Get Text Block by Position

Applies to TestComplete 14.72, last modified on April 22, 2021

In your tests, you may need to get a text block residing in the desired area of your tested application to simulate user actions on it:

By using search preference

When obtaining the needed text block by using the OCR.Recognize.BlockByText method or the OCR Action operation, you can specify the area where the text block will be searched for (for example, at the bottom of the tested object). Usually, you use this approach to distinguish several text blocks that contain the same text.

In keyword tests

For an OCR Action operation, specify your search preference:

Select your search preference for an OCR action

Click the image to enlarge it.

In scripts

When calling the OCR.Recognize.BlockByText method, specify the SelectionPreference parameter. For example:

OCR.Recognize(testedObject).BlockByText("substring", spNearestToCenter)

By coordinates

Sometimes, specifying search preferences when getting a text block may not be enough. For example, there can be a situation where you cannot specify the preferred search area explicitly. In this case, you can specify the search area by its coordinates:

  1. Get an image of the needed search area by its coordinates. To do this, use the Picture method that all onscreen objects have.

  2. Use the OCR.Recognize method to recognize the text the obtained image contains.

  3. Get the needed block of the recognized text by using either the Block property or the BlockByText method.

  4. TestComplete cannot simulate user actions on text blocks obtained from images. It can simulate user actions only on text blocks it obtains from onscreen objects.

    As a workaround, you can do the following:

    1. Get the screen area to which the text block corresponds. To do this, you can use the TextBlock.Bounds property that contains information on the text block size and location.

    2. Simulate user actions on the screen area. You can simulate coordinate-based actions like clicks, double-clicks, touches (for mobile applications), and so on.

In scripts

The code below contains the GetTextBlockByCoordinates routine that obtains a tested onscreen object, the coordinates of the search area and the text to search for. The routine returns the Rect object that describes the area within the tested object where the text block resides.

The returned Rect object can be used to simulate user actions on the appropriate area of the tested onscreen object containing the needed text. The Main routine shows how to simulate a click in the center of the returned area.

JavaScript, JScript

function GetTextBlockByCoordinates(anObject, x, y, width, height, aText)
{
  var res = Utils.Rect;
  // Get an image of the specified area of an onscreen object
  var pic = anObject.Picture(x, y, width, height);
  if (pic != null)
  {
  // Recognize the text in the image
  var obj = OCR.Recognize(pic);
  if (obj.FullText != "")
  {
      if (aqString.Find(obj.FullText, aText, 0, false ) > -1)
      {
        // Use the BlockByText method to find the specified text
        // in the text recognized in the image

        var block = obj.BlockByText(aText);
        res.Left = block.Bounds.Left + x;
        res.Top = block.Bounds.Top + y;
        res.Bottom = res.Top + block.Bounds.Height;
        res.Right = res.Left + block.Bounds.Width;

        return res;

      }
    }
    return null;
  }
  return null;

}

function Main()
{
  var textToSearch = "substring";

  // Get an onscreen object
  var p = Sys.WaitProcess("MyApp").WaitWindow("Window", "*", -1, 1000);

  // Get info on the area that contains the needed substring
  // and resides at the specified coordinates
  var b = GetTextBlockByCoordinates(p, 50, 250, 300, 50, textToSearch);

  if (b != null)
  {
    // Simulate user actions
    …
    // Simulate a click in the center of the area
    p.Click(b.Left + b.Width / 2, b.Top + b.Height / 2);

    …

  }

}

Python

def GetTextBlockByCoordinates(anObject, x, y, width, height, aText):
  res = Utils.Rect;
  # Get an image of the specified area of an onscreen object
  pic = anObject.Picture(x, y, width, height);
  if pic != None:
    # Recognize the text in the image
    obj = OCR.Recognize(pic)
    if obj.FullText != "":
      if aqString.Find(obj.FullText, aText, 0, False ) > -1:
        # Use the BlockByText method to find the specified text
        # in the text recognized in the image
               
        block = obj.BlockByText(aText)
        res.Left = block.Bounds.Left + x
        res.Top = block.Bounds.Top + y
        res.Bottom = res.Top + block.Bounds.Height
        res.Right = res.Left + block.Bounds.Width
        
        return res
        
    return None
  return None

def Main():
  textToSearch = "substring"
  
  # Get an onscreen object
  p = Sys.WaitProcess("MyApp").WaitWindow("Window", "*", -1, 1000)
  
  # Get info on the area that contains the needed substring
  # and resides at the specified coordinates
  b = GetTextBlockByCoordinates(p, 50, 270, 320, 55, textToSearch);
  
  if b != None:
    # Simulate user actions

    # Simulate a click in the center of the area
    p.Click(b.Left + b.Width / 2, b.Top + b.Height / 2)

VBScript

Function GetTextBlockByCoordinates(anObject, x, y, width, height, aText)
  Set GetTextBlockByCoordinates = Nothing
  ' Get an image of the specified area of an onscreen object
  Set pic = anObject.Picture(x, y, width, height)
  If Not pic Is Nothing Then
    ' Recognize the text in the image
    Set obj = OCR.Recognize(pic)
    If obj.FullText <> "" Then
      If aqString.Find(obj.FullText, aText, 0, False ) > -1 Then
        ' Use the BlockByText method to find the specified text
        ' in the text recognized in the image

        Set res = Utils.Rect

        Set block = obj.BlockByText(aText)
        res.Left = block.Bounds.Left + x
        res.Top = block.Bounds.Top + y
        res.Bottom = res.Top + block.Bounds.Height
        res.Right = res.Left + block.Bounds.Width

        Set GetTextBlockByCoordinates = res

      End If
    End If
  End If
End Function

Sub Main
  textToSearch = "substring"

  ' Get an onscreen object
  Set p = Sys.WaitProcess("MuApp").WaitWindow("Window", "*", -1, 1000)

  ' Get info on the area that contains the needed substring
  ' and resides at the specified coordinates
  Set b = GetTextBlockByCoordinates(p, 50, 250, 300, 50, textToSearch)

  If Not b Is Nothing Then
    ' Simulate user actions
    …
    ' Simulate a click in the center of the area
    Call p.Click(b.Left + b.Width / 2, b.Top + b.Height / 2)
    …

  End If

End Sub

DelphiScript

function GetTextBlockByCoordinates(anObject, x, y, width, height, aText);
var res, pic, obj, block;
begin
  result := nil;
  // Get an image of the specified area of an onscreen object
  pic := anObject.Picture(x, y, width, height);
  if not (pic = nil) then
  begin
    // Recognize the text in the image
    obj := OCR.Recognize(pic);
    if obj.FullText <> '' then
    begin
      if aqString.Find(obj.FullText, aText, 0, false ) > -1 then
      begin
        // Use the BlockByText method to find the specified text
        // in the text recognized in the imag

        res := Utils.Rect;

        block := obj.BlockByText(aText);
        res.Left := block.Bounds.Left + x;
        res.Top := block.Bounds.Top + y;
        res.Bottom := res.Top + block.Bounds.Height;
        res.Right := res.Left + block.Bounds.Width;

        result : = res;
      end;
    end;

  end;
end;

procedure Main();
var textToSearch, p, b;
begin
  textToSearch := 'substring';

  // Get an onscreen object
  p := Sys.WaitProcess('MyApp').WaitWindow('Window', '*', -1, 1000);

  // Get info on the area that contains the needed substring
  // and resides at the specified coordinates
  b := GetTextBlockByCoordinates(p, 50, 250, 300, 50, textToSearch);

  if not (b = nil) then
  begin
    // Simulate user actions
    …
    // Simulate a click in the center of the area
    p.Click(b.Left + b.Width / 2, b.Top + b.Height / 2);
    …

  end;

end;

C++Script, C#Script

function GetTextBlockByCoordinates(anObject, x, y, width, height, aText)
{
  var res = Utils["Rect"];
  // Get an image of the specified area of an onscreen object
  var pic = anObject["Picture"](x, y, width, height);
  if (pic != null)
  {
  // Recognize the text in the image
  var obj = OCR["Recognize"](pic);
  if (obj["FullText"] != "")
  {
      if (aqString["Find"](obj["FullText"], aText, 0, false ) > -1)
      {
        // Use the BlockByText method to find the specified text
        // in the text recognized in the image

        var block = obj["BlockByText"](aText);
        res.Left = block["Bounds"]["Left"] + x;
        res.Top = block["Bounds"]["Top"] + y;
        res.Bottom = res["Top"] + block["Bounds"]["Height"];
        res.Right = res["Left"] + block["Bounds"]["Width"];

        return res;

      }
    }
    return null;
  }
  return null;

}

function Main()
{
  var textToSearch = "substring";

  // Get an onscreen object
  var p = Sys["WaitProcess"]("MyApp")["WaitWindow"]("Window", "*", -1, 1000);

  // Get info on the area that contains the needed substring
  // and resides at the specified coordinates
  var b = GetTextBlockByCoordinates(p, 50, 250, 300, 50, textToSearch);

  if (b != null)
  {
    // Simulate user actions
    …
    // Simulate a click in the center of the area
    p["Click"](b["Left"] + b["Width"] / 2, b["Top"] + b["Height"] / 2);

    …

  }

}

In keyword tests

  1. Copy the GetTextBlockByCoordinates function code from the example above to a script unit in your project in TestComplete.

  2. Call the function from your keyword test by using the Run Script Routine operation.

  3. Save the area the function returns to a variable.

    To learn how to get the results of a routine called from a keyword test, see Checking Operation Result. To learn how to set a variable value in a keyword test, see Keyword Test Variables.

  4. Simulate user actions on the area, for example, by using the Run Code Snippet operation.

Get a screen area that contains the needed text

As an alternative, copy the Main routine from the example above to a script unit and modify it to work with your tested objects and coordinates. After that, you can call that routine from your keyword test.

Important

TestComplete can also recognize text in images loaded from a file. However, it cannot simulate user actions on recognized text blocks. To simulate user actions on screen areas specified by images you load from files, we recommend that you perform Image-Based testing. See Image-Based Testing.

See Also

About Optical Character Recognition
Picture Object
Image-Based Testing
Get Controls With No Text Contents

Highlight search results