fix(ai): collapse blank-line runs in extracted page text

apps/server/src/collaboration/collaboration.util.ts

@@ -57,6 +57,7 @@ import {
   JSONContent,
 } from '@tiptap/core';
 import { generateHTML, generateJSON } from '../common/helpers/prosemirror/html';
+import { collapseBlankLines } from '../common/helpers';
 // @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
@@ -146,7 +147,7 @@ export function htmlToJson(html: string) {
 }
 
 export function jsonToText(tiptapJson: JSONContent) {
-  return generateText(tiptapJson, tiptapExtensions);
+  return collapseBlankLines(generateText(tiptapJson, tiptapExtensions));
 }
 
 export function jsonToNode(tiptapJson: JSONContent) {

apps/server/src/common/helpers/index.ts

@@ -1,4 +1,5 @@
 export * from './utils';
+export * from './text.utils';
 export * from './nanoid.utils';
 export * from './file.helper';
 export * from './constants';

apps/server/src/common/helpers/text.utils.spec.ts

@@ -0,0 +1,14 @@
+import { collapseBlankLines } from './text.utils';
+
+describe('collapseBlankLines', () => {
+  it.each([
+    ['a\n\n\n\nb', 'a\n\nb'],
+    ['a\n\nb', 'a\n\nb'],
+    ['a\nb', 'a\nb'],
+    ['\n\n\n\na\n\n\n', '\n\na\n\n'],
+    ['no newlines', 'no newlines'],
+    ['', ''],
+  ])('collapses %j to %j', (input, expected) => {
+    expect(collapseBlankLines(input)).toBe(expected);
+  });
+});

apps/server/src/common/helpers/text.utils.ts

@@ -0,0 +1,3 @@
+export function collapseBlankLines(text: string): string {
+  return text.replace(/\n{2,}/g, '\n\n');
+}