The underlying actions support in Chrome has been extended since chrome.automation was initially written.
This bug tracks the work to adopt the same support.
As a result, chrome.automation clients will have two ways to call actions:
1. via specific bindings such as
AutomationNode.prototype.focus()
AutomationNode.prototype.doDefault()
AutomationNode.prototype.setSelection(long start, long end)
etc
2. AutomationNode.prototype.performAction(string actionType)
This alternate version only takes an actionType. It won't allow for clients to call actions that require additional arguments.
Comment 1 by bugdroid1@chromium.org
, Apr 9 2018