Pandoc Markdown

Pandoc Markdown is a CommonMark superset supported by the universal document converter Pandoc. It adds YAML metadata blocks, fenced divs, bracketed spans, superscript/subscript, strikethrough, footnotes, and heading IDs.

Package: relational-text/pandocNamespace: org.pandoc.facet

Functions

import { from, to } from 'relational-text/registry'

from('pandoc', input: string): Document

Parse a Pandoc Markdown string into a RelationalText Document.

const doc = from('pandoc', `---
title: My Document
author: Alice
---

## Introduction {#intro}

This has ^superscript^ and ~subscript~ and ~~strikethrough~~.

::: aside
A fenced div with class "aside".
:::

See the note.[^1]

[^1]: Footnote content here.
`)

Supports:

• YAML front matter (--- blocks) — each key stored as a metadata block with { key, value } attrs
• Headings with optional {#id} attribute suffix
• Fenced divs (::: class … :::) — stored as div blocks with { class } attr
• Footnote refs ([^key]) and footnote definitions ([^key]: text)
• Bracketed spans ([text]{.class #id})
• Standard CommonMark headings, code blocks, blockquotes, lists, HR, paragraphs
• Inline marks: **strong**, *italic*, _italic_, ~~strikethrough~~, ^superscript^, ~subscript~, `code`

to('pandoc', doc: Document): string

Render a RelationalText Document to Pandoc Markdown.

const md = to('pandoc', doc)

Metadata blocks are assembled into a YAML front-matter block at the top. Automatically applies lenses targeting org.pandoc.facet via lensGraph.autoTransform().

ensurePandocLexicon(): void

Explicitly register the Pandoc lexicon. Called automatically by from('pandoc', ...) / to('pandoc', ...) on first use.

Feature Mapping

Inline Marks

Pandoc syntax	Feature name	RT hub name
**text**	strong	bold
*text* / _text_	emphasis	italic
~~text~~	strikethrough	strikethrough
^text^	superscript	superscript
~text~	subscript	subscript
`text`	code-span	code
[text](url)	link	link
	image	image
[text]{.class}	span	—
[^key]	footnote-ref	—

Block Elements

Pandoc syntax	Feature name	Attrs
# … ###### (+ optional {#id})	heading	{ level: 1–6, id?: string }
```lang ```	code-block	{ language?: string, id?: string }
::: class … :::	div	{ class?: string }
[^key]: text	footnote-def	{ key }
> text	blockquote-marker	—
- text	bullet-list-marker / list-item-text	—
1. text	ordered-list-marker / list-item-text	—
---	horizontal-rule	—
key: value (front matter)	metadata	{ key, value }

Examples

Import

import { from } from 'relational-text/registry'

const doc = from('pandoc', `---
title: My Paper
---

## Results {#results}

The formula: x^2^ + y^2^ = z^2^.

- First item
- Second item with ~~deleted~~ text

See the note.[^note]

[^note]: Footnote text here.
`)

Export

import { from, to } from 'relational-text/registry'

const doc = from('markdown', '## Heading\n\n**bold** and _italic_\n\n1. one\n2. two')
const pd = to('pandoc', doc)
// "## Heading\n\n**bold** and *italic*\n\n1. one\n2. two"

Cross-format

import { from, to } from 'relational-text/registry'

const doc = from('pandoc', '## Title {#t}\n\nHello **world**.')
const html = to('html', doc)
// '<h2>Title</h2>\n<p>Hello <strong>world</strong>.</p>\n'

Notes

• YAML front matter is parsed as flat key: value pairs. Nested YAML structures are stored as literal strings.
• Fenced divs (::: class) produce div blocks; on export, div blocks are wrapped with :::.
• Footnote refs ([^key]) are stored as inline footnote-ref entities. Footnote defs ([^key]: text) are footnote-def blocks. Both round-trip faithfully; cross-format conversions that don't support footnotes will drop them.
• The pandoc-to-relationaltext lens maps the core CommonMark subset to RT hub names. Pandoc-specific features (fenced divs, spans, footnotes, metadata) are not mapped to RT hub equivalents.