Capybara Selenium Webdriver: Headless Chrome (with file downloads!) & Headless Firefox

Brewfile

tap "caskroom/cask"

cask "google-chrome"
cask "firefox"

brew "chromedriver"
brew "geckodriver"

chrome.rb

require 'capybara/rspec'
require 'selenium/webdriver'

options = Selenium::WebDriver::Chrome::Options.new
options.add_preference(:download, prompt_for_download: false,
                                  default_directory: '/tmp/downloads')

options.add_preference(:browser, set_download_behavior: { behavior: 'allow' })

Capybara.register_driver :chrome do |app|
  Capybara::Selenium::Driver.new(app, browser: :chrome, options: options)
end

Capybara.register_driver :headless_chrome do |app|
  options.add_argument('--headless')
  options.add_argument('--disable-gpu')
  options.add_argument('--window-size=1280,800')

  driver = Capybara::Selenium::Driver.new(app, browser: :chrome, options: options)

  ### Allow file downloads in Google Chrome when headless!!!
  ### https://bugs.chromium.org/p/chromium/issues/detail?id=696481#c89
  bridge = driver.browser.send(:bridge)

  path = '/session/:session_id/chromium/send_command'
  path[':session_id'] = bridge.session_id

  bridge.http.call(:post, path, cmd: 'Page.setDownloadBehavior',
                                params: {
                                  behavior: 'allow',
                                  downloadPath: '/tmp/downloads'
                                })
  ###

  driver
end

Capybara.javascript_driver = ENV['GUI'] ? :chrome : :headless_chrome

firefox.rb

require 'capybara/rspec'
require 'selenium/webdriver'

Capybara.register_driver :firefox do |app|
  Capybara::Selenium::Driver.new(app, browser: :firefox)
end

Capybara.register_driver :headless_firefox do |app|
  options = Selenium::WebDriver::Firefox::Options.new
  options.headless! # added on https://github.com/SeleniumHQ/selenium/pull/4762

  Capybara::Selenium::Driver.new app,
    browser: :firefox,
    options: options
end

Capybara.javascript_driver = ENV['GUI'] ? :firefox : :headless_firefox

Gemfile

source 'https://rubygems.org'

gem 'selenium-webdriver'
gem 'capybara'
gem 'rspec'

The section on how to download attachments from chromedriver headless in ruby was very useful. Thanks 👏

thank you for this gist, very useful to me across multiple projects!

BIG BIG THANK YOU!!!!!

Was anybody able to get this to work for multiple tests/scenarios?

In my case I had to run the code that sets Page.setDownloadBehavior before every test that needed to do a headless download. It would download to the path specified on the first test but then on subsequent tests the download worked but instead was downloaded to the project root folder rather than the directory in downloadPath.

In a Before(@javascript) I just run the code that makes the http call to bridge and now it works for every test.

I did check that the session_id was the same between test runs but it seems that thedownloadPath setting is lost somewhere.

Also I'm not sure if it is a newer way but I used

driver.browser.execute_cdp("Browser.setDownloadBehavior", { behavior: "allow", downloadPath: download_path }) instead.

It uses execute_cdp_cmd and also I thought that Browser.setDownloadBehavior might be more appropriate than Page.

That being said whether I used the exact code above or my changes the downloadPath setting was lost after the first test.

Was anybody able to get this to work for multiple tests/scenarios?

@ozzyaaron we've hit the same issues here, the downloadPath seems to be lost between tests.
Clearly not ideal but we've worked around the issue by having the session destroyed and rebuilt for each example in the suite.

  after do
    Capybara.current_session.quit
  end

Thank you bbonamin and dancristianb for saving me tons of time as I update an old Rails app's webdrivers gem from 4.2.0 - with the tap syntax - to 5.3.1 - with this one! 🥳👏👏👏

You might want to use the new headless mode in Chrome:

options.add_argument('--headless=new')

Also don't forget other default driver options.