markdown-it-gladest

A Markdown-it plugin to render LaTeX mathematical formulas using the Typst typesetting system. It leverages a high-performance native Rust addon to generate SVG or PNG images from your math blocks.

Features

• Renders inline ($ ... $) and block ($$ ... $$) LaTeX math.
• Powered by Typst for accurate and beautiful mathematical typesetting.
• Uses a native Rust addon for fast rendering.
• Outputs math as embedded SVG (default) or PNG images.
• Configurable PPI (Pixels Per Inch) for PNG output.
• Customizable fonts for body text and mathematical formulas.
• Easy integration with Markdown-it.
• Includes basic error handling display for rendering issues.

Installation

npm install markdown-it @fuuck/markdown-it-gladest
# or
yarn add markdown-it @fuuck/markdown-it-gladest
# or
bun add markdown-it @fuuck/markdown-it-gladest

Note: This package relies on a native Node.js addon built with Rust. Pre-compiled binaries for common platforms (Windows, macOS, Linux) are typically provided. If a pre-compiled binary is not available for your specific platform/architecture, you might need a Rust toolchain installed to build the addon during installation.

Usage

Basic Usage

import { writeFileSync } from "node:fs";
import MarkdownIt from "markdown-it";
import markdownItGladest from "@fuuck/markdown-it-gladest";

// Initialize Markdown-it
const md = new MarkdownIt().use(markdownItGladest, {
  // Plugin options (optional)
  format: "svg", // 'svg' (default) or 'png'
  ppi: 600, // Pixels Per Inch for PNG rendering (default: uses Typst default)
});

// Your Markdown content with LaTeX
const markdownContent = `
# Normal Distribution

The probability density function (PDF) is:

$$
f(x) = \\frac{1}{\\sigma\\sqrt{2\\pi}} e^{-\\frac{1}{2}\\left(\\frac{x-\\mu}{\\sigma}\\right)^2}
$$

Where $\\mu$ is the mean and $\\sigma$ is the standard deviation.
The standard normal distribution has $\\mu = 0$ and $\\sigma = 1$, denoted as $Z \\sim N(0,1)$.
An important transformation is standardization: $Z = \\frac{X-\\mu}{\\sigma}$.
`;

// Render the Markdown
const htmlResult = md.render(markdownContent);

// Save or use the result
writeFileSync("output.html", htmlResult);
console.log("Rendered HTML saved to output.html");

/*
Expected output in output.html (structure may vary slightly):

<h1>Normal Distribution</h1>
<p>The probability density function (PDF) is:</p>
<div class="gladst-block"><img src="data:image/svg+xml;base64,..." alt="f(x) = \\frac{1}{\\sigma\\sqrt{2\\pi}} e^{-\\frac{1}{2}\\left(\\frac{x-\\mu}{\\sigma}\\right)^2}" /></div>
<p>Where <span class="gladst-inline"><img src="data:image/svg+xml;base64,..." alt="\\mu" /></span> is the mean and <span class="gladst-inline"><img src="data:image/svg+xml;base64,..." alt="\\sigma" /></span> is the standard deviation.
The standard normal distribution has <span class="gladst-inline"><img src="data:image/svg+xml;base64,..." alt="\\mu = 0" /></span> and <span class="gladst-inline"><img src="data:image/svg+xml;base64,..." alt="\\sigma = 1" /></span>, denoted as <span class="gladst-inline"><img src="data:image/svg+xml;base64,..." alt="Z \\sim N(0,1)" /></span>.
An important transformation is standardization: <span class="gladst-inline"><img src="data:image/svg+xml;base64,..." alt="Z = \\frac{X-\\mu}{\\sigma}" /></span>.</p>
*/

Font Configuration

You can customize fonts used for rendering mathematical formulas by specifying system fonts or font files:

Using System Fonts

import MarkdownIt from "markdown-it";
import markdownItGladest from "@fuuck/markdown-it-gladest";

const md = new MarkdownIt().use(markdownItGladest, {
  format: "svg",
  fonts: {
    // Configure body font (used for text elements)
    bodyFont: {
      system: "Times New Roman",
    },
    // Configure math font (used for mathematical symbols)
    mathFont: {
      system: "Latin Modern Math",
    },
  },
});

Using Font Files

import MarkdownIt from "markdown-it";
import markdownItGladest from "@fuuck/markdown-it-gladest";

const md = new MarkdownIt().use(markdownItGladest, {
  format: "svg",
  fonts: {
    // Use custom font files
    bodyFont: {
      file: "/path/to/your/custom-font.ttf",
    },
    mathFont: {
      file: "/path/to/your/math-font.otf",
    },
  },
});

Mixed Font Configuration

import MarkdownIt from "markdown-it";
import markdownItGladest from "@fuuck/markdown-it-gladest";

const md = new MarkdownIt().use(markdownItGladest, {
  format: "png",
  ppi: 300,
  fonts: {
    // Use system font for body text
    bodyFont: {
      system: "Georgia",
    },
    // Use custom font file for math
    mathFont: {
      file: "/usr/share/fonts/opentype/lmodern/lmmath-regular.otf",
    },
  },
});

const content = `
Using custom fonts for better typography:

$$
\\sum_{i=1}^{n} \\frac{1}{i^2} = \\frac{\\pi^2}{6}
$$

The Euler's identity: $e^{i\\pi} + 1 = 0$.
`;

console.log(md.render(content));

Options

You can pass an options object when enabling the plugin with .use():

• format:

  • Type: 'svg' | 'png'
  • Default: 'svg'
  • Description: Specifies the output image format for the rendered math formulas. SVG is generally recommended for scalability and quality, while PNG might be needed for specific compatibility reasons.

• ppi:

  • Type: number | null
  • Default: null (uses the default PPI configured in the underlying Typst rendering engine)
  • Description: Sets the Pixels Per Inch for PNG rendering. Higher values result in larger, more detailed images. This option is ignored if format is 'svg'. Invalid values (e.g., non-positive numbers) will also cause it to fall back to the default.

• fonts:

  • Type: FontConfig | undefined
  • Default: undefined (uses default Typst fonts)
  • Description: Font configuration object that allows customizing the fonts used for rendering.

  FontConfig Properties:

  • bodyFont: Font configuration for body text elements
    • system: Use a system font by name (string)
    • file: Use a font file by path (string)
  • mathFont: Font configuration for mathematical symbols
    • system: Use a system font by name (string)
    • file: Use a font file by path (string)

  Important: You cannot specify both system and file for the same font type. Choose one approach per font.

Syntax

• Inline Math: Use single dollar signs ($). Example: This is inline math: $E = mc^2$.
• Block Math: Use double dollar signs ($$). These will be rendered as block elements (typically centered on their own line). Example:

  $$
  \int_a^b f(x) dx = F(b) - F(a)
  $$

The content between the delimiters is treated as LaTeX code and passed to Typst for rendering. Make sure to escape literal dollar signs in your Markdown text using a backslash: \$.

How It Works

1. Parsing: Markdown-it parses the input text. markdown-it-gladest registers inline and block rules to detect $ ... $ and $$ ... $$ sequences.
2. Tokenization: When math delimiters are found, custom tokens (gladst_inline_math, gladst_block_math) are generated, containing the LaTeX code within them.
3. Rendering: The plugin provides renderer functions for these custom tokens.
4. Native Call: The renderer function calls the renderLatex function exported by the native Rust addon, passing the LaTeX code, the delimiter type ($ or $$), and the configured options (format, ppi).
5. Typst Execution: The Rust addon invokes the Typst library to parse the LaTeX input and render it into the desired format (SVG or PNG).
6. HTML Generation: The Rust addon returns an HTML string, typically an <img> tag with the image data embedded as a Base64 data URI in the src attribute. If an error occurs during Typst rendering, an error message wrapped in a <span> or <div> is returned instead.
7. Output: The plugin wraps the returned HTML in a <span> (for inline) or <div> (for block) with appropriate CSS classes (gladst-inline or gladst-block) and inserts it into the final HTML output generated by Markdown-it.

Error Handling

If the Typst engine encounters an error while rendering a formula (e.g., invalid LaTeX syntax), the plugin will output an error message directly in the HTML instead of an image. This message will be wrapped in a <span class="gladst-error-inline"> or <div class="gladst-error-block"> and will often include the original formula and a summary of the error (potentially in the title attribute for hover details). Check your browser's developer console for more detailed error messages logged by the plugin.

License

This package MIT licensed.