Consider creating a more robust comparison system for tests

[ssnepenthe]

The ultimate goal would be that tests can still pass even with minor/irrelevant changes to the page.

This is mostly looking at images - I have observed at least the following:

• Cache buster as part of the path (scripss networks) or query string
• Switching from http to https or from either to protocol-relative URLs (epicurious)
• Host changes completely from sharding across multiple domains (general mills)

[ssnepenthe]

Quick checklist of the minimum I would like to handle:

• Similar string comparison
  • "2 sprigs of fresh sage leaves" should match "2 sprigs of sage leaves"
  • Needs to work for arrays of strings as well
• Similar image comparison
  • Should allow for scheme changes (http => https)
  • Should allow for domain changes (cdn1.example.com => cdn2.example.com)
  • Should handle cache busters in query string (example.com/image.jpg?ac5fd3e)
  • Probably a little more difficult, but would be nice to be able to handle cache buster in the filename (example.com/image.ac5fd3e.jpg)