Add locator strategy to ExtractData annotation

[tobiasstamann]

ExtractData might become flaky if there's a spinner or something similar present on the website or if data is loaded dynamically after page has loaded.

It might be good to have the opportunity to have the same locator strategy support like with actions

[tobiasstamann]

Currently, driver.findElements() is used to find all root elements related with data extraction.

This can cause problems if elements are loaded dynamically after page invocation. In this case it's likely that extrated data is represented by an empty list. Only way to prevent this is to use ecplicit pause times, which can be quite flaky.