fix editor converter (docmost#1647)

Philipinho · web-flow · commit 313503037601 · 2025-09-30T16:07:19.000+01:00
diff --git a/apps/server/package.json b/apps/server/package.json
@@ -62,7 +62,7 @@
     "class-validator": "^0.14.1",
     "cookie": "^1.0.2",
     "fs-extra": "^11.3.0",
-    "happy-dom": "^15.11.6",
+    "happy-dom": "^18.0.1",
     "jsonwebtoken": "^9.0.2",
     "kysely": "^0.28.2",
     "kysely-migration-cli": "^0.4.2",
diff --git a/apps/server/src/collaboration/collaboration.util.ts b/apps/server/src/collaboration/collaboration.util.ts
@@ -35,11 +35,10 @@ import {
   Subpages,
 } from '@docmost/editor-ext';
 import { generateText, getSchema, JSONContent } from '@tiptap/core';
-import { generateHTML } from '../common/helpers/prosemirror/html';
+import { generateHTML, generateJSON } from '../common/helpers/prosemirror/html';
 // @tiptap/html library works best for generating prosemirror json state but not HTML
 // see: https://github.com/ueberdosis/tiptap/issues/5352
 // see:https://github.com/ueberdosis/tiptap/issues/4089
-import { generateJSON } from '@tiptap/html';
 import { Node } from '@tiptap/pm/model';
 
 export const tiptapExtensions = [
diff --git a/apps/server/src/common/helpers/prosemirror/html/generateHTML.ts b/apps/server/src/common/helpers/prosemirror/html/generateHTML.ts
@@ -1,21 +1,29 @@
-import { Extensions, getSchema, JSONContent } from '@tiptap/core';
-import { DOMSerializer, Node } from '@tiptap/pm/model';
-import { Window } from 'happy-dom';
+import { type Extensions, type JSONContent, getSchema } from '@tiptap/core';
+import { Node } from '@tiptap/pm/model';
+import { getHTMLFromFragment } from './getHTMLFromFragment';
 
+/**
+ * This function generates HTML from a ProseMirror JSON content object.
+ *
+ * @remarks **Important**: This function requires `happy-dom` to be installed in your project.
+ * @param doc - The ProseMirror JSON content object.
+ * @param extensions - The Tiptap extensions used to build the schema.
+ * @returns The generated HTML string.
+ * @example
+ * ```js
+ * const html = generateHTML(doc, extensions)
+ * console.log(html)
+ * ```
+ */
 export function generateHTML(doc: JSONContent, extensions: Extensions): string {
+  if (typeof window !== 'undefined') {
+    throw new Error(
+      'generateHTML can only be used in a Node environment\nIf you want to use this in a browser environment, use the `@tiptap/html` import instead.',
+    );
+  }
+
   const schema = getSchema(extensions);
   const contentNode = Node.fromJSON(schema, doc);
 
-  const window = new Window();
-
-  const fragment = DOMSerializer.fromSchema(schema).serializeFragment(
-    contentNode.content,
-    {
-      document: window.document as unknown as Document,
-    },
-  );
-
-  const serializer = new window.XMLSerializer();
-  // @ts-ignore
-  return serializer.serializeToString(fragment as unknown as Node);
+  return getHTMLFromFragment(contentNode, schema);
 }
diff --git a/apps/server/src/common/helpers/prosemirror/html/generateJSON.ts b/apps/server/src/common/helpers/prosemirror/html/generateJSON.ts
@@ -1,21 +1,55 @@
-import { Extensions, getSchema } from '@tiptap/core';
-import { DOMParser, ParseOptions } from '@tiptap/pm/model';
+import type { Extensions } from '@tiptap/core';
+import { getSchema } from '@tiptap/core';
+import { type ParseOptions, DOMParser as PMDOMParser } from '@tiptap/pm/model';
 import { Window } from 'happy-dom';
 
-// this function does not work as intended
-// it has issues with closing tags
+/**
+ * Generates a JSON object from the given HTML string and converts it into a Prosemirror node with content.
+ * @remarks **Important**: This function requires `happy-dom` to be installed in your project.
+ * @param {string} html - The HTML string to be converted into a Prosemirror node.
+ * @param {Extensions} extensions - The extensions to be used for generating the schema.
+ * @param {ParseOptions} options - The options to be supplied to the parser.
+ * @returns {Promise<Record<string, any>>} - A promise with the generated JSON object.
+ * @example
+ * const html = '<p>Hello, world!</p>'
+ * const extensions = [...]
+ * const json = generateJSON(html, extensions)
+ * console.log(json) // { type: 'doc', content: [{ type: 'paragraph', content: [{ type: 'text', text: 'Hello, world!' }] }] }
+ */
 export function generateJSON(
   html: string,
   extensions: Extensions,
   options?: ParseOptions,
 ): Record<string, any> {
-  const schema = getSchema(extensions);
+  if (typeof window !== 'undefined') {
+    throw new Error(
+      'generateJSON can only be used in a Node environment\nIf you want to use this in a browser environment, use the `@tiptap/html` import instead.',
+    );
+  }
 
-  const window = new Window();
-  const document = window.document;
-  document.body.innerHTML = html;
+  const localWindow = new Window();
+  const localDOMParser = new localWindow.DOMParser();
+  let result: Record<string, any>;
 
-  return DOMParser.fromSchema(schema)
-    .parse(document as never, options)
-    .toJSON();
+  try {
+    const schema = getSchema(extensions);
+    let doc: ReturnType<typeof localDOMParser.parseFromString> | null = null;
+
+    const htmlString = `<!DOCTYPE html><html><body>${html}</body></html>`;
+    doc = localDOMParser.parseFromString(htmlString, 'text/html');
+
+    if (!doc) {
+      throw new Error('Failed to parse HTML string');
+    }
+
+    result = PMDOMParser.fromSchema(schema)
+      .parse(doc.body as unknown as Node, options)
+      .toJSON();
+  } finally {
+    // clean up happy-dom to avoid memory leaks
+    localWindow.happyDOM.abort();
+    localWindow.happyDOM.close();
+  }
+
+  return result;
 }
diff --git a/apps/server/src/common/helpers/prosemirror/html/getHTMLFromFragment.ts b/apps/server/src/common/helpers/prosemirror/html/getHTMLFromFragment.ts
@@ -0,0 +1,54 @@
+import type { Node, Schema } from '@tiptap/pm/model';
+import { DOMSerializer } from '@tiptap/pm/model';
+import { Window } from 'happy-dom';
+
+/**
+ * Returns the HTML string representation of a given document node.
+ *
+ * @remarks **Important**: This function requires `happy-dom` to be installed in your project.
+ * @param doc - The document node to serialize.
+ * @param schema - The Prosemirror schema to use for serialization.
+ * @returns A promise containing the HTML string representation of the document fragment.
+ *
+ * @example
+ * ```typescript
+ * const html = getHTMLFromFragment(doc, schema)
+ * ```
+ */
+export function getHTMLFromFragment(
+  doc: Node,
+  schema: Schema,
+  options?: { document?: Document },
+): string {
+  if (options?.document) {
+    const wrap = options.document.createElement('div');
+
+    DOMSerializer.fromSchema(schema).serializeFragment(
+      doc.content,
+      { document: options.document },
+      wrap,
+    );
+    return wrap.innerHTML;
+  }
+
+  const localWindow = new Window();
+  let result: string;
+
+  try {
+    const fragment = DOMSerializer.fromSchema(schema).serializeFragment(
+      doc.content,
+      {
+        document: localWindow.document as unknown as Document,
+      },
+    );
+
+    const serializer = new localWindow.XMLSerializer();
+    result = serializer.serializeToString(fragment as any);
+  } finally {
+    // clean up happy-dom to avoid memory leaks
+    localWindow.happyDOM.abort();
+    localWindow.happyDOM.close();
+  }
+
+  return result;
+}
