Suppose that I want to support the following custom elements:

1. Valid: <element-one attribute-one="1"></element-one>
2. Valid: <element-two attribute-two="2"></element-two>
3. Invalid: <element-one attribute-two="2"></element-one>
4. Invalid: <element-two attribute-one="1"></element-two>

I can get close with this:

{
  CUSTOM_ELEMENT_HANDLING: {
    tagNameCheck: (tagName) => tagName.match(/^element-(one|two)$/),
    attributeNameCheck: (attr) => attr.match(/^attribute-(one|two)$/),
    allowCustomizedBuiltInElements: false,
  },
}

This would allow my valid examples to pass, but it would also allow my invalid examples to pass.

If the function signature of the attributeNameCheck could receive a second argument, representing the tagName, then the configuration could be defined as follows:

{
  CUSTOM_ELEMENT_HANDLING: {
    tagNameCheck: (tagName) => tagName.match(/^element-(one|two)$/),
    attributeNameCheck: (attr, tagName) => {
      if (tagName === 'element-one') {
        return ['attribute-one'].includes(attr);
      } else if (tagName === 'element-two') {
        return ['attribute-two'].includes(attr);
      } else {
        return false;
      }
    },
    allowCustomizedBuiltInElements: false,
  },
}

Is this potentially a feature that you'd be willing to support? Are there any security concerns or performance concerns that would make this unacceptable?