Avoid doing two MBID lookups for each Tidal artist

[derat]

sources/online/tidal/tidal.go currently looks up both https://tidal.com/artist/<id> and https://listen.tidal.com/artist/<id> in MusicBrainz to try to find a given Tidal artist's MBID, since I've seen relationships using both forms.

I think that this code in the MB server tries to normalize Tidal URLs to just use tidal.com:

https://github.com/metabrainz/musicbrainz-server/blob/5856335ba8a224e7010f7b3a6f26aca686f1a06f/root/static/scripts/edit/URLCleanup.js#L4600-L4604

I'm guessing that the listen.tidal.com relationships could've been entered before the server started normalizing the URLs (which 37d4ea835e014649afc686ba7022f61971e328e3 implemented, I think). Some relevant links:

• https://tickets.metabrainz.org/browse/MBS-10621
• https://community.metabrainz.org/t/adding-tidal-to-streaming-page/499485

It'd be nice to only do a single lookup, but I don't know if there's any way to do so besides cleaning up all of the relationships in the DB.

[derat]

For better or worse, I'm embarking on the task of cleaning up the database:

• https://tickets.metabrainz.org/browse/MBBE-71
• https://community.metabrainz.org/t/registering-a-bot-to-canonicalize-old-tidal-urls/625741
• https://github.com/derat/mbbot

It looks like there are a bit over 12,000 URLs, so it'll take a while.

[derat]

I stopped looking up listen.tidal.com URLs in 5f3791fdde36b063e8039d25af75d15d65c5ba88.