Spreewald

Spreewald is a collection of useful steps for cucumber. Feel free to fork.

You can find a list of all contained steps at the end of this README.

Installation

Add this line to your application's Gemfile:

gem 'spreewald'

And then execute:

$ bundle

Or install it yourself as:

$ gem install spreewald

Supported versions

Spreewald is currently tested against and Ruby 2.6.6 with Rails 6 and Capybara 3.

Usage

Require all Spreewald steps by putting

require 'spreewald/all_steps'

into either your support/env.rb or step_definitions/spreewald_steps.rb.

Steps are grouped into a number of categories. If you only want to use a subset of Spreewald's steps, you can pick and choose single categories by requiring single files like this:

require 'spreewald/email_steps'
require 'spreewald/web_steps'

Have a look at our recommended Capybara defaults as they impact the behavior of many Spreewald steps.

Spreewald binary

Spreewald comes with a binary that prints a list of all Cucumber steps from Spreewald and your project. It will filter the list by any string you pass it. Example usage:

spreewald # lists all steps
spreewald mail # lists all steps that contain "mail"

The binary also prints paths and selectors; run spreewald --help to learn more.

Waiting for page load

Spreewald's web steps are all aware that you might run them with a Selenium/Capybara webdriver, and wait for the browser to finish loading the page, if necessary.

This is done by rerunning any assertions until they succeed or a timeout is reached.

We consider a couple of potential exceptions as "retriable", including

Capybara::ElementNotFound, (R)Spec::Expectations::ExpectationNotMetError, Capybara::Poltergeist::ClickFailed

You can add your own error class with

ToleranceForSeleniumSyncIssues::RETRY_ERRORS << 'MyCustomError'

You can achieve this in your own steps by wrapping them inside a patiently do block, like

Then /^I should see "([^\"]*)" in the HTML$/ do |text|
  patiently do
    page.body.should include(text)
  end
end

More info here.

Overriding steps

Thanks to cucumber_priority you can override any step definition from Spreewald with your own pattern. Cucumber will not raise Cucumber::Ambiguous if your custom steps collide with a Spreewald step.

If you are upgrading from Capybara 2 to Capybara 3, you might see failing tests with Spreewald steps like I should see. This is caused by a breaking change in Capybara's Finders that accept a :text option. To activate Capybara 2 behavior globally in your project, enable this flag:

Capybara.default_normalize_ws = true

This will affect all Spreewald steps that are using Capybara's :text option.

Furthermore, we recommend setting Capybara's matching strategy to :prefer_exact. This will positively affect Spreewald steps as it prevents the Capybara::Ambiguous error in the edge case when two fields are matching the given name, but one of the matches includes the name only as a substring,

Capybara.match = :prefer_exact

If you want Spreewald to match fields, links and buttons against the aria-label attribute, enable the following global configuration in Capybara 2.8+: 

Capybara.enable_aria_label = true

Contributing

Testing

Gemika is used to manage Gemfiles for different ruby versions.

Spreewald has different kind of tests:

  • Unit tests live in spec/
  • Test applications for integration tests with various Capybara versions live in tests/.
  • Tests for the Spreewald binary live in features/

Run all tests for your current ruby version with rake or rake matrix:tests. To bundle use rake matrix:install first.

Please note that you need geordi to run the integration tests.

If you would like to contribute:

  • Fork the repository
  • Push your changes with specs
  • Make sure rake passes
  • Regenerate the "Steps" section of this Readme with rake update_readme, if needed
  • Make a pull request

This README

The "Steps" section is autogenerated by rake update_readme from comments in the step definitions.

Steps

browser_tab_steps.rb

  • When I open ... in a new browser tab

Opens the page in a new browser tab and switches to it.

  • When I close the browser tab

Closes the current browser tab and switches back to the first tab.

  • When I switch to the new(ly opened)? browser tab

Waits for the new browser tab to appear, then switches to it.

  • When I switch( back)? to the previous browser tab

Changes the browser context to the second-last browser tab.

  • When I may open a new browser tab

Required for the check whether a new browser tab was opened or not.

  • Then I should( not)? have opened a new browser tab

Example (positive expectation):

  When I may open a new browser tab
    And I click on "Open link in new browser tab"
  Then I should have opened a new browser tab

Example (negative expectation):

  When I may open a new browser tab
    And I click on "Open link in current browser tab"
  Then I should not have opened a new browser tab

  • Then there should be (\d+) browser tabs?

  • Then there should be at least (\d+) browser tabs?

development_steps.rb

  • Then it should work...?

Marks scenario as pending, optionally explained with a reason.

  • Then console

Pauses test execution and opens an IRB shell with current context. Does not halt the application- under-test.

  • AfterStep @slow-motion

Waits 2 seconds after each step

  • AfterStep @single-step

Waits for a keypress after each step

email_steps.rb

  • When I clear my e?mails

  • Then (an?|no)( HTML| plain-text|) e?mail should have been sent with:

Example:

  Then an email should have been sent with:
    """
    From: [email protected]
    Reply-To: [email protected]
    To: [email protected]
    CC: [email protected]
    BCC: [email protected]
    Subject: The subject may contain "quotes"
    Attachments: image.jpg, attachment.pdf

    This is the message body. You can use * as a wildcard to omit the rest
    of a line *
    Or you can omit multiple lines if the asterisk is the only
    character in a single line, like this:
    *

    """

You may skip lines in the header. Please note: In older versions of Spreewald, unmentioned body lines were skipped. Now you have to use the asterisk explicitly to omit lines in the body.

  • When I follow the (first|second|third)? link in the e?mail

Please note that this step will only follow HTTP and HTTPS links. Other links (such as mailto: or ftp:// links) are ignored.

  • Then no e?mail should have been sent

  • Then I should see "..." in the( HTML| plain-text|) e?mail

Checks that the last sent email includes some text

  • Then show me the e?mail( header)?s

Print all sent emails to STDOUT (optionally only the headers).

  • Then show me the e?mail( header)?s with:

Print a subset of all sent emails to STDOUT This uses the same syntax as Then an email should have been sent with:

frame_steps.rb

  • When ... inside the ... iframe

You can append inside the [name or number] iframe to any other step. Then the step will operate inside the given iframe. Examples:

  Then I should see "Kiwi" inside the 1st iframe
  Then I should see "Cherry" inside the fruits iframe
  When I press "Save" inside the 2nd iframe
  • When I switch to the ... iframe

This step will switch to the iframe identified by its name or number. All further steps will operate inside the iframe. To switch to operating on the main page again, use the step "I switch back to the whole page". Examples:

  When I switch to the 1st iframe
  When I switch to the fruits iframe

Please note: This step is only available for Capybara >= 3.

  • When I switch back to the whole page

This step can be used to switch back to the whole page if you switched to operating inside an iframe before (step I switch to the ... iframe).

Please note: This step is only available for Capybara >= 3.

session_steps.rb

  • When ... in the browser session "..."

You can append in the browser session "name" to any other step to execute the step in a different browser session.

You may need to update other steps to allow multiple sessions (e.g. your authentication steps have to support multiple logged in users). More details here.

table_steps.rb

  • Then I should( not)? see a table with (exactly )?the following rows( in any order)?:?

Check the content of tables in your HTML.

See this article for details.

time_steps.rb

Steps to travel through time

This uses Timecop or Active Support 4.1+ to stub Time.now / Time.current. The user is responsible for including one of the two gems.

Please note that the two approaches branch. While ActiveSupport will freeze the time, Timecop will keep it running.

  • When the (date|time) is "?(\d4-\d2-\d2( \d1,2:\d2)?)"?

Example:

  Given the date is 2012-02-10
  Given the time is 2012-02-10 13:40
  • When the time is "?(\d1,2:\d2)"?

Example:

  Given the time is 13:40
  • When it is (\d+|an?|some|a few) (seconds?|minutes?|hours?|days?|weeks?|months?|years?) (later|earlier)

Example:

  When it is 10 minutes later
  When it is a few hours earlier

web_steps.rb

Most of cucumber-rails' original web steps plus a few of our own.

Note that cucumber-rails deprecated all its steps quite a while ago with the following deprecation notice. Decide for yourself whether you want to use them:

This file was generated by Cucumber-Rails and is only here to get you a head start These step definitions are thin wrappers around the Capybara/Webrat API that lets you visit pages, interact with widgets and make assertions about page content.

If you use these step definitions as basis for your features you will quickly end up with features that are:

  • Hard to maintain
  • Verbose to read

A much better approach is to write your own higher level step definitions, following the advice in the following blog posts:

  • When ... within ...

You can append within [selector] to any other web step, even multiple times. Be aware that within will only look at the first element that matches. If this is a problem for you following links, you might want to have a look at the 'When I follow "..." inside any "..."'-step.

Example:

  Then I should see "some text" within ".page_body"

  • Given I am on ...

  • When I go to ...

  • Then I should be on ...

  • When I press "..."

  • When I follow "..."

  • When I fill in "..." (with|for) "..."

Fill in text field

  • When I fill in "..." (with|for):

Fill in text field with multi-line block You can use a doc string to supply multi-line text

Example:

  When I fill in "some field" with:
  """
  Apple
  Banana
  Pear
  """
  • When I fill in "..." (with|for) '...'

Fill in text field

  • When I select "..." from "..."

Select from select box

  • When I check "..."

Check a checkbox

  • When I uncheck "..."

Uncheck a checkbox

  • When I choose "..."

Select a radio button

  • When I attach the file "..." to "..."

Attach a file to a file upload form field

  • Then I should( not)? see "..."

Checks that some text appears on the page

Note that this does not detect if the text might be hidden via CSS

  • Then I should( not)? see /.../

Checks that a regexp appears on the page

Note that this does not detect if the text might be hidden via CSS

  • Then I should( not)? see a field "..."

Checks for the existance of an input field (given its id or label)

  • Then I should( not)? see the (number|amount) ([-\d,.]+)( ...)?

Use this step to test for a number or money amount instead of a simple Then I should see

Checks for an unexpected minus sign, correct decimal places etc.

See here for details

  • Then I should( not)? see '...'

Like Then I should see, but with single instead of double quotes. In case the expected string contains quotes as well.

  • Then I should( not)? see "..." in the HTML

Check that the raw HTML contains a string

  • Then I should see an error

Checks that status code is 400..599

  • Then I should( not)? see (an|the) element "..."

Check that an element with the given selector is present on the page.

Example:

  Then I should see an element ".panel"
  Then I should see the element ".panel"
  Then I should not see an element ".sidebar"
  Then I should not see the element ".sidebar"
  • Then I should( not)? see (an|the) element for ...

Check that an element with the given selector alias is present on the page.

Example:

  Then I should see an element for the panel
  Then I should see the element for the panel
  Then I should not see an element for the sidebar
  Then I should not see the element for the sidebar
  • Then I should see in this order:?

Checks that these strings are rendered in the given order in a single line or in multiple lines

Example:

  Then I should see in this order:
    | Alpha Group |
    | Augsburg    |
    | Berlin      |
    | Beta Group  |
  • Then I should( not)? see a link labeled "..."

Checks that the page contains a link with a given text or title attribute.

  • Then the "..." field should( not)? contain "..."

Checks that an input field contains some value (allowing * as wildcard character)

  • Then the "..." field should( not)? contain:

Checks that a multiline textarea contains some value (allowing * as wildcard character)

  • Then I should see a form with the following values:

Checks that a list of label/value pairs are visible as control inputs.

Example:

  Then I should see a form with the following values:
    | E-mail | [email protected]   |
    | Role   | Administrator |
  • Then the "..." field should have the error "..."

Checks that an input field was wrapped with a validation error

  • Then the "..." field should( not)? have an error

  • Then the "..." checkbox should( not)? be checked?

  • Then the radio button "..." should( not)? be (checked|selected)

  • Then I should have the following query string:

Example:

  I should have the following query string:
    | locale        | de  |
    | currency_code | EUR |

Succeeds when the URL contains the given locale and currency_code params

  • Then show me the page

Open the current Capybara page using the launchy or capybara_screenshot gem

  • Then I should get a response with content-type "..."

Checks Content-Type HTTP header

  • Then I should get a download with filename "..."

Checks Content-Disposition HTTP header

Attention: Doesn't work with Selenium, see https://github.com/jnicklas/capybara#gotchas

  • Then "..." should( not)? be selected for "..."

Checks that a certain option is selected for a text field

  • Then nothing should be selected for "..."

  • Then "..." should( not)? be an option for "..."

Checks for the presence of an option in a select

  • Then the window should be titled "..."

  • When I reload the page

  • Then (the tag )?"..." should be visible

Checks that an element is actually present and visible, also considering styles. Within a selenium test, the browser is asked whether the element is really visible In a non-selenium test, we only check for .hidden, .invisible or style: display:none

More details here

  • Then (the tag )?"..." should be hidden

Checks that an element is actually present and hidden, also considering styles. Within a selenium test, the browser is asked whether the element is really hidden. In a non-selenium test, we only check for .hidden, .invisible or style: display:none

  • When I click on "..."

Click on some text that might not be a link.

Example:

  When I click on "Collapse"
  • When I click on the element "..."

Click on an element with the given selector.

Example:

  When I click on the element ".sidebar"
  • When I click on the element for ...

Click on the element with the given selector alias.

Example:

  When I click on the element for the sidebar
  • Then "..." should link to "..."

Use this step to check external links.

Example:

  Then "Sponsor" should link to "http://makandra.com/"

Don't forget the trailing slash. Otherwise you'll get the error expected: /http:\/\/makandra.com(\?[^\/]*)?$/ got: "http://makandra.com/" (using =~)

  • When I follow "..." inside any "..."

Click a link within an element matching the given selector. Will try to be clever and disregard elements that don't contain a matching link.

Example:

  When I follow "Read more" inside any ".text_snippet"

  • Then I should( not)? see "..." inside any "..."

  • When I fill in "..." with "..." inside any "..."

  • When I confirm the browser dialog

  • When I cancel the browser dialog

  • When I enter "..." into the browser dialog

  • Then the "..." (field|button|checkbox|radio button) should( not)? be disabled

Tests that an input, button, checkbox or radio button with the given label is disabled.

  • Then the "..." field should( not)? be visible

Tests that a field with the given label is visible.

  • When I perform basic authentication as ".../..." and go to ...

Performs HTTP basic authentication with the given credentials and visits the given path.

More details here.

  • When I go back

Goes to the previously viewed page.

  • Then the "..." select should( not)? be sorted

Tests whether a select field is sorted. Uses Array#natural_sort, if defined; Array#sort else.