fix: restore depth guard in getParentBlockInfo to prevent RangeError (blo-1103) by hedi-ghodhbane · Pull Request #2585 · TypeCellOS/BlockNote

hedi-ghodhbane · 2026-03-23T09:55:06Z

Summary

PR #2126 introduced a regression in getParentBlockInfo (mergeBlocks.ts). The depth guard that prevented calling $pos.before(0) on the top-level document node was removed.

When pressing Delete at the end of the last block, $pos.depth is 1, so depth becomes 0, and $pos.before(0) throws:

RangeError: There is no position before the top-level node

This PR restores the guard by returning undefined when depth < 1, before the .before() call.

Test plan

Press Delete at the end of the last block in the editor → should be a no-op, no error
Press Delete in an empty editor → should be a no-op, no error
Forward-delete between blocks should still work as expected (merging blocks)

Fixes #2584

Summary by CodeRabbit

Bug Fixes
- Prevented an edge-case failure when merging blocks at the document root by ensuring no invalid parent resolution occurs for top-level positions, improving stability during block merges.
Tests
- Added a unit test covering the top-level block case to prevent regressions and validate correct behavior when no parent block exists.

PR TypeCellOS#2126 removed the depth guard that prevented calling `$pos.before(0)` on the top-level document node. When pressing Delete at the end of the last block, `$pos.depth` is 1, so `depth` becomes 0, and `$pos.before(0)` throws "RangeError: There is no position before the top-level node". This restores the guard by returning undefined when depth < 1, before the `.before()` call. Fixes TypeCellOS#2584

vercel · 2026-03-23T09:55:11Z

@hedi-ghodhbane is attempting to deploy a commit to the TypeCell Team on Vercel.

A member of the Team first needs to authorize it.

coderabbitai · 2026-03-23T09:55:22Z

📝 Walkthrough

Walkthrough

A depth guard was added to getParentBlockInfo so it returns undefined when the computed depth is less than 1, preventing calls to $pos.before(0) that previously threw a RangeError for top-level positions.

Changes

Cohort / File(s)	Summary
Merge logic `packages/core/src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.ts`	Added a guard: return `undefined` when computed `depth < 1` to avoid calling `$pos.before(0)` and throwing a RangeError at top-level positions.
Tests `packages/core/src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.test.ts`	Imported `getParentBlockInfo` and added a test asserting it returns `undefined` for a top-level block position (verifies `$pos.depth - 1 < 1` before assertion).

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Poem

🐰
I nibbled at depth where the root once bled,
Placed a gentle guard where a RangeError fled.
Now Delete hops lightly, no crash in sight,
Merges stay calm through day and night. 🥕

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description check	✅ Passed	The description provides a clear summary of the regression, explains the root cause, and outlines the fix with test plan, but lacks some template sections like Impact and Checklist.
Linked Issues check	✅ Passed	The PR fully addresses issue `#2584` by restoring the depth guard in getParentBlockInfo and adding a test case verifying the fix prevents the RangeError.
Out of Scope Changes check	✅ Passed	All changes are directly related to fixing the regression in getParentBlockInfo; the test file update is a necessary part of addressing the linked issue.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Title check	✅ Passed	The title accurately describes the main change: restoring a depth guard in getParentBlockInfo to prevent a RangeError. This is directly supported by the code changes (5 lines added as a guard check) and the PR objectives.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

nperez0111 · 2026-03-23T12:10:08Z

@hedi-ghodhbane thanks for contributing this, ideally we'd have a test case which proves this behavior

pkg-pr-new · 2026-03-23T12:16:34Z

Open in StackBlitz

@blocknote/ariakit

npm i https://pkg.pr.new/@blocknote/ariakit@2585

@blocknote/code-block

npm i https://pkg.pr.new/@blocknote/code-block@2585

@blocknote/core

npm i https://pkg.pr.new/@blocknote/core@2585

@blocknote/mantine

npm i https://pkg.pr.new/@blocknote/mantine@2585

@blocknote/react

npm i https://pkg.pr.new/@blocknote/react@2585

@blocknote/server-util

npm i https://pkg.pr.new/@blocknote/server-util@2585

@blocknote/shadcn

npm i https://pkg.pr.new/@blocknote/shadcn@2585

@blocknote/xl-ai

npm i https://pkg.pr.new/@blocknote/xl-ai@2585

@blocknote/xl-docx-exporter

npm i https://pkg.pr.new/@blocknote/xl-docx-exporter@2585

@blocknote/xl-email-exporter

npm i https://pkg.pr.new/@blocknote/xl-email-exporter@2585

@blocknote/xl-multi-column

npm i https://pkg.pr.new/@blocknote/xl-multi-column@2585

@blocknote/xl-odt-exporter

npm i https://pkg.pr.new/@blocknote/xl-odt-exporter@2585

@blocknote/xl-pdf-exporter

npm i https://pkg.pr.new/@blocknote/xl-pdf-exporter@2585

commit: 55be6bd

…0.47.1 https://sonarly.com/issue/17525?type=bug Pressing the Delete (forward-delete) key at the end of the last block in any BlockNote rich text editor throws an unhandled RangeError, crashing the editor. This is caused by a regression in BlockNote 0.47.1's `getParentBlockInfo` function. Fix: Pinned `@blocknote/mantine`, `@blocknote/react`, `@blocknote/xl-docx-exporter`, and `@blocknote/xl-pdf-exporter` back to exact version `0.47.0` in `packages/twenty-front/package.json`. BlockNote 0.47.1 introduced a regression in `@blocknote/core` (upstream PR #2126, "handle more delete key cases") that removed a depth guard in `getParentBlockInfo()`. This causes `$pos.before(0)` to be called when the cursor is at the end of the last top-level block and the user presses Delete, throwing `RangeError: There is no position before the top-level node`. The dependabot commit `6a2e0182ab` bumped these from exact `"0.47.0"` to `"^0.47.1"`, pulling in the buggy release. This pin restores the exact versions from the original deliberate upgrade (commit `4ed09a3feb`). The server-side `@blocknote/server-util` is left unchanged at `^0.47.1` since it's a headless package that doesn't use ProseMirror's view layer or keyboard handlers — the regression only affects browser-side editing. Once BlockNote publishes a fix (tracked in upstream issue [#2584](TypeCellOS/BlockNote#2584), fix PR [#2585](TypeCellOS/BlockNote#2585)), these packages can be upgraded again.

hedi-ghodhbane · 2026-03-23T14:47:37Z

@nperez0111 I'm not sure what you mean by a test case. You mean a record / demo of the issue? or a Unit test?

nperez0111 · 2026-03-23T14:49:40Z

@hedi-ghodhbane I mean more of a unit test which proves that this does what it intended to. There is a test file next to the one you modified: https://github.com/hedi-ghodhbane/BlockNote/blob/4bbfd90ee0749fa038a500e043021d19bbb43916/packages/core/src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.test.ts

Verifies that getParentBlockInfo returns undefined when called on a top-level block (depth < 1), preventing the RangeError that was thrown by $pos.before(0). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

hedi-ghodhbane · 2026-03-23T15:03:34Z

@nperez0111 Added a unit test in 41bd6683b that verifies getParentBlockInfo returns undefined for top-level blocks instead of throwing.

Without the depth guard (reproduces the bug):

× |@blocknote/core| > Test mergeBlocks > getParentBlockInfo returns undefined for top-level block
FAIL  RangeError: There is no position before the top-level node
    884|  throw new RangeError("There is no position before the top-…
 ❯ Module.getParentBlockInfo src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.ts:20:32
 Test Files  1 failed (1)
      Tests  1 failed | 11 passed (12)

coderabbitai

🧹 Nitpick comments (1)

packages/core/src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.test.ts (1)

80-88: Tighten this regression test by asserting the depth-guard precondition.

Right now the test only checks the return value. Please also assert that the resolved position is in the depth < 1 guard path so the test can’t pass for an unrelated undefined case.

Suggested test hardening

   const beforePos = getPosBeforeSelectedBlock();
   const doc = getEditor()._tiptapEditor.state.doc;
+  const resolved = doc.resolve(beforePos);
+  expect(resolved.depth - 1).toBeLessThan(1);
   const result = getParentBlockInfo(doc, beforePos);

   expect(result).toBeUndefined();

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In
`@packages/core/src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.test.ts`
around lines 80 - 88, The test should also assert the resolved position meets
the depth guard to ensure we exercise the "top-level" branch: after computing
beforePos via getPosBeforeSelectedBlock() and doc from
getEditor()._tiptapEditor.state.doc, resolve the position (use
doc.resolve(beforePos) or the same helper used in getParentBlockInfo) and assert
resolvedPos.depth < 1 before calling getParentBlockInfo; keep the existing
expect(result).toBeUndefined() to validate the return value for the top-level
case (functions referenced: getParentBlockInfo, getPosBeforeSelectedBlock,
getEditor).

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In
`@packages/core/src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.test.ts`:
- Around line 80-88: The test should also assert the resolved position meets the
depth guard to ensure we exercise the "top-level" branch: after computing
beforePos via getPosBeforeSelectedBlock() and doc from
getEditor()._tiptapEditor.state.doc, resolve the position (use
doc.resolve(beforePos) or the same helper used in getParentBlockInfo) and assert
resolvedPos.depth < 1 before calling getParentBlockInfo; keep the existing
expect(result).toBeUndefined() to validate the return value for the top-level
case (functions referenced: getParentBlockInfo, getPosBeforeSelectedBlock,
getEditor).

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 82547264-af75-450d-a460-65416afc1ed9

📥 Commits

Reviewing files that changed from the base of the PR and between 4bbfd90 and 41bd668.

📒 Files selected for processing (1)

packages/core/src/api/blockManipulation/commands/mergeBlocks/mergeBlocks.test.ts

Verify that the resolved position's depth is < 1 before calling getParentBlockInfo, ensuring we exercise the top-level branch. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

hedi-ghodhbane · 2026-03-24T08:44:25Z

Coderabbit comment fixed :)

nperez0111

Thanks @hedi-ghodhbane!

vercel · 2026-03-24T08:59:20Z

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

Project	Deployment	Actions	Updated (UTC)
blocknote	Error		Mar 24, 2026 9:01am
blocknote-website	Ready	Preview	Mar 24, 2026 9:01am

YousefED requested a review from matthewlipski March 23, 2026 10:02

test: add unit test for getParentBlockInfo depth guard

41bd668

Verifies that getParentBlockInfo returns undefined when called on a top-level block (depth < 1), preventing the RangeError that was thrown by $pos.before(0). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

coderabbitai bot reviewed Mar 23, 2026

View reviewed changes

test: assert resolved position depth triggers the guard

55be6bd

Verify that the resolved position's depth is < 1 before calling getParentBlockInfo, ensuring we exercise the top-level branch. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

nperez0111 approved these changes Mar 24, 2026

View reviewed changes

vercel bot had a problem deploying to Preview – blocknote March 24, 2026 09:01 Failure

vercel bot deployed to Preview – blocknote-website March 24, 2026 09:01 View deployment

YousefED changed the title ~~fix: restore depth guard in getParentBlockInfo to prevent RangeError~~ fix: restore depth guard in getParentBlockInfo to prevent RangeError (blo-1103) Mar 25, 2026

Uh oh!

Conversation

hedi-ghodhbane commented Mar 23, 2026 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Summary by CodeRabbit

Uh oh!

vercel bot commented Mar 23, 2026

Uh oh!

coderabbitai bot commented Mar 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Estimated code review effort

Poem

Uh oh!

nperez0111 commented Mar 23, 2026

Uh oh!

pkg-pr-new bot commented Mar 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

hedi-ghodhbane commented Mar 23, 2026

Uh oh!

nperez0111 commented Mar 23, 2026

Uh oh!

hedi-ghodhbane commented Mar 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

coderabbitai bot left a comment

Choose a reason for hiding this comment

Uh oh!

hedi-ghodhbane commented Mar 24, 2026

Uh oh!

nperez0111 left a comment

Choose a reason for hiding this comment

Uh oh!

vercel bot commented Mar 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

hedi-ghodhbane commented Mar 23, 2026 •

edited by coderabbitai bot

Loading

coderabbitai bot commented Mar 23, 2026 •

edited

Loading

pkg-pr-new bot commented Mar 23, 2026 •

edited

Loading

hedi-ghodhbane commented Mar 23, 2026 •

edited

Loading

vercel bot commented Mar 24, 2026 •

edited

Loading