enscribe-core

The inward-pointing shared foundation that enscribe's parser, interpreter, and future output generators depend on.

enscribe-core depends on nothing internal — it is the root of the package dependency graph. The parser (remark-enscribe), the interpreter (enscribe-interpreter), and any future output generator (enscribe-jats-export, etc.) all depend on it.

By design, enscribe-core is filesystem-free and browser-safe — its modules do not import fs, path, url, or any other Node-only built-in, and they do not carry heavy runtime dependencies. This is what lets the shared foundation be bundled into a future client-side build of enscribe.

The architectural decision is described in the architecture Phase 0 report that designed this package. The contents grow over several extraction slices; the first slice (this one) moves the pure-data, zero-logic-risk modules (dsl-registry, sigil-mapping) here from remark-enscribe. Later slices add tag-shape builders, error-node shapes, the centralized tree walkers, the numbering registry, the colon-id helper, and the file.data plugin-bus key constants.

What is not in enscribe-core:

• Build-time-only code (the Peggy grammar compiler; the filesystem vocabulary loader).
• Output-stage-specific code (the HTML attribute mapper; the future JATS attribute mapper).
• Heavy runtime dependencies (KaTeX, citation-js, tippy, etc.).