v5 api: initial schemas, types & basic routes

[zoriya]

This is the first PR for the v5's api (see #597).

Checklist for this PR:

• Database schemas for:
  • shows (movies | series)
  • entries (episode | movie | special | extra)
  • videos
  • seasons
• Api types & examples for:
  • movies
  • series
  • episodes
  • specials
  • extras
  • movies contained in series (called movie-entry here)
  • seasons
• Api setup:
  • Authorization
  • Error-Problems
  • Clean validations errors (readable)
  • Data seeding for tests (use examples)
  • Clean swagger
• Routes to fetch
  • a movie via id/slug
    • languages
      • query user's requested language
      • fallback if we are missing a field not planned
      • allow a preferOriginal setting for images
    • with all translations (if requested)
    • with all videos
  • all movies
    • w/ arbitrary filters
    • keyset pagination
    • sorts
    • w/ video (at least if we have one or not, idk if we include all videos at all time)
  • get a random movie (or sort by random)

Future PRs will be:

• Search (with postgres as the backing db, not meilisearch)
• Images handling
  • Download images
  • Scale images (low/medium/high)
  • Calculate blurhash
  • Serve images via id
  • Serve image via /movie/id/poster & co (& support language header here)
• Other get routes (collections, seasons, series, episodes, extra...)
• Mixed routes (shows, news...)
• User specific information (watch-status, history)
• Other metadata (studios, staff...)
• Scanner's workflow (see Document scanner's v5 workflow #678 for the spec)

[zoriya]

It now has all the features i want for a first valid version, i'll add some tests & merge soon