So i work with automated UI testing using sikuli. as of right now I use screen grabs to dictate what the code should do (normal sikuli stuff) the problem is: whenever the Devs change the UI (e.i. change the font) for the application i'm testing, I have to go through and re-take all the pictures i have in my tests or else they fail. is there a way around this. One of my peers mentioned something about the "DOM" I'm not sure what that is, but he said I could use that instead of the bit map that sikuli uses. Thanks for any help in advanced
EDIT: for example I have code that says onAppear(picture of text, do this) so when the text of the app changes, the picture of the text no longer matches. adding regions wouldn't help. Although the text appears in the same place every time, sikuli still wouldn't recognize it.
At the moment I an trying to learn sikuli how to recognize certain words. I am doing okey with it. But some words he is bad a regonizing. (Dark background, white words, cetrain boxes, etc.) I know the sikuli team is working on impoving that in the next version.
At the moment I (mostly) also use screens, but the word recognizing is comming and will make it a lot more easy. Next version hopefully. :-)