Checkpoint: Batch-Converting a Fixtures Folder

The problem: prove the pipeline works on more than one hand-picked file

Lessons 1-5 each ran the pipeline on one fixture at a time, chosen for you. This checkpoint has no new Markdoc API, it's a small script that puts the pieces together on a whole folder: every top-level .md file in fixtures/, converted to HTML, written to disk.

The code, piece by piece

const mdFiles = fs
  .readdirSync(FIXTURES_DIR, { withFileTypes: true })
  .filter((entry) => entry.isFile() && entry.name.endsWith(".md"))
  .map((entry) => entry.name)
  .sort();

Only top-level .md files, not fixtures/partials/. Partial files (Lesson 11) aren't meant to be converted standalone, they're meant to be included inside other documents via config.partials.

function convertFile(filename) {
  const ast = Markdoc.parse(source);
  const frontmatter = ast.attributes.frontmatter ? yaml.load(...) : {};
  const config = { variables: { frontmatter, name: "reader", showBanner: true } };
  const renderTree = Markdoc.transform(ast, config);
  const html = Markdoc.renderers.html(renderTree);
  return { nodeCount: countNodes(ast), html };
}

The exact Lesson 1-5 pipeline, applied per file: hello.md has no frontmatter (ast.attributes.frontmatter is undefined, so frontmatter falls back to {}), article.md does, and its {% $frontmatter.title %} references resolve the same way Lesson 5 demonstrated.

function countNodes(node) {
  return 1 + node.children.reduce((sum, child) => sum + countNodes(child), 0);
}

A plain recursive count over the Ast (not the render tree), used only for this lesson's summary table, not something Markdoc itself exposes.