Capturing logcat on scenario fail

[Brad Thompson]

Hey everyone,

So I'm trying to capture logs from android device on fail.  To do this I have the following:

After do |scenario|

  if scenario.failed?

     adb logcat -t 500 > Feature_name_Scenario_name.txt

  end

end

-----------------

So this seems like a good idea but what I'm finding is Ruby's memory is going crazy when I run this.  Specifically I run our tests through Jenkins.  I believe this is causing memory on the jenkins slave to go wild.

Has anyone tried grabbing logcat files on failure?

Any ideas around this?

Thanks,
Brad

[Brad Thompson]

I see in the operations.rb file we have a method:

      def get_logcat_cmd(tag)

        # returns raw logcat output for our tag

        # filtering out everthing else

        "#{adb_command} logcat -d -v raw #{tag}:* *:S"

      end

can this be used? what is tag?

[Brad Thompson]

I got this to work so if anyone else wants to do this.

In the After Scenario hook put:

run_adb_command("logcat -t 3000 > #{logcat_output_filename}")

then I have a method:

def run_adb_command(cmd)

  full_cmd = "#{default_device.adb_command} #{cmd}"

  raise "failed to run #{full_cmd}" unless system(full_cmd)

end