Allow more checksum computations in Content model

This does not change the default behavior.

• Content: Allow to get more hash than the default ones Only if it's supported by the model.

• hashutil: Open more checksum algorithms. Those are the ones supported by the integrity field [1]. This will be needed for the future new content/directory loaders.

• [1] https://w3c.github.io/webappsec-subresource-integrity/#grammardef-hash-algo

Related to T3781

Test Plan

I'll check the coverage and maybe add some missing tests in due time

---

Migrated from D8582 (view on Phabricator)