Roadmap

This document outlines longer-term ideas for multimark: features that would expand its capabilities beyond the current render-to-string API.

Planned — Near Term

Enhancements that build directly on existing infrastructure.

Type Stubs / py.typed Marker

Add a py.typed marker and inline type annotations for better IDE support and static analysis.

  • PEP 561 compliant py.typed marker file in the package
  • Full type annotations on all public functions and the Options enum
  • Improved autocomplete, hover docs, and type checking in editors

Planned — Medium Term

Features that expand the scope of what multimark can produce.

Streaming / Incremental API

Expose the cmark parser object for large documents, allowing feed-based parsing without buffering the entire input in memory.

  • Parser class wrapping cmark_parser_new / cmark_parser_feed / cmark_parser_finish
  • Extensions attachable at parser creation time
  • Render from the finished AST node with any renderer
  • Useful for processing very large Markdown files or streaming input

AST Access

Let users walk the parsed tree from Python (not just render to a string).

  • parse() function returning a Node wrapper around the cmark AST
  • Tree traversal: node.children, node.parent, node.next, node.prev
  • Node type introspection: node.type, node.literal, node.url, node.title
  • Mutation support: insert, remove, and replace nodes before rendering
  • Enables custom transformations, inspections, and selective rendering

Feedback & Contributions

Have ideas for features not listed here? Open an issue with the enhancement label. Contributions to any planned item are welcome so check existing issues first to avoid duplication.

This roadmap is a living document. It is updated as features ship and new priorities emerge.