We have a basicMarkupParser that runs first and includes the Nostr-y stuff, and that can run before Asciidoc _or_ advancedMarkupParser.

Always two steps. Second step ignores the parts that are already parsed.