[cucumber] automation and historical

Testing RSpec
1

Hi guys,

Can someone point me to information regarding:

  1. running cucumber as an automated task decoupled from CI. Assuming
    cron or similar.
  2. parsing and storing results so they can be graphed, etc.

Many thanks in advance,

Tim

2

Hate to “glom on”…I was curious about the exit status from cucumber…

I did a rake features to a log file and this is the last bit:

35 steps passed
2 steps failed
4 steps skipped
1 steps pending

exit status was 0 (echo $?)

Do you guys parse this from the log or what?

Many thanks,

Tim

3

what kind of graphs are you thinking? Can they be plugged into rcumber?

4

On Mon, Dec 22, 2008 at 11:52 AM, Tim W. [email protected] wrote:

Tim

rspec-users mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/rspec-users

Hi,

Well, you can just run “rake features” from cron…

As for storing the results for graphing, you could make a custom
listener that stores the results in a CSV list. You know
features_run,scenarios_run,passing_steps, etc etc. Then you have a
little script that runs over the CSV file and feeds the data into a
graphing tool/lib. You could actually do all of that from within a
formatter too if you want…just tally up the results and then at the
end build a graph from it.

Pat

5

On Mon, Dec 22, 2008 at 8:55 PM, Tim W. [email protected] wrote:

Really??

Try running rake features in examples/i18/en
Try running it again - this time with an error added to calculator.rb

First time I get 0, second 1.

Please confirm.

Aslak

6

Hi Aslak,

Yes, that produced 1. I’ll try it again tomorrow. So, if anything
fails the error code will be not 0. Is it always 1 or the number of
failing tests or anything else?

Thanks again,

Tim

On Mon, Dec 22, 2008 at 2:04 PM, aslak hellesoy

7

The standard stuff, number of tests, passing, by type, daily results
summary…things that show a consistent attention to test automation.

Let me play with rcumber. Thanks, Tim

8

Den 23. des… 2008 kl. 01.46 skrev “Tim W.” [email protected]:

Hi Aslak,

Yes, that produced 1. I’ll try it again tomorrow. So, if anything
fails the error code will be not 0. Is it always 1 or the number of
failing tests or anything else?

Always 1, but I like your idea.