Usage of block in hooks.rb

[NaviHan]

Hi all

Im just looking for some good documentation which explains the usage of blocks in hooks.rb

For example "Before", "After" , "AfterStep" etc...

Thanks in advance

[Lakshya Kapoor]

You are confusing Cucumber hooks with Watir hooks.

Watir only has the After hook - https://www.rubydoc.info/gems/watir/Watir/AfterHooks

Anyway, here is how you can use the 3 hooks you mentioned:

Before: Do something before the test starts (before first step). For example: Launch browser, log into the application, or to setup Watir's AfterHook to check for errors (see the link above).

After: Do something after the test finishes (after last step). For example: Log out of the application, close browser, or data cleanup.

AfterStep: Do something after each step in the Scenario. For example: Take a screenshot if the step fails.

You can read more here - https://docs.cucumber.io/cucumber/api/#hooks

[NaviHan]

Thanks Laskhya. Im aware of the use of the hooks. But rathet confused by the syntax.. Here are the blocks in my projects hooks.rb

Sometimes "scenario" is passed, Sometimes "scenario" and "block" are passed. Its quite confusing how exactly its doing?

Before do
 @browser = browser
end
Before do|scenario|
 ENV['HTTP_PROXY'] = ENV['http_proxy'] = nil
 DataMagic.load('order_details.yml')
 DataMagic.load_for_scenario(scenario)
end
After do |scenario|
 @browser.execute_script("javascript:localStorage.clear();")
 @browser.cookies.clear
 if (scenario.passed?)
 puts 'Scenario Passed !'
 else
 puts 'Scenario Failed !'
 end
end
AfterStep('@screen') do
 filename = DateTime.now.strftime("%d%b%Y%H%M%S")
 @browser.screenshot.save ("C:/Users/lohit.kotian/Documents/Automation/screenshots-cucumber/#{filename}.png")
 #@browser.screenshot.save ("C:/Jenkins/workspace/screenshots-cucumber/#{filename}.png")
end
Around('@multipletimes') do |scenario, block|
 $counter = 0
 $total_times_to_run = 6
 while $counter < $total_times_to_run do
 block.call
 puts("Total times scripts were repeated = #$counter" )
 $counter +=1
 end
end
After do |scenario|
 take_screenshot(@browser, scenario)
end

[Lakshya Kapoor]

Sorry, I assumed otherwise because you posted in the Watir group :).

I haven't used the Around hook before, so I decided to figure it out and explain it here. Hopefully someone more experienced can chime in if my answer is incomplete or incorrect. 

So, it appears that the Around hook is specifically designed to provide access to the scenario (Cucumber::Core::Test::Case) object and a block (Proc). The block is what holds the execution logic/steps in your Scenario (source) AND the Before/After hooks. If you don't call the block, the Before/After hooks and the Scenario steps will not execute. This establishes the execution order as Around hook > Before/After hook > Scenario steps. You can test this by printing something from the Before/After hooks and commenting out block.call in your Around hook. You'll see that nothing is printed, including the Scenario steps, until you uncomment block.call.

Now that we know the Before/After hooks are invoked by the block from the Around hook, there is no further use of the block once it is called. Therefore, the Before/After hooks only allow us to work with the scenario object.

Hope this explains the reasoning behind the difference in the syntax.

[Lakshya Kapoor]

Also, I would recommend joining the Cucumber channels on Slack (https://cucumber.io/support#slack) if you have more questions.

[Lakshya Kapoor]

Since we can't edit our posts, here is a quick update to my answer: Technically, the correct execution order is Around hook > Before hook > Scenario > After hook.

[NaviHan]

Thanks Lakshya for the detailed explanation.

So if I want to execute a scenario 3 times, here is the order of execution

iteration 1 -> Before Hook   Scenario Steps  After Hook

iteration 2 -> Before Hook   Scenario Steps  After Hook

iteration 3 -> Before Hook   Scenario Steps  After Hook

Or is it

Before Hook

iteration 1 -> Scenario Steps

iteration 2 -> Scenario Steps

iteration 3 -> Scenario Steps

After Hook

Also what is the difference between

Before do
 @browser = browser
end




and




Before do |scenario|
 @browser = browser
end




On Tuesday, 29 January 2019 16:47:10 UTC+11, NaviHan wrote:

[Lakshya Kapoor]

You are welcome!

The order of execution will always be:

iteration 1 -> Before Hook   Scenario Steps  After Hook

iteration 2 -> Before Hook   Scenario Steps  After Hook

iteration 3 -> Before Hook   Scenario Steps  After Hook

This is also a good practice, so that your Scenarios always start from a clean state - new browser instance, new login session, etc.

The difference between the two Before hooks you shared is that you are getting access the scenario object in the second one. This object allows you to get info such as Scenario title, description, tags, pass/fail status in After hook, and so on. This doc will help you figure out all the methods you can use on that object - https://www.rubydoc.info/github/cucumber/cucumber-ruby/

[NaviHan]

Hi Lakshya

Here is the hooks.rb used in our project..

Eventhough the the browser is assigned to instance variable @browser we still get to see all the details you mentioned. So Im not quite getting the difference between the two Before hooks

require 'rubygems'

require 'page-object'

require 'watir'

require 'page-object/page_factory'

require 'log4r'

require 'cucumber'

require 'selenium-webdriver'

require 'fig_newton'

require 'watir-screenshot-stitch'

require 'mini_magick'

args = ['--allow-running-insecure-content']

browser = Watir::Browser.new :chrome, options: {args: args}

$code_retry = 30

Before do

 @browser = browser

end

Before do|scenario|

        ENV['HTTP_PROXY'] = ENV['http_proxy'] = nil

        DataMagic.load_for_scenario(scenario)

        @browser.window.maximize

        @browser.driver.manage.timeouts.implicit_wait = 0

        @browser.cookies.clear

        @browser.driver.manage.window.maximize

        PageObject.default_element_wait=(10)

        PageObject.javascript_framework = :jquery

        #Selenium::WebDriver.logger.level = :info

end #before scenario

# "after all"

at_exit do

 browser.close

end

# need to check this code for the full page screenshots as a part of the enhancement

  def take_screenshot(browser, scenario)

    time = Time.now.strftime("%Y-%m-%d_%H%M")

    opts = {:page_height_limit => 5000}

    if scenario.failed?

      scenario_name = scenario.name.gsub(/[^\w\-]/, ' ')

      screenshot_path =  "#{scenario_name}" + "_failure_" + time

      @browser.screenshot.save_stitch("./screenshots/#{screenshot_path}.png", browser, opts)

    end

  end

On Tuesday, 29 January 2019 16:47:10 UTC+11, NaviHan wrote:

[Lakshya Kapoor]

There is no difference in functionality from a Hooks perspective. They both work the same way, except that the in the second one you are getting access to the Scenario details (object).

Actually, you can come both the Before hooks to achieve the same result:

Before do|scenario|
       @browser = browser

       ENV['HTTP_PROXY'] = ENV['http_proxy'] = nil

       DataMagic.load_for_scenario(scenario)

       @browser.window.maximize

       @browser.driver.manage.timeouts.implicit_wait = 0

       @browser.cookies.clear

       @browser.driver.manage.window.maximize

       PageObject.default_element_wait=(10)

       PageObject.javascript_framework = :jquery

       #Selenium::WebDriver.logger.level = :info

end #before scenario