Add new EventSchema and SchemaRegistry API

[Zsailer]

This proceeds the work in #2. The review in that PR became focused on "categories" and "redactionPolicies". I'm open this PR here to keep that discussion in tact while moving the review of the basic API design (minus the categories stuff) here.

Highlights

• Adds EventSchema API for validating schemas and events.
• Adds SchemaRegistry to cache validators for all registered schemas. This prevents a memory leak I have been seeing this library.
• Validates all schemas against a meta schema to enforce Jupyter Events schema structure.
• Removed "categories" handling (for now), since this is more of a telemetry use-case.
• Defaults to emitting all events (opposite of Jupyter Telemetry).

Public API changes

• "categories" is no longer used for redacting sensitive information. I
• "allowed_schemas", "allowed_categories", and "allowed_properties", have been removed from the EventLogger.
  • All registered event types will be emitted.
  • The expectation is that consumers will use filters in the logging handlers to listen for specific events.
• EventLogger.record_event has been renamed to EventLogger.emit.

[Zsailer]

This PR is ready for final review.

The only test that's failing is the "check-links" test in the documentation, which is failing because the new docs refer to links that will be created once this PR is merged.

[blink1073]

APIs look good to me, just one minor question

[Zsailer]

Thanks, @blink1073! I addressed various comments from @kevin-bates in #2 too. I'll likely merge this here work on getting the documentation rendering. Then, I'll shift to integrating with Jupyter Server.

[kevin-bates]

@Zsailer - this looks good. Sorry about focusing on #2 rather than this PR. Looks like the majority of comments are not applicable in this version (nice!). This is the only comment that's tbd, but that's something that we can look into later and/or doesn't affect the service overall.

[Zsailer]

@kevin-bates I think I addressed that comment in this commit: 8d43c5b

[kevin-bates]

Thanks @Zsailer. It looks like that warning will (also) be issued if there are no handlers, yet have registered (and matching) schema/version pairs - which, I suspect, will be common and probably not worthy of a warning. I was thinking only the missing schema/version pair scenario would be logged.

[Zsailer]

Ah, yes. Thanks, Kevin! Fixed in 4659ddf