@curvenote/schema

Schema for interactive scientific writing, with translations to MyST flavoured markdown, LaTeX, and HTML.

Overview & Goals

• Provide a typed schema for writing reactive scientific documents using @curvenote/components
  • Uses Web Components in the rendered HTML output for non-standard components
  • Uses standard html for all other compnents, with no styling enforced
• Interoperability with CommonMark markdown and MyST
  • Through fromMarkdown and toMarkdown methods
• Provide components for WYSIWYG editing of reactive documents
  • See @curvenote/editor or curvenote.com for the editor!

Choices

• The internal representation for the library is a ProseMirror Document structure (that can be rendered as JSON)
• markdown-it is used parse and tokenize markdown content

Schema

The schema has Nodes and Marks where Nodes are basically a block of content (paragraph, code, etc.), and Marks are inline modifications to the content (bold, emphasis, links, etc.). See the ProseMirror docs for a visual explanation.

Overview of Nodes

• Basic Markdown
  • text
  • paragraph
  • heading
  • blockquote
  • code_block
  • image
  • horizontal_rule
  • hard_break
  • ordered_list
  • bullet_list
  • list_item
• Presentational Components
  • callout
  • aside
  • math
  • equation
• Reactive Components
  • variable
  • display
  • dynamic
  • range
  • switch

Overview of Marks

• link
• code
• em
• strong
• superscript
• subscript
• strikethrough
• underline
• abbr

Simple Example

This moves from markdown --> JSON --> HTML. The JSON is the intermediate representation for @curvenote/editor.

import { Schema, nodes, marks, fromMarkdown, toHTML } from '@curvenote/schema';
import { JSDOM } from 'jsdom';

const schema = new Schema({ nodes, marks });

const content = '# Hello `@curvenote/schema`!';
const doc = fromMarkdown(content, schema);

console.log(doc.toJSON());
>> {
    "type": "doc",
    "content": [
      {
        "type": "heading",
        "attrs": { "level": 1 },
        "content": [
          { "type": "text", "text": "Hello " },
          {
            "type": "text",
            "marks": [ { "type": "code" } ],
            "text": "@curvenote/schema"
          },
          { "type": "text", "text": "!" }
        ]
      }
    ]
  }

// Assuming we are in node, just use `document` if in a browser!
const { document } = new JSDOM('').window;

// Now move the document back out to html
const html = toHTML(doc, schema, document);

console.log(html);
>> "<h1>Hello <code>@curvenote/schema</code>!</h1>"

Roadmap

• Integrate other @curvenote/components as nodes
• Improve equation and start to go to/from various MyST syntax for this
• Add figure properties (name, width, caption etc.)
• Provide citations, probably bring in a bibtex parser
  • Introduce citation and reference component to curvenote/components or article
• Add overlaping roles/directives with MyST (e.g. see executablebooks/meta#70) for pointers
  • Add the necessary pieces to curvenote/components that are not basic html (MyST uses sphinx for the heavy lifting, cross-refs etc.)
• Provide other sereializers from the document strucutre (e.g. latex or simple html without curvenote/components, possibly idyll)

See also:

• Idyll Lang has a different markdown-like serialization with very similar base components to curvenote - see curvenote/article#8 for a comparison.