February 22, 2010, 1:52 PM — Computer users with rudimentary skills will be able to program via screen shots rather than lines of code with a new graphical scripting language called Sikuli that was devised at the Massachusetts Institute of Technology.
With a basic understanding of Python, people can write programs that incorporate screen shots of graphical user interface (GUI) elements to automate computer work.
One example given by the authors of a paper about Sikuli is a script that notifies a person when his bus is rounding the corner so he can leave in time to catch it.
The script would pull together visual elements of a GPS-driven bus-tracking application that a transit authority might make available online. First, the user would box and capture a map image of the corner that the bus will turn to trigger the notification. That image is pasted into a line of code in Sikuli Script Editor that would look like this:1: street_corner=find( ).
The image of the street corner to be found would be pasted inside the set of parentheses.
Then the script would command looking for the pointer that indicates a bus's location. That line of code would look like this:2: while not street_corner.inside().find( ).similar(0.7):
A captured image of the pointer icon would be pasted in the second set of parentheses. The script seeks out the image of the bus pointer as pasted, and the pointer image that was boxed and cut necessarily includes some background. But as the pointer moves around the map, the background changes, so there will be no exact match for the image as pasted. To account for the differences the "similar.(0.7)" command indicates that the script should find images that are 70% similar to the icon pasted in the line of code.
The script tries to find the bus icon within the target area every 60 seconds, and that is written "sleep(60)". When the bus icon enters the target area, it triggers this response, scripted as: "popup("The bus is arriving!")".
Sikuli -- which means God's eye in the language of the Huichol Indians in Mexico -- also has a search function. Users paste in an icon from a program they are working with and the search engine will find sites that tell more about its function.
Specifying the visual search is actually faster than specifying a search based on keywords, say the researchers, Tom Yeh, Tsung-Hsiang Chang and Robert C. Miller.