dorfteich/apps/api/src/pages/yjs-content.ts

import {
  docToHtml,
  docToMarkdown,
  docToPlainText,
  editorSchema,
  extractOutline,
  OutlineEntry,
  extractWikilinkSlugs,
} from '@dorfteich/shared';
import { Node } from 'prosemirror-model';
import { prosemirrorJSONToYXmlFragment, yXmlFragmentToProseMirrorRootNode } from 'y-prosemirror';
import * as Y from 'yjs';

/**
 * The Yjs XmlFragment name the editor binds to (TipTap's collaboration
 * extension defaults to "default", #25) — api, web, and collab (#35) must
 * all agree on this or Yjs states become unreadable across them.
 */
const FRAGMENT_NAME = 'default';

/** Thrown for state bytes that are not a well-formed Yjs update for this schema. */
export class InvalidPageStateError extends Error {}

function docFromState(state: Uint8Array): Node {
  const ydoc = new Y.Doc();
  try {
    Y.applyUpdate(ydoc, state);
    return yXmlFragmentToProseMirrorRootNode(ydoc.getXmlFragment(FRAGMENT_NAME), editorSchema);
  } catch (error) {
    throw new InvalidPageStateError(error instanceof Error ? error.message : 'invalid Yjs state');
  } finally {
    ydoc.destroy();
  }
}

/**
 * Encode a ProseMirror document as the initial Yjs state a page is created
 * with. Used both for a fresh empty page and for importing a converted
 * document (#63) as a page's starting content — the editor binds to the same
 * {@link FRAGMENT_NAME}, so an opening client sees exactly this document.
 */
export function docToState(doc: Node): Uint8Array<ArrayBuffer> {
  const ydoc = new Y.Doc();
  try {
    const fragment = ydoc.getXmlFragment(FRAGMENT_NAME);
    prosemirrorJSONToYXmlFragment(editorSchema, doc.toJSON(), fragment);
    // Copy into a plain ArrayBuffer-backed view — yjs's own return type is
    // the wider `Uint8Array<ArrayBufferLike>`, which Prisma's Bytes input
    // (`Uint8Array<ArrayBuffer>`) does not accept directly.
    return new Uint8Array(Y.encodeStateAsUpdate(ydoc));
  } finally {
    ydoc.destroy();
  }
}

/** A fresh Yjs state containing a single empty paragraph. */
export function emptyPageState(): Uint8Array<ArrayBuffer> {
  return docToState(editorSchema.node('doc', null, [editorSchema.node('paragraph')]));
}

export interface DerivedPageContent {
  plainText: string;
  markdown: string;
  html: string;
  outline: OutlineEntry[];
  /** fileIds of every `image` node currently embedded in the document
   * (issue #31) — `PagesService.saveState` uses this to keep
   * `Attachment.pageId` pointed at whichever page's content actually
   * embeds the file, which is what the trash-purge job uses to find a
   * purged page's files. */
  imageFileIds: string[];
  /** Outgoing wikilink target slugs (issue #117): pages created through the
   * api (imports, phantom-create) seed their `page_links` rows from this —
   * collab, the content writer, rewrites them on every later save. */
  wikilinkSlugs: string[];
}

function imageFileIdsOf(doc: Node): string[] {
  const ids: string[] = [];
  doc.descendants((node) => {
    if (node.type.name === 'image' && typeof node.attrs.fileId === 'string') {
      ids.push(node.attrs.fileId);
    }
  });
  return ids;
}

/**
 * Decodes a page's Yjs state into the derived representations stored in
 * `page_content_cache` (issue #23). The collab server (#35) will decode
 * the same way and call the same shared derivation functions (#24).
 */
export function deriveContent(state: Uint8Array): DerivedPageContent {
  const doc = docFromState(state);
  return {
    plainText: docToPlainText(doc),
    markdown: docToMarkdown(doc),
    html: docToHtml(doc),
    outline: extractOutline(doc),
    imageFileIds: imageFileIdsOf(doc),
    wikilinkSlugs: extractWikilinkSlugs(doc),
  };
}