enscribe-interpreter

The enscribe interpreter. Takes an mdast tree containing enscribeTag nodes (produced by remark-enscribe) and emits a self-contained Layer 1 HTML string.

enscribeInterpreter is a unified plugin. It registers a chain of mdast transforms — discovery, structural transformation, numbering, ref / cite resolution, note placement, bibliography — and a compiler that bridges mdast to hast (via mdast-util-to-hast, with a custom enscribeTag handler) and serializes to HTML. CSS, fonts, and conditional JavaScript for hover previews are inlined into the output so documents render correctly from file:// and offline.

Usage

import { unified } from 'unified';
import remarkParse from 'remark-parse';
import remarkEnscribe from '@enscribejs/remark';
import { enscribeInterpreter } from '@enscribejs/interpreter';

const result = await unified()
  .use(remarkParse)
  .use(remarkEnscribe)
  .use(enscribeInterpreter)
  .process(source);

console.log(String(result)); // HTML string

Options: katexCss, hoverPreviewMode, assetsDir. See notes/specs/interpreter.md §12 for details.

Documentation

• STATUS.md — current project state.
• notes/specs/interpreter.md — interpreter architecture: plugin chain, handler dispatch, schema dispatch, asset injection, error handling.
• notes/specs/pipeline.md — pipeline stages, ordering, dependencies, data-flow examples.
• BACKLOG-ROADMAP.md — open work.