If you want to automate anything under Windows, I would recommend you Automa - new lightweight GUI automation tool. It simply "finds" UI elements - all you need to do is to provide their names as human user would see them. And it's actually easier to use than Sikuli - you don't have take any screenshots. For instance, if you want to automatically sign in to Windows Live Messenger, you can run the following script:
start("messenger")
write("my.email@domain.com", into="Email address")
write("secret", into="Password")
click("Sign in")
or, similarly, in order to type in credentials and automatically log in to Facebook, all you need to do is execute the following commands:
start("firefox")
write("facebook.com", into="Location")
press(ENTER)
write("my.email@domain.com", into="Email or Phone")
write("secret", into="Password")
click("Log in")
The tool works on most Windows applications (including web browsers), no matter what technology they are written in. It's written in Python and can be used from within any Python application through importable api library.
About Automa vs Sikuli: Automa is as high-level as Sikuli, but does not require screenshots. This has several advantages:
- Automa scripts are easier to store in a version control system.
- Automa scripts being purely text based makes them easier to maintain: Imagine a label changing somewhere. This label might appear in several Sikuli screenshots, which you will then have to re-take. With Automa, all you have to do is a simple search-and-replace across files.
- It is more stable with respect to changes in the user interface - imagine a colour changing or a button moving by just a few pixels.
- Other nice features such as being able to write the scripts/tests before the application exists - you can always write a script saying click "here", type this, click "there", already from the spec of a feature/program. With screenshots that isn't possible.
Disclosure: I work on Automa.
