Same, here. Currently you would be required to scrape the website and extract the DOM elements from the HTML. And then parse those. I am very curious if we can get an answer from somewhere. I've found nothing in meta.