Method to extract URLs from text without normalizing

[jage]

Currently we have Twingly::URL::Normalizer.normalize() which can extract URLs, but it will normalize the URLs at the same time. We need a method that can just extract URLs from text with no normalization.

[dentarg]

Me and @jage talked some about this issue, from another issue:

Uhm, to just add it, not hard. To not create a mess, pretty hard. We already have an extract_url method in the normalizer. So this should be in the url class, which now returns an object with a .url accessor ... it should probably just delegate everything except .domain etc ...

@jage suggested we could try some mob programming refactoring twingly-url, WDYT @walro @roback?

[roback]

👍 That would be a good idea.

[walro]

Sure