Option #1 Selenium-WebDriver gem

First install the selenium-webdriver gem:

gem install selenium-webdriver

Then start using it:

require 'selenium-webdriver'
width = 1024
height = 728
driver = Selenium::WebDriver.for :firefox
driver.navigate.to 'http://google.com'
driver.execute_script %Q{
  window.resizeTo(#{width}, #{height});
}
driver.save_screenshot('/tmp/screenshot.png')
driver.quit

Reference.

Option #2 Selenium Server and the Selenium Gem

I was using this option for quite some time, when it stopped working I started using the selenium-webdriver gem instead.

First create a separate Firefox profile:

firefox -ProfileManager -no-remote

Find the path to the profile by executing this code in the JavaScript error console in Firefox:

Components.classes["@mozilla.org/file/directory_service;1"].getService( Components.interfaces.nsIProperties).get("ProfD", Components.interfaces.nsIFile).path

The path looks somewhat like this on OSX:

/Users/christian/Library/Application Support/Firefox/Profiles/xxx

Now download the Selenium Server. Then start a Selenium server:

java -jar ~/Downloads/selenium-server-standalone-2.0b1.jar -firefoxProfileTemplate "/Users/christian/Library/Application Support/Firefox/Profiles/9eb0mn2d.screenshots/" -singleWindow

Next install the Selenium Ruby client:

gem install selenium-client

Take a screenshot with Ruby:

require 'rubygems'
require 'selenium'

@selenium = Selenium::SeleniumDriver.new("localhost", 4444, "*firefox", "http://www.cnn.com", 10000);
@selenium.start
@selenium.open "/US/"
@selenium.wait_for_page_to_load(60)
@selenium.send(:sleep, 7) # Wait for Flash and other crap to load
@selenium.capture_entire_page_screenshot('/tmp/screenshot.png', '');
@selenium.stop

Want more? Read the Selenium RC and Selenium client documentation

To change the accept-language header that Chrome sends with each request in Capybara tests, add the ‘selenium-webdriver (3.142.6)’ preference to the driver’s options:

  opts.add_preference('intl.accept_languages', 'sv')

Example:

Capybara.register_driver :chrome do |app|
  caps = Selenium::WebDriver::Remote::Capabilities.chrome(loggingPrefs: { browser: 'ALL' })
  opts = Selenium::WebDriver::Chrome::Options.new
  opts.add_preference('intl.accept_languages', 'sv')
  chrome_args = %w[--no-sandbox --disable-popup-blocking --enable-features=NetworkService,NetworkServiceInProcess --window-size=1920,1080]
  chrome_args.each { |arg| opts.add_argument(arg) }
  Capybara::Selenium::Driver.new(app, browser: :chrome, options: opts, desired_capabilities: caps)
end

Tested with:

• capybara (3.30.3)
• selenium-webdriver (3.142.6)

You an enable the chromedriver log to debug and fix chromedriver and Selenium errors such as “unknown error: net::ERR_CONNECTION_REFUSED”.

First start the chromedriver in the background:

nohup chromedriver --port=12345 --verbose --log-path=/tmp/chromedriver.log &

Then configure Selenium to use the chromedriver instance:

Capybara::Selenium::Driver.new(@app, browser: :chrome, options: ..., url: 'http://localhost:12345')

Next run the tests and inspect the chromedriver log.

In my case the issue was that an unreachable URL, see the unreachableUrl property in the output below:

[1659101781.048][DEBUG]: DevTools WebSocket Event: Page.frameNavigated B30A4B9FE0908FDACE0522341267520A {
   "frame": {
      "adFrameStatus": {
         "adFrameType": "none"
      },
      "crossOriginIsolatedContextType": "NotIsolated",
      "domainAndRegistry": "",
      "gatedAPIFeatures": [  ],
      "id": "B30A4B9FE0908FDACE0522341267520A",
      "loaderId": "595BC7FEFD2C1ECCE31157FF7972F889",
      "mimeType": "text/html",
      "secureContextType": "InsecureScheme",
      "securityOrigin": "://",
      "unreachableUrl": "http://localhost:3030/",
      "url": "chrome-error://chromewebdata/"
   },
   "type": "Navigation"
}

Thanks: https://makandracards.com/makandra/301588-how-to-enable-chromedriver-logging