On Tuesday 04 April 2006 10:46 pm, Kyle Shank wrote:
> The RemoteTestRunner issue is a result of the new Rails integration
> testing. They define a default_test in
> ActionController::IntegrationTest that gets picked up by the test
> runner because it extends Test::Unit::TestCase.
But they don't do define that. The following snippet is from my installed gem
for action_pack-1.12.0. It contains the only two references to
'default_test' in the entire gem.
class IntegrationTest < Test::Unit::TestCase
# Work around a bug in test/unit caused by the default test being named
# as a symbol (:default_test), which causes regex test filters
# (like "ruby test.rb -n /foo/") to fail because =~ doesn't work on
# symbols.
def initialize(name) #:nodoc:
super(name.to_s)
end
# Work around test/unit's requirement that every subclass of TestCase have
# at least one test method. Note that this implementation extends to all
# subclasses, as well, so subclasses of IntegrationTest may also exist
# without any test methods.
def run(*args) #:nodoc:
return if @method_name == "default_test"
super
end
I can't claim to fully understand their comments, but there is NOT a method
called default_test. It looks like it's saying that if you create a method
called 'default_test' in your derived class, it will not run though. But
test methods are supposed *start* with "test_", not end with "_test".
If it were my decision, I'd be very reluctant to complicate the UI (or limit
the RemoteTestRunner) to support some wierd rails behavior. If necessary, an
option on RemoteTestRunner would be reasonable. Of course, there could be
some other problem with RemoteTestRunner.
David
> What happens is that
> if there are 9 real tests that all pass the bar will be green but the
> view will show 9/10 tests with an empty block for the last one and
> this default_test will appear in the test hierarchy. As a hack I do a
> check for the ActionController::IntegrationTest and bypass when it
> attempts to get picked up. I haven't committed the code or anything
> but there needs to be a way to configure the test launcher to exclude
> certain classes or just implement the hack.
>
> On 4/4/06, Christopher Williams <[EMAIL PROTECTED]> wrote:
> > All,
> >
> > > I think an 0.8.0 release (at least a release candidate) is do-able.
> > > We already have a lot of functionality since the 0.7.0 release, and
> > > it's been working fine for me (I patched bugs with the new browsing
> > > stuff locally, but until just now SF.net's CVS was down so I just
> > > checked them in).
> > >
> > > The only bug I'd like to see fixed for an 0.8.0 is the Fixnum
> > > syntax highlighting - we highlight numbers inside identifiers (like
> > > method names) when we shouldn't. Beyond that I'd encourage everyone
> > > to try out a copy of the nightly build and make sure what's in
> > > there is stable - particularly Ruby browsing, handling of RJS/RXML,
> > > the new class wizard, template insertion via content assist, code
> > > formatting and the goto matching bracket action.
> > >
> > > There's release related tasks I'd like to see happen as well, but
> > > they should be able to be done post-release - particularly, I'd
> > > like to use RadRails experience to help us create some single-
> > > download releases of 0.8.0 - so you pick the download for your
> > > platform and get the bare minimum stuff you need to run RDT (plus
> > > maybe some branding like splash loader screen, icon, etc).
> >
> > An additional couple notes:
> >
> > We should probably set up our plugins to enforce that they require
> > the 3.1.x versions of eclipse plugins (we'd need to figure out which
> > plugins and versions exactly) - so we can avoid our users trying RDT
> > out on 3.2 and having all sorts of odd bugs.
> >
> > Next, I think immediately after this release we should target Eclipse
> > 3.2. They've released their last milestone build and the final 3.2
> > should be coming by June.
> >
> > Thanks,
> > Chris
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting language
> that extends applications into web and mobile media. Attend the live
> webcast and join the prime developer group breaking into this new coding
> territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid0944&bid$1720&dat1642
> _______________________________________________
> Rubyeclipse-development mailing list
> Rubyeclipse-development@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/rubyeclipse-development
-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
Rubyeclipse-development mailing list
Rubyeclipse-development@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rubyeclipse-development
