Hi, I have a webpage (using Ajax calls) in which the HTMLTD contains an HTMLEditBox. And the InnerText of HTMLTD changes dynamically with value entered in edit-box.
For e.g. if we enter ‘1’ in the edit box, the HTMLTD innertext displays as
06-Greenville”. You may refer image attached.
Here the last 3 lines are dynamically generated based on the value in Editbox (i.e. Innertext is dynamically changing). I use data driven automation & want TestPartner to click on one of the value (part of inner text) dynamically generated.
Presently TP records the value selection as a click -
HTMLEditBox("Name =TerritoryCode").Click 40, 53.
Is there any way I can tell TP to click on the particular part of InnerText (and not just click based on pixel position)?
I tried many combinations below, but none of them seems to be working.Some of them are:
A. When using wildcard in the Innertext, TestPartner is not able to recognize the changing values (Richland or Sumter etc). It only recognizes the part ‘TerritoryCode’.
B. TestPartner not able to identify changing territory values through BitMap or any other check.
I just have an opinion that it is really unmanageable to automate some functionalities of a webpage having Ajax Calls , because there are some instances where only 1 frame is refreshed/loaded and not the whole page. In this case the automation tool is not able to capture the changed property. Because the changed property/field was already hidden there somewhere in the frame. Any automation tool (whether TestPartner/QTP) recognize only ‘Page Load Changes’ and NOT the ‘Frame Load Changes’. (I have mentioned similar thing in my earlier post as well - http://www.sqaforums.com/showflat.ph...e=0#Post530381 )
However I would like to be proved wrong [img]/images/graemlins/smile.gif[/img]. Thanks in advance.
Re: Ajax based automation/challenge for TestingTools?
Interesting challenge unfortunately I don't have access to this type of environment [img]/images/graemlins/frown.gif[/img]. I think you could prove your theory by simply verifying if it works after the entire page is reloaded. If so, I would call tech support and ask if this is known behavior or a defect.
Now, I think you should be able to get around this problem by using the Microsoft HTML object library as a reference. You should be able to use the specific object type to fire the event that you are looking for thus avoiding having to rely on coordinate based clicks. Specifically look at the FireEvent method. Also look at the other methods and properties available in the library as it will probably give you a few more ideas on how to make your scripts more robust.