This pull request is being automatically deployed with Vercel (learn more).
To see the status of your deployment, click below or on the icon next to each commit.

🔍 Inspect: https://vercel.com/fbopensource/lexical/7GSjfmz5djcMVxqfr77ppNfN6pnU
✅ Preview: https://lexical-git-child-conversion-fbopensource.vercel.app

New after API should cover forChild usage, or is there any new use cases you think won't work?

  u: (domNode: Node) => {
    return {
      node: null,
      after: (lexicalNodes) => {
        for (let i = 0; i < lexicalNodes.length; i++) {
          if ($isTextNode(lexicalNodes[i])) {
            lexicalNodes[i].toggleFormat('underline');
          }
        }
        return lexicalNodes;
      },
    };
  },

New after API should cover forChild usage, or is there any new use cases you think won't work?

  u: (domNode: Node) => {
    return {
      node: null,
      after: (lexicalNodes) => {
        for (let i = 0; i < lexicalNodes.length; i++) {
          if ($isTextNode(lexicalNodes[i])) {
            lexicalNodes[i].toggleFormat('underline');
          }
        }
        return lexicalNodes;
      },
    };
  },

Technically, it will, but in cases where the TextNodes are not direct children of the tag applying the formatting, we’d miss them. You’d have to visit every node in the subtree, not just the direct children. One example is:

<u><a>Hello</a></u>

In this case. I was thinking forChild is more convenient and performant than traversing the entire subtree after the transformation is complete.

Ah, got it! I think I was confused by the name, assuming it's only for immediate child nodes. Maybe can adjust it to reflect functionality

ul: {
  node: LexicalNode,
  after(Array<LexicalNode>) => Array<LexicalNode>,  <-- post process immediate children, returned value used as children
  forChild(LexicalNode) => void                     <-- recurse into tree, returned value ignored
}

Maybe other folks have ideas, or it can be merged as is too

Ah, got it! I think I was confused by the name, assuming it's only for immediate child nodes. Maybe can adjust it to reflect functionality

ul: {
  node: LexicalNode,
  after(Array<LexicalNode>) => Array<LexicalNode>,  <-- post process immediate children, returned value used as children
  forChild(LexicalNode) => void                     <-- recurse into tree, returned value ignored
}

Maybe other folks have ideas, or it can be merged as is too

Yea I used to call it "childConversion", which is also not obvious. I really hate the term "conversion" for some reason.

Hopefully some documentation will help. I am going to leave this open for a bit to give others a chance to weigh in, for sure.