How to mark a step fail in Cucumber?

Question

I'm checking to see if there is any error message in log files. If an error message found in a log file, then I use 'raise' statement to print out the founding and continue checking the next log file. With an error message is found in a log file, Cucumber still indicated the step passed. I'd like to know how to mark a step fail, so it will print it at end of run (for example: 2 scenarios (1 failed, 1 passed) 4 steps (1 failed, 3 passed)). Any help would be appreciated!

                 Scenario: Running rake test
                   Given the system is installed
                   Then I run the rake test



        logs_all = s.sudo "egrep -i '#{error_message}' #{log_file}"
        logs_all.each do |hostname, logs|
           unless logs.empty?
            puts line, "Unhappy logs on #{hostname}", line, logs
            happy = false
           end

          begin
            raise "Unhappy logs found! in #{log_file}" unless happy
          rescue StandardError => error
            puts error.message
          end

        end

Answer 1

First I would recommend reading http://pragprog.com/book/hwcuc/the-cucumber-book or at least http://pragprog.com/book/achbd/the-rspec-book which has a whole part on Cuccumber.

Cucumber tests must pass, not fail. If one fails, the other may be skipped, which is not what we want when running a test suite. But you can check that a step raises an error if a log is not empty.

A scenario could be :

Feature: Logs
Scenario: Checking non-empty log files
    Given there are logged errors
    When I check if logs are empty
    Then an error should be raised

Run cucumber , it will suggest the necessary steps :

You can implement step definitions for undefined steps with these snippets:

Given /^there are logged errors$/ do
  pending # express the regexp above with the code you wish you had
end

When /^I check if logs are empty$/ do
  pending # express the regexp above with the code you wish you had
end

Then /^an error should be raised$/ do
  pending # express the regexp above with the code you wish you had
end

Put these into a .../features/step_definitions/xxx_steps.rb file, then start replacing pending with real code. You can't use the block of code you posted. You'll start writing RSpec tests. A first step could be writing a method which reads the logs. A second method detects if any log is not empty. The last spec will check that an error is raised.

HTH