Revert [Breaking Change][lexical] Bug Fix: Commit updates on editorSetRootElement(null) (#7023)

[potatowagon]

S482409: Editors on Meta's www were crashing with error: Got unexpected null or undefined on reloading the editor.

the root cause is a nullthrows at

  const parentElementDOM = nullthrows(
    editor.getElementByKey(parentElementNode.getKey()),
  );

MLCFloatingParagraphStylesPlugin.react.js in the www integration layer. More investigation needed to determine whats the best approach to handle null being returned to parentElementDOM.

PR7023 is suspected to cause the change, and carries more risk then expected. Reverting the sync diff with PR7023 mitigates the issue. sorry @etrepum to unblock the syncs, may we revert this PR first?

[vercel]

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

Name	Status	Preview	Comments	Updated (UTC)
lexical	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Jan 15, 2025 8:48am
lexical-playground	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Jan 15, 2025 8:48am

[github-actions]

size-limit report 📦

Path	Size
lexical - cjs	28.99 KB (-0.12% 🔽)
lexical - esm	28.81 KB (-0.1% 🔽)
@lexical/rich-text - cjs	37.96 KB (-0.01% 🔽)
@lexical/rich-text - esm	30.93 KB (+0.1% 🔺)
@lexical/plain-text - cjs	36.51 KB (-0.07% 🔽)
@lexical/plain-text - esm	28.15 KB (-0.03% 🔽)
@lexical/react - cjs	39.77 KB (+0.01% 🔺)
@lexical/react - esm	32.26 KB (+0.01% 🔺)

[etrepum]

I think the real question here is why is an editor setting the root element to null and also expecting the associated DOM to be non-null? I don't have a strong opinion on whether the reconciliation here should be immediate or not but having setRootElement(nonNull) work significantly differently than setRootElement(null) is confusing behavior and should at least be documented and tested as the expected semantics if we are to leave this in place.