Skip to content
Ruby-GNOME2 integration testing library
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
lib
.document
.gitignore
LICENSE
README.md
Rakefile
VERSION
gutkumber.gemspec

README.md

Gutkumber

by Daniel Lucraft (dan@fluentradical.com)

Description

A set Ruby-GNOME2 automation helpers. Comes with Cucumber formatters and steps for writing features.

Features

  • Interact with dialogs using Cucumber steps:

     When I click on "OK" in the "Save changes?" dialog
    
  • Make keypresses:

     When I press "Shift+K"
    
  • and mouse clicks:

     When I right click on ....
    
  • Type into TextEntrys or TextViews

     When I type "hello"
    

Install & Requirements

Clone from git://github.com/danlucraft/gutkumber.git Requires Cucumber >= 0.3

Usage

Gutkumber requires that the Gtk main loop NOT be started, in order that it can control the flow of events between steps. Therefore the code that starts the main loop should be modified to look like this:

unless $gutkumber
  Gtk.main_with_queue(100)
end

Other than that, you can load your application normally, from your Cucumber env.rb file.

When you run your Cucumber tests, you must specify one of the Gtk formatters, like this:

--format Cucumber::Formatter::GtkFormatter
--format Cucumber::Formatter::GtkProgressFormatter

Writing a testable Ruby-GNOME2 application

In normal Gtk, Dialog#run blocks. This is no good for testing. To see why, imagine these (imaginary) steps:

When I open the "Save" dialog
And I fill in a filename of "Foo.rb"
And I click on "Save"

Since the Dialog#run method blocks (makes your application do nothing until it receives some input into the dialog), the second two steps will never be reached.

Therefore Gutkumber overrides Dialog#run to just show the Dialog, and not block.

So if you use Dialog#run in your application, then the implication is that you MUST make sure that ALL effects from the Dialog are handled in response blocks. The following will not work anymore:

response = dialog.run
# ... do stuff based on response

Previously the other stuff would only be run after the dialog closed, but in the testing environment, dialog.run will just show the dialog and return nil immediately. Change this code to:

dialog.run do |response|
  # ... do stuff based on response.
end
# absolutely nothing can go here.

This is not ideal. But I haven't been able to come up with any other way to make testing dialogs possible.

License

Gutkumber is copyright 2009 Daniel B. Lucraft and contributors. It is licensed under the GPL2.

You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.