feat: allow direct pdf archives

[fmartingr]

This pull request aims to provide PDF archives for bookmarks by directly downloading the file.

Notable changes

• Logic for archivers now uses the ArchiverDomain
  • ArchiverDomain initializes a list of archivers that try to match the content type, if it does, that archiver is used. If an error is raised, the next archiver is used and so on.
• Added PDFArchiver: Just downloads the file.
• Added WARCArchiver as a "catch all" (current default)
• Added database columns bookmark.archiver (warc/pdf) and archive_path.
  • Bookmark has an archive if archive_path != ""

Pending

• [ ] Migrate the archive_path for existing bookmarks: iterate over the bookmarks, check if the archive exists and set the path and the archiver to warc.
• [ ] Refactor EPub logic separate from the archiver
• [x] Allow different archivers in archive page

Closes #929