Ordered provider verification

task.rb

require 'pact/provider/proxy/tasks'

Pact::ProxyVerificationTask.new :userapi_service do | task |
  task.pact_url './spec/pacts/userapi_service_consumer-userapi_service_provider.json'
  task.provider_base_url 'https://userapi.com/'
end

Pact::ProxyVerificationTask.new :otherapi_service do | task |
  task.pact_url './spec/pacts/otherapi_service_consumer-otherapi_service_provider.json'
  task.provider_base_url 'https://otherapi.com'
end

desc('Run the provider integration tests, in the specified order, ')
task(:provider_verify_ordered) do
  interactions_to_verify= [['registration success','user'],['get_offer','other'],['issue offer','offer'],['payment','user'],['get extra offers','other']]

  error = false
  interactions_to_verify.each do |interaction, s|
    begin
      service = (s == 'user' ? 'userapi_service' : 'otherapi_service')
      puts "Checking interaction #{interaction} on #{service}"
      ENV['PACT_DESCRIPTION'] = interaction
      Rake::Task["pact:verify:#{service}"].execute
    rescue StandardError => e
      error = true
      puts e.inspect
      puts "Checking interaction #{interaction} failed\n\n"
    end
  end
end

A few pointers: Catching Exception in Ruby can cause problems, it's better to catch StandardError, because catching Exception will stop you being able to send signals to the script like Ctl+C to stop it. See this post here for more information: http://daniel.fone.net.nz/blog/2013/05/28/why-you-should-never-rescue-exception-in-ruby/

Also, the script you had would not return an error code if it failed, because it was catching the SystemExit exception when one of the tasks failed. One way to fix this, if you actually wanted to continue would be to have an error flag that got set to true, and then fail if error at the end of the script. But having changed the script to not rescue Exception, this will now stop if one of the tasks fails - I assume there is no point continuing with a later step if the first step has failed.