Steal feature tags from Noto

@twardoch had the idea of using Noto as a base for shaping "best practices" for other scripts. As a first approximation, we could scrape the Noto fonts for each script to determine which OpenType features they use, and use that to generate a list of expected features in incoming fonts.

No files, tests, or entry points are named in the issue. Start by locating how Shaperglot currently handles scripts and font features, then determine how Noto fonts could be scraped per script and how the resulting expected-feature lists would be represented and checked.

Written by the indexing model from the issue text.