Refrerring to https://en.wikipedia.org/wiki/Uniform_R ... IdentifierExtract URLs. This filter lists http://, https://, ftp:// and gopher:// URLs one per line (using a DOS line feed, CR/LF).
it's apparent that only four schemes are presently supported.
For example, Chrome has special URLs that load application-specific pages instead of websites or files on disk.
It would seem sensible therefore to permit the user to extend this filter to facilitate the extraction of URLs that make use of such further schemes.