Page 1 of 1

Object repository Vs. Data driven technique

Posted: Fri Aug 07, 2009 2:19 pm
by cancerion
Hi,
Can you please tell me the benefit of using object repository over Data driven technique?

Thanks,
Sachin

Re: Object repository Vs. Data driven technique

Posted: Fri Aug 07, 2009 5:40 pm
by Ciege
Sorry for not answering your question directly, but...

Personally I do not use the repository (yet). Although I've had a good long talk about it with the great guys at Ranorex. I think it's because of my automation roots. I was "brought up" using automation tools that did not have repositories and I think it just stuck with me.

I find it more elegant to write methods and functions to search for/find objects at runtime. In this way, I create an automation framework that I can use with all of my different automation tests and suites. I never need to build or rebuild a repo, my framework (should) always find the objects whenever I need it to.

In the end my framework is quite large and very complex and my actual test code for any given test is quite small. It is also very easy to teach others to use the framework API to do their bidding for them and not need them to learn (until they are ready) how to do the heavy lifting.

I separate my framework code into different applicable DLLs.
One for "normal" desktop apps (includes methods like ClickPushbutton, WaitForWindow, CheckBoxClick, etc...)
Another for IE apps (includes methods like DOMckPushbutton, WaitForDOMObjectReady, DOMCheckBoxClick, etc...)
And other for Excel integration (includes methods like ReadCell, WriteCell, ReadDataRange, etc...)

All in all it works for me... But you may be wired differently than I am and may want to explore what works best for you. The nice thing with Ranorex is you have that choice to do what you want to do...

Re: Object repository Vs. Data driven technique

Posted: Tue Aug 11, 2009 9:20 am
by Support Team
With the repository you're able to prepare logical named objects which are accessible directly in code. That way you'll achieve more readable test automation code automatically. The big advantage instead of doing attribute based search all the time in code is to maintain all these attributes in an externally defined repository file. The choice really depends on the type of use case. But especially when working with recordings in combination with user code actions it's really easy to extend or to change existing recordings by modifying the repository.

kind regards,

Christoph,
Ranorex Support Team