Failing before :all doesn't move progress bar in html reports

Hello.

I’ve found one bug when failing before :all block doesn’t move html
report’s
progress bar correctly.

You can test it out easily by doing something like this:

describe “bad bad bad” do

before :all do
true.should be_false
end

it “empty” do
end

end

Now run this spec with html formatter and you should get 100% width to
rspec-header div, but it has width of 0%. When you add some other
describe
blocks with specs then it would be something below 100%, but should be
100%
if everything is finished.

I’ve fixed this in my custom htmlformatter by adding these lines into
example_failed method prior invoking super:

def example_failed example, counter, failure

#…
#some other stuff
#…

if example.description == "before(:all)"
  @example_number +=

@options.example_groups[example_group_number-1].examples.size
end
super

end

If there’s any better solution, please let me know :slight_smile:

Best Regards,
juuser.

View this message in context:
http://www.nabble.com/failing-before-%3Aall-doesn't-move-progress-bar-in-html-reports-tp19059450p19059450.html
Sent from the rspec-users mailing list archive at Nabble.com.

Maybe you shouldn’t be placing spectations in a before or after filter,
it
doesn’t make much sense.

It makes sense to me. Let’s say that you want to initialize object in
your
before :all, but this fails.

For example, in my case I’m using RSpec for automated web testing with
Watir
and I’m logging into web application from before :all and if this fails
then
it’s logical that there’s no point to execute any further specs and
carry on
with next describe block, but html report won’t be correct.

Maurício Linhares-3 wrote:

Maybe you shouldn’t be placing spectations in a before or after filter, it
doesn’t make much sense.


View this message in context:
http://www.nabble.com/failing-before-%3Aall-doesn't-move-progress-bar-in-html-reports-tp19059450p19059735.html
Sent from the rspec-users mailing list archive at Nabble.com.

On Tue, Aug 19, 2008 at 5:55 PM, Maurício Linhares
[email protected] wrote:

Maybe you shouldn’t be placing spectations in a before or after filter, it
doesn’t make much sense.

I disagree. They can be useful for verifying assumptions about
fixture data. At any rate, a bug’s a bug.

juuser: can you please submit a ticket @ http://rspec.lighthouseapp.com
?

Pat