Proposal: new API to get text content across shadowRoot

[woody-li]

What problem are you trying to solve?

Currently, we already have HTML-related API: getHTML, setHTML, etc.
But there's no way to get text content across shadowRoot.

In some cases, we need to get real text content across element shadowRoots.
Such as a customized tooltip, check content in the auto-test frame, etc.

What solutions exist today?

No response

How would you solve it?

How about adding a new API, like getText or getTextContent?
Add an option like checkVisibility to deal with hidden elements. (The difference between innerText and textContent)

Anything else?

No response

[annevk]

How would this be different from the textContent getter?

[woody-li]

@annevk You can not use testContent to get real text content across shadowRoot.

[annevk]

You'll have to elaborate as this proposal is very much unclear to me.

[rniwa]

Perhaps you mean a new API like Node.prototype.getTextContent which takes an optional argument of shadow roots like getHTML?

[woody-li]

@rniwa Yeah, it is.

This proposal is meant to resolve how to get text content across shadowRoot.
Both innerText and textContent cannot get text across shadow DOM.

For example:

<div id="box">
  // some other elements.
  <my-element>
    #shadow-fragment
        <span>Some test content</span>
  </my-element>
</div>

In some cases, I want to get the real text content of the box element.
I have to query every element, check whether it's a custom element, and collect its inner elements again and again.
Then, use all elements to map a list of text content.

---

This proposal is meant to create a new API like getTextContent, using it to get the real text content.
And I think it should contain an options object argument.
One option for deciding whether to get text across shadowRoot.
One option is to get innerText or textContent.

---

This is my first proposal here.
Please let me know if there's any problem.
Thanks.

[zcorpan]

Such as a customized tooltip, check content in the auto-test frame, etc.

Can you elaborate on the use cases, please? It's difficult to evaluate an API proposal without understanding what problem is being solved.