Problem using cucumber with selenium

Testing RSpec
1

I’m trying to use Cucumber with Selenium to test my app. I have a
method
called “login_as” defined as follows:

def login_as(user, password)
visit new_session_path
fill_in “username”, :with => user.username
fill_in “password”, :with => password
click_button(‘Go’)
response.body.should =~ /Signed in as <a
href="[^"]+">#{user.username}</a>/
end

This works fine when I’m not using Selenium, but it fails when I do use
Selenium. Weirder still is that if I use Selenium but put a “debugger”
step
before the “click_button” step, and I execute the last two lines through
the
console, everything works. When it fails, the error seems to indicate
that
nothing on the page has changed – that is, that “response” is still the
old
response from before I executed the “click_button” step. My best guess
is
that the code perhaps isn’t waiting for the browser to actually load the
new
page before looking at the response. Does anyone know how to fix this
issue?

Thanks…

2

found the answer via some more searching; it appears this is a bug with
webrat that hasn’t been fixed yet; if i add
“selenium.wait_for_page_to_load(5)”
after the click_button command, things work fine.

i’ve opened up a new ticket in the webrat lighthouse acct.

3

2009/4/20 Barun S. [email protected]

found the answer via some more searching; it appears this is a bug with
webrat that hasn’t been fixed yet; if i add
“selenium.wait_for_page_to_load(5)”
after the click_button command, things work fine.

I hit a similar problem and ended up putting the following in my feature
setup (it would need something similar for the response method but I’m
not
actually using that).
Doing it this way means you don’t have to put any extras in your
features to
work around the bug so once it’s fixed you can just remove the hack :slight_smile:
module Webrat
class SeleniumSession
def response_body
sleep 0.2
selenium.get_html_source
end
end
end

I’ve not experimented with the sleep length I just tried it with 0.2 and
it
worked…

– roovo