Chrome Automation of existing instance

[plastix]

Hi all

Can IUIAutomation automate existing instances of Chrome to find objects (by ID, for example) within iframes and set text of inputs within them ?  I need to basically poll existing Chrome instances regularly to see if a login form is loaded (which will itself be loaded in an iframe potentially within any number of tabs/instances). Chrome will have accessibility 'on', but there are restrictions that won't permit the installation of any non-mainstream extensions etc (othewise would have looked at the _Chrome_UDF). I started dabbling with Selenium (via Selenium Wrapper ) and it worked well spawning an instance of Chrome browser etc - but I later realised that it can't attach to an existing instance of Chrome... I've looked at Junkew's IUIAutomation framework and could see might be feasible to manipulate controls by click etc - but not necessarily at a DOM / DOM tree level...

Anyone got any experience / success with this specific issue ? TIA 

 

[computergroove]

If you can get the control of the chrome instance I don't see why you couldn't do this. You could also make the shortcut for chrome start a script that starts chrome.

[junkew]

No problem you just can attach to running instance with iuiautomation. With tab accessibility of chrome you can see the default exposed tree. By entering javascript lines thru the addressbar the sky will be the limit on what you can see in chrome but also other browsers.

Edited May 5, 2015 by junkew