Search code examples
rubyseleniumcapybaratumblrpoltergeist

Ruby: Phantom.js blocked on specific site?


I am using capybara poltergeist to automate a small script on tumblr.com

My script works fine with my chrome driver.. And my poltergeist driver loads all other websites just fine, but for some reason throws a Capybara::Poltergeist::StatusFailError when I try to load tumblr.

Steps for reproduction:

$ brew install phantomjs
$ gem install capybara
$ gem install poltergeist
$ gem install selenium-webdriver
$ irb


require 'capybara/poltergeist'

module Drivers
  class Poltergeist < Capybara::Poltergeist::Driver
    def needs_server?
      false
    end
  end
end

Capybara.register_driver :poltergeist_errorless do |app|
  Drivers::Poltergeist.new(app, js_errors: false, timeout: 10000, phantomjs_options: ['--load-images=no', '--ignore-ssl-errors=yes'])
end

session = Capybara::Session.new(:poltergeist_errorless)
session.visit('https://google.com') # This works fine
session.visit('https://tumblr.com') # This does not work?

I tried to set all of my headers to look my google chrome's request, but that also does not seem to fix it. Does anyone have any suggestions?


Solution

  • The problem is related to phantomjs SSL handshake fail. You can take my gist and run with phantomjs, you will see:

    [cut]
    = onResourceError()
      - unable to load url: "https://www.tumblr.com/"
      - error code: 6, description: SSL handshake failed
    = onResourceReceived()
      id: 3, stage: "end", response: {"contentType":null,"headers":[],"id":3,"redirectURL":null,"stage":"end","status":null,"statusText":null,"time":"2014-09-16T12:06:05.547Z","url":"https://www.tumblr.com/"}
    = onLoadFinished()
      status: fail
    DONE WITH  fail WebPage(name = "WebPage")
    

    Checking a little bit a workaround is to use --ssl-protocol=any in phantom, so your code will become:

    Capybara.register_driver :poltergeist_errorless do |app|
      Drivers::Poltergeist.new(app, js_errors: false, timeout: 10000, phantomjs_options: ['--load-images=no', '--ignore-ssl-errors=yes', '--ssl-protocol=any'])
    end
    

    To work.

    References: