HTML & SSR

@plim/html renders a Plim document model to an HTML string with no DOM APIs - for server previews, SEO, transactional emails, and edge runtimes. It’s the read-only, server-side counterpart to @plim/editor. serializeToHTML accepts a DocumentNode, EditorState, Snapshot, or BlockNode[]; output is escaped by default (only the raw_html block is emitted verbatim - sanitize untrusted input).

See the @plim/html API reference for serializeToHTML, BlockRenderer, and the rendering context.

import { serializeToHTML, type BlockRenderer } from '@plim/html';

// Fragment by default; pass { document: true } for a full <!doctype html> page.
const fragment = serializeToHTML(snapshot);

// Override or add per-block / per-mark renderers - everything else falls back to the defaults.
const callout: BlockRenderer = (node, ctx) =>
  `<aside${ctx.attr('class', ctx.classFor('callout'))}>${ctx.renderInline(node.text)}</aside>`;

const page = serializeToHTML(doc, { document: true, classPrefix: 'plim-', blocks: { callout } });

Unknown blocks render as a neutral <div data-block-type="..."> (or your onUnknownBlock); unknown marks pass their inner HTML through unchanged, so the package stays decoupled from @plim/collaboration’s commentMark and any custom marks.