Grab all text visible on a Web page with Ruby and Ferrum

grab.rb

# notice the cheating technique of selecting all,
# copying to clipboard, then reading the
# clipboard back via JavaScript(!!)
# there's also some stuff to rip content out of
# shadow roots which can be useful if a page
# is doing dynamic rendering
#
# MIT licensed, (c) 2022 Peter Cooper

require 'ferrum'

def get_browser
  browser = Ferrum::Browser.new(slowmo:0.1)
  context = browser.contexts.create
  [browser, context, context.create_page]
end

def close_browser(browser, context, page)
  context.dispose
  browser.quit
end

def get_page(url)
  browser, context, page = get_browser
  page.go_to(url)
  browser.mouse.move(x: rand(400), y: rand(400))
  shadows = page.evaluate_func <<~JS
    function() {
      let res = '';
      for (let el of document.getElementsByTagName('*'))
        if (el.shadowRoot) res += el.shadowRoot.innerHTML + " \\n";
      return res;
    }
  JS
  page.keyboard.type([:Ctrl, 'a'],[:Ctrl, 'c'])
  contents = page.evaluate %{window.getSelection().toString()}
  contents = contents + shadows
  close_browser(browser, context, page)
  return contents
end

p get_page(.... whatever URL you like here ....)