Cucumber features as documentation

Cucumber allows for prose in features and scenarios. Example:

Feature: Cancel account

  There are several ways to cancel a user account. Admins need to 
  do it in complex cases, but normally, users can do it themselves.
  
  Scenario: User cancels his own account
    
    Users should be able to cancel an account themselves, so the 
    admins do not need to do it.
    
    Given a user account for "willy@astor.de"
    When I sign in as "willy@astor.de"
    And I follow "Cancel account"
    Then I should see "Account canceled"...

Firefox: Focus-sensitive Selenium tests do not work reliably in parallel test execution

This is a problem when using Selenium with Firefox. We recommend using ChromeDriver for your Selenium tests.


Firefox will not trigger focus/blur events when its window is not focused. While this makes sense in standard usage, it breaks in parallel test execution.

Please do not rely on focus events in your tests. The linked card has an example of how to build passing tests that deal with focus/blur events.

MetricsGraphics.js

MetricsGraphics.js is a library built on top of D3 that is optimized for visualizing and laying out time-series data. It provides a simple way to produce common types of graphics in a principled, consistent and responsive way. The library currently supports line charts, scatterplots and histograms as well as features like rug plots and basic linear regression.

New Firefox and gem versions for our Selenium testing environment (Ubuntu 14.04+)

Firefox 5.0.1, which we were using for most Rails 2.3 projects, does not run on Ubuntu 14.04 any more. Here is how to update affected projects.

  1. Update (or create) .firefox-version with the content: 24.0
    If you haven't installed Firefox 24 yet, the next time you run tests with Geordi, it will tell you how to install it.

  2. On a Rails 2 project:

Heads up: JavaScript does not like big numbers

In a JavaScript console, type this:

> 9112347935156469760
9112347935156470000

Ooops. And that's not a float!

This occurs because JavaScript uses double precision floats to store numbers.

So according to IEEE floating point definition only numbers between -(2^53 - 1) (-9007199254740991) and 2^53 - 1 (9007199254740991) can safely be represented in JavaScript.

Note that ECMAScript 6 will probably also offer [Number.MAX_SAFE_INTEGER](https://developer.mozilla.org/en-US/docs/W...

HTML5: Allow to choose multiple files with a vanilla file picker

Modern browsers natively suppport file pickers that allow the user to choose multiple files at once. To activate this feature, set the multiple attribute:

<input type="file" name="images[]" multiple />

Or in a Rails view:

<%= file_field_tag "images[]", multiple: true %>

This works in IE10+.

Make sure that the field name ends in [] so your server-side code will parse the incoming files into an array. Obviously this naming convention is not compatible with default Rails nested attribute setters, so you'll need to write a form ...

How to fix: Font too small when reading e-mails in Thunderbird

In Thunderbird, you can set custom font faces and sizes for reading plain-text e-mails. However, Thunderbird sometimes "randomly" does not respect your choices.

This is actually not a bug, but a rather weird feature: Fonts are defined per encoding of e-mails.

Some e-mails will be considered Unicode, some Western (ISO 8859-1), and some maybe yet another encoding.

The advanced font settings dialog by default just opens on "Western". Choose a different encoding from the "Fonts for" dropdown menu and you'll see that your custom font sett...

Capybara will not find links without an href attribute

Capybara and most assistive technology will fail to find <a> tags that are missing an href attribute. This will probably happen to you every now and then on JavaScript-heavy applications.

An example would be an AngularJS application where the following HTML actually works. [1]

<a ng-click="hello()">Hello</a>

Capybara will fail to find that link, even though looking it up via the DOM shows it:

>> find_link("Hello")
Capybara::...

The Curious Case of the Flip-Flop

The flip-flop operator is a hotly contested feature of Ruby. It's still struggling to find an idiomatic use case, except for a few very rarely needed things. It's not something you'll likely reach for on a daily, weekly or even monthly basis. The only thing you really need to know about it is what it does, and that's only in case you encounter it in someone else's code. Many even go as far to say not to use the flip-flop operator, that it only adds confusion.

My brain just melted.

Testing drag&drop with Selenium

When using jQueryUI's Sortable plugin (either directly or via Angular's ui.sortable), you might struggle testing your nice drag&drop GUI since Selenium webdriver does not support native dragging events.

But jQueryUI uses jquery.simulate for their testing, so why shouldn't you? There is even an extension to it that makes testing drag & drop quite easy.

Here is what you need:

  1. jquery.simulate.js
  2. [`jquery.simula...

Interacting with a Microsoft Exchange server from Ruby

Microsoft Exchange service administrators can enable Exchange Web Services (EWS) which is a rather accessible XML API for interacting with Exchange. This allows you to read and send e-mails, create appointments, invite meeting attendees, track responses, manage to-do tasks, check user availability and all other sorts of things that are usually only accessible from Outlook.

You can implement an EWS by hand-rolling your XML (the [docs](http://msdn.microsoft.com/en-us/...

Feedjira

Great gem to consume RSS feeds. I was missing some features on Ruby's RSS::Parser that I found in Feedjira:

  • Speed
  • Does not break on slightly malformed RSS feeds (like a missing length attribute on an <enclosure> tag on gizmodo.de's feed)
  • It automatically resolves Feedburner-mangled URLs (hooray!)

The GitHub project has only a minimalistic readme. You can find its documentation on their homepage.

Iterating over a Ruby Hash while tracking the loop index

You know each_with_index from arrays:

    ['hello', 'universe'].each_with_index do |value, index|
      puts "#{index}: #{value}"
    end
    # 0: hello
    # 1: universe

This also works on hashes. However, mind the required syntax:

    { hello: 'universe', foo: 'bar' }.each_with_index do |(key, value), index|
      puts "#{index}: #{key} => #{value}"
    end
    # 0: hello => universe
    # 1: foo => bar

The reason is that each_with_index yields 2 elements to the block, and you need to deconstruct the first el...

Cucumber: Skipping steps in a scenario outline, based on the current example

In Cucumber, scenario outlines help avoiding tests that are basically the same, except for a few variables (such as different inputs). So far, nothing new.

The problem

Now what if your test should (or should not) do something, like filling in a field only for some tests?

    Scenario Outline: ...
      When I open the form
        And I fill in "Name" with "<name>" # <= we want to do this only occasionally
      Then everybody should be happy
      
    Examples:
      | name  |
      | Alice |
      | Bob   |

You could o...

RSpec 3 no longer chooses a spec's type based on its directory

While RSpec 1 and 2 decided that specs inside spec/model are model specs, and those inside spec/features are feature specs (and so on), RSpec 3 will no longer do that by default.

This will result in errors such as missing routing helpers, etc.

There are 2 ways to fix this:

  • Explicitly set the type on each spec. For example:

    describe '...', type: 'feature' do
      # ...
    end
    
  • Add this to your spec_helper.rb (inside the RSpec.configure block) to restore the old behavior:

    ...

Test if all your favicons exist

When you don't only have a favicon.ico in your project but also PNGs of different sizes and backgrounds, you should test if all those files are actually reachable.

Here are a few selectors to get you started:

    'link[rel~="icon"]' # regular ones, matches "shortcut icon" and "icon"
    'link[rel="apple-touch-icon"]' # iOS
    'meta[content][name="msapplication-TileImage"]' # IE11
    'meta[content][name^="msapplication-square"]' # IE11

A s...

Use byebug on Ruby 2+

The debugger gem does not seem to be properly working on Ruby 2. Use byebug instead!

Byebug is a simple to use, feature rich debugger for Ruby 2. It uses the new TracePoint API for execution control and the new Debug Inspector API for call stack navigation, so it doesn't depend on internal core sources. It's developed as a C extension, so it's fast. And it has a full test suite so it's reliable. Note that byebug works only for ruby 2.0.0 or newer. For...

Device sizing and network throttling are coming to Chrome DevTools

See screenshot here.

This is great news because network throttling is very painful in Linux.

The features are already in Chrome Canary, so expect them to come to your Chrome sources soon.

How to test bundled applications using Aruba and Cucumber

Aruba is an extension to Cucumber that helps integration-testing command line tools.

When your tests involve a Rails test application, your tool's Bundler environment will shadow that of the test application. To fix this, just call unset_bundler_env_vars in a Cucumber Before block.

Previously suggested solution

Put the snippet below into your tool's features/support/env.rb -- now any command run through Aruba (e.g. via #run_simple) will have a clean Bundler envir...

Cucumber step to manually trigger javascript events in Selenium scenarios (using jQuery)

When you cannot make Selenium trigger events you rely on (e.g. a "change" event when filling in a form field), trigger it yourself using this step:

When /^I manually trigger a (".*?") event on (".*?")$/ do |event, selector|
  page.execute_script("jQuery(#{selector}).trigger(#{event})")
end

Note that this only triggers events that were registered through jQuery. Events registered through CSS or the native Javascript registry will not trigger.

Tearing Down Capybara Tests of AJAX Pages

An all-in-approach to fix the problem of pending AJAX requests dying in the browser when the server ends a test or switches scenarios.


We were able to work around this issue in most projects by doing this instead:

After '@javascript' do
  step 'I wait for the page to load'
end

Hover an element with Capybara < 2

You need this awkward command:

page.driver.browser.action.move_to(page.find(selector).native).perform

Note that there are better ways for newer Capybaras.

mattheworiordan/capybara-screenshot

Using this gem, whenever a Capybara test in Cucumber, Rspec or Minitest fails, the HTML for the failed page and a screenshot (when using capybara-webkit, Selenium or poltergeist) is saved into $APPLICATION_ROOT/tmp/capybara.

Link via Binärgewitter Podcast (German).

Atomic Grouping in regular expressions

A little-known feature of modern Regexp engines that help when optimizing a pattern that will be matched against long strings:

An atomic group is a group that, when the regex engine exits from it, automatically throws away all backtracking positions remembered by any tokens inside the group.