), is because of the deletion of objects in other features. So I'm a little confused that you're having problems with it.
#2, I'm always getting one datetime delta story fail using 0.9.8.1 (curiosly enough, it passed with 0.9.8) - and sometimes a second one, although that's rare. I'll try slowing it down a bit more, see if that gets around it. And when I have a chance, I'll look into merging your Windows-related changes - always appreciated (and if you send me a pull request, I won't forget, because I keep not-yet-done requests aside from general archiving). Cheers -- Pat On 27/07/2009, at 5:00 PM, rob-twf wrote: > > Just thought I'd post a quick update to say that I've fixed it :) The > problem was down to the private/protected stubbing not working > correctly: the method_at_any_level? method of the stubber class was > using symbols to check method names which is fine for 1.9 but not for > 1.8 and also the test cases weren't actually testing with private or > protected methods which is why they still passed. > > I've sent a pull request to Pat, if you want to check it out in the > meantime my fork is here: http://github.com/thewebfellas/not-a-mock/tree > > I also have a fork of ThinkingSphinx that has patches for some other > test problems (ensuring files are closed/deleted when necessary and > fixing the cucumber features so they don't hang on Windows waiting for > Sphinx to start): http://github.com/thewebfellas/thinking-sphinx/tree > > There are two remaining problems: > > 1. The sti_searching feature has two failures: "Then I should get as > many results as there are animals" and "Then I should get as many > results as there are cats" both actually return 1 less result than > expected. > > It seems to be because a cat is deleted in the deleting_instances > scenarios but the db/indexes aren't being rebuilt between scenarios. > Running the feature on its own, rather than as part of the full suite > removes the failure - as does adding a Before block to the > post_database code that rebuilds the database, fixture data and > indexes before each scenario. The problem with the latter is it also > means running the features then takes a loooong time, but it does > ensure each scenario gets a clean slate to work with... one for more > consideration I think :) > > 2. I sometimes get failures in the datetime_deltas scenarios (lines 37 > and 55). > > I've tracked these down to the 'And I wait for Sphinx to catch up' > step not waiting long enough for my clunky old PC to actually rotate > the indexes. Increasing the sleep time did help, but again slowed the > features down. This one is probably not worth worrying too much about. > > Rob > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Thinking Sphinx" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/thinking-sphinx?hl=en -~----------~----~----~----~------~----~------~--~---
