[lexical-playground] Bug Fix: Normalize collapsible content children by rohan-patnaik · Pull Request #8539 · facebook/lexical

rohan-patnaik · 2026-05-22T07:11:34Z

Description

This PR fixes the collapsible playground crash when invalid inline or empty content is placed directly under CollapsibleContentNode.

CollapsibleContentNode is a shadow root, so its children need to be block elements. The content transform now wraps direct inline children in paragraphs, leaves existing block children unchanged, and normalizes empty content with a paragraph so the repro state stays editable.

Closes #8530

Test plan

Before

The issue repro could load a collapsible-content node with no block child. After dirtying that content node, later editor operations could throw because the content node did not have a valid block structure.

After

corepack pnpm exec vitest --project unit --run packages/lexical-playground/__tests__/unit/CollapsibleContainerNode.test.ts
corepack pnpm exec eslint packages/lexical-playground/src/plugins/CollapsibleExtension/index.ts packages/lexical-playground/__tests__/unit/CollapsibleContainerNode.test.ts
corepack pnpm exec prettier --check packages/lexical-playground/src/plugins/CollapsibleExtension/index.ts packages/lexical-playground/__tests__/unit/CollapsibleContainerNode.test.ts

meta-cla · 2026-05-22T07:11:40Z

Hi @rohan-patnaik!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at [email protected]. Thanks!

vercel · 2026-05-22T07:11:40Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
lexical	Ready	Preview, Comment	May 25, 2026 4:44am
lexical-playground	Ready	Preview, Comment	May 25, 2026 4:44am

mayrang · 2026-05-22T08:36:57Z

The issue's repro link starts with a collapsible-content that has no children at all (children: []) inside a normal container. Neither path in this PR's transform touches an empty content, so it stays empty after normalization — and Shift+Enter still throws Expected node CollapsibleContentNode of type collapsible-content to have a block ElementNode ancestor. Reproduced on main, and confirmed the throw goes away once the same transform also handles the empty-content case. A regression test loading the repro state would catch this directly.

rohan-patnaik · 2026-05-22T09:02:07Z

@mayrang thanks for catching the empty-content repro case. I addressed it in d896b1c by normalizing empty collapsible-content to contain a paragraph, and added a regression test that loads the serialized repro shape before dirtying the content node.

etrepum · 2026-05-22T15:29:16Z

The title and description of this PR doesn't match the pull request template https://raw.githubusercontent.com/facebook/lexical/refs/heads/main/.github/pull_request_template.md

rohan-patnaik · 2026-05-22T16:22:32Z

@etrepum thanks for pointing that out. I updated the title and PR body to follow the template format. Could you please take another look and let me know if this reads correctly now?

…nt-inline-children

etrepum · 2026-05-22T16:34:44Z

Title and description look good, no need to update the branch unless there's a conflict. As a first time contributor this will require a maintainer to approve the CI checks to run again which will slow down your review.

levensta · 2026-05-22T17:00:40Z

+const $wrapInlineContentChildren = (node: CollapsibleContentNode) => {
+  if (node.isEmpty()) {
+    node.append($createParagraphNode());
+    return;
+  }
+
+  let paragraph: ReturnType<typeof $createParagraphNode> | null = null;
+
+  for (const child of node.getChildren()) {
+    if ($isBlockElementNode(child)) {


$isBlockElementNode checks only ElementNode, although there are block decorators too

I also think it makes sense to export $wrapInlineNodes to lexical/utils and reuse it
https://github.com/facebook/lexical/blob/main/packages/lexical/src/LexicalSelection.ts#L3328-L3366

Thanks, that makes sense. I updated this in ec9e073: moved the inline wrapping helper into LexicalUtils, re-exported it through lexical / @lexical/utils, and reused it in the collapsible transform.

I also added a reg test covering the case you called out: block decorator children stay unwrapped, while inline content is still wrapped into paragraphs.

Checked locally too:

vitest --project unit --run packages/lexical-playground/__tests__/unit/CollapsibleContainerNode.test.ts

tsc --noEmit

I think the $wrapInlineContentChildren wrapper is not needed, it checks exactly the same thing as $wrapInlineNodes

ok I guess I agree that the pre-check duplicates what $wrapInlineNodes already does.

The only extra behavior in $wrapInlineContentChildren is the empty content case, where we add a paragraph. $wrapInlineNodes([]) would return no children, so removing that handling would regress the empty-content repro.

Would you prefer that I keep a very small wrapper for the empty case and always call $wrapInlineNodes, or inline that logic directly in the transform?

etrepum · 2026-05-23T05:45:32Z

  );
 }

+export function $wrapInlineNodes(nodes: LexicalNode[]) {


This new export should have documentation, an explicit return type, and ideally some tests. I know this is code just copied from elsewhere but the LineBreakNode special case seems a bit suspicious. Not sure that's really the correct behavior in all cases. I'm not sure that it's correct for [LineBreakNode, TextNode] to be indistinguishable from [TextNode] after wrapping.

That makes sense. I removed the new public export in e53545c and kept this PR scoped to the collapsible fix instead.

The wrapping logic is now local to the playground collapsible transform, and the net PR diff is back to the collapsible extension plus its regression tests. This keeps the empty-content fix and the block-decorator coverage without adding a new utility API.

etrepum · 2026-05-24T05:51:42Z

+    return;
+  }
+
+  let paragraph: ReturnType<typeof $createParagraphNode> | null = null;


Suggested change

let paragraph: ReturnType<typeof $createParagraphNode> | null = null;

let paragraph: ElementNode | null = null;

etrepum · 2026-05-24T05:55:45Z

+    }
+
+    if (paragraph === null) {
+      paragraph = $createParagraphNode();


Suggested change

paragraph = $createParagraphNode();

paragraph = child.createParentNode();

rohan-patnaik · 2026-05-25T04:43:39Z

@etrepum thanks, I pushed 90cdb1a with the wrapping cleanup.

Used ElementNode | null for the local wrapper type and switched the non-empty wrapping path to child.createParentElementNode().
Used the existing createParentElementNode() API since createParentNode() is not available on these nodes.

Checked locally:

vitest --project unit --run packages/lexical-playground/__tests__/unit/CollapsibleContainerNode.test.ts
tsc --noEmit

fix(playground): wrap inline collapsible content

444bd37

rohan-patnaik requested review from acywatson, etrepum, fantactuka, ivailop7, potatowagon and zurfyx as code owners May 22, 2026 07:11

vercel Bot deployed to Preview – lexical-playground May 22, 2026 07:12 View deployment

vercel Bot deployed to Preview – lexical May 22, 2026 07:12 View deployment

fix(playground): normalize empty collapsible content

d896b1c

meta-cla Bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label May 22, 2026

vercel Bot deployed to Preview – lexical May 22, 2026 09:03 View deployment

vercel Bot deployed to Preview – lexical-playground May 22, 2026 09:03 View deployment

etrepum added the extended-tests Run extended e2e tests on a PR label May 22, 2026

rohan-patnaik changed the title ~~Fix collapsible content inline child normalization~~ [lexical-playground] Bug Fix: Normalize collapsible content children May 22, 2026

Merge remote-tracking branch 'origin/main' into fix-collapsible-conte…

4510f0e

…nt-inline-children

vercel Bot deployed to Preview – lexical May 22, 2026 16:23 View deployment

vercel Bot deployed to Preview – lexical-playground May 22, 2026 16:23 View deployment

levensta reviewed May 22, 2026

View reviewed changes

fix: reuse inline wrapping for collapsible content

ec9e073

vercel Bot deployed to Preview – lexical-playground May 22, 2026 17:58 View deployment

vercel Bot deployed to Preview – lexical May 22, 2026 17:59 View deployment

etrepum reviewed May 23, 2026

View reviewed changes

fix: keep collapsible wrapping local

e53545c

vercel Bot deployed to Preview – lexical-playground May 23, 2026 20:30 View deployment

vercel Bot deployed to Preview – lexical May 23, 2026 20:30 View deployment

etrepum reviewed May 24, 2026

View reviewed changes

fix: use node parent for collapsible wrapping

90cdb1a

vercel Bot deployed to Preview – lexical-playground May 25, 2026 04:44 View deployment

vercel Bot deployed to Preview – lexical May 25, 2026 04:44 View deployment

etrepum approved these changes May 25, 2026

View reviewed changes

etrepum added this pull request to the merge queue May 25, 2026

Merged via the queue into facebook:main with commit eeefefc May 25, 2026
42 checks passed

etrepum mentioned this pull request May 28, 2026

v0.45.0 #8580

Merged

levensta mentioned this pull request May 31, 2026

Bug: "Children of root nodes must be elements or decorators" invariant isn't enforced #7625

Closed

	let paragraph: ReturnType<typeof $createParagraphNode> \| null = null;
	let paragraph: ElementNode \| null = null;

	paragraph = $createParagraphNode();
	paragraph = child.createParentNode();

Uh oh!

Conversation

rohan-patnaik commented May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Test plan

Before

After

Uh oh!

meta-cla Bot commented May 22, 2026

Action Required

Process

Uh oh!

vercel Bot commented May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mayrang commented May 22, 2026

Uh oh!

rohan-patnaik commented May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

etrepum commented May 22, 2026

Uh oh!

rohan-patnaik commented May 22, 2026

Uh oh!

etrepum commented May 22, 2026

Uh oh!

levensta May 22, 2026

Choose a reason for hiding this comment

Uh oh!

rohan-patnaik May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

levensta May 22, 2026

Choose a reason for hiding this comment

Uh oh!

rohan-patnaik May 22, 2026

Choose a reason for hiding this comment

Uh oh!

etrepum May 23, 2026

Choose a reason for hiding this comment

Uh oh!

rohan-patnaik May 23, 2026

Choose a reason for hiding this comment

Uh oh!

etrepum May 24, 2026

Choose a reason for hiding this comment

Uh oh!

etrepum May 24, 2026

Choose a reason for hiding this comment

Uh oh!

rohan-patnaik commented May 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

rohan-patnaik commented May 22, 2026 •

edited

Loading

vercel Bot commented May 22, 2026 •

edited

Loading

rohan-patnaik commented May 22, 2026 •

edited

Loading

rohan-patnaik May 22, 2026 •

edited

Loading

rohan-patnaik commented May 25, 2026 •

edited

Loading