capybara assert attributes of an element

问题:

I'm using RSpec2 and Capybara for acceptance testing.

I would like to assert that link is disabled or not in Capybara. How can I do this?

回答1:

How are you disabling the link? Is it a class you're adding? An attribute?

# Check for a link that has a "disabled" class:
page.should have_css("a.my_link.disabled")
page.should have_xpath("//a[@class='disabled']")

# Check for a link that has a "disabled" attribute:
page.should have_css("a.my_link[disabled]")
page.should have_xpath("//a[@class='disabled' and @disabled='disabled']")

# Check that the element is visible
find("a.my_link").should be_visible
find(:xpath, "//a[@class='disabled']").should be_visible

The actual xpath selectors may be incorrect. I don't use xpath often!

回答2:

Another simple solution is to access the HTML attribute you are looking for with []:

find('#my_element')['class']
# => "highlighted clearfix some_other_css_class"

find('a#my_element')['href']
# => "http://example.com

# or in general, find any attribute, even if it does not exist
find('a#my_element')['no_such_attribute']
# => ""

Note that Capybara will automatically try to wait for asynchronous requests to finish, but it may not work in some cases:

• http://www.elabs.se/blog/53-why-wait_until-was-removed-from-capybara

Here is one workaround if you are having trouble with assertions on elements that are updated asynchronously:

• http://robots.thoughtbot.com/automatically-wait-for-ajax-with-capybara

回答3:

It was a bit messy to find out the correct xpath, here is the correct one,
using capybara 0.4.1.1

# <a href="/clowns?ordered_by=clumsyness" class="weep">View Clowns</a>  

page.should have_xpath("//a[@class='weep'][@href='/clowns?ordered_by=clumsyness']", :text => "View Clowns")

If you only have a link without a class, use

page.should have_link('View Clowns', :href => '/clowns?ordered_by=clumsyness')

Something like this will sadly not work:

page.should have_link('This will not work!', :href => '/clowns?ordered_by=clumsyness', :class => "weep")

The class option will be ignored.

回答4:

I recommend using have_link and find_link(name)[:disabled] in two separate assertions. While performing the second assertion alone is simpler, this makes error messages about missing links look nicer, making your test results easier to read.

expect(page).to have_link "Example"
expect(find_link("Example")[:disabled]).to be false

Note that "Example" can be changed to the name or id of the link.

回答5:

page.should have_link('It will work this way!', {:href => '/clowns?ordered_by=clumsyness', :class => "smile"})

have_link expects a hash of options which is empty if you do not provide any. You can specify any attributes the link should have - just make sure you pass all the options in ONE hash.

Hope this helps

PS: For attributes like data-method you have to pass the attribute name as a string since the hyphen breaks the symbol.

回答6:

Whenever possible, you should try to use the Capybara provided wrappers which will work more consistently across drivers.

For the particular case of disabled, a wrapper was introduced in 2.1: https://github.com/jnicklas/capybara/blob/fc56557a5463b9d944207f2efa401faa5b49d9ef/History.md#version-210

If you use it, you will get sensible results on both RackTest and Poltergeist:

HTML:

<input type="text" id="disabled-false"            ></div>
<input type="text" id="disabled-true"     disabled></div>
<input type="text" id="disabled-js-true"          ></div>
<input type="text" id="disabled-js-false" disabled></div>
<script>
  document.getElementById('disabled-js-true').disabled = true
  document.getElementById('disabled-js-false').disabled = false
</script>

Tests:

!all(:field, 'disabled-false',    disabled: false).empty? or raise
 all(:field, 'disabled-false',    disabled: true ).empty? or raise
 all(:field, 'disabled-true',     disabled: false).empty? or raise
!all(:field, 'disabled-true',     disabled: true ).empty? or raise
 all(:field, 'disabled-js-true',  disabled: true ).empty? or raise
 all(:field, 'disabled-js-false', disabled: false).empty? or raise

Capybara.current_driver = :poltergeist
!all(:field, 'disabled-false',    disabled: false).empty? or raise
 all(:field, 'disabled-false',    disabled: true ).empty? or raise
 all(:field, 'disabled-true',     disabled: false).empty? or raise
!all(:field, 'disabled-true',     disabled: true ).empty? or raise
!all(:field, 'disabled-js-true',  disabled: true ).empty? or raise
!all(:field, 'disabled-js-false', disabled: false).empty? or raise

Note how by using this instead of CSS selectors, the Javascript tests will work without any changes if you start using a Js capable driver.

Runnable test file here.

回答7:

bowsersenior, thanks for a hint

Another simple solution is to access the HTML attribute you are looking for with []

Here is an example:

let(:action_items) { page.find('div.action_items') }

it "action items displayed as buttons" do
  action_items.all(:css, 'a').each do |ai|
    expect(ai[:class]).to match(/btn/)
  end
end

回答8:

Using Rspec3's syntax i did it this way:

expect(page).not_to have_selector(:link_or_button, 'Click here')

回答9:

Simply you can use page.has_css? method

page.has_css?('.class_name') 

this will return true if element exists.

Do some action based on validation.

page.has_css?('.class_name') do
  #some code
end

回答10:

According to the docs you can use the [attribute] accessor syntax:

find('#selector')['class'] => "form-control text optional disabled"

For disabled, you could also do this:

expect(find('#selector').disabled?).to be(true)