Currently, we check for domains by doing a simple string match against e.g. news.com. However, there is the issue that if a citation is referencing bestnews.com, it would still match with news.com despite being another domain.

We should improve our domain matching by ensuring we are checking against the full URL (this likely means appending to the beginning of and doing two checks for . and // [as in http(s)://] on each domain). This will likely also require separating out the string matching for non-domains (e.g. /opinions/).