v0.3.19

The Programmable Typesetter for the Web

Postext takes semantic markdown and applies centuries-old editorial layout rules — Knuth-Plass justification, orphan prevention, figure placement — producing publication-grade pages rendered to HTML, canvas, or print-ready PDF.

Get StartedView on GitHub

What Is Postext

Editorial Layout,
Computed for the Web

CSS gives you responsive layout — flexbox, grid, columns. But it cannot balance columns, prevent orphans, flow text around figures, or place footnotes at the bottom of a column. These are editorial decisions that have been refined over centuries of print typesetting.

Postext bridges that gap. Feed it semantic markdown and a configuration object. It measures text 300–600× faster than the DOM using @chenglou/pretext, then computes the complete page geometry — every line, float, and caption — and renders it through HTML, canvas, or print-ready PDF backends.

Input

# Chapter One

The quick brown fox jumps over
the lazy dog. A paragraph with
a {{figure:hero-img}} reference
and a footnote.[^1]

[^1]: Additional context placed
at column bottom automatically.

Output

Capabilities

What Print Typesetters Knew All Along

Knuth-Plass Justification

The same optimal line-breaking algorithm TeX uses, with TeX-pattern hyphenation and orphan control — evenly set text blocks with no rivers of whitespace, beyond anything CSS can express.

Resources & Cross-References

Figures, SVG diagrams, and tables float to the top or bottom of columns and pages — the way print floats work — with typed numbering and :ref cross-references that always stay in sync.

Table & Caption Styling

Tables typeset with independent body and header typography, and captions that follow your configuration — numbered, styled, and placed alongside the resource they describe.

Single-Ink Diagrams

Recolor SVG diagrams to tints of one ink by luminance, so every illustration matches your design — ready for spot-colour printing.

LaTeX Math

Inline and display equations written in LaTeX, measured and typeset by the engine so they sit on the baseline grid alongside the text.

Web & Print-Ready PDF

One layout engine, three renderers. The same computed pages render to HTML, canvas, or print-ready PDF — with bookmarks and CMYK colour.

API

One Function, Full Control

Call buildDocument with your content and a configuration object. Get back the complete page geometry, ready to render with renderToHtml — or the canvas and PDF backends.

import { buildDocument, renderToHtml } from "postext";

const doc = buildDocument(
  { markdown },
  {
    layout: { layoutType: "double" },
    bodyText: {
      textAlign: "justify",
      hyphenation: { enabled: true, locale: "en-us" },
    },
  }
);

const html = renderToHtml(doc);

Workflow

Three Steps to Publication-Grade Layout

  1. Write Semantic Markdown

    Author your content in enriched markdown with resource references and footnote markers. No layout concerns — just content.

  2. Configure Layout Rules

    Define columns, typography rules, resource placement strategies, and section overrides in a PostextConfig object.

  3. Render to HTML, Canvas, or PDF

    Call buildDocument and receive the complete layout geometry — then render it with the HTML, canvas, or print-ready PDF backend.

Installation

Start in Under a Minute

# Install with your package manager
pnpm add postext
// Quick start
import { buildDocument, renderToHtml } from "postext";

const doc = buildDocument({
  markdown: "# Hello World\n\nYour content here.",
});

const html = renderToHtml(doc);