Text Search and OCR / Sikuli's Text Recognition

Hi,

You can use text recognition like that:

Settings.OcrTextRead = true;
Region sc = new Screen(0);
sc.findText("text").highlight(1);

Text recognition doesn't always work as expected. In many cases the text will not be recognized.
To see what is exactly recognized on your current screen, you can try to read the whole screen and check whether the text you are trying to find was detected:

Region sc = new Screen(0);
System.out.println(sc.text());

It is possible to fine tune the Tesseract to better recognize your specific font. You can try and Google it.

Since she talks about Jython:

Settings.OcrTextRead = True

When running scripts in the IDE only, one can also set in the preferences -> more options the respective options.

Thank You Eugene and RaiMan

I am running my scripts on IDE only. I have this line of code which works:

findText("text name")

However, how do I click the text on my screen without using a screenshot?

In your opinion what are the advantages and disadvantages of sikuli's text recognition

click(findText("text name"))

since findText("text name") returns the match region where the text was found.

-- advantage
internally uses Tesseract ;-)

-- disadvantage
still a very poor and restricted implementation.

-- conclusion
- fine if it works with your fonts and language
- bad if it does not work (but some chances to make it work - see Eugene's tip)

Thanks a lot Raimund :)