Reading a dynamic text relative to an image

see feature grow()
http://sikulix-2014.readthedocs.io/en/latest/region.html#extend-regions-and-create-new-regions-based-on-existing-regions

Thanks for your response. I tried grow(), and it is better, but for some reason, it is reading a strange additional character and removing a character. Basically, with the example below, I am opening up Wordpad and based on the image of "Home" menu on top, I am trying to read "View" menu into the value variable. For some reason, without the addition of grow(), it was reading "Vie" and now with the addition of grow, it is reading "Vie
E" instead of "View".

Not sure why "w" is not being read and not sure where "E" on the second line is coming from. Any thoughts?

[code]
import java.util.Iterator;

import org.sikuli.basics.Debug;
import org.sikuli.basics.Settings;
import org.sikuli.script.*;

public class TestSikuli {

    public static void main(String[] args) throws InterruptedException {
            Screen s = new Screen();
            Debug.setDebugLevel(3);

     Settings setting = new Settings();
   setting.OcrTextSearch = true;
   setting.OcrTextRead = true;

            try{
             App region = App.open("C:\\Program Files\\Windows NT\\Accessories\\wordpad.exe");
             Region appRegion = region.focusedWindow();
             appRegion.highlight(1);

             // get the project folder absolute path
             String workingDir = System.getProperty("user.dir");

             ImagePath.add(TestSikuli.class.getCanonicalName() + "/images");
             String imagesPath = workingDir + "\\images\\";

             System.out.println("images path is " + imagesPath);

             Screen objScreen = new Screen();

             objScreen.type("Hello, Sikuli!");

             Iterator<Match> matchedObjects = objScreen.findAllText("Sikuli");
             Match singleMatch = null;

             if(matchedObjects.hasNext()) {
              System.out.println("------------------- Inside if #1 -----------------------");
              singleMatch = matchedObjects.next();
              System.out.println("------------------- Found object is : " + singleMatch.text());
             }

             Iterator<Match> matchedObjects1 = objScreen.findAllText("Paint");
             Match singleMatch1 = null;

             if(matchedObjects1.hasNext()) {
              System.out.println("------------------- Inside if #2 -----------------------");
              singleMatch1 = matchedObjects1.next();
              System.out.println("------------------- Found object is : " + singleMatch1.text());
             }

             objScreen.wait(imagesPath + "home.PNG").offset(50, 0).grow(25,0).click();

             String value = objScreen.wait(imagesPath + "home.PNG").offset(50, 0).grow(25,0).text();

             System.out.println("------------------- Value is -----------------------" + value);

            }
            catch(FindFailed e){
                    e.printStackTrace();
            }

    }
}
[/code]

Might be due to the known weaknesses of the text feature.

To check wether the text you want to have can be read correctly, just run this oneliner in the SikuliX IDE:

uprint(selectRegion().text())

this will allow you to select a region containing text and watch the result.

take care, that at best only the area containing the text is selected.

You simply have to live with the results.
May be your experiences help you, to make your script better.