Support client-rendered pages

[rosszurowski]

It'd be cool if new-work supported sites rendered on the client-side. I tried to add Dan Holdsworth's site to my list, but since it's rendered on the client with Angular, newwork can't detect any changes.

Some ideas on ways to handle this:

• Execute the client scripts before fetching the URL (I imagine this could get pretty messy)
• Allow defining a "source" URL that's different from the site URL. His site uses a JSON API to fetch the content (this won't work for every client-rendered site)

[s3ththompson]

Yep, this is definitely on the list. As you say, the easiest way to do this would be to allow an alternate source URL (overlap with #2 actually). Alternatively, I've been trying to wait for headless Chrome libraries to mature a bit so that we can scrape HTML after waiting for the page to render clientside (however, this will slow down scraping significantly).