Unable to click link using selenium webdriver - ruby

Question

I'm trying to get an automated Google search to click on the first link. So far I have not been successful and was wondering if someone could assist me. The search results populate, although the act of clicking the first link fails every time.
require "selenium-webdriver"
driver = Selenium::WebDriver.for :firefox
driver.navigate.to "http://google.com"
element = driver.find_element(:name, 'q')
element.send_keys "translate"
element.submit
resultlink = driver.find_Element(:link, "Google Translate")
resultlink.click

Answer 1

How about you try locating the First link using CSS selectors, something like this:
driver.find_element(:css, "#rso li:nth-child(1) div > h3 > a").click
where the 1 in the brackets (after nth-child) refers to the first search result.
Also I may be wrong, but try :link_text instead of :link, something like this :
resultlink = driver.find_element(:link_text, "Google Translate")
resultlink.click

Answer 2

If you watch this while it's happening, you might notice that it's failing before the results load. This is probably the single most annoying aspect of automation: timing.
I tried adding sleep(5) before defining the element and it worked. However, sleeps are generally bad, so you should instead give selenium a little leeway to find the element before deciding it doesn't exist. You do this through implicit waits. For example:
driver.manage.timeouts.implicit_wait = 5 #time in seconds
This sets the maximum time that selenium will allow for an element to load. If it finds it sooner, it will continue right away. For this reason, it is far more efficient than sleep. More info is available in the documentation. Set this any time before you need to find your element. Once set, this will apply for the remainder of your test. It's a good idea in general to allow for slight delays and/or network hiccups.

Answer 3

First, if you are learning Selenium, don't use any of the Google pages to start with. They look simple but are extremely tricky and complex under the hood. Find another website to automate please. It is against Google's user agreement anyway.
Then I can provide you working code. Note Google search results may render differently in different browsers, and you also need to use WebDriverWait for waiting.
require 'selenium-webdriver'
driver = Selenium::WebDriver.for :firefox
driver.navigate.to "http://google.com"
element = driver.find_element(:name, 'q')
element.send_keys "translate"
element.submit
wait = Selenium::WebDriver::Wait.new(:timeout => 10)
wait.until {
driver.find_element(:css , 'h3 > a')
}
# click first result
# driver.find_element(:xpath , '(.//h3/a)[1]').click
results = driver.find_elements(:css , 'h3 > a')
results.each { |result|
if result.attribute('textContent') == 'Google Translate'
result.click
break
end
}
(.//h3/a)[1] means the first result. In Firefox, results don't have unique data-href for identifying, so you need to use index.
Otherwise, you can loop through all result links for a link that its attribute textContent equals Google Translate. Note the link text for it is actually Google <em>Translate</em>, so using text() in XPath might not work.
If you find the solution above is too much to take in, it proves you shouldn't start learning Selenium using Google pages in the first place. ;)