"Extract URL" seems to work only if a valid URL is found in the link in the tag.
However, some links may be something else than http, for example itpc://...
(iTunes). In this case, Extract URL gives an error.

Could you add a module "Extract link" that also works with any link, not just
valid URLs?