Search code examples
rubyweb-scrapingcapybarapoltergeist

How to debug capybara not clicking links?

I'm trying to use capybara for web scraping. Namely, making it click the link, which is a span actually, but has onclick handler attached:

#!/usr/bin/env ruby
require 'rubygems'
require 'bundler/setup'
require 'capybara/dsl'

include Capybara::DSL
Capybara.default_driver = :poltergeist

visit 'http://example.com'
p current_url
sleep 10
first('#some-id').click
sleep 10
p current_url

But judging from current_url it fails. For now I'm using poltergeist driver. When I run this code against the other (test) page with span having onclick handler attached, it works. How can I figure out what's the issue?

Solution

  • What I did. Installed selenium-webdriver. Switched to driver :selenium:

    Capybara.default_driver = :selenium

    And added byebug after clicking the link. And now I see that clicking the link redirects to the other site, but also opens new window. However, capybara reports url didn't change.

    UPD The issue was that clicking the link opened two new windows (in terms of capybara, from user point of view only one more window appeared). That's why capybara reported the same url. And with Thomas Walpole's help, I was able to get to the new url of initial window:

    within_window(Capybara.current_session.windows[-2]) { p current_url }

    See this comment for alternative ways of choosing the window.