category_urls() gets ".com" and ".co.uk", as well as "http://" and "https://" URL variants

For some major UK news sources, which also have a ".com" variant, running category_urls() just produces the other variant (".co.uk"). This is also true of "http://" and "https://".

e.g. for "https://www.bbc.co.uk", the resulting categories are:

• https://www.bbc.co.uk
• https://www.bbc.co.uk (Not a typo, this URL is repeated)
• http://www.bbc.co.uk
• http://www.bbc.com
• https://www.bbc.com
• https://www.bbc.co.uk/accessibility

It also appears as if the original URL is also added into this list, as it's repeated.

Each one of these sites has its own categories (e.g. "https://www.bbc.co.uk/news" and "https://www.bbc.co.uk/sport"), but these are not found when using category_urls().